The absolute calibration of HJ-1B thermal infrared channel is achieved by 7 times cross-calibration of Qinghai Lake in 2010, between the HJ-1B thermal infrared channel and MODIS 31, 32 channels. At the same time, the intercept of calibration coefficients is calculated. First, the spatial match are completed between the HJ-1B IRS image and MODIS 31, 32 channels image. Then, the spectral response of HJ-1B IRS thermal channel and MODIS 31, 32 channels are used for the spectral match. Based on the SST algorithms of MODIS, the MODTRAN4.0 is used to simulate the radiative transfer results under the condition of variety surface temperature of the Qinghai Lake, the atmospheric profile, observational geometry combined with the spectral response functions of the remote sensors. The quantitative linear relationship can be established with the simulation results. The gain and offset of cross-calibration are 52.7788 1/(W/(m2•sr•μm)), -55.4137. The result needs further validation of longer sequence.

With the development of Web services technology, the number of service increases rapidly, and it becomes a challenge task that how to efficiently discovery the services that exactly match the user's requirements from the large scale of services library. Many semantic Web services discovery technologies proposed by the recent literatures only focus on the keyword-based or primary semantic based service's matching. This paper studies the rules and rule reasoning based service matching algorithm in the background of large scale services library. Firstly, the formal descriptions of semantic web services and service matching is presented. The services' matching are divided into four levels: Exact, Plugin, Subsume and Fail and their formal descriptions are also presented. Then, the service matching is regarded as rule-based reasoning issues. A set of match rules are firstly given and the related services set is retrieved from services ontology base through rule-based reasoning, and their matching levels are determined by distinguishing the relationships between service's I/O and user's request I/O. Finally, the experiment based on two services sets show that the proposed services matching strategy can easily implement the smart service discovery and obtains the high service discovery efficiency in comparison with the traditional global traversal strategy.

3D urban area model has a variety aspect of application. The most exist methods follows a three step pipeline. In this paper, we follow the pipeline to reconstruct the 3D model of urban area. An improved covariance analysis method is proposed to remove the irrelevant part of LiDAR. This preprocessing method is more robust than the original method for the low resolution LiDAR data. Also, a classification method based on the Markov Random Field on the connected component is proposed in this paper. The experiment shows our method can reconstruct 3D model of urban area from low resolution LiDAR data.

Drought has been a frequently happened type of disaster in China, and it has caused massive losses to people's lives. Especially the drought happened in Shandong province in the late 2010, which was recognized as the severest in the past five hundred years in some areas. Evaluation must be done in order to make proper rescue plans. Instead of collecting data site by site, remote sensing is an efficient way to acquire data in a large area, which is very helpful for drought identification. Some normal ways in remote sensing for drought analysis are explained and compared in this paper, and then the VSWI method is chosen to evaluation the drought in Shandong province. Because of its free data policy and wide availability, the data sets acquired by Terra-MODIS are chosen to identify the drought severity in Shandong province. From the drought severity level images we can see that almost the whole area of Shandong province was lack of water except the Weishan Lake and eastern coastline regions where large area of water exists. The southwest region, including Heze and Jining, is in moderate drought condition, where it is used to be an important grain-producing area. This drought condition will inevitably put a negative effect on its grain production. The central and southern areas were in severe drought condition, but fortunately these areas are of hills and mountains, so the drought will only affect the lives of residents. The northern parts, including Dezhou and Bingzhou areas, were also in severe drought condition, and these regions are also important for grain-producing, so the severe drought disaster will lead to a sharp grain output cut. This analysis results will not only shed light on the rescue process, but also give the government some clues on how to maintain the grain supply safety.

The difficulty existing in synthetic aperture radar (SAR) image classification is large amounts of unpredictable and inestimable speckle, leading to degradation of the image quality and concealing important objectives of interest. By exploiting an efficient image features extraction technique, bag-of-visual-words (BOV) for its ability of 'midlevel' feature representation, and a new developed non-local (NL-) means denosing method suitable for multiplicative speckle, we present a novel and effective BOV framework for SAR image classification. Compared with the other two representative algorithms, the experimental results show that the proposed algorithm has obtained more satisfactory and cogent classification performance and performed more robustness to SAR speckle.

This paper presents our initial research on calibrating developed Multi-vision Oblique Photogrammetry System (MOPS). First, 2D point control from so-called Radial Alignment Constrain (RAC) by Tsai's two-stage calibration is outlined and its potential issues to camera calibration accuracy are discussed as well. Second, we make well study on image radial distortion when principle point parameters are ignored and with perspective invariant of space line, one new radial distortion correctness algorithm is proposed based on 2D line control. Third, procedure of Tsai's two-stage calibration is modified to corporate with proposed radial distortion correctness algorithm and new steps to calibrate digital cameras are summarized in detail. Finally, calibration test was implemented on selected digital camera Nikon P5100 by means of 2D chessboard as reference object. The comparison of proposed approach with Tsai's two-stage calibration is also given in this paper and valuable conclusions are conducted as well.

A new method, sparse representation based spectral clustering (SC) with Nyström method, is proposed for synthetic aperture radar (SAR) image segmentation. Different from the conventional SC, this proposed technique is developed by using the sparse coefficients which obtained by solving l1 minimization problem to construct the affinity matrix and the Nyström method is applied to alleviate the segmentation process. The advantage of our proposed method is that we do not need to select the scaling parameter in the Gaussian kernel function artificially. We apply the proposed method, k-means and the classic spectral clustering algorithm with Nyström method to SAR image segmentation. The results show that compared with the other two methods, the proposed method can obtain much better segmentation results.

Automatic generation of DEM from LIDAR point clouds is attractive to photogrammetry community. This paper explores the possibility of using Multi-Scale SVM technique to classify untextured Lidar data into ground points and non-ground points so that DEM can be generated efficiently. First, irregular LIDAR point clouds are rasterized and a set of features including local height variation, min/max slope, plane flatness/direction and laser return intensity are generalized as well. Second, we establish Multi-Scale SVM classification levels by implementing SVM classier at different scale-space of Lidar data and one defined conditional probabilistic model is computed to make final classification. Finally, adaptive medium filter is implemented to smooth the isolated ground points mixed with little non-ground points and because the removal of non-ground points left quite a lot "blank holes", we further triangulate smoothed non-ground points to generate DEM automatically. The experimental results prove to be quite significant for real applications.

Spectral unmixing in hyperspectral remote sensing image has been widely researched in the last two decades. N-FINDR algorithm is one of the most classical and commonly-used endmember extraction algorithms. Nevertheless, it is a timeconsuming task that cannot meet the time requirement of many applications. In order to make N-FINDR computationally feasible, we consider parallel implementation of N-FINDR algorithm on hybrid multiple-core CPU and GPU parallel platform. First, a multi-core CPU-based parallel N-FINDR algorithm is considered based on a modified N-FINDR with two improvements. And by using the increasing programmability and parallelism of commodity GPU, a GPU-based parallel N-FINDR is presented. Finally, by taking advantages of the capability of the aforementioned algorithms, a hybrid multiple-core CPU and GPU parallel N-FINDR is proposed by using a virtual thread technique and an adaptive algorithm in which the computational load can be adaptively adjusted according to the capability of CPU and GPU. In experiment, our proposed parallel N-FINDR algorithms improved the accuracy of the original N-FINDR algorithm, and most importantly, they greatly improved the performance of N-FINDR algorithm.

Because the terrain of mountain glacier is usually very rugged, it is hard to measure glaciers and estimated their changes in larger area by conventional measuring method. With fast development of remote sensing technique, synthetic aperture radar (SAR) interferometry is used for glacier monitoring with the ability of all-time and all-weather. Although interferometric coherence is a very good index to glacier, it is difficult to distinguish glacier area from non-glacier area when their coherence is similar. In this case, interferometric phase can play an important role to identify glacier. In this paper, phase texture analysis method is proposed to extract glacier. 8 texture features were analyzed based on co-occurrence matrix (COM), including mean, variance, homogeneity, contrast, dissimilarity, entropy, second moment, and correlation. Among them, variance, contrast and dissimilarity can distinguish glacier from non-glacier clearly most, so they are chosen for RGB combination. Then the RGB combination image is classified into several land covers by maximum likelihood classification (MLC). With post-classification processing, glacier area can be extracted accurately. Landsat TM images validate the proposed method.

Three key techniques in registration of CBERS-02B image are introduced in this paper. Including an improved image matching strategy aided by a voting algorithm, the voting algorithm directly estimates the transformation parameters of image instead of finding the one to one corresponding, making the image matching much more efficient and reliable. Beside that, the method of handling the interior deformation problem caused by the incorrect splicing of three sub CCD arrays is introduced. At last, a virtual combination of satellite image technique is introduced, the adjacent satellite images in the same orbit are virtually combined to form a large image which is taken as the orientation unit, purposed on accelerating the orientation process speed and decreasing the probability of orientation failure. With the techniques mentioned above, the two main problems about CBERS-02B image: the inaccurate ephemeris parameters and the poor internal accuracy caused by the incorrect splicing of three sub CCD arrays are solved successfully, the absolute positioning accuracy reaches 1:10000 mapping accuracy and the internal accuracy reaches sub-pixel after the orientation process. It indicates that extensive usage of Chinese high resolution satellite imagery may become possible and practical.

In this paper, we address the problem of extracting the edge of power lines from aerial images, which is a critical step for the identification of the components equipped on power lines and for the diagnosis of the broken-strands on power lines. As for the problem, a novel idea is proposed based on the fact that the textural differentiation of Dissimilarity distributions between power line and background in a square window can be represented by some points of local peaks on the statistical curves such that the edge of power lines can be detected. Based on the novel idea, an algorithm named Square Window wIth Four symmeTrical axiSes (SWIFTS) is developed. Experiments and analyses demonstrate that the SWIFTS algorithm has better performance than other classical method in terms of extracting the edge of power lines, preserving the bent information of the edge at the point of broken-strand.

By the advantages on its' early time, high resolution and cheap price compared with other high resolution satellite images, CORONA image can be used for change detection studies, DEM extraction, tectonic studies and other applications. CORONA image does not digitally record surface information like other current remote sensing images, but through the exposure on the film. Because of the recording form, it contains serious brightness deviation between different images. In addition, the panoramic photography of CORONA generates the geometric distortion. In this paper the region of Yili river in Xinjiang province was selected as the experiment area. Through the study on the pre-process of CORONA images, it finally improves the quality of image that could satisfy the need of GIS mapping and provide foundation for the follow-up process of image analysis and information extraction.

The fusion of multi-source data is one of the most promising techniques for improved classification of remote sensing images. This paper utilized the images by fusing the high-resolution optical and SAR images for land use classification. The high-resolution optical and SAR images present the land features for many aspects and supplement information to each other because of their different imaging modes and wavebands, so classical methods such as Hue - Intensity -Saturation(HIS) transform based, Brovey(color normalized),Principal Component Substitution(PCS) approaches and wavelet-based method are used in fusing process, then the fused results been evaluated through the calculation of some quantitative index(Mean grey value, Standard error, Entropy, Definition). The wavelet-based fused image shows better effect than others. A MLC(maximum likelihood classification) method was employed to the wavelet-based fused image ,Classification accuracy was assessed using high-resolution aerial orthophotos. The overall accuracy for six classes(Settlement, River, Agricultural field, Country road, Tree nurseries, Bare land) was found to be 94.36% with kappa coefficient of 0.92.

In this paper, a simulation method is introduced to generate synthetic aperture radar (SAR) image based on ray tracing algorithm. 3D models of buildings, which are triangulated and described with vectors, are introduced into the simulator and then the simulated images can be generated under different viewing configuration. The simulation consists of three steps -modeling of the scene, tracing and generation of high resolution SAR images. 3D models of man-made objects are illuminated by a virtual antenna whose signal is simplified by rays sent to the objects and back to the sensor. Then the intensity map of rays is gridded into the SAR images. In the end, two buildings, one with a plane roof and the other with a gable roof, are imported into the simulator under different viewing configuration. The effects of layover, shadow and double bounce are simulated correctly in geometrically. So the simulator can be used for some interpret complex SAR images which are composed of buildings.

The relationship between image quality and optical Q is sensitive to the system design parameters such as modulation transfer function (MTF), signal-to-noise ratio (SNR), and ground sampled distance (GSD). In this paper, two concepts about Q and NIIRS for remote sensing imager quality analysis and design are presented, and each part physics mean in model and formula are illustrated separately, the influence of Q on the image quality of remote sensing system is specific analyzed, the system performance of Q = 1 and Q = 2 of remote sensing base on NIIRS are compared, the interaction between remote sensing image quality and Q is obtained based on NIIRS by image simulations and comparing with each other simulation results, the optimum design of remote sensors are mainly discussed. Finally, a method to improve the image quality scale is posed for remote sensing systems, the image quality scale of remote sensing systems can be improved by increasing along-scan sampling. Comparisons with preliminary results show that the interaction between image quality and Q is very sensitive to the system design parameters although more validation experiments are needed.

Cloud is the common phenomenon in the obtaining and application of the visible-light remote sensing image, and these images are usually degraded by the turbid medium (e.g., fog, cloud) in the atmosphere. This article researched the defogging techniques of the single satellite remote sensing image on the basis of the dark channel prior and the imaging model of the cloud images, and proposed a new approach to calculate the airlight in view of problems of asymmetric thickness of cloud in the satellite remote sensing images and unreasonable estimation of the airlight in the existing defogging algorithms. In my approach, cloud thickness information was added to effectively avoid the phenomenon of 'over-treatment' and 'under-treatment'. Finally, the feature of scientific and effective were validated by the successful testament of respectively using medium-resolution and high-resolution satellite remote sensing data.

Lakes in arid regions of Central Asia act as essential components of regional water cycles, providing sparse but valuable water resource for the fragile ecological environments and human lives. Lakes in Central Asia are sensitive to climate change and human activities, and great changes have been found since 1960s. Mapping and monitoring these inland lakes would improve our understanding of mechanism of lake dynamics and climatic impacts. ICESat/GLAS satellite laser altimetry provides an efficient tool of continuously measuring lake levels in these poorly surveyed remote areas. An automated mapping scheme of lake level changes is developed based on GLAS altimetry products, and the spatial and temporal characteristics of 9 typical lakes in Central Asia are analyzed to validate the level accuracies. The results show that ICESat/GLAS has a good performance of lake level monitoring, whose patterns of level changes are the same as those of field observation, and the max differences between GLAS and field data is 3cm. Based on the results, it is obvious that alpine lakes are increasing greatly in lake levels during 2003-2009 due to climate change, while open lakes with dams and plain endorheic lakes decrease dramatically in water levels due to human activities, which reveals the overexploitation of water resource in Central Asia.

Feathering is a most widely used method in seamless satellite image mosaicking. A simple but effective algorithm - double regions growing (DRG) algorithm, which utilizes the shape content of images' valid regions, is proposed for generating robust feathering-line before feathering. It works without any human intervention, and experiment on real satellite images shows the advantages of the proposed method.

The building recognition from high resolution remote sensing images is an internationally advanced research field. But there are still many difficult to be solved. In this paper, in order to extract and recognize the special building from the high resolution remote sensing images, a new algorithm is presented. First, a self-adaptive Average Absolute Difference Maximum edge enhancement algorithm is presented to enhance the edge of the building and suppress the background at one time. Second, the locally self-adaptive segmentation algorithm is implemented to obtain the binary image. Third, the thinning algorithm is implanted to obtain the single pixel edge of the building and a pruning algorithm is necessary in order to reduce the computation times in the following process. The following step is obtained by the Hough transformation to make the edge of the building into polygon. Finally, an Interior angle chain (IAC) is proposed to recognize the building with different shape. The experimental results demonstrate that the new algorithm can extract the special shape building quickly and accurately.

With the technological development background of "3S", the research area is Yachi demonstrate area in Bijie city of Guizhou. For traditional methods, NDVI is always used to extract Vegetation information from remote sensing image. Based on theory of dimidiate pixel model,the study region defined NDVISsoil and NDVISveg,vegetation, the inversion dimidiate pixel model was established based on vegetation index(NDVI) to extract vegetation coverage, vegetation coverage can be calculated by using the model of quantifying vegetation fraction from normalized difference vegetation,then acquired relationships between NDVI Serial Images and landscape pattern karst rock-desertification, finally yachi demonstrate area gainer karst rocky desertification landscape pattern of distribution. Using this methods of karst rocky desertification landscape pattern research has significantly improved the accuracy of classification, providing a scientific basis to ecological restoration of karst rocky desertification and other integrated control.

In this paper, the support vector machine (SVM) algorithm was utilized to tackle the classification of high resolution images from airborne digital sensor systems. Firstly, the original image was classified using SVM of four common types of kernel functions, namely linear, polynomial, RBF and sigmoid function, and the SVM with RBF kernel function can achieve the most satisfactory result. On the other hand, Getis-Ord Gi, one type of local spatial statistics, had been calculated with varying lags from 1 to 10. When classifying Gi image with lag of 3 using SVM of the RBF kernel function, an overall accuracy of 95.66% was achieved, which is more satisfactory than the result from the original image. The result shows that Gi images with lags less than the variogram range can be used instead of the original multi-spectral image to improve classification accuracy between features with similar spectral characteristics like trees and lawns, as a result, to increase the overall classification accuracy.

Dynamic monitoring forest is critical for forest management and protection, while existing satellites hardly meet the requirements of the temporal and spatial resolution for forest mapping. HJ-1 A/B, recently launched by China, can form the 700m-wide multi-spectrum image with 30-m resolution of any location every 2 days, also leading to large data to process. In this study, seven-feature approach, utilizing Principal Components, NDVIs, and DEM, is developed to map forest effectively, promising potential applications of HJ-1 A/B in land resources.

Aiming at limitations of existing multiresolution analysis (MRA) fusion methods, this paper proposes a new fusion method which combines curvelet and wavelet transform. Curvelet transform processes edges better than wavelet transform does. While wavelet transform handles smooth area better than curvelet transform does. As an image often includes more than one feature, the proposed method is conducted on the basis of region segmentation and use Àtrous wavelet transform (ATWT) to fuse smooth areas and fast discrete curvelet transform (FDCT) to fuse areas with edges. Furthermore, an optimal objective function defined based on a balance between spectral preservation and spatial resolution improvement is put forward to search optimal segmentation threshold. The optimal fusion result can be obtained by fusion processing through the optimal segmentation threshold. Landsat TM multispectral (MS) images and SPOT Panchromatic (Pan) image covering a region of Wuhan in Hubei province are tested to assess this proposed method. Visual evaluation and statistics analysis are employed to assess the quality of fused images of different methods. The proposed method demonstrates best results among methods being tested in this study. So by combining attributes of both transforms, it is possible to get better image fusion result than by using wavelet and curvelet individually.

Traditional shadow detection methods are usually detected shadow areas by the single threshold in shadow feature map. This leads to the detection results susceptible to affect by noise, and some special target (high-bright objects and green vegetation etc.) susceptible to misdetection. In this paper, a shadow detection method is proposed based on pulse coupled neural network (PCNN). The model can ignore small differences of pixels values in one area, because the network output is not only associated with the pixel brightness but also associated with pixel spatial location. Firstly, a new shadow feature map is build. Then PCNN model is applied to get optimal detection result with max entropy. The experimental results showed that the proposed model performed better than the single threshold models.

Because of the influence of atmosphere, the characteristics of the sensor and the image recording system, the quality of on-orbit obtained satellite images maybe get lower. The radiation distortion is caused by various factors, such as the blurring of the imaging system, non-linear amplitude response, vignetting and shadows, transmission noise, atmospheric interference, the changes of surface exposure and the radiance changes with the angle. According to the actual characteristics of high-resolution satellite images, we presented a radiation correction method which solved the image distortion problem. The radiation correction can provide a good basis for the image geometric correction, image fusion, etc. By the actual obtaining data of satellite on-orbit experiments, this program solves the problems of chrominance, mosaics, correction, drop-out lines, and noise existed in the images which are obtained by the raw data after decryption and decompression. The radiation correction can provide a good basis for the image mapping process, image fusion, etc.

The technology of Mask image dodging based on Fourier transform is a good algorithm in removing the uneven luminance within a single image. At present, the difference method and the ratio method are the methods in common use, but they both have their own defects .For example, the difference method can keep the brightness uniformity of the whole image, but it is deficient in local contrast; meanwhile the ratio method can work better in local contrast, but sometimes it makes the dark areas of the original image too bright. In order to remove the defects of the two methods effectively, this paper on the basis of research of the two methods proposes a balance solution. Experiments show that the scheme not only can combine the advantages of the difference method and the ratio method, but also can avoid the deficiencies of the two algorithms.

Here we analyzed many kinds of Net primary productivity (NPP) calculating methods and chose the Monteith model based on the Light Utility Efficiency(LUE) to estimate the NPP value of Heilongjiang province in 2003. In this model, the NOAA/AVHRR data, climate data and eradiation data were used to calculate the NPP with the support of GIS technology. Then the spatial change of NPP was analyzed and the spatial map of NPP was drawn. The result showed that the mean value of NPP in the study area is 329.2gC / (m²·a). There is a consistent trend between the spatial distribution of NPP and temperature and rainfall. The NPP values of southeastern and central area are higher than that of western drought area and Daxingan Mountains because of their better humidity and heat. The NPP values of different vegetation have significant differences. They decrease in the sequence of forest, shrub, cropland and grassland. The NPP values have significantly positive correlation with mean annual temperature, rainfall and NDVI.

In the paper, the parameters of hyperspectral data of the environment disaster reduction satellites have been introduced, firstly. Then the pro-processing methods for hyperspectral data have been elaborated according to the characteristics of the hyperspectral sensor of the environment disaster reduction satellites. After analysis the problems existing in the pre-processing of hyperspectral data, the hyperspectral data have been employed to classify the land features. The experimental evaluation shows that the performance of classifying the hyperspectral data of environment disaster reduction satellites is excellent.

The traditional fog detection methods based on remote sensing mainly used polar-orbiting satellite data (MODIS, AVHRR) to establish the fog detection model, but the transit time are always later and the time resolution are about one day, so they cannot be good to meet the requirements of fog detection. FY2E (a geostationary satellite) data will be chosen to build the day fog detection model for its high time resolution (one hour) and relatively rich spectrum. In this paper object-oriented thinking and texture differences between fog and cloud were introduced to the fog detection model. According to the simulation results of streamer radiative transfer model based on FY2E data, Snow Separation Index (SSI) will be built to extract snow from fog and Cloud Separation Index (CSI) will be built to extract low clouds from fog. A day fog detection model for FY2E data will be built based on object-oriented thinking and several characteristic parameters. The experiments shows that the fog detection model proposed in this paper achieved good results.

Registration is an essential preprocessing procedure for extensive application of remote sensing images. As a robust feature-based descriptor used for image registration, SIFT shows great superiority in describing graphical structure with its invariant characteristic to illumination, rotation and scale changes. However, it is not well performed under low contrast and weak edge conditions, which has a high appearance with remote sensing images. In this paper, a novel remote sensing image registration is proposed to cope with issues that SIFT hardly performs an acceptable accurate feature point matching. Frequency based and spatial based algorithms, which namely are Log Gabor wavelet transformation and discrete probability relaxation respectively, are introduced as a compound strategy to figure out mismatching candidates. Experiment results show that the proposed method could get lower root mean square error (RMSE) and higher correct matching ratio (CMR) than other typical methods utilized to compare with. As proposed compound outlier removal strategy is out of relying on transformation modal, it is considerably feasible and well performed without RANSAC algorithm. Furthermore, this method is invariant to shift, rotation, illumination and scale change dues to SIFT participation.

Karst rocky desertification is a significant environmental and ecological problem in Southwest China. In this paper, the spectral information, spatial context and topography information were utilized to synthetically discriminate the Karst rocky desertification degree, which are derived from The SPOT satellite imagery and DEM. By the back-propagation neural network, we proposed the classification model structure and classified the rocky desertification levels in Du'an County of Guangxi province, China. The results verified the classification model of Karst rocky desertification degree is efficient and accurate.

A new image fusion approach based on the modeling of shearlet coefficients with normal inverse gaussian model is proposed. The approach focus on the fusion of noisy images. Based on the statistical model and additive white gaussian noise, an subband adaptive shrinkage function is derived by using the maximum a posteriori rule. And then, the new scheme for shearlet-domain image fusion is proposed by incorporating the adaptive shrinkage rule into the fusion scheme. Experimental results show the proposed method perfor very well with noisy images, outperform other conventional methods.

An adaptive High-frequency Information Fusion Algorithm of Radar and Optical Images is proposed in this paper, in order to improve the resolution of the radar image and reserve more radar information. Firstly, Hough Transform is adopted in the process of low-resolution radar image and high-resolution optical image registration. The implicit linear information is extracted from two different heterogeneous images for better result. Then NSCT transform is used for decomposition and fusion. In different decomposition layers or in the same layer with different directions, fusion rules are adaptive for the high-frequency information of images. The ratio values of high frequency information entropy, variance, gradient and edge strength are calculated after NSCT decomposition. High frequency information entropy, variance, gradient or edge strength, which has the smallest ratio value, is selected as an optimal rule for regional fusion. High-frequency information of radar image could be better retained, at the same time the low-frequency information of optical image also could be remained. Experimental results showed that our approach performs better than those methods with single fusion rule.

Nowadays, with digital cameras and mass storage devices becoming increasingly affordable, each day thousands of pictures are taken and images on the Internet are emerged at an astonishing rate. Image retrieval is a process of searching valuable information that user demanded from huge images. However, it is hard to find satisfied results due to the well known "semantic gap". Image classification plays an essential role in retrieval process. But traditional methods will encounter problems when dealing with high-dimensional and large-scale image sets in applications. Here, we propose a novel multi-manifold classification model for image retrieval. Firstly, we simplify the classification of images from high-dimensional space into the one on low-dimensional manifolds, largely reducing the complexity of classification process. Secondly, considering that traditional distance measures often fail to find correct visual semantics of manifolds, especially when dealing with the images having complex data distribution, we also define two new distance measures based on path-based clustering, and further applied to the construction of a multi-class image manifold. One experiment was conducted on 2890 Web images. The comparison results between three methods show that the proposed method achieves the highest classification accuracy.

According to the characteristic of multi-digital cameras system, the paper presents the concept of System's Interior Orientation Parameters (SIOP) and verifies that the SIOP acquired from calibration in the inside field can effectively improve the accuracy and efficiency of data processing of multi-digital cameras system. Importantly, the paper builds a solution model of SIOP and verifies the feasibility of this solution model by experiment. Above all, the major research in the paper gives an answer to the question that how to calibrate the multi-digital cameras system efficiently and accurately.

Texture feature of image is one of the most important factors in the processing of information extraction from satellite scene image. In this paper the texture feature analysis was introduced in the processing of the classification of the objects in coastal zone. During the texture analysis process, how to extract effectively the texture features is the key factor. In the experiment of coastal classification, this paper introduced a method of a set of texture features selection based on step-by-step discriminance. Texture is described by Gray level co-occurrence matrix in this study, and there are 192 texture features to describe the characteristics of coastal objects. With the features selection method presented by this paper, five values were chosen as the representatives to classify the object texture feature. By means of the neural networks the object classification mode based on the texture features was defined and the object classifications of the southern coast of Laizhou Bay were carried out. Results show the step-by-step discriminance not only can decrease the dimension of the texture feature database, but also ensure and improve the accuracy of the classification, and the classification accuracy was up to 83.4%. The neural networks mode is the most effective method to account for the classification of the typical objects in coastal zone.

A new method of change detection in SAR images based on spectral clustering is presented in this paper. Spectral clustering is employed to extract change information from a pair images acquired on the same geographical area at different time. Watershed transform is applied to initially segment the big image into non-overlapped local regions, leading to reduce the complexity. Experiments results and system analysis confirm the effectiveness of the proposed algorithm.

In recent years, many researches are about building 3D object model in the fields of computer vision and photogrammetry, and camera calibration becomes the key problem. For long focal length digital camera calibration, there's its particularity. In this paper, aiming at long focal length lens, a camera calibration method based on point-line combination with vanishing points is proposed. This approach overcame the demerit of the conventional calibration theory with vanishing points; therefore the precision of calibration parameters became better.

Drought is one of the major environmental disasters in China. With the development of geographical information system and remote sensing technology, the real time monitoring drought over the large areas can be achieved. In this paper, the modified apparent thermal inertia (ATI) method was applied to monitor the drought while the traditional method was mainly used in bare soil and sparse vegetation. Taking Hebei plain as the study area, this paper calculated the apparent thermal inertia of all pixels firstly, and then got the 1km resolution classification result aggregated from 30m CCD data of HJ-1 satellite. Through analyzing the fine resolution classification result, the pure pixels were mask and the apparent thermal inertia was calculated. At last, the ATI values of all pixels were interpolated from the ATI of pure pixels via ARCGIS software. The application of the model to the announcement from Meteorological Bureau has given consistent results.

For most remote sense image applications, variations in solar illumination conditions, atmospheric scattering and absorption, and detector performance need to be normalized, especially in time series analysis such as change detection. For the purpose of radiometric correction, two levels of radiometric correction, absolute and relative, have been developed for remote sense imagery. In this paper, we select the Fast Line-of-sight Atmospheric Analysis of Spectral Hypercubes (FLAASH) algorithm as the Atmospheric correction method, and compare it with an automatic method for relative radiometric normalization based on a linear scale invariance of the multivariate alteration detection (MAD) transformation. The performances of both methods are compared using a landsat TM image pairs, the results from the two techniques have been compared both visually and using a measure of the fit based on standard error statistic.

Radiometric difference, misregistration error and the determination of classification threshold for difference image seriously influenced the detection accuracy of traditional pixel-level change detection algorithms, and it is difficult to get the true changes of interest from various kinds of detected changes. Therefore, a novel change detection algorithm is proposed to detect changes of man-made objects in remote sensing images. Large-size images are divided into overlapping and multi-scale sub-images, and three kinds of multi-scale structural features (including interscale or intrascale features), such as central-shift moments, gradient-magnitude features, gradient-orientation features and line-length features are extracted by support vector machine (SVM) classification. Experimental results demonstrate the feasibility and effectiveness of the proposed algorithm.

3D objects model plays an important role in 3D Geographic Information System (3DGIS). However, 3D object modeling is a costly and time-consuming work. In this paper, a simple 3D modeling approach is presented. It's based on epipolar image, and using constraints from regular geometry shape to construct a realistic 3D model interactively and fast. The result can satisfy fast modeling and analysis requirement in some case of emergency rescue. Prototype software from experiment shows the approach is efficient and promising.

Clustering analysis groups data objects based on information only found in the data that describes the objects and the relationships. As it is a spatial method, more research are focused on remote sensing application recently. This paper presents comparison of two classic cluster algorithms used in hyperspectral remote sensing image classification and the results showed that the classification of maximum likelihood algorithm is better than ISODATA algorithm.

The fusion method for the wide range resolution images will contribute to take the advantage of high time-resolution of MODIS data and high spatial-resolution of TM data, which will provide the time-series information matching the crop growth. The paper test the wavelet transform model from wavelet basis, decomposition level and fusion rule. By evaluating the quality of fusion images from several indexes, the paper analyzed the impact of fusion quality of MODIS and TM images from the parameter setting of wavelet transform. According to the comparison of many experiments, the study chose decomposition level 4, BIOR 6.8 of wavelet basis and high-replace-low of fusion rule. The study showed that the fusion method of wavelet transform could reserve the spectral feature of time-series information and enhance the spatial resolution from 250 meter to 30 meter. The time-series fusing images could be applied for crop monitoring.

This paper focuses on the topography measurement of island with satellite remote sensing. In this paper, remote sensing stereo images from Cartosat-1 satellite are used to survey the topographic information of island. Ground control points (GCPs) data are collected through on-site measurement. Then, some distributed GCPs are selected for remote sensing image processing. The digital elevation information (DEM) is extracted from Cartosat-1 remote sensing data by DEM extraction processing. And then, contour, slope and aspect are calculated based on DEM information. Finally, some ground control points is selected to validate the accuracy of topographic information. The results show that the accuracy of topographic information obtained 4.32m at horizontal position and 6.24m at vertical elevation accuracy.

The recently-emerged compressive sensing (CS) theory goes against the Nyquist-Shannon (NS) sampling theory and shows that signals can be recovered from far fewer samples than what the NS sampling theorem states. In this paper, to solve the problems in image fusion step of the full-scene image mosaic for the multiple images acquired by a low-altitude unmanned airship, a novel information mutual complement (IMC) model based on CS theory is proposed. IMC model rests on a similar concept that was termed as the joint sparsity models (JSMs) in distributed compressive sensing (DCS) theory, but the measurement matrix in our IMC model is rearranged in order for the multiple images to be reconstructed as one combination. The experimental results of the BP and TSW-CS algorithm with our IMC model certified the effectiveness and adaptability of this proposed approach, and demonstrated that it is possible to substantially reduce the measurement rates of the signal ensemble with good performance in the compressive domain.

The paper introduces a remote inspecting method of surface vibrator by double concerned laser. This inspecting method with two lasers has higher inspecting accuracy. It is designed for picking up higher précising sound vibrating information. Sound vibrating information are researched by analyzing the obtained opti-electric information. Its detecting precision is as high as laser' wavelength' level. It is designed on the basis of picking up sound technology derived from Michelson interference method.

In this paper, the different methods of orientating different resolution satellite images with the existing vector maps are introduced. To moderate high resolution satellite image, such as SPOT5 image, the vector road maps are often used for the orientation, and the vector lines in the maps usually represent the road central lines. And to higher resolution satellite image, such as QuickBird image, the vector lines in the maps usually represent the road edges, instead. Beside that, the different extend of details of the images makes it necessary to handle them with different methods. Because in very high resolution satellite images a lot of disturbing image features will be extracted along with the wanted one. A voting algorithm is employed to solve the problem, the approach is based on the previous work where an edge-based voting strategy was studied. The voting algorithm has the advantages of globally optimal and robust to noise. It can directly estimate the transformation parameters, meanwhile providing the potential matches of edge points, these matches can then be used in calculating the accurate orientation parameters, and providing the chance of change detection since the unchanged objects can be marked out with this method.

Soil moisture is the important indicator of climate, hydrology, ecology, agriculture and other parameters of the land surface and atmospheric interface. Soil moisture plays an important role on the water and energy exchange at the land surface/atmosphere interface. Remote sensing can provide information on large area quickly and easily, so it is significant to do research on how to monitor soil moisture by remote sensing. This paper presents a method to assess soil moisture status using Landsat TM data over Three Gorges area in China based on TVDI. The potential of Temperature- Vegetation Dryness Index (TVDI) from Landsat TM data in assessing soil moisture was investigated in this region. After retrieving land surface temperature and vegetation index a TVDI model based on the features of Ts-NDVI space is established. And finally, soil moisture status is estimated according to TVDI. It shows that TVDI has the advantages of stability and high accuracy to estimating the soil moisture status.

We investigate how to better use mutual information (MI) to select bands for hyperspectral image classification with less human intervention. Mutual information effectively measures the statistical dependence between two random variables. By modeling ground truth (e.g., a reference map) as one of the two random variables, MI can be used to find the spectral bands that contribute most to image classification. Extending our earlier work, we propose a sliding window model and apply mutual information to construct the estimated reference map, which need less human intervention. Experiments on the AVIRIS 92AV3C data set show that the proposed approach outperformed the benchmark methods, removing up to 55% of bands without significant loss of classification accuracy, compared to the 40% from that using the reference map accompanied with the data set. Meanwhile, its performance is found to be much robust to accuracy degradation when bands are cut off beyond 60%, revealing a better agreement in the mutual information estimation.

Drought is one of major nature disaster in the world and China. China has a vast territory and very different spatio-temporal distribution weather condition. Therefore, drought disasters occur frequently throughout China, which may affect large areas and cause great economic loss every year. In this paper, geostationary meteorological remote sensing data, FY-2C/D/E VISSR and three quantitative remotely sensed models including Cloud Parameters Method (CPM), Vegetation Supply Water Index (VSWI), and Temperature Vegetation Dryness Index (TVDI) have been used to dynamically monitor severe drought in southwest China from 2009 to 2010. The results have effectively revealed the occurrence, development and disappearance of this drought event. The monitoring results can be used for the relevant disaster management departments' decision-making works.

The situation of geological disaster prevention and control in our country is very serious. The application of remote sensing technique in geological disaster monitoring has been more than 30 years, and which has got many successful experiences. Remote sensing technology applications has shown enormous potential in investigation, evaluation, forecast and warning, and relief aspects of geological hazard research, and has been more and more attention to, for remote sensing technology is an high technology means with the ability of big range, all-weather, and dynamic monitoring the spatio-temporal changes of disasters. In the 2008 Wenchuan earthquake disaster, remote sensing technology played a huge role in the emergency response of the earthquake damage. But it has also exposed some problems that urgent needs to solve, the most critical problem of which is lacking of fast and efficient methods for disaster information extraction from remote sensing image data. This problem is also the key technical problems needs to solve and improve in the work of remote sensing application for geological hazard emergency investigation and evaluation. An interactive interpretation method platform based on Grab Cut segmentation was proposed in this paper, and used in interaction extraction of geological hazard information from remote sensing image. Based on ArcGIS Engine second development environment and in the .net framework, the interactive interpretation method utilize Grab Cut segmentation algorithm was developed. Because of Grab Cut image segmentation algorithm exploiting the texture and edge features, using this method can obtain ideal interpretation results and better improve interpreting efficiency, with less interactive steps.

The paper analyzes and researches the possibility and measure of photoelectric-inspect of sound frequency by phase modulating. Analyzing the 4-Frame Phase Shifting analyzing method used in sound frequency photoelectric-inspect. It is verified that vibrator film and the position of exploring instrument determines inspecting precision. This step directly influences sound frequency spectrum and dynamic range. A kind of vibrator film choosing reference gist has been brought up. This inspecting method can be used in sound information analyzing.

In this paper, a optimized algorithm to recognize and remove hazes and clouds from remotely sensed images of Landsat MSS/TM/ETM+ over land has been proposed. This algorithm uses only the image feature to automatically recognize and remove contamination of hazes and clouds which will prevent satellite image from assessing land surface variables. The hazes and clouds can be detected on the base of the reflectance difference with the other regions, likes thermal spectrum region. Based on both fourth tasseled cap parameter and a haze optimized transformation(HOT) as a measure of haze/cloud spatial density for single Landsat MSS/TM/ETM+ image, haze and clouds can be quantitatively recognized and removed. The performance of the proposed algorithm is demonstrated experimentally. This method can be used for atmospheric corrections to improve landscape change detection.

A new adaptive remote sensing image fusion algorithm based on Directionlet transform is proposed in this paper. Directionlet transform introduces multi-directional anisotropic basic function based on integer lattice which can effectively capture the anisotropic characteristics of images. For the high-frequency multi-directional coefficients, a contrast-based criterion is adopted and the low-pass coefficients are then selected adaptively based on a weighted criterion. Experimental results show that the proposed remote sensing fusion algorithm outperforms the wavelet-based fusion algorithm, both in terms of visual effect as well as quantized objective evaluation index, which proves that this algorithm is a feasible and effective remote sensing image fusion method.

Satellite measurements of the distribution of the global atmospheric CO₂ would get its continuous change. The atmospheric infrared sounder (AIRS) enables us to monitor the global distribution and transport of middle troposphere CO₂. Mount merapi is an active strato-volcano located on the border between central java and yogyakarta, indonesia. The AIRS data were acquired from 15th october to 15th november in 2008, 2009 and 2010 to monitor the temporal-spatial distribution of erupted CO₂ from the volcano. Mid-tropospheric CO₂ concentration would increase gradually and reach its peak in one day from eruption. The dispersal range of erupted CO₂ was -7.540831±7.5°, 110.444817±5° in the graticules centering around gunung merapi. Having a high correlation with the eruptions, the mid-troposphere CO₂ concentration of 2010 showed different trend comparing with 2008 and 2009 trend. The 4-day CO₂ concentration data of 2010 over the volcano tended to increase by 2.9 ppmv and 4.1 ppmv comparing with that of 2009 and 2008 respectively. These observations provide the evidence that extensive release of CO₂ occurs during the volcano eruption time and using the AIRS CO₂ products to monitor the temporal-spatial distribution of erupted CO₂ from volcanoes is possible.

SLHF (Surface Latent Heat Flux) is an atmospheric parameter, which can describe the heat released by phase changes and dependent on meteorological parameters such as surface temperature, relative humidity, wind speed etc. There is a sharp difference between the ocean surface and the land surface. Recently, many studies related to the SLHF anomalies prior to earthquakes have been developed. It has been shown that the energy exchange enhanced between coastal surface and atmosphere prior to earthquakes can increase the rate of the water-heat exchange, which will lead to an obviously increases in SLHF. In this paper, two earthquakes in 2010 (Haiti earthquake and southwest of Sumatra in Indonesia earthquake) have been analyzed using SLHF data by STD (standard deviation) threshold method. It is shows that the SLHF anomaly may occur in interpolate earthquakes or intraplate earthquakes and coastal earthquakes or island earthquakes. And the SLHF anomalies usually appear 5-6 days prior to an earthquake, then disappear quickly after the event. The process of anomaly evolution to a certain extent reflects a dynamic energy change process about earthquake preparation, that is, weak-strong-weak-disappeared.

The laser scanning confocal microscope (LSCM) offers several advantages over conventional optical microscopy, but most LSCM work is qualitative analysis and it is very hard to achieve quantitative detection directly with the changing of the fluorescent intensity. A new real time sensor system for the antibody-antigen interaction detection was built integrating with a LSCM and a wavelength-dependent surface plasmon resonance (SPR) sensor. The system was applied to detect the bonding process of human IgG and fluorescent-labeled affinity purified antibody in real time. The fluorescence images changing is well with that of SPR wavelengths in real time, and the trend of the resonance wavelength shift with the concentrations of antibody is similar to that of the fluorescent intensity changing. The results show that SPR makes up the short of quantificational analysis with LSCM with the high spatial resolution. The sensor system shows the merits of the of the LSCM and SPR synergetic application, which are of great importance for practical application in biosensor and life science for interesting local interaction.

Fingerprint recognition is one of the most important technologies in biometric identification and has been wildly applied in commercial and forensic areas. Fingerprint classification, as the fundamental procedure in fingerprint recognition, can sharply decrease the quantity for fingerprint matching and improve the efficiency of fingerprint recognition. Most fingerprint classification algorithms are based on the number and position of singular points. Because the singular points detecting method only considers the local information commonly, the classification algorithms are sensitive to noise. In this paper, we propose a novel fingerprint classification algorithm combining the local and global information of fingerprint. Firstly we use local information to detect singular points and measure their quality considering orientation structure and image texture in adjacent areas. Furthermore the global orientation model is adopted to measure the reliability of singular points group. Finally the local quality and global reliability is weighted to classify fingerprint. Experiments demonstrate the accuracy and effectivity of our algorithm especially for the poor quality fingerprint images.

Global change research requires high spatial, multi-scale, and consistent remote sensing data. Advanced land observing Satellite (ALOS) implemented a systematic data observation program, providing free polarimetric Phased Array L-band Synthetic Aperture Radar (PALSAR) mosaics to support global change research. The study took Mangrove Nature Reserve of Dongzhai Harbor in Hainan Province as an example, investigating the performance of PALSAR mosaics to map mangrove distribution in this area. In addition, backscattering characteristics of different objects were also compared, which proved the monitoring capability of PALSAR mosaics. The study indicates that PALSAR mosaics is a reliable data source for monitoring mangrove and is helpful to regional-scale research while high spatial resolution is guaranteed.

For the purpose of improving the efficiency of GIS data-updating over urban areas, a new approach of aerial triangulation (AT) with ground control points auto extracted from the existing AT results is introduced in the paper. The existing AT results mean the existing aerial imagery with interior and exterior orientation parameters. Through a multi-views image matching between the new aerial photos and old photos, enormous ground control points could be auto extracted and their ground coordinates could be calculated by the forward-intersecting of old aerial photos. The efficiency and accuracy of the method have been tested and analysed with several actual aerial imagery blocks. The experiment results presented in this paper shows that the method could not only extract enormous GCPs automatically which lead to a greatly improvement of the efficiency of AT over urban areas, but also achieve the same level of accuracy of the old AT results.

Taking Shandong province as study area, based on CCD and IRS data obtained by HJ-1 secondary planet, recurring to the VSWI, this work studied the agriculture drought distribution on Mar. 7th, 2011, using geographic information system technology the monitoring results were suitably expressed and it is coincident with the late average rainfall of the cities of this province. This result not only provides a scientific basis in assessing the drought disaster and making mitigation measures, but also affirms the effect of HJ-1 satellite data in drought monitoring.

Water is the basic needs for human society, and the determining factor of stability of ecosystem as well. There are lots of lakes on Tibet Plateau, which will lead to flood and mudslide when the water expands sharply. At present, water area is extracted from TM or SPOT data for their high spatial resolution; however, their temporal resolution is insufficient. MODIS data have high temporal resolution and broad coverage. So it is valuable resource for detecting the change of water area. Because of its low spatial resolution, mixed-pixels are common. In this paper, four spectral libraries are built using MOD09A1 product, based on that, water body is extracted in sub-pixels utilizing Multiple Endmembers Spectral Mixture Analysis (MESMA) using MODIS daily reflectance data MOD09GA. The unmixed result is comparing with contemporaneous TM data and it is proved that this method has high accuracy.

Group Animation is a difficult technology problem which always has not been solved in computer Animation technology, All current methods have their limitations. This paper put forward a method: the Motion Coordinate and Motion Speed of true fish group was collected as sample data, reverse cloud generator was designed and run, expectation, entropy and super entropy are gotten. Which are quantitative value of qualitative concept. These parameters are used as basis, forward cloud generator was designed and run, Motion Coordinate and Motion Speed of two-dimensional fish group animation are produced, And two spirit state variable about fish group : the feeling of hunger, the feeling of fear are designed. Experiment is used to simulated the motion state of fish Group Animation which is affected by internal cause and external cause above, The experiment shows that the Group Animation which is designed by this method has strong Realistic.

In this paper, we reviewed the principle, data, methods and steps in suspended sediment research by using remote sensing, summed up some representative models and methods, and analyzes the deficiencies of existing methods. Combined with the recent progress of remote sensing theory and application in water suspended sediment research, we introduced in some data processing methods such as atmospheric correction method, adjacent effect correction, and some intelligence algorithms such as neural networks, genetic algorithms, support vector machines into the suspended sediment inversion research, combined with other geographic information, based on Bayesian theory, we improved the suspended sediment inversion precision, and aim to give references to the related researchers.

Yushu Earthquake of magnitude 7.1 Richter in 2010 has brought a huge loss of personal lives and properties to China. National Disaster Reduction Center of China implemented the disaster assessment by using remote sensing images and field investigation. Preliminary judgment of disaster scope and damage extent was acquired by change detection. And the building region of hard-hit area Jiegu town was partitioned into 3-level grids in airborne remote sensing images by street, type of use, structure, and about 685 girds were numbered. Hazard assessment expert group were sent to implement field investigation according to each grid. The housing damage scope and extent of loss were defined again integrated field investigation data and local government reported information. Though remote sensing technology has played an important role in huge disaster monitoring and assessment, the automatic capability of disaster information extraction flow, three-dimensional disaster monitoring mode and bidirectional feedback mechanism of products and services should still be further improved.

Hybrid opto-digital joint transform correlator (HODJTC) is effective for image motion measurement, but it is different from the traditional joint transform correlator (JTC) because it only has one process of optical Fourier transform and the intermediate joint power spectrum (JPS) is directly input to a digital unit to compute the image motion. In original HODJTC, the correlation image is obtained by using a global discrete Fourier transform operator. In fact, the location of cross-correlation peak corresponding to zero image motion is known and the maximum image shift can also be estimated. Thus, the local cross-correlation image can be directly obtained by adopting a local Fourier transform operator instead. With the pixel-level location of cross-correlation peak is obtained, the up-sampling technique is introduced to relocate the peak in even higher accuracy. Both the experiments and simulations demonstrate that the RMS (root-mean-square) error of motion determination can be controlled below 0.05 pixels, which is a very promising result for real application.

Spectral imaging technology have been used mostly in remote sensing, but have recently been extended to new area requiring high fidelity color reproductions like telemedicine, e-commerce, etc. These spectral imaging systems are important because they offer improved color reproduction quality not only for a standard observer under a particular illuminantion, but for any other individual exhibiting normal color vision capability under another illuminantion. A possibility for browsing of the archives is needed. In this paper, the authors present a new spectral image browsing architecture. The architecture for browsing is expressed as follow: (1) The spectral domain of the spectral image is reduced with the PCA transform. As a result of the PCA transform the eigenvectors and the eigenimages are obtained. (2) We quantize the eigenimages with the original bit depth of spectral image (e.g. if spectral image is originally 8bit, then quantize eigenimage to 8bit), and use 32bit floating numbers for the eigenvectors. (3) The first eigenimage is lossless compressed by JPEG-LS, the other eigenimages were lossy compressed by wavelet based SPIHT algorithm. For experimental evalution, the following measures were used. We used PSNR as the measurement for spectral accuracy. And for the evaluation of color reproducibility, ΔE was used.here standard D65 was used as a light source. To test the proposed method, we used FOREST and CORAL spectral image databases contrain 12 and 10 spectral images, respectively. The images were acquired in the range of 403-696nm. The size of the images were 128*128, the number of bands was 40 and the resolution was 8 bits per sample. Our experiments show the proposed compression method is suitable for browsing, i.e., for visual purpose.

We propose a fast multiscale face detector that boosts a set of SVM-based hierarchy classifiers constructed with two heterogeneous features, i.e. Multi-block Local Binary Patterns (MB-LBP) and Speeded Up Robust Features (SURF), at different image resolutions. In this hierarchical architecture, simple and fast classifiers using efficient MB-LBP descriptors remove large parts of the background in low and intermediate scale layers, thus only a small percentage of background patches look similar to faces and require a more accurate but slower classifier that uses distinctive SURF descriptor to avoid false classifications in the finest scale. By propagating only those patterns that are not classified as background, we can quickly decrease the amount of data need to be processed. To lessen the training burden of the hierarchy classifier, in each scale layer, a feature selection scheme using Binary Particle Swarm Optimization (BPSO) searches the entire feature space and filters out the minimum number of discriminative features that give the highest classification rate on a validation set, then these selected distinctive features are fed into the SVM classifier. We compared detection performance of the proposed face detector with other state-of-the-art methods on the CMU+MIT face dataset. Our detector achieves the best overall detection performance. The training time of our algorithm is 60 times faster than the standard Adaboost algorithm. It takes about 70 ms for our face detector to process a 320×240 image, which is comparable to Viola and Jones' detector.

There are regional and periodic droughts in Chongqing, which impacted seriously on agricultural production and people's lives. This study attempted to monitor the drought in Chongqing with complex terrain using MODIS data. First, we analyzed and compared three remote sensing methods for drought monitoring (time series of vegetation index, temperature vegetation dryness index (TVDI), and vegetation supply water index (VSWI)) for the severe drought in 2006. Then we developed a remote sensing based drought monitoring model for Chongqing by combining soil moisture data and meteorological data. The results showed that the three remote sensing based drought monitoring models performed well in detecting the occurrence of drought in Chongqing on a certain extent. However, Time Series of Vegetation Index has stronger sensitivity in time pattern but weaker in spatial pattern; although TVDI and VSWI can reflect inverse the whole process of severe drought in 2006 summer from drought occurred - increased - relieved - increased again - complete remission in spatial domain, but TVDI requires the situation of extreme drought and extreme moist both exist in study area which it is more difficult in Chongqing; VSWI is simple and practicable, which the correlation coefficient between VSWI and soil moisture data reaches significant levels. In summary, VSWI is the best model for summer drought monitoring in Chongqing.

Computer vision technology has been increasingly used for automatically observing crop growth state, but as one of the key parameters in the field of agro-meteorological observation, crop canopy height is still measured manually in the actual observation process up to now. In order to automatically measure the height based on the forward-and-downward-looking image in the existing monocular vision observation system, a novel method is proposed, that is, to measure the canopy height indirectly by the solving algorithm for the actual height of the vertical objects (SAAH) with the help of the intelligent sensor device. The experiment results verified the feasibility and validity of our method, and that the method could meet the actual observation demand.

Based on the radical defects of chromaticity color reproduction technology and current color management technology, a new theoretical and technology way of multi-spectral color reproduction was put forward. And the essence of multi-spectral color reproduction technology was deeply analyzed and a quite complete frame of theoretical technology was sketched. Then it analyzed seven key technologies of multi-spectral color reproduction deeply, that was, the design of multi-spectral color reproduction system, the acquisition of multi-spectral image, the fusion of multi-spectral image, the reconstruction of spectral reflectance, dimension reduction of multi-spectral image, algorithm and model development of spectrum color separation and the spectral color management technology. Meanwhile, it pointed out the approach and prospect of the study on multi-spectral color reproduction technology.

This paper analyzed the evolution of drought and the spectral response of the crop at different growing seasons focuses on the irrigated agricultural areas of northen Henan using the HJ-1 data and MODIS data,associated with relevant meteologic data, regional geographical data and the social economic data.The Spatial and temporal distribution of the risk of disaster-causing factors and the fragility of the disaster-affected body was conducted and the comprehensive index of agricultral drought risk was built up.Then, trend of the agricultural drought was analyzed and the irrigated agricultural drought risk class was performed and the possible hazard and influence of agricultural drought and the performance of appropriate strategy to reduce agricultral drought have been estimated.At last,verification of the results and improvement of the model have been carried out supported by the historic cases, expert system and the on-site investigation data.

Purpose: To detect the defects during the high speed process of web printing, such as smudges, doctor streaks, pin holes, character misprints, foreign matters, hazing, wrinkles, etc., which are the main infecting factors to the quality of printing presswork. Methods: A set of novel machine vision system is used to detect the defects. This system consists of distributed data processing with multiple linear cameras, effective anti-blooming illumination design and fast image processing algorithm with blob searching. Also, pattern matching adapted to paper tension and snake-moving are emphasized. Results: Experimental results verify the speed, reliability and accuracy of the proposed system, by which most of the main defects are inspected at real time under the speed of 300 m/min. Conclusions: High speed quality inspection of large-size web requires multiple linear cameras to construct distributed data processing system. Also material characters of the printings should also be stressed to design proper optical structure, so that tiny web defects can be inspected with variably angles of illumination.

Purpose: To detect and identify all kinds of visible defection on floating glass production line in real time, such as bubbles, tin ash, impurities, stone, craterlet, thread, distortion, etc., which is extremely important for glass process management, grading and stowing. Methods: A set of machine vision system is used to detect the defects. In this system, a novel illumination technique and sensor unit based on time-variale LED raster is developed to obtain both the distortion and deformation features together. The resulting defects are determined by a RLE-based image processing algorithm and transferred to subsequent marking or cutting devices. Results: Real experiments illustrated the stability and effectiveness of this system, by which most of the main defects are inspected at real time under the speed of 30 m/min. With 5 m glass width, the inspection precisions are 0.1 mm/pixel both in direction of width and length. Applications verify the speed, reliability and accuracy of the proposed method. Conclusions: Quality inspection of floating glass at real time requires multiple linear cameras to construct distributed data processing system. Also material characters of the glass should be stressed to design proper optical structure, so that the glass defects will be inspected successfully. Using this system, the quality of floating glass can be improved effectively.

On April 14th, 2010, a large earthquake (Ms=7.1) shocked the Yushu county in the Qinghai Province at western China. It caused severe damages and fatalities. The purpose of this paper is to estimate the building damages caused by the earthquake based on pre and post earthquake ALOS SAR data. The distribution of the buildings in the Yushu city was extracted from the intensity images of the SAR data. Phase coherence between two scenes of SAR data were calculated to assess the building damages. The results show that the buildings close to the surface rupture caused by the earthquake are broken severely. The further to the surface rupture, the less damages of the buildings. It indicates that avoiding the potential rupture zone in the urban planning is necessary to mitigate the earthquake hazards.

This article explored the method of photo-based 3D modeling for the arc bridge structures by ordinary digital camera. Firstly, a series of processes had been studied by using ordinary digital camera that included the camera calibration, data acquisition, data management, and 3D orientation, setting scale and textures, etc., then the 3D model from photos can be built. The model can be measured, edited and close to the real structures. Take an interior masonry arch bridge as an example, build 3D model through the processes above by using camera HP CB350. The 3D model can be integrated with the loading conditions and material properties, to provide the detailed data for analyzing the structure. This paper has accumulated the experience in data acquisition and modeling methods. The methods can be applied to other structural analysis, and other conditions of 3D modeling with fast and economic advantages.

Remote Sensing surveillance constitutes an important component of oil spill disaster management system, but subject to monitoring accuracy and ability, which suffered from resolution, environmental conditions, and look-alikes. So this article aims to provide information of identification and distinguishing of look-alikes for optical sensors, and then improve the monitoring precision. Although limited by monitoring conditions of the atmosphere and night, optical satellite remote sensing can provide the intrinsic spectral information of the film and the background sea, then affords the potentiality for detailed identification of the film thickness, oil type classification (crude/light oil), trends, and sea surface roughness by multi-type data products. This paper focused on optical sensors and indicated that these false targets of sun glint, bottom feature, cloud shadow, suspend bed sediment and surface bioorganic are the main factors for false alarm in optical images. Based on the detailed description of the theory of oil spill detection in optical images, depending on the preliminary summary of the feature of look-alikes in visible-infrared bands, a discriminate criteria and work-flow for slicks identification are proposed. The results are helpful to improve the remote sensing monitoring ability and the contingency planning.

Optical information hiding system has many features such as high processing speed, high parallel, high encryption dimension and high speed of optical transformation and related operations, more advantages than digital method in some way. But it has not adequate security, and enough combination with techniques of digital image processing. So on basis of analyzing existing image hiding and analyzing techniques, we give out the idea. We should adopt idea of virtual optics on the way of all-digital simulation to do research of optical image hiding and analyzing methods based on optical image processing technique especially technique of double random phase encoding and digital holography.

At present, the most technology of counting money is to use the money counter in financial fields. The paper presents a new method for automatic counting paper money which is based on image processing technology. Firstly, the paper money image is acquired by CCD. After analyzing the feature of image, we find that in Cr-space the edge of each paper money is enhanced. Then we use the north-west sobel operator for filtering and north sobel operator for detecting edge. Although the image-processed better highlight the edge of each paper money, the edge is rough and its variance is high. It is hardly to threshold the image for getting the single-pixel edge linked. After Different segmentation algorithm was been used for deriving the edge of paper money, we find the Two-dimensional Histogram θ-division algorithm is suitable for our purpose. The experimental result is proved satisfied. The detecting rate reached 100% in controlled environment for RMB. However, if we want to detect other kinds of paper money such as dollar, there also have several problems to be solved.

An airborne laser remote sensing technology is proposed to detect natural gas pipeline leakage in helicopter which carrying a detector, and the detector can detect a high spatial resolution of trace of methane on the ground. The principle of the airborne laser remote sensing system is based on tunable diode laser absorption spectroscopy (TDLAS). The system consists of an optical unit containing the laser, camera, helicopter mount, electronic unit with DGPS antenna, a notebook computer and a pilot monitor. And the system is mounted on a helicopter. The principle and the architecture of the airborne laser remote sensing system are presented. Field test experiments are carried out on West-East Natural Gas Pipeline of China, and the results show that airborne detection method is suitable for detecting gas leak of pipeline on plain, desert, hills but unfit for the area with large altitude diversification.

Analysis and simulate international relationship by using high-tech cluster of information technology is important and meaningful to the modern research of strategic work and the strategic decision-making. This paper use Multi-Agents system to model the international relationship with macrocosmic level and microcosmic level. For the microcosmic level shows activities and the rules of correlative activities of countries, propose a reinforcement learning algorithm to improve cooperative efficiency among agents, which based on common goal and joint rewards to insure agents to learn cooperative behavior. A prototype and an experiment were given.

This paper studied on change detection algorithm of high resolution (HR) Synthetic Aperture Radar (SAR) images based on multi-scale texture features. Firstly, preprocessed multi-temporal Terra-SAR images were decomposed by 2-D dual tree complex wavelet transform (DT-CWT), and multi-scale texture features were extracted from those images. Then, log-ratio operation was utilized to get difference images, and the Bayes minimum error theory was used to extract change information from difference images. Lastly, precision assessment was done. Meanwhile, we compared with the result of method based on texture features extracted from gray-level cooccurrence matrix (GLCM). We had a conclusion that, change detection algorithm based on multi-scale texture features has a great more improvement, which proves an effective method to change detect of high spatial resolution SAR images.

In this paper, more than 1000 SAR (ENVISAT, ERS-2) and optical (MODIS, HJ-1A/1B) images are used to analyze the distribution, sources and propagation of internal waves in South China Sea. The distribution of internal waves shows that, internal waves mainly occur in three regions of South China Sea: 1) Northern South China Sea, 2) Western South China Sea, 3) Southern South China Sea. Internal waves are observed all year in the South China Sea, most during summer, least in winter. The sources and propagation of internal waves are obtained from the distribution of internal waves. 1) in Northern South China Sea, most internal waves origin in Luzon Strait and propagate westward across Dongsha Island, then some internal waves turn into northwestward, others continue to propagate westward.2) in Western South China Sea, internal waves generate at the shelf, and propagate to the shore or off the shore.3) in Southern South China Sea, internal waves generate at the shelf. Some internal waves propagate to the Kalimantan shore, some propagate southwestward.

Vibrator shell is a part with complex structure. While the vibrator shell is designed and manufactured by traditional sand casting process, more than 80% castings are found the defects of porosity, shrinkage and pouring-shortage at the top. Aiming to the problems in traditional sand casting, this paper focused on the improvement of castings structure and the optimization of casting process. Designing process bar in the gate-channel region which is connected with the gate in castings is used to improve the castings structure, and low speed filling and solidification under high pressure are adopted to optimize the casting process which is finished by self-made four-column type hydraulic machine equipped. It can be seen that the castings quality can be greatly improved by process improvement.

The fractional vegetation cover (Fvc) can be characterized by strong temporal dynamics that need to be considered for environmental recovery assessment in the Wenchuan earthquake disaster area. This study is to test the capacity of satellite remote sensing images in estimating fractional vegetation cover in the Sichuan disaster area. Two different spectral mixture analysis methods including N* and N*2 transformed from NDVI are applied for fractional vegetation cover estimation using Landsat TM image. High-resolution fractional vegetation cover production from airborne images is used for validation. The result shows that N* has the non linear relationship with airborne fractional vegetation cover values in this study area and N*2 takes a better performance (R²=0.993). In total, retrieved Fvc values from Landsat TM image have a strong relationship to those from airborne image, and can be used in environmental dynamics assessment in the Wenchuan Earthquake disaster area.

In this paper the significant wave height (SWH) and the wind field products of ECMWF re-analysis data are used to derive the location of typhoon center, to analyze the temporal and spatial features of the SWH induced by typhoons, and to study the relationships between the SWH and wind speed. The results are compared with merged SWH data from several satellite altimeters (GFO, TOPEX/Poseidon, Jason-1 and Envisat) and wind vectors from QuikSCAT. Typhoon eyes are observed by using SAR and MODIS data. It is shown that (1) the spatial distribution of wind fields from ECMWF re-analysis data is almost in accordance with that of wind fields from QuikSCAT; (2) the spatial distribution of SWH from ECMWF re-analysis data is almost in accordance with that of SWH from merged SWH data; (3) the distribution of higher wind speed and higher wave height are consistent with the SWH and the wind field product of ECMWF re-analysis data; (4) the centers of typhoon waves lag behind the centers of typhoons; (5) the top of typhoon move faster that the bottom in the case of Saomai.

In this paper, a homography-based predictive control strategy is presented for the visual servoing problem. Visual servoing play an important role in robot research domain. Nowadays, advanced control laws like Model Predictive Control have been used to solve the visual servoing problems. The extention of MPC into visual servoing has many advantagies. But most of these methods depend on the depth information and pricise intrinsic parameters of camera. In order to avoid the measurement of 3D structure, we thought the homography-based predictive control approach for visual servoing.

Spaceborne synthetic aperture radar (SAR) plays an important role in sea surface wind field. Upwelling is a common oceanic phenomenon which influences the wind retrieving from SAR imagery. It smoothes the sea surface and its backscatter cross section in SAR images is low. The sea surface wind retrieved directly from SAR imagery in upwelling region will be underestimated. In this article, using sea surface temperature (SST) and chlorophyll2-a derived from EOS MODIS, the low backscatter features in ENVISAT ASAR image are analyzed along the Zhejiang Coast in the East China Sea. And CMOD4 algorithm is adopted to retrieve sea surface wind speed, using wind directions from interpolated NCEP / NCAR reanalysis data. Results show that wind speed is negatively biased due to the low Normalized Radar Cross Section (NRCS) associated with the Zhejiang Coastal Upwelling. Based on statistical method, the quantitative relationship between SAR-derived wind speeds and high resolution numerical meteorological model results is established. Then using this quantitative relationship, the underestimated wind speeds in upwelling region are corrected and the impact of coastal upwelling on SAR wind retrieval is quantitatively evaluated.

Based on daily MODIS Enhanced Vegetation Index (EVI), monthly MODIS EVI was calculated, and was analyzed by Principal Component Analysis (PCA) further, whose first three components were inputted as the first three parameters for classification. The fourth, fifth classification parameters were obtained by applying kriging interpolation, principal component analysis, and data re-sampling to temperature, precipitation, bio-temperature and DEM data. Based on these five parameters, the Inner Mongolia grassland was classified by using ISODATA. The results showed that the grassland classification of MODIS EVI provided more information than that of AVHRR NDVI and MODIS NDVI. This method could identify five classes of Inner Mongolia grassland clearly, and sub-classes of temperate meadow, temperate steppe and temperate desert steppe were also distinguished.

With the Landsat-TM of 1993 and 2009, the incremental construction land vector of Changhai county (Changshan islands) was extracted by means of noise removing, Principal Component Analysis (PCA), Color Composing, Classifying and Difference value arithmetic. With methods of overlay, buffer and geographic statistic analysis in MapInfo applied to construction land layer, the spatial distribution and the influencing factors were analyzed. The results showed: from 1993 to 2009, the construction land in Changhai county increased to some extent, and the population growth and the economic growth are the main promoting factors; The transport corridor have spatial attraction to the incremental construction land; the distribution of construction land has the trend to the infrastructural facilities such as harbors or dock; the industrial distribution also influences on the distribution of incremental construction land.

Wild fire of forest and grassland is one of the major natural disasters in the world. The fire of forest and grassland occurs frequently due to global changing and uneven precipitation distribution, which may affect large areas and cause great economic loss. With the advantages of wide-coverage, high spatio-temporal resolution, remote sensing technology has been used for the wild fire monitoring activities. In this paper, the fire point detection model and burned area monitoring model, for HJ-1A/B data, have been developed and applied for the southeast Australian wild fire monitoring in 2009. The application shows that the models and the data are efficient for the wild fire monitoring.

Walsh-Haar function system that was first introduced by us is a new kind of function systems, and has a good global / local property. This function system is called Walsh ordering function system since its generation kernel functions belong to Walsh ordering Walsh function system. We worked out a recursive property of the matrix WH KR ^m+1 WH corresponding to the first KR ^m+1 Walsh-Haar functions in Walsh-Haar function system, and proved that Walsh-Haar function system is perfect and orthogonal similar to Walsh function system and Haar function system. Thus, discrete Walsh-Haar transformation (DW-HT) is an orthogonal transformation that can be widely used in signal processing. In this paper, using the recursive property of the matrix WH KRm+1 WH and the fast algorithm of discrete Walsh transformation in Walsh ordering, we have designed a fast algorithm of Walsh ordering (k, k-1) type DW-HT based on the bisection technique. As one of its applications, we use it to detect image edges. Compare with some edge-detecting methods, the method in this paper detects more details of image edge. The idea and method used to design the fast algorithm in this paper can be used to design fast algorithms of other ordering (k, k-1) type DW-HTs and other discrete orthogonal transformations.

A new naked eyes stereoscopic display optical system for multi-users was proposed in this paper, different from the conventional naked eyes stereoscopic display monitors with sophistication technology and bad uniformity of illumination. According to the binocular parallax technology, Fresnel lens and polaroid sheet were chosen to focus the rays on the observers' eyes. Freeform lens were used in this system to enhance illumination uniformity of pictures and efficiency of the system. The results of simulation showed that the users can obtain stereoscopic effect at their each places, uniformity of illumination on LCD was above 0.6. Therefore, this system had a simple and compact structure, and the quality of pictures users obtain was better than system without freeform lenses.

To avoid the limitation of common route planning method to blindly pursue higher Machine Intelligence and autoimmunization, this paper presents a human-machine cooperation route planning method. The proposed method includes a new A* path searing strategy based on dynamic heuristic searching and a human cooperated decision strategy to prune searching area. It can overcome the shortage of A* algorithm to fall into a local long term searching. Experiments showed that this method can quickly plan a feasible route to meet the macro-policy thinking.

Active learning is typically limited by the small sample problem which makes the resulting classifiers perform poorly, especially in the initial stages. To overcome this problem, in this paper, a novel framework - graph-incorporated active learning - is proposed, in which the selection pool is regarded as a graph. Its graph structure is applied to both improve sample selection criterion and provide the learner enough pseudo-labeled samples. By comparing with the state-of-theart technique, the experiments on benchmark datasets show that the improvement of the proposed method is significant, i.e., it can solve the small problem well. The framework is combined with, but is not limited to, SVM.

Low link utility, RTT unfairness and unfairness of Multi-Bottleneck network are the existing problems in the present network congestion control algorithms at large. Through the analogy of network congestion control with the "El Farol Bar" problem, we establish a congestion control model based on minority game(MG), and then present a novel network congestion control algorithm based on the model. The result of simulations indicates that the proposed algorithm can make the achievements of link utility closing to 100%, zero packet lose rate, and small of queue size. Besides, the RTT unfairness and the unfairness of Multi-Bottleneck network can be solved, to achieve the max-min fairness in Multi-Bottleneck network, while efficiently weaken the "ping-pong" oscillation caused by the overall synchronization.