This PDF file contains the editorial “Guest Editorial: Special Section on Image and Video Databases” for JEI Vol. 7 Issue 02

Although content based retrieval of images is increasingly common, the use of media content as a basis for navigation has received relatively little attention. In this article we describe our recent development of facilities in the MAVIS/Microcosm architecture for generic link authoring and following from nontext media, and in particular, the use of shape and texture for content based navigation from images. Applications from a product catalog and an archeological collection are presented, together with an outline of an image viewer providing rapid delineation of object shapes in images when authoring or following links. The current relationship between Microcosm and the World Wide Web is summarized and the performance of the MAVIS/Microcosm system is described.

In this article, we propose two novel histogram-based techniques that are robust to the changes in image illumination levels. Given a query image and an image database, current histogram-based techniques retrieve similar images acquired under similar illumination levels. However, these techniques fail when images are acquired under varying illumination conditions. First, we propose employing moments of the image histogram that are invariant to scaling and translation of image gray levels. Second, we propose comparing the parameters of histograms of the wavelet subbands for indexing. These parameters are modified appropriately to counter the effect of changes in illumination. The proposed techniques can be combined to further improve the indexing efficiency. The techniques are computationally inexpensive and can also be easily integrated within a wavelet-based image coder.

To keep pace with the increased popularity of digital video as an archival medium, the development of techniques for fast and efficient analysis of video streams is essential. In particular, solutions to the problems of storing, indexing, browsing, and retrieving video data from large multimedia databases are necessary to allow access to these collections. Given that video is often stored efficiently in a compressed format, the costly overhead of decompression can be reduced by analyzing the compressed representation directly. In earlier work, we presented compressed domain parsing techniques which identified shots, subshots, and scenes. In this article, we present efficient key frame selection, feature extraction, indexing, and retrieval techniques that are directly applicable to MPEG compressed video. We develop a frame type independent representation which normalizes spatial and temporal features including frame type, frame size, macroblock encoding, and motion compensation vectors. Features for indexing are derived directly from this representation and mapped to a low-dimensional space where they can be accessed using standard database techniques. Spatial information is used as primary index into the database and temporal information is used to rank retrieved clips and enhance the robustness of the system. The techniques presented enable efficient indexing, querying, and retrieval of compressed video as demonstrated by our system which typically takes a fraction of a second to retrieve similar video scenes from a database, with over 95% recall.

In this article we describe the results of a study on similarity evaluation in image retrieval using color, object orientation, and relative position as content features, in a framework oriented to image repositories where the semantics of stored images are limited to a specific domain. The focus is not on a complete description of image content, which is supposed to be known to some extent, but on the extraction of simple and immediate features that can assure, through their combination, automated image analysis and efficient retrieval. Relevance feedback is introduced as an effective way to improve retrieval accuracy. A simple prototype system is also introduced that computes feature descriptors and allows users to enter queries, browse the retrieved images, and refine the results through relevance feedback analysis.

To guarantee security and privacy in image and video archival applications, efficient bulk encryption techniques are necessary which are easily implementable in soft- and hardware and are able to cope with the vast amounts of data involved. Experience has shown that block-oriented symmetric product ciphers constitute an adequate design paradigm for resolving this task, since they can offer a very high level of security as well as very high encryption rates. In this contribution we introduce a new product cipher which encrypts large blocks of plaintext by repeated intertwined application of substitution and permutation operations. While almost all of the current product ciphers use fixed (predefined) permutation operations on small data blocks, our approach involves parametrizable (keyed) permutations on large data blocks (whole images) induced by specific chaotic systems (Kolmogorov flows). By combining these highly unstable dynamics with an adaption of a very fast shift register based pseudo-random number generator, we obtain a new class of computationally secure product ciphers which offer many features that make them superior to contemporary bulk encryption systems when aiming at efficient image and video data encryption.

Digital watermarking techniques sign images by introducing changes that are imperceptible to the human eye but easily recoverable by a computer program. Generally, the watermark is a number which identifies the owner of the image. The locations in the image where the watermark is embedded are determined by a secret key. Doing so prevents possible pirates from easily removing the digital watermark. Furthermore, it should be possible to recover the embedded watermark from an altered image. Possible alterations of signed images include blurring, compression, and geometrical transformations such as rotation and translation. These alterations are referred to as attacks. A new method based on amplitude modulation is presented. Single watermark bits are multiply embedded by modifying pixel values in the blue channel. These modifications are proportional to the luminance and either additive or subtractive, depending on the value of the bit. This new method has shown to be resistant to both classical attacks, such as filtering, and geometrical attacks. Moreover, the watermark can be recovered without the original image.

The linear unsharp masking technique used in image contrast enhancement is modified in this article by introducing a control term expressed as a rational function of the local input data. In this way noise amplification is avoided and, at the same time, overshoot effects on sharp edges are limited. Experimental results support the validity of the method, even as a preprocessor for interpolation systems.

In this article a structure-adaptive approach to the evaluation of image local properties for adaptive filtering is described. The adaptive procedure is based on selection of the most homogeneous neighborhood region from several possible structuring regions by the principle of maximum posterior probability. Then, an optimal evaluation of the pixel value at this point is performed involving pixels from the determined neighborhood region and the symmetric structuring region. The trimmed mean filters are used for the robust evaluation of local properties during estimation of object and background intensities when the supposed additive noise has a mixed conditional distribution, e.g., normal distribution with outliers. A time-efficient scheme for fast implementation of this method is proposed as well.

A new cryptographic method for encrypting still images is proposed. This method is designed on the basis of the base-switching lossless image compression algorithm. The method therefore simultaneously possesses both image encryption and lossless compression abilities. A given image is first partitioned into nonoverlapping fixed-size subimages, and each subimage will then have its own base value. These subimages are then encoded and encrypted one by one according to the base values. By choosing the function to encrypt the base value, there are (128!)^t [or (128!)^3t] possible ways to encrypt a gray-scaled (color) image if t layers are used in the encryption system. The theoretical analysis needed to support the proposed encryption method is provided, and the experimental results are also presented.

We present two improved methods for the codebook search in this article. We call them the improved eigenvector method (IEVM) and the improved triangle inequality elimination (ITIE), respectively. IEVM is a full-search (FS) equivalent method, but ITIE is not. In these two methods we propose some techniques, such as the orthogonal checking and the fixed range search, to speed up their performance. According to our experimental results, IEVM is an efficient method. It is faster than FS, TIE, and EVM. Its execution time and number of operations per pixel were 4.86% and 4.76% of those in FS, respectively. As for ITIE, it is always faster than IEVM even if we limit its peak signal-to-noise ratio degradation to be no greater than 3% of the image quality of FS. Under this limitation, the execution time and the number of operations per pixel of ITIE were only 3.08% and 3.05% of those in FS, respectively.

We present a novel technique for segmentation of a JPEG-compressed document based on block activity. The activity is measured as the number of bits spent to encode each block. Each number is mapped to a pixel brightness value in an auxiliary image which is then used for segmentation. We introduce the use of such an image and show an example of a simple segmentation algorithm, which was successfully applied to test documents. The document is segmented into characteristics regions labeled as background, halftones, text, graphics, and continuous tone images. The key feature of the proposed framework is that the desired region can be identified and cropped (or removed) from the compressed data without decompressing the image.

As digital signal processors (DSPs) become more advanced, many real-time recognition problems will be solved with completely integrated solutions. In this paper a methodology which is designed for today’s DSP architectures and is capable of addressing applications in real-time color object recognition is presented. The methodology is integrated into a processing structure called raster scan video processing which requires a small amount of memory. The small amount of memory required enables the entire recognition system to be implemented on a single DSP. This auto-associative segmentation approach provides a means for desaturated color images to be segmented. The system is applied to the problem of stop sign recognition in realistically captured outdoor images.

This PDF file contains the editorial “Book Rvw: Techniques and Standards for Image, Video and Audio Coding," by K. R. Rao and J. J. Hwang for JEI Vol. 7 Issue 02

This PDF file contains the editorial “Techniques and Standards for Image, Video and Audio Coding” for JEI Vol. 7 Issue 02