This PDF file contains the front matter associated with SPIE Proceedings Volume 8063, including the Title Page, Copyright information, Table of Contents, and the Conference Committee listing.

The issue of privacy in video surveillance has drawn a lot of interest lately. However, thorough performance analysis and validation is still lacking, especially regarding the fulfillment of privacy-related requirements. In this paper, we first review recent Privacy Enabling Technologies (PET). Next, we discuss pertinent evaluation criteria for effective privacy protection. We then put forward a framework to assess the capacity of PET solutions to hide distinguishing facial information and to conceal identity. We conduct comprehensive and rigorous experiments to evaluate the performance of face recognition algorithms applied to images altered by PET. Results show the ineffectiveness of naïve PET such as pixelization and blur. Conversely, they demonstrate the effectiveness of more sophisticated scrambling techniques to foil face recognition.

Watermarking is known to be a very difficult task. Robustness, Distortion, Payload, Security, Complexity are many constraints to deal with. When applied to a video stream, the difficulty seems to be growing in comparison to image watermarking. Many classical non malicious manipulations of a compressed stream may suppress the embedded information. For example, a simple re-compression of a DVD movie (MPEG2 compression) to a DivX movie will defeat most of the current state-of-the-art watermarking systems. In this talk, we will expose the different techniques in order to watermark a video compressed stream. Before, we will present the H.264/AVC standard which is one of the most powerful video-compression algorithms. The discussion about video watermarking will be illustrated with H.264 streams. The specific example of traitor tracing will be presented. Deadlocks will be discussed, and then the possible extensions and the future applications will conclude the presentation.

A novel technology that significantly enhances security and trust in wireless and wired communication networks has been developed. It is based on integration of a novel encryption mechanism and novel data packet structure with enhanced security tools. This novel data packet structure results in an unprecedented level of security and trust, while at the same time reducing power consumption and computing/communication overhead in networks. As a result, networks are provided with protection against intrusion, exploitation, and cyber attacks and posses self-building, self-awareness, self-configuring, self-healing, and self-protecting intelligence.

Over the past several years there has been an apparent shift in research focus in the area of digital steganography and steganalysis - a shift from primarily image based methods to a new focus on broader multimedia techniques. More specifically the area of digital audio steganography is of prime interest. We introduce a new high capacity, covert channel data embedding and recovery system for digital audio carrier files using a key based encoding and decoding method. It will be shown that the added information file is interleaved within the carrier file and is fully indexed allowing for segmented extraction and recovery of data at chosen start and stop points in the sampled stream. The original audio quality is not affected by the addition of this covert data. The embedded information can also be secured by a binary key string or cryptographic algorithm and resists statistical analytic detection attempts. We will also describe how this new method can be used for data compression and expansion applications in the transfer and storage of digital multimedia to increase the overall data capacity and security.

Smart sensors can gather all kind of information and process it. Cameras are still dominating and smart cameras can offer services for face recognition or person tracking. Operators are building collaborations to cover a larger area, to save costs and to add more and different sensors. Cryptographic methods may achieve integrity and confidentiality between operators, but not trust. Even if a partner or one of his sensors is authenticated, no statements can be made about the quality of the sensor data. Hence, trust must be established between the partners and their sensors. Trust can be built based on past experience. A reputation system collects opinions of operators about the behavior of sensors and calculates trust based on these opinions. Many reputation systems have been proposed, e.g., for authentication of files in peer-topeer networks. This work presents a new reputation system, which is designed to calculate the trustworthiness of smart sensors and smart sensor systems. A new trust model, including functions to calculate and update trust on past experiences, is proposed. When fusing information of multiple sensors, it cannot always be reconstructed, which information led to a bad result. Hence, an approach for fair rating is shown. The proposed system has been realized in a Service-Oriented Architecture for easy integration in existing smart sensor systems, e.g., smart surveillance systems. The model itself can be used in every decentralized heterogeneous smart sensor network.

In this paper, we discuss the effects of uncommon time constraints to the stopping criteria for managing timecritical decision-making in human-machine interaction. The continued growth of multimedia signal processing demands real-time solutions supporting a human-machine interactive decision process. We provide a methodology that locates best alternatives in human-machine interactive decision-making by including human perception in the effects of uncommon time constraints, instead of the prevailing heuristics. The proposed methodology contributes to the stopping criteria of such as an entry-exit control procedure using a monitor device by introducing the alert and confirm built-in functions into human-machine interaction for mitigating the misapprehension of decision makers that originates from human perception.

Threshold operators are conventionally used in wavelet-based denoising applications. Different thresholding schemes have been suggested to achieve improved balance between mitigating various signal distortions and preserving signal details. In general, these state-of-the-art threshold operators are nonlinear shrinkage functions such as well known "soft" and "hard" thresholds and their hybrids. Recently a nonlinear polynomial threshold has been introduced which integrates several known approaches and can be optimized using a least squares technique. While significantly improving the performance - this approach is computationally intensive and is not flexible enough for band-adaptive processing. In this paper an adaptive least mean squared (LMS) optimization approach is proposed and studied which drastically reduces computational load and is convenient for band-adaptive denoising scenarios. The approach is successfully applied to 1D and 2D signals, and the results demonstrate improved performance in comparison with the conventional methods.

Fingerprint recognition is one of the most commonly used forms of biometrics and has been widely used in daily life due to its feasibility, distinctiveness, permanence, accuracy, reliability, and acceptability. Besides cost, issues related to accuracy, security, and processing time in practical biometric recognition systems represent the most critical factors that makes these systems widely acceptable. Accurate and secure biometric systems often require sophisticated enhancement and encoding techniques that burdens the overall processing time of the system. In this paper we present a comparison between common digital and optical enhancementencoding techniques with respect to their accuracy, security and processing time, when applied to biometric fingerprint systems.

This paper aims to provide a remote fingerprint object authentication protocol dedicated to anti-counterfeiting applications. The corresponding security model is given. The suggested scheme is based on an Elliptic Curve Cryptography (ECC) encryption adding a mechanism to control integrity at the verification stage. The privacy constraint useful in many applications leads us to embed a Private Information Retrieval scheme in the protocol. As in a previous SPIE presentation, we begin with an optical reader. We drastically lower the amount of computation made at this stage.

This paper offers an innovative image processing technique (smart data compression) for some Department of Defense and Government users, who may be disadvantaged in terms of network and resource availability as they operate at the tactical edge. Specifically, we propose using the concept of autonomous anomaly detection to significantly reduce the amount of data transmitted to the disadvantaged user. The primary sensing modality is hyperspectral, where a national asset is expected to fly over the region of interest acquiring and processing data in real time, but transmitting only the corresponding data of scene anomalies, their spatial relationships in the imagery, range and navigational direction. Results from a proof of principle experiment using real hyperspectral imagery are encouraging.

Hypercomplex approaches are seeing increased application to signal and image processing problems. The use of multicomponent hypercomplex numbers, such as quaternions, enables the simultaneous co-processing of multiple signal or image components. This joint processing capability can provide improved exploitation of the information contained in the data, thereby leading to improved performance in detection and recognition problems. In this paper, we apply hypercomplex processing techniques to the logo image recognition problem. Specifically, we develop an image matcher by generalizing classical phase correlation to the biquaternion case. We further incorporate biquaternion Fourier domain alpha-rooting enhancement to create Alpha-Rooted Biquaternion Phase Correlation (ARBPC). We present the mathematical properties which justify use of ARBPC as an image matcher. We present numerical performance results of a logo verification problem using real-world logo data, demonstrating the performance improvement obtained using the hypercomplex approach. We compare results of the hypercomplex approach to standard multi-template matching approaches.

Video compression algorithms such as H.264 offer much potential for parallel processing that is not always exploited by the technology of a particular implementation. Consumer mobile encoding devices often achieve real-time performance and low power consumption through parallel processing in Application Specific Integrated Circuit (ASIC) technology, but many other applications require a software-defined encoder. High quality compression features needed for some applications such as 10-bit sample depth or 4:2:2 chroma format often go beyond the capability of a typical consumer electronics device. An application may also need to efficiently combine compression with other functions such as noise reduction, image stabilization, real time clocks, GPS data, mission/ESD/user data or software-defined radio in a low power, field upgradable implementation. Low power, software-defined encoders may be implemented using a massively parallel memory-network processor array with 100 or more cores and distributed memory. The large number of processor elements allow the silicon device to operate more efficiently than conventional DSP or CPU technology. A dataflow programming methodology may be used to express all of the encoding processes including motion compensation, transform and quantization, and entropy coding. This is a declarative programming model in which the parallelism of the compression algorithm is expressed as a hierarchical graph of tasks with message communication. Data parallel and task parallel design patterns are supported without the need for explicit global synchronization control. An example is described of an H.264 encoder developed for a commercially available, massively parallel memorynetwork processor device.

Methods for embedment of stream cipher rules into compressive Elias-type entropy coders are presented. Such systems have the ability to simultaneously compress and encrypt input data; thereby providing a fast and secure means for data compression. Procedures for maintaining compressive performance are articulated with focus on further compression optimizations. Furthermore, a novel method is proposed which exploits the compression process to hide cipherstream information in the case of a known plaintext attack. Simulations were performed on images from a variety of classes in order to grade and compare the compressive and computational costs of the novel system relative to traditional compression-followed-by-encryption methods.

Secure wireless connectivity between mobile devices and financial/commercial establishments is mature, and so is the security of remote authentication for mCommerce. However, the current techniques are open for hacking, false misrepresentation, replay and other attacks. This is because of the lack of real-time and current-precise-location in the authentication process. This paper proposes a new technique that includes freshly-generated real-time personal biometric data of the client and present-position of the mobile device used by the client to perform the mCommerce so to form a real-time biometric representation to authenticate any remote transaction. A fresh GPS fix generates the "time and location" to stamp the biometric data freshly captured to produce a single, real-time biometric representation on the mobile device. A trusted Certification Authority (CA) acts as an independent authenticator of such client's claimed realtime location and his/her provided fresh biometric data. Thus eliminates the necessity of user enrolment with many mCommerce services and application providers. This CA can also "independently from the client" and "at that instant of time" collect the client's mobile device "time and location" from the cellular network operator so to compare with the received information, together with the client's stored biometric information. Finally, to preserve the client's location privacy and to eliminate the possibility of cross-application client tracking, this paper proposes shielding the real location of the mobile device used prior to submission to the CA or authenticators.

Face recognition is one of the most desirable biometric-based authentication schemes to control access to sensitive information/locations and as a proof of identity to claim entitlement to services. The aim of this paper is to develop block-based mechanisms, to reduce recognition errors that result from varying illumination conditions with emphasis on using error correction codes. We investigate the modelling of error patterns in different parts/blocks of face images as a result of differences in illumination conditions, and we use appropriate error correction codes to deal with the corresponding distortion. We test the performance of our proposed schemes using the Extended Yale-B Face Database, which consists of face images belonging to 5 illumination subsets depending on the direction of light source from the camera. In our experiments each image is divided into three horizontal regions as follows: region1, three rows above the eyebrows, eyebrows and eyes; region2, nose region and region3, mouth and chin region. By estimating statistical parameters for errors in each region we select suitable BCH error correction codes that yield improved recognition accuracy for that particular region in comparison to applying error correction codes to the entire image. Discrete Wavelet Transform (DWT) to a depth of 3 is used for face feature extraction, followed by global/local binarization of coefficients in each subbands. We shall demonstrate that the use of BCH improves separation of the distribution of Hamming distances of client-client samples from the distribution of Hamming distances of imposter-client samples.

This paper proposes a simplied fast method to estimate the head pose in monocular images. The head pose is estimated comparing the position of the landmark points localized in the face, which correspond to pre-selected points of a 3D face model. To localize these points on a face image (such as the eyes, mouth and nose) we use Active Shape Models. These points are detected in the 2D space, and next the 3D face model is adjusted by geometric transformations to the 2D points on the face image to estimate the head angular position in the 3D space. Our preliminary experimental results are encouraging, since they show that our simplied approach has a competitive accuracy when compared to more sophisticated methods available in the literature.

In this paper, we proposed a new skin recognition approach for human identification. An automatic skin segmentation method is developed to detect the location of eyes and the skin area. A 1D Log-Gabor filter is used for skin feature extraction and template generation. And the center area of the skin template is matched with the database. The experimental results show that the proposed skin recognition method can achieve better performance compared to face recognition systems when it is used on the CASIA-Iris-Distance.

Biometrics identifies/verifies a person using his/her physiological or behavioral characteristics. It is becoming an important ally for law enforcement and homeland security. However, there are some safety and privacy concerns: biometric based systems can be accessed when users are under threat, reluctant or even unconscious states. In this paper, we introduce a new method which can identify a person and detect his/her willingness. Our experimental results show that the new approach can enhance the security by checking the consent signature while achieving very high recognition accuracy.

Tattoo segmentation is challenging due to the complexity and large variance in tattoo structures. We have developed a segmentation algorithm for finding tattoos in an image. Our basic idea is split-merge: split each tattoo image into clusters through a bottom-up process, learn to merge the clusters containing skin and then distinguish tattoo from the other skin via top-down prior in the image itself. Tattoo segmentation with unknown number of clusters is transferred to a figureground segmentation. We have applied our segmentation algorithm on a tattoo dataset and the results have shown that our tattoo segmentation system is efficient and suitable for further tattoo classification and retrieval purpose.

The aim of this article is to give a practical overview of certain commercial software for investigation of the new iPhone 4. It is demonstrated how important data stored in the iPhone are investigated. Different cases of investigations are presented that are well-suited for forensics lab work.

FPGA devices with embedded DSP and memory blocks, and high-speed interfaces are ideal for real-time video processing applications. In this work, a hardware-software co-design approach is proposed to effectively utilize FPGA features for a prototype of an automated video surveillance system. Time-critical steps of the video surveillance algorithm are designed and implemented in the FPGAs logic elements to maximize parallel processing. Other non timecritical tasks are achieved by executing a high level language program on an embedded Nios-II processor. Pre-tested and verified video and interface functions from a standard video framework are utilized to significantly reduce development and verification time. Custom and parallel processing modules are integrated into the video processing chain by Altera's Avalon Streaming video protocol. Other data control interfaces are achieved by connecting hardware controllers to a Nios-II processor using Altera's Avalon Memory Mapped protocol.

Digital steganographic algorithms hide secret messages in seemingly innocent cover objects, such as images. Steganographic algorithms are rapidly evolving, reducing distortions, and making detection of altered cover objects by steganalysis algorithms more challenging. The value of current steganographic and steganalysis algorithms is difficult to evaluate until they are tested on realistic datasets. We propose a system approach to steganalysis for reliably detecting steganographic objects among a large number of images, acknowledging that most digital images are intact. The system consists of a cascade of intrinsic image formations filters (IIFFs), where the IIFFs in the early stage are designed to filter out non-stego images based on real world constraints, and the IIFFs in the late stage are designed to detect intrinsic features of specific steganographic routines. Our approach makes full use of all available constraints, leading to robust detection performance and low probability of false alarm. Our results based on a large image set from Flickr.com demonstrate the potential of our approach on large-scale real-world repositories.

The aim of this article is to show forensic investigation methods for mobile phones to students in a university forensic lab environment. Students have to learn the usefulness of forensic procedures to ensure evidence collection, evidence preservation, forensic analysis, and reporting. Open source tools as well as commercial forensic tools for forensic investigation of modern mobile (smart) phones are used. It is demonstrated how important data stored in the mobile device are investigated. Different scenarios of investigations are presented that are well-suited for forensics lab work in university.

This paper proposes a palmprint identification system using Finite Ridgelet Transform (FRIT) and Bayesian classifier. FRIT is applied on the ROI (region of interest), which is extracted from palmprint image, to extract a set of distinctive features from palmprint image. These features are used to classify with the help of Bayesian classifier. The proposed system has been tested on CASIA and IIT Kanpur palmprint databases. The experimental results reveal better performance compared to all well known systems.

Biometric systems such as face recognition must address four key challenges: efficiency, robustness, accuracy and security. Isometric projection has been proposed as a robust dimension reduction technique for a number of applications, but it is computationally demanding when applied to high dimensional spaces such as the space of face images. On the other hand, wavelet transforms have shown to provide an efficient tool for facial feature representation and face recognition with significant reduction in dimension. In this paper, we propose a hybrid approach that combines the efficiency and robustness of wavelet transforms with isometric projections for face features extraction in the transformed domain to be used for recognition. We shall compare the recognition accuracy of our approach with the accuracy of other commonly used projection techniques in the wavelet domain such as PCA and LDA. The security of biometric templates is addressed by adopting a lightweight random projection technique as an add-on subsystem. The results are based on experiments conducted on a publicly available benchmark face database.

Although biometric authentication is perceived to be more reliable than traditional authentication schemes, it becomes vulnerable to many attacks when it comes to remote authentication over open networks and raises serious privacy concerns. This paper proposes a biometric-based challenge-response approach to be used for remote authentication between two parties A and B over open networks. In the proposed approach, a remote authenticator system B (e.g. a bank) challenges its client A who wants to authenticate his/her self to the system by sending a one-time public random challenge. The client A responds by employing the random challenge along with secret information obtained from a password and a token to produce a one-time cancellable representation of his freshly captured biometric sample. The one-time biometric representation, which is based on multi-factor, is then sent back to B for matching. Here, we argue that eavesdropping of the one-time random challenge and/or the resulting one-time biometric representation does not compromise the security of the system, and no information about the original biometric data is leaked. In addition to securing biometric templates, the proposed protocol offers a practical solution for the replay attack on biometric systems. Moreover, we propose a new scheme for generating a password-based pseudo random numbers/permutation to be used as a building block in the proposed approach. The proposed scheme is also designed to provide protection against repudiation. We illustrate the viability and effectiveness of the proposed approach by experimental results based on two biometric modalities: fingerprint and face biometrics.

This paper presents a palmprint based verification system using SIFT features and Lagrangian network graph technique. We employ SIFT for feature extraction from palmprint images whereas the region of interest (ROI) which has been extracted from wide palm texture at the preprocessing stage, is considered for invariant points extraction. Finally, identity is established by finding permutation matrix for a pair of reference and probe palm graphs drawn on extracted SIFT features. Permutation matrix is used to minimize the distance between two graphs. The propsed system has been tested on CASIA and IITK palmprint databases and experimental results reveal the effectiveness and robustness of the system.

The PreNotiS (preventive notification system) was proposed to address the current lack in consumer prevention and disaster informatics systems. The underscore of this letter is to propose PreNotiS as a provision of trusted proxies of information sourcing to be integral to the disaster informatics framework. To promote loose coupling among subsystems, PreNotiS has evolved into a model-view-controller (MVC) architecture via object-oriented incremental prototyping. The MVC specifies how all subsystems and how they interact with each other. A testing framework is also proposed for the PreNotiS to verify multiple concurrent user access which might be observable during disasters. The framework relies on conceptually similar self-test modules to help with serviceability.

This paper demonstrates wavelet-denoising approach using polynomial threshold operators in 3-dimensional applications. This paper compares the efficacy of different denoising algorithms on 3D biomedical images using 3D wavelet transform. The denoising mechanism is demonstrated by mitigating noise of different variances using polynomial thresholding. Our approach is to apply a parameterized threshold and optimally choose the parameters for high performance noise suppression depending on the nature of the images and noise. Comparative studies in the wavelet domain conclude that the presented method is viable for 3D applications. It also confirms the feasibility in using the polynomial threshold operators as a wavelet-polynomial threshold based interpolation filter. The filter applied to assist three spatial-based interpolation algorithms (i.e. Nearest-neighbor, Bilinear, and Bicubic) and to a spectral wavelet-based interpolation algorithm. Simulation shows that the denoising using polynomial threshold operators mitigates distortions for the interpolation.

Image similarity measures are crucial for image processing applications which require comparisons to ideal reference images in order to assess performance. The Structural Similarity (SSIM), Gradient Structural Similarity (GSSIM), 4-component SSIM (4-SSIM) and 4-component GSSIM (4-GSSIM) indexes are motivated by the fact that the human visual system is adapted to extract local structural information. In this paper, we propose a new measure which enhances the gradient information used for quality assessment. An analysis of the proposed image similarity measure using the LIVE database of distorted images and their corresponding subjective evaluations of visual quality illustrate the improved performance of the proposed metric.

Hands-on experiments with electronic devices have been recognized as an important element in the field of engineering to help students get familiar with theoretical concepts and practical tasks. The continuing increase the student number, costly laboratory equipment, and laboratory maintenance slow down the physical lab efficiency. As information technology continues to evolve, the Internet has become a common media in modern education. Internetbased remote laboratory can solve a lot of restrictions, providing hands-on training as they can be flexible in time and the same equipment can be shared between different students. This article describes an on-going remote hands-on experimental radio modulation, network and mobile applications lab project "eComLab". Its main component is a remote laboratory infrastructure and server management system featuring various online media familiar with modern students, such as chat rooms and video streaming.

This paper presents a new image contrast enhancement method based on the complex Ensemble Empirical Mode Decomposition (EEMD) and alpha-rooting approaches. The new scheme also takes advantage of Fourier transform properties and complex Intrinsic Mode Functions (IMFs). The experimental results show that the images obtained by this combination are more appealing than the best results yielded by classical alpha-rooting technique. This algorithm is tested on the medical images, hazy aerial images and other low contrast images.

This paper presents a color image enhancement algorithm based on shifting the histogram of logarithmic DCT coefficients of the luminance component. The novelty of this work lies in utilizing both the spatial domain histogram concept and transform domain enhancement techniques to perform the enhancement. This paper also demonstrates a quantitative measurement based upon contrast entropy to choose the optimal parameters and determines the effectiveness of the method. In addition, different color spaces including HSV, YCbCr, and Principle Component Analysis (PCA) have been examined for the algorithm to determine the best visual results. We also present a comprehensive review of four commonly used color image enhancement techniques, such as mutiscale retinex with color restoration, multi contrast enhancement, multi contrast enhancement with dynamic compression and color image enhancement by scaling, and compare them with the proposed histogram shifting techniques. Computer simulations and analysis show that the histogram shifting method outperforms the commonly used methods for most images.

The common treatments of melancholia are psychotherapy and taking medicines. The psychotherapy treatment which this study focuses on is limited by time and location. It is easier for psychiatrists to grasp information from clinical manifestation but it is difficult for psychiatrists to collect information from patients' daily conversations or emotion. To design a system which psychiatrists enable to capture patients' daily symptoms will show great help in the treatment. This study proposes to use fuzzy data mining algorithm to find association rules among keywords segmented from patients' daily voice/text messages to assist psychiatrists extract useful information before outpatient service. Patients of melancholia can use devices such as mobile phones or computers to record their own emotion anytime and anywhere and then uploading the recorded files to the back-end server for further analysis. The analytical results can be used for psychiatrists to diagnose patients' degrees of melancholia. Experimental results will be given to verify the effectiveness of the proposed methodology.

In recent years, driven by the development of steganalysis methods, steganographic algorithms have been evolved rapidly with the ultimate goal of an unbreakable embedding procedure, resulting in recent steganographic algorithms with minimum distortions, exemplified by the recent family of Modified Matrix Encoding (MME) algorithms, which has shown to be most difficult to be detected. In this paper we propose a compressed sensing based on approach for intrinsic steganalysis to detect MME stego messages. Compressed sensing is a recently proposed mathematical framework to represent an image (in general, a signal) using a sparse representation relative to an overcomplete dictionary by minimizing the l1-norm of resulting coefficients. Here we first learn a dictionary from a training set so that the performance will be optimized using the KSVD algorithm; since JPEG images are processed by 8x8 blocks, the training examples are 8x8 patches, rather than the entire images and this increases the generalization of compressed sensing. For each 8x8 block, we compute its sparse representation using OMP (orthogonal matching pursuit) algorithm. Using computed sparse representations, we train a support vector machine (SVM) to classify 8x8 blocks into stego and non-stego classes. Then given an input image, we first divide it into 8x8 blocks. For each 8x8 block, we compute its sparse representation and classify it using the trained SVM. After all the 8x8 blocks are classified, the entire image is classified based on the majority rule of 8x8 block classification results. This allows us to achieve a robust decision even when 8x8 blocks can be classified only with relatively low accuracy. We have tested the proposed algorithm on two datasets (Corel-1000 dataset and a remote sensing image dataset) and have achieved 100% accuracy on classifying images, even though the accuracy of classifying 8x8 blocks is only 80.89%.

In this paper the possibilities of forensic investigation of CD, DVD and Blu-ray Discs is presented. It is shown which information can be read by using freeware and commercial software for forensic examination. It particulary it describes the visualization of hidden content and the possibility to find out information about the burning hardware used for writing.