This PDF file contains the front matter associated with SPIE Proceedings Volume 6579, including the Title Page, Copyright information, Table of Contents, Introduction, and the Conference Committee listing.

We consider dense networks of surveillance cameras capturing overlapped images of the same scene from different viewing directions, such a scenario being referred to as multi-view. Data compression is paramount in such a system due to the large amount of captured data. In this paper, we propose a Multi-view Distributed Video Coding approach. It allows for low complexity / low power consumption at the encoder side, and the exploitation of inter-view correlation without communications among the cameras. We introduce a combination of temporal intra-view side information and homography inter-view side information. Simulation results show both the improvement of the side information, as well as a significant gain in terms of coding efficiency.

In this paper, we present a memory-efficient, contour-based, region-of-interest (ROI) algorithm designed for ultra-low-bit- rate compression of very large images. The proposed technique is integrated into a user-interactive wavelet-based image coding system in which multiple ROIs of any shape and size can be selected and coded efficiently. The coding technique compresses region-of-interest and background (non-ROI) information independently by allocating more bits to the selected targets and fewer bits to the background data. This allows the user to transmit large images at very low bandwidths with lossy/lossless ROI coding, while preserving the background content to a certain level for contextual purposes. Extremely large images (e.g., 65000 X 65000 pixels) with multiple large ROIs can be coded with minimal memory usage by using intelligent ROI tiling techniques. The foreground information at the encoder/decoder is independently extracted for each tile without adding extra ROI side information to the bit stream. The arbitrary ROI contour is down-sampled and differential chain coded (DCC) for efficient transmission. ROI wavelet masks for each tile are generated and processed independently to handle any size image and any shape/size of overlapping ROIs. The resulting system dramatically reduces the data storage and transmission bandwidth requirements for large digital images with multiple ROIs.

This paper presents a software-only, real-time video coder/decoder (codec) with super-resolution-based enhancement for ultra-low-bit-rate compression. The codec incorporates a modified JPEG2000 core and interframe predictive coding, and can operate with network bandwidths as low as 500 bits/second. Highly compressed video exhibits severe coding artifacts that degrade visual quality. To lower the level of noise and retain the sharpness of the video frames, we build on our previous work in super-resolution-based video enhancement and propose a new version that is suitable for real-time video coding systems. The adopted super-resolution-based enhancement uses a constrained set of motion vectors that is computed from the original (uncompressed) video at the encoder. Artificial motion is also added to the difference frame to maximize the enhancement performance. The encoder can transmit either the full set of motion vectors or the constrained set of motion vectors depending upon the available bandwidth. At the decoder, each pixel of the decoded frame is assigned to a motion vector from the constrained motion vector set. L₂-norm minimization super-resolution is then applied to the decoded frame set (previous frame, current frame, and next frame). A selective motion estimation scheme is proposed to prevent ghosting, which otherwise would result from the super-resolution enhancement when the motion estimation fails to find appropriate motion vectors. Results using the proposed system demonstrate significant improvements in the quantitative and visual quality of the coded video sequences.

While bandwidth and storage capabilities are increasing with advances in technology, so are the demands of users as quantities of data to transmit and store grow simultaneously. The need for effective signal compression is always present, as reduction in data size is required while maintaining minimal signal destruction. This paper presents a new signal compression scheme that uses coordinate logic transforms in combination with Boolean minimized representations. The processing of the coordinate logic transforms on signal data helps reduce unnecessary signal complexity allowing more effective data reduction during the Boolean minimized form encoding. The coordinate logic transforms are an alternate method for calculating coordinate logic filters which simplify the computational complexity through algorithmic parallelism. In this work, the combination of coordinate logic transforms with a Boolean minimized form encoding scheme allows for a new compression technique that is applicable to both binary and grayscale input images.

In this paper, we propose an intelligent bird information retrieval system which aims to construct a mobility-learning activity under the up-to-date wireless technology. The system consists of a Tablet PC and PDAs with wireless networking capabilities. The PDA is equipped with a friendly retrieval interface and a good learning environment. In our system, users only need to click the buttons or input the keywords to retrieve bird information. Besides, users can discuss or share their information and knowledge via the wireless network. Our system saves bird information in four categories including "Introduction," "Images," "Sound," "Streaming Media," and "Ecological Memo." The integral knowledge helps users understand more about birds. Data mining and fuzzy association rules are applied to recommend users those birds they may be interested in. A streaming server on the Tablet PC is built to provide the streaming media for PDA users. By this way, PDA users can enjoy the multimedia from Tablet PC in real time without downloading completely. Finally, the system is a perfect tool for outdoor teaching and can be easily extended to provide navigation and touring services for national parks or museums.

Malicious nodes can seriously impair the performance of wireless ad hoc networks as a result of different actions such as packet dropping. Secure routes are shortest paths on which every node on the route is trusted even if unknown. Secure route discovery requires the adoption of mechanisms of associating trust to nodes. Most existing secure route discovery mechanisms rely on shared keys and digital signature. In the absence of central nodes that act as certification authority, such protocols suffer from heavy computational burden and are vulnerable to malicious attacks. In this paper we shall review existing techniques for secure routing and propose to complement route finding with creditability scores. Each node would have a credit list for its neighbors. Each node monitors its neighbors' pattern of delivering packets and regularly credits are reviewed and updated accordingly. Unlike most existing schemes the focus of our work is based on post route discovery stage, i.e. when packets are transmitted on discovered routes. The level of trust in any route will be based on the credits associated with the neighbors belonging to the discovered route. We shall evaluate the performance of the proposed scheme by modifying our simulation system so that each node has a dynamic changing "credit list" for its neighbors' behavior. We shall conduct a series of simulations with and without the proposed scheme and compare the results. We will demonstrate that the proposed mechanism is capable of isolating malicious nodes and thereby counteracting black hole attacks.

This paper presents a new method which enhances the quality-of-service (QoS) and hence the response time and queuing delay of real-time interactive multimedia over the Internet. A service class based on differentiated services mechanism has been developed. Evaluation of response time under different traffic conditions has been conducted via simulation. Specifically, the impact on routers performance at the boundary of a DS-enabled domain was evaluated using OPNET and the results are presented. Since audio and video traffic have different needs, priority schemes for different types of interactive multimedia traffic have been studied to provide control and predictable service, and therefore, better quality of service.

Cellular communications constitute a significant portion of the global telecommunications market. Therefore, the need for secured communication over a mobile platform has increased exponentially. Steganography is an art of hiding critical data into an innocuous signal, which provide answers to the above needs. The JPEG is one of commonly used format for storing and transmitting images on the web. In addition, the pictures captured using mobile cameras are in mostly in JPEG format. In this article, we introduce a switching theory based steganographic system for JPEG images which is applicable for mobile and computer platforms. The proposed algorithm uses the fact that energy distribution among the quantized AC coefficients varies from block to block and coefficient to coefficient. Existing approaches are effective with a part of these coefficients but when employed over all the coefficients they show there ineffectiveness. Therefore, we propose an approach that works each set of AC coefficients with different frame work thus enhancing the performance of the approach. The proposed system offers a high capacity and embedding efficiency simultaneously withstanding to simple statistical attacks. In addition, the embedded information could be retrieved without prior knowledge of the cover image. Based on simulation results, the proposed method demonstrates an improved embedding capacity over existing algorithms while maintaining a high embedding efficiency and preserving the statistics of the JPEG image after hiding information.

Modern steganography is a secure communication of information by embedding a secret-message within a "cover" digital multimedia without any perceptual distortion to the cover media, so the presence of the hidden message is indiscernible. Recently, the Joint Photographic Experts Group (JPEG) format attracted the attention of researchers as the main steganographic format due to the following reasons: It is the most common format for storing images, JPEG images are very abundant on the Internet bulletin boards and public Internet sites, and they are almost solely used for storing natural images. Well-known JPEG steganographic algorithms such as F5 and Model-based Steganography provide high message capacity with reasonable security. In this paper, we present a method to increase security using JPEG images as the cover medium. The key element of the method is using a new parametric key-dependent quantization matrix. This new quantization table has practically the same performance as the JPEG table as far as compression ratio and image statistics. The resulting image is indiscernible from an image that was created using the JPEG compression algorithm. This paper presents the key-dependent quantization table algorithm and then analyzes the new table performance.

There are several security issues tied to multimedia when implementing the various applications in the cellular phone and wireless industry. One primary concern is the potential ease of implementing a steganography system. Traditionally, the only mechanism to embed information into a media file has been with a desktop computer. However, as the cellular phone and wireless industry matures, it becomes much simpler for the same techniques to be performed using a cell phone. In this paper, two methods are compared that classify cell phone images as either an anomaly or clean, where a clean image is one in which no alterations have been made and an anomalous image is one in which information has been hidden within the image. An image in which information has been hidden is known as a stego image. The main concern in detecting steganographic content with machine learning using cell phone images is in training specific embedding procedures to determine if the method has been used to generate a stego image. This leads to a possible flaw in the system when the learned model of stego is faced with a new stego method which doesn't match the existing model. The proposed solution to this problem is to develop systems that detect steganography as anomalies, making the embedding method irrelevant in detection. Two applicable classification methods for solving the anomaly detection of steganographic content problem are single class support vector machines (SVM) and Parzen-window. Empirical comparison of the two approaches shows that Parzen-window outperforms the single class SVM most likely due to the fact that Parzen-window generalizes less.

One of the great challenges of the existing watermarking methods is their limited resistance to the extensive geometric attacks. Geometric attacks can be decomposed into two classes: global distortion such as rotations and translations and local distortion such as the StirMark attack. We have found that the weakness of multiple watermark embedding methods that were initially designed to resist geometric attacks is the inability to withstand the combination of geometric attacks. In this paper, the watermark is used in the gray-scale authentication image. We propose a robust image watermarking scheme that can withstand the geometric attacks by using local tri-mesh feature points. Our proposed method can resynchronize the attacked images and is independent of the embedding and authentication process. The geometric invariant scheme is combined with the complementary modulation embedding strategy to enhance the resistance of geometric attacks. The experimental results verify that the proposed scheme is effective for geometric attacks.

In this contribution, we present a novel technique for imperceptible and robust watermarking of digital images. It is based on the host image 2-th level decomposition using the Fibonacci-Haar Transform (FHT) and on the Singular Value Decomposition (SVD) of the transformed subbands. The main contributions of this approach are the use of the FHT for hiding purposes, the flexibility in data hiding capacity, and the key-dependent secrecy of the used transform. The experimental results show the effectiveness of the proposed approach both in perceived quality of the watermarked image and in robustness against the most common attacks.

In this paper we introduce a new chaotic stream cipher Mmohocc which utilizes the fundamental chaos characteristics. The designs of the major components of the cipher are given. Its cryptographic properties of period, auto- and cross-correlations, and the mixture of Markov processes and spatiotemporal effects are investigated. The cipher is resistant to the related-key-IV, Time/Memory/Data tradeoff, algebraic, and chosen-text attacks. The keystreams successfully passed two batteries of statistical tests and the encryption speed is comparable with RC4.

In the last decade a lot of efforts have been devoted to the development of biometrics-based authentication systems. In this paper we propose a signature-based biometric authentication system, where watermarking techniques are used to embed some dynamic signature features in a static representation of the signature itself, stored either in a centralized database or in a smartcard. The user authentication can be performed either by using some static features extracted from the acquired signature or by using both the aforementioned static features together with the dynamic features embedded in the enrollment stage. A multi-level authentication system, which is capable to provide various degree of security, is thus obtained. The proposed watermarking techniques are tailored to images with sharp edges, like a signature picture, in order to obtain a robust embedding method while keeping intact the original structure of the host signal. Experimental results show the two different levels of security which can be reached when either static features or both static and dynamic features are employed in the authentication process.

Receiver operating characteristic (ROC) curve is widely used in biometric identification. It is a plot of the detection power virus false alarm rate. It is an objective measure of accuracy. Positive biometrics identification is one-to-many match. ROC curve has been served as a "golden" criterion in measuring the accuracy of biometrics system for positive biometric identification. However, in this paper, we will analyze the problems of using ROC curve as the sole criterion in positive biometrics identification. From the view of detection and estimation theory, ROC curve only took concerns of system variance, and would not be able to detect the system bias, which could give wrong conclusion in evaluating system accuracy across multiple databases. ROC curve does not reflect the cost function, the database size, the quality of the image, and many other factors that are important in system performance and accuracy. We will use iris recognition as an example to discuss these issues. At the end, we will discuss some possible solutions to solve these problems.

In this paper, a novel authentications system combining biometric cryptosystems with digital watermarking is presented. One of the main vulnerabilities of the existing data hiding systems is the public knowledge of the embedding domain. We propose the use of biometric data, minutiae fingerprint set, for generating the encryption key needed to decompose an image in the Tree structured Haar transform. The uniqueness of the biometrics key together with other, embedded, biometric information guarantee the authentication of the user. Experimental tests show the effectiveness of the proposed system.

A new computationally efficient framework for vehicle tracking on a mobile platform is proposed. The principal component of the framework is the log-polar transformation applied to video frames captured from a standard uniformly sampled format camera. The log-polar transformation provides two major benefits to real-time vehicle tracking from a mobile vehicle platform moving along a single or multi-lane road. First, it significantly reduces the amount of data required to be processed since it collapses the original Cartesian video frames into log-polar images with much smaller dimensions. Second, the log-polar transformation is capable of mitigating perspective distortion due to its scale invariance property. This second aspect is of interest for vehicle tracking because the target vehicle appearance is preserved for all distances from the observer (camera). This works however only if the center of log-polar transformation is coincident with the vanishing point of perspective view. Therefore, a road following algorithm is proposed to keep the center of log-polar transform on the vanishing point at every video frame compensating for the carrying vehicle movements. Since the algorithm is intended to be used in the mobile embedded devices, it is developed to achieve both mathematical simplicity and algorithmic efficiency while avoiding computationally expensive mathematical functions. The use of trigonometric and exponential functions is minimized comparing to the log-Hough transform traditionally used in log-polar space. This new algorithm focuses on straight radial line fragments, thus shifting its mathematical engine to the linear equations' domain.

This paper presents a method of image enhancement using an adaptive thresholding method based on the human visual system. We utilize a number of different enhancement algorithms applied selectively to the different regions of an image to achieve a better overall enhancement than applying a single technique globally. The presented method is useful for images that contain various regions of improper illumination. It is also practical for correcting shadows. This thresholding system allows various enhancement algorithms to be used on different sections of the image based on the local visual characteristics. It further allows the parameters to be tuned differently for the specific regions, giving a more visually pleasing output image. We demonstrate the algorithm and present results for several high quality images as well as lower quality images such as those captured using a cell phone camera. We then compare and contrast our method to other state-of-the-art enhancement algorithms.

The wavelet transform is a powerful tool for image and video processing, useful in a range of applications. This paper is concerned with the efficiency of a certain fast-wavelet-transform (FWT) implementation and several wavelet filters, more suitable for constrained devices. Such constraints are typically found on mobile (cell) phones or personal digital assistants (PDA). These constraints can be a combination of; limited memory, slow floating point operations (compared to integer operations, most often as a result of no hardware support) and limited local storage. Yet these devices are burdened with demanding tasks such as processing a live video or audio signal through on-board capturing sensors. In this paper we present a new wavelet software library, HeatWave, that can be used efficiently for image/video processing/analysis tasks on mobile phones and PDA's. We will demonstrate that HeatWave is suitable for realtime applications with fine control and range to suit transform demands. We shall present experimental results to substantiate these claims. Finally this library is intended to be of real use and applied, hence we considered several well known and common embedded operating system platform differences; such as a lack of common routines or functions, stack limitations, etc. This makes HeatWave suitable for a range of applications and research projects.

In this paper, a new watermarking scheme using elemental images in integral imaging for protection of information is proposed. Elemental images, which have an effect of distributing information of three-dimensional objects, are used as a watermark. The elemental images watermark with depth information of embedded patterns provides us to reconstruct entire embedded pattern with partial information. To show the usefulness of the proposed scheme, we carry out preliminary experiments and show some experimental results.

In this paper we propose an Orthogonal Frequency Division Multiplexing Ultra Wide Band (OFDM-UWB) system that introduces encryption, mutual authentication, and data integrity functions, at the physical layer, without impairing spectral efficiency. Encryption is performed by rotating the constellation employed in each band by means of a pseudorandom phase-hopping sequence. Authentication and data integrity, based on encrypted-hash, are directly coupled with Forward Error Correction (FEC). Dependence of the phase hopping sequence on the transmitted message deny the use of the phase hopping obtained by means of known and chosen plaintext attacks for decryption of further messages. Moreover, since phase hopping generation keys change very rapidly they are also difficultly detectable from a hypothetic man in the middle. Computer simulations confirm the superior performance, even in terms of BER, to a standard PSKOFDM system, due to the FEC capabilities of encrypted hash.

Measurement of image similarity is important for a number of image processing applications. Image similarity assessment is closely related to image quality assessment in that quality is based on the apparent differences between a degraded image and the original, unmodified image. Automated evaluation of image compression systems relies on accurate quality measurement. Current algorithms for measuring similarity include mean squared error (MSE), peak signal-to-noise ratio (PSNR), and structural similarity (SSIM). They have some limitations: such as consistent, accuracy and incur greater computational cost. In this paper, we show that a modified version of the measurement of enhancement by entropy (EME) can be used as an image similarity measure, and thus an image quality measure. Until now, EME has generally been used to measure the level of enhancement obtained using a given enhancement algorithm and enhancement parameter. The similarity-EME (SEME) is based on the EME for enhancement. We will compare SEME to existing measures over a set of images subjectively judged by humans. Computer simulations have demonstrated its promise through a set of examples, as well as comparison to both subjective ratings and state-of-the-art objective methods on a database of images compressed with JPEG.

In recent years there has been an increased interest in audio steganography and watermarking. This is due primarily to two reasons. First, an acute need to improve our national security capabilities in light of terrorist and criminal activity has driven new ideas and experimentation. Secondly, the explosive proliferation of digital media has forced the music industry to rethink how they will protect their intellectual property. Various techniques have been implemented but the phase domain remains a fertile ground for improvement due to the relative robustness to many types of distortion and immunity to the Human Auditory System. A new method for embedding data in the phase domain of the Discrete Fourier Transform of an audio signal is proposed. Focus is given to robustness and low perceptibility, while maintaining a relatively high capacity rate of up to 172 bits/s.

An empirical method for Canny filter optimization is explored and applied to the problem of measuring rigid motion between targets for nanometer motion detection. Operating with an image space pixel size of 3 μm, we are able to obtain static target localization to 6 nm at 2σ variation. To discriminate target roughness from sub-pixel measurement noise we use a Laplacian filter method. To extend the resolution beyond the limits of a single sub-pixel sample we use multiple adjacent edge locations along a single target to statistically reduce the overall resolution. With sufficient samples we obtain near .001 pixels resolving power. Even at this resolution we have not reached the limits of sampling which are possible from simultaneously sampling sets of parallel lines allowing for future refinement of method to localize well below .001 pixels.