This PDF file contains the front matter associated with SPIE Proceedings Volume 7351, including the Title Page, Copyright information, table of Contents, Introduction (if any), and the Conference Committee listing.

Detecting dim targets in infrared imagery remains a challenging task. Several techniques exist for detecting bright, high contrast targets such as CFAR detectors, edge detection, and spatial thresholding. However, these approaches often fail for detection of targets with low contrast relative to background clutter. In this paper we exploit the transient capture capability and directional filtering aspect of wavelets to develop a wavelet based image enhancement method. We develop an image representation, using wavelet filtered imagery, which facilitates dim target detection. We further process the wavelet-enhanced imagery using the Michelson visibility operator to perform nonlinear contrast enhancement prior to target detection. We discuss the design of optimal wavelets for use in the image representation. We investigate the effect of wavelet choice on target detection performance, and design wavelets to optimize measures of visual information on the enhanced imagery. We present numerical results demonstrating the effectiveness of the approach for detection of dim targets in real infrared imagery. We compare target detection performance to performance obtained using standard techniques such as edge detection. We also compare performance to target detection performed on imagery enhanced by optimizing visual information measures in the spatial domain. We investigate the stability of the optimal wavelets and detection performance variation, across perspective changes, image frame sample (for frames extracted from infrared video sequences), and image scene content types. We show that the wavelet-based approach can usually detect the targets with fewer false-alarm regions than possible with standard approaches.

Tracking moving objects is a critical step for smart video surveillance systems. Despite the complexity increase, multiple camera systems exhibit the undoubted advantages of covering wide areas and handling the occurrence of occlusions by exploiting the different viewpoints. The technical problems in multiple camera systems are several: installation, calibration, objects matching, switching, data fusion, and occlusion handling. In this paper, we address the issue of tracking moving objects in an environment covered by multiple un-calibrated cameras with overlapping fields of view, typical of most surveillance setups. Our main objective is to create a framework that can be used to integrate objecttracking information from multiple video sources. Basically, the proposed technique consists of the following steps. We first perform a single-view tracking algorithm on each camera view, and then apply a consistent object labeling algorithm on all views. In the next step, we verify objects in each view separately for inconsistencies. Correspondent objects are extracted through a Homography transform from one view to the other and vice versa. Having found the correspondent objects of different views, we partition each object into homogeneous regions. In the last step, we apply the Homography transform to find the region map of first view in the second view and vice versa. For each region (in the main frame and mapped frame) a set of descriptors are extracted to find the best match between two views based on region descriptors similarity. This method is able to deal with multiple objects. Track management issues such as occlusion, appearance and disappearance of objects are resolved using information from all views. This method is capable of tracking rigid and deformable objects and this versatility lets it to be suitable for different application scenarios.

A modified multiframe image restoration algorithm of degraded images is described in this paper, which is based on the method developed by V. Katkovnik. The projection gradient algorithm based on anisotropic LPA-ICI filtering, being proposed by V. Katkovnik, could only restore images contaminated by Gaussian noisy, also it was too complicated and time consuming. By improving Katkovnik's cost function and applying constraints on image intensity value in iteration, we reached a new multiframe recursive iteration restoration scheme in frequency domain. This method is suitable for reconstruction of images degenerated by both Gaussian noise and Poissonian noise, as well by mixed noise. Experimental results demonstrated that this method works efficiently, and could well restore images blurred heavily by multi-noise.

In imaging applications the prevalent effects of atmospheric turbulence comprise image dancing and image blurring. Suggestions from the field of image processing to compensate for these turbulence effects and restore degraded imagery include Motion-Compensated Averaging (MCA) for image sequences. In isoplanatic conditions, such an averaged image can be considered as a non-distorted image that has been blurred by an unknown Point Spread Function (PSF) of the same size as the pixel motions due to the turbulence and a blind deconvolution algorithm can be employed for the final image restoration. However, when imaging over a long horizontal path close to the ground, conditions are likely to be anisoplanatic and image dancing will effect local image displacements between consecutive frames rather than global shifts only. Therefore, in this paper, a locally operating variant of the MCA-procedure is proposed, utilizing Block Matching (BM) in order to identify and re-arrange uniformly displaced image parts. For the final restoration a multistage blind deconvolution algorithm is used and the corresponding deconvolution results are presented and evaluated.

In this paper, a technique is presented to alleviate ghosting artifacts in the decoded video sequences for low-bit-rate video coding. Ghosting artifacts can be defined as the appearance of ghost like outlines of an object in a decoded video frame. Ghosting artifacts result from the use of a prediction loop in the video codec, which is typically used to increase the coding efficiency of the video sequence. They appear in the presence of significant frame-to-frame motion in the video sequence, and are typically visible for several frames until they eventually die out or an intra-frame refresh occurs. Ghosting artifacts are particularly annoying at low bit rates since the extreme loss of information tends to accentuate their appearance. To mitigate this effect, a procedure with selective in-loop filtering based on motion vector information is proposed. In the proposed scheme, the in-loop filter is applied only to the regions where there is motion. This is done so as not to affect the regions that are devoid of motion, since ghosting artifacts only occur in high-motion regions. It is shown that the proposed selective filtering method dramatically reduces ghosting artifacts in a wide variety of video sequences with pronounced frame-to-frame motion, without degrading the motionless regions.

Indoor Positioning Systems using WLANs have become very popular in recent years. These systems are spawning a new class of applications like activity recognition, surveillance, context aware computing and location based services. While Global Positioning System (GPS) is the natural choice for providing navigation in outdoor environment, the urban environment places a significant challenge for positioning using GPS. The GPS signals can be significantly attenuated, and often completely blocked, inside buildings or in urban canyons. As the performance of GPS in indoor environments is not satisfactory, indoor positioning systems based on location fingerprinting of WLANs is being suggested as a viable alternative. The Indoor WLAN Positioning Systems suffer from several phenomena. One of the problems is the continual availability of access points, which directly affects the positioning accuracy. Integrity monitoring of WLAN localization, which computes WLAN positioning with different sets of access points is proposed as a solution for this problem. The positioning accuracy will be adequate for the sets which do not contain faulty or the access points which are offline, while the sets with such access points will fail and they will report random and inaccurate results. The proposed method identifies proper sets and identifies the rogue access points using prediction trajectories. The combination of prediction and correct access point set selection provides a more accurate result. This paper discusses about integrity monitoring method for WLAN devices and followed by how it monitors and developing the application on mobile platforms.

Traditional iris recognition algorithms can work well for the frontal iris images. However, when the gaze of an eye changes with respect to the camera lens, many times the size, shape, and detail of iris patterns will change as well and cannot be matched to enrolled images using traditional methods. Additionally, the transformation of off-angle eyes to polar coordinates becomes much more challenging and noncooperative iris algorithms will require a different approach. In this paper, we propose a new approach for iris recognition. This new method does not require polar transformation, affine transformation or highly accurate segmentation to perform iris recognition. Our research results using the remote non-cooperative iris Image database show that the proposed method works well on frontal look images and off-angle images as well.

The first step in a facial recognition system is to find and extract human faces in a static image or video frame. Most face detection methods are based on statistical models that can be trained and then used to classify faces. These methods are effective but the main drawback is speed because a massive number of sub-windows at different image scales are considered in the detection procedure. A robust face detection technique based on an encoded image known as an "integral image" has been proposed by Viola and Jones. The use of an integral image helps to reduce the number of operations to access a sub-image to a relatively small and fixed number. Additional speedup is achieved by incorporating a cascade of simple classifiers to quickly eliminate non-face sub-windows. Even with the reduced number of accesses to image data to extract features in Viola-Jones algorithm, the number of memory accesses is still too high to support realtime operations for high resolution images or video frames. The proposed hardware design in this research work employs a modular approach to represent the "integral image" for this memory-intensive application. An efficient memory manage strategy is also proposed to aggressively utilize embedded memory modules to reduce interaction with external memory chips. The proposed design is targeted for a low-cost FPGA prototype board for a cost-effective face detection/recognition system.

This paper presents a low-cost method for providing biometric verification for applications that do not require large database sizes. Existing portable iris recognition systems are typically self-contained and expensive. For some applications, low cost is more important than extremely discerning matching ability. In these instances, the proposed system could be implemented at low cost, with adequate matching performance for verification. Additionally, the proposed system could be used in conjunction with any image based biometric identification system. A prototype system was developed and tested on a small database, with promising preliminary results.

Face recognition has been widely used to automatically identify and verify a person. In this paper, we proposed a new approach based on orthogonal subspace projection (OSP) to identification of human faces. In linear mixture model of the face images, the OSP faces of the training images are calculated by using orthogonal subspace projection approach and the signal-to noise ratio maximization. And the weight parameter of the input image is obtained to do face recognition.

This paper introduces a new recursive sequence called the truncated P-Fibonacci sequence, its corresponding binary code called the truncated Fibonacci p-code and a new bit-plane decomposition method using the truncated Fibonacci pcode. In addition, a new lossless image encryption algorithm is presented that can encrypt a selected object using this new decomposition method for privacy protection. The user has the flexibility (1) to define the object to be protected as an object in an image or in a specific part of the image, a selected region of an image, or an entire image, (2) to utilize any new or existing method for edge detection or segmentation to extract the selected object from an image or a specific part/region of the image, (3) to select any new or existing method for the shuffling process. The algorithm can be used in many different areas such as wireless networking, mobile phone services and applications in homeland security and medical imaging. Simulation results and analysis verify that the algorithm shows good performance in object/image encryption and can withstand plaintext attacks.

We present a statistical footprint-based method to characterize several symmetric cryptographic primitives as they are used in lightweight digital image encryption. In particular, using spatial-domain histogram and frequency-domain image analysis techniques, we identify a number of metrics from the encrypted images and use them to contrast the security performance of different cryptographic primitives. For each of the metrics, the best performing cryptographic primitive is identified. Complementary primitives are then combined to result in a product cipher with better cryptographic performance.

This paper presents a novel approach to compose discrete unitary transforms that are induced by input signals which are considered to be generators of the transforms. Properties and examples of such transforms, which we call the discrete heap transforms are given. The transforms are fast, because of a simple form of decomposition of their matrices, and they can be applied for signals of any length. Fast algorithms of calculation of the direct and inverse heap transforms do not depend on the length of the processed signals. In this paper, we demonstrate the applications of the heap transforms for transformation and reconstruction of one-dimensional signals and two-dimensional images. The heap transforms can be used in cryptography, since the generators can be selected in different ways to make the information invisible; these generators are keys for recovering information. Different examples of generating and applying heap transformations over signals and images are considered.

Fingerprint recognition applications as means of identity authentication that deals with accuracy and security are becoming more acceptable in areas such as financial transactions, access to secured buildings, commercial driver license and identity check at entry borders, to mention a few. This paper presented a new approach of using two patterns of the same person, intelligently fused, to form a new unique pattern of the same person. The Laplacian pyramid (LP) level 7 image fusion approach and the logical "OR" and logical "AND" operators for the decision fusion approach were tested with respect to their performance in accuracy, security and processing speed of the recognition system. The concept of receiver operator characteristic (ROC) curve to indicate any improvement in accuracy and security of the process was used. Finally, an overall comparison and analysis of performance between traditional systems that used a single pattern and our proposed system that used two fused fingerprint patterns in the biometric system was presented.

Edge detection is an important preprocessing task which has been used extensively in image processing. As many applications heavily rely on edge detection, effective and objective edge detection evaluation is crucial. Objective edge map evaluation measures are an important means of assessing the performance of edge detectors under various circumstances and in determining the most suitable edge detector or edge detector parameters. Quantifiable criteria for objective edge map evaluation are established relative to a ground truth, and the weaknesses and limitations in the Pratt's Figure of Merit (FOM), the objective reference-based edge map evaluation standard, are discussed. Based on the established criteria, a new reference-based measure for objective edge map evaluation is presented. Experimental results using synthetic images and their ground truths show that the new measure for objective edge map evaluation outperforms Pratt's FOM visually as it takes into account more features in its evaluation.

It is presented a robust three dimensional scheme using fuzzy and directional techniques in denoising video color images contaminated by impulsive random noise. This scheme estimates a noise and movement level in local area, detecting edges and fine details in an image video sequence. The proposed approach cares the chromaticity properties in multidimensional and multichannel images. The algorithm was specially designed to reduce computational charge, and its performance is quantified using objective criteria, such as Pick Signal Noise Relation, Mean Absolute Error and Normalized Color Difference, as well visual subjective views. Novel filter shows superiority rendering against other well known algorithms found in the literature. Real-time analysis is realized on Digital Signal Processor to outperform processing capability. The DSP was designed by Texas Instruments for multichannel processing in the multitask process, and permits to improve the performance of several tasks, and at the same time enhancing processing time and reducing computational charge in such a dedicated hardware.

Recent advances in biometric technology have pushed towards more robust and reliable systems. We aim to build systems that have low recognition errors and are less affected by variation in recording conditions. Recognition errors are often attributed to the usage of low quality biometric samples. Hence, there is a need to develop new intelligent techniques and strategies to automatically measure/quantify the quality of biometric image samples and if necessary restore image quality according to the need of the intended application. In this paper, we present no-reference image quality measures in the spatial domain that have impact on face recognition. The first is called symmetrical adaptive local quality index (SALQI) and the second is called middle halve (MH). Also, an adaptive strategy has been developed to select the best way to restore the image quality, called symmetrical adaptive histogram equalization (SAHE). The main benefits of using quality measures for adaptive strategy are: (1) avoidance of excessive unnecessary enhancement procedures that may cause undesired artifacts, and (2) reduced computational complexity which is essential for real time applications. We test the success of the proposed measures and adaptive approach for a wavelet-based face recognition system that uses the nearest neighborhood classifier. We shall demonstrate noticeable improvements in the performance of adaptive face recognition system over the corresponding non-adaptive scheme.

We present a novel algorithm for direct image registration, based on a generic description of the geometric transformation. The direct image registration algorithm consists of minimizing the intensity discrepancy between images. We propose the Gauss - Newton algorithm for the solution of this minimization problem. The method solves the optical flow equation iteratively to reduce the cost function. Registration was successfully performed on images taken from both aerial and ground platforms.

Privacy and security are vital concerns for practical biometric systems. The concept of cancelable or revocable biometrics has been proposed as a solution for biometric template security. Revocable biometric means that biometric templates are no longer fixed over time and could be revoked in the same way as lost or stolen credit cards are. In this paper, we describe a novel and an efficient approach to biometric template protection that meets the revocability property. This scheme can be incorporated into any biometric verification scheme while maintaining, if not improving, the accuracy of the original biometric system. However, we shall demonstrate the result of applying such transforms on face biometric templates and compare the efficiency of our approach with that of the well-known random projection techniques. We shall also present the results of experimental work on recognition accuracy before and after applying the proposed transform on feature vectors that are generated by wavelet transforms. These results are based on experiments conducted on a number of well-known face image databases, e.g. Yale and ORL databases.

One of the basic challenges to robust iris recognition is iris segmentation. This paper proposes the use of an artificial neural network and a feature saliency algorithm to better localize boundary pixels of the iris. No circular boundary assumption is made. A neural network is used to near-optimally combine current iris segmentation methods to more accurate localize the iris boundary. A feature saliency technique is performed to determine which features contain the greatest discriminatory information. Both visual inspection and automated testing showed greater than 98 percent accuracy in determining which pixels in an image of the eye were iris pixels when compared to human determined boundaries.

The H.264 video compression standard, aka MPEG 4 Part 10 aka Advanced Video Coding (AVC) allows new flexibility in the use of video in the battlefield. This standard necessitates encoder chips to effectively utilize the increased capabilities. Such chips are designed to cover the full range of the standard with designers of individual products given the capability of selecting the parameters that differentiate a broadcast system from a video conferencing system. The SmartCapture commercial product and the Universal Video Stick (UVS) military versions are about the size of a thumb drive with analog video input and USB (Universal Serial Bus) output and allow the user to select the parameters of imaging to the. Thereby, allowing the user to select video bandwidth (and video quality) using four dimensions of quality, on the fly, without stopping video transmission. The four dimensions are: 1) spatial, change from 720 pixel x 480 pixel to 320 pixel x 360 pixel to 160 pixel x 180 pixel, 2) temporal, change from 30 frames/ sec to 5 frames/sec, 3) transform quality with a 5 to 1 range, 4) and Group of Pictures (GOP) that affects noise immunity. The host processor simply wraps the H.264 network abstraction layer packets into the appropriate network packets. We also discuss the recently adopted scalable amendment to H.264 that will allow limit RAVC at any point in the communication chain by throwing away preselected packets.

Nowadays, Wireless and mobile communications technologies are the most important areas, which are rapidly expanding either in horizontal or vertical directions. WiMAX is trying to compete with WiFi in coverage and data rate, while the inexpensive WiFi still very popular in both personal and business use. Efficient bandwidth usage, Multi-Standard convergence and Wireless Mesh Networks (WMN) are the main vertical tends in the wireless world. WiMAX-WiFi convergence as an ideal technology that provides the best of both worlds: WiMAX new features and the low cost of the WiFi. In order to create a heterogeneous network environment, differences between the two technologies have been investigated and resolved. In the Multi-Carrier WiMAX-WiFi Convergence, the mismatch between the fixed WiMAXOFDM (N_fft=256) and the WiFi-OFDM (N_fft=64) has been confirmed as a physical layer issue that will never be solved as MAC layer problem; therefore the current proposal is how to build what we called the "Convergence-Bridge". This bridge is like an extra thin layer, which is responsible for harmonizing the mismatch. For the WiFi-OFDM physical layer, the paper has selected the IEEE 802.11n OFDM standard while it is being developed. The proposal does not suggest changing the standard itself but modifying some functions to be configurable. The IEEE 802.11 standard has fixed the configurations for WiFi mode only, while our proposal is to set up these functions for WiFi and WiMAX modes.

In this first part of the latest latency-information theory (LIT) and applications paper series powerful and fast 'knowledge-unaided' power-centroid (F-KUPC) radar is revealed. More specifically, it is found that for real-world airborne moving target indicator radar subjected to severely taxing environmental conditions F-KUPC radar approximates the signal to interference plus noise ratio (SINR) radar performance derived with more complex knowledge-aided power-centroid (KAPC) radar. KAPC radar was discovered earlier as part of DARPA's 2001-2005 knowledge-aided sensor signal processing expert reasoning (KASSPER) Program and outperforms standard priorknowledge radar schemes by several orders of magnitude in both the compression of sourced intelligence-space of priorknowledge, in the form of SAR imagery, and the compression of processing intelligence-time of the associated clutter covariance processor, while also yielding an average SINR radar performance that is approximately 1dB away from the optimum. In this paper, it is shown that the average SINR performance of significantly simpler F-KUPC radar emulates that of KAPC radar and, like KAPC radar, outperforms a conventional knowledge-unaided sample covariance matrix inverse radar algorithm by several dBs. The matlab simulation programs that were used to derive these results will become available in the author's Web site.

Since its introduction more than six decades ago by Claude E. Shannon information theory has guided with two performance bounds, namely source-entropy H and channel capacity C, the design of sourced intelligence-space compressors for communication systems, where the units of intelligence-space are 'mathematical' binary digit (bit) units of a passing of time uncertainty nature. Recently, motivated by both a real-world radar problem treated in the first part of the present paper series, and previous uncertainty/certainty duality studies of digital-communication and quantizedcontrol problems by the author, information theory was discovered to have a 'certainty' time-dual that was named latency theory. Latency theory guides with two performance bounds, i.e. processor-ectropy K and sensor consciousness F the design of processing intelligence-time compressors for recognition systems, where the units of intelligence-time are 'mathematical' binary operator (bor) units of a configuration of space certainty nature. Furthermore, these two theories have been unified to form a mathematical latency-information theory (M-LIT) for the guidance of intelligence system designs, which has been successfully applied to real-world radar. Also recently, M-LIT has been found to have a physical LIT (P-LIT) dual that guides life system designs. This novel physical theory addresses the design of motion life-time and retention life-space compressors for physical signals and also has four performance bounds. Two of these bounds are mover-ectropy A and channel-stay T for the design of motion life-time compressors for communication systems. An example of a motion life-time compressor is a laser system, inclusive of a network router for a certainty, or multi-path life-time channel. The other two bounds are retainer-entropy N and sensor scope I for the design of retention life-space compressors for recognition systems. An example of a retention life-space compressor is a silicon semiconductor crystal, inclusive of a leadless chip carrier for an uncertainty, or noisy life-space sensor. The eight performance bounds of our guidance theory for intelligence and life system designs will be illustrated with practical examples. Moreover, a four quadrants (quadrants I and III for the two physical theories and quadrants II and IV for the two mathematical ones) LIT revolution is advanced that highlights both the discovered dualities and the fundamental properties of signal compressors leading to a unifying communication embedded recognition (CER) system architecture.

Iris recognition algorithms depend on image processing techniques for proper segmentation of the iris. In the Ridge Energy Direction (RED) iris recognition algorithm, the initial step in the segmentation process searches for the pupil by thresholding and using binary morphology functions to rectify artifacts obfuscating the pupil. These functions take substantial processing time in software on the order of a few hundred million operations. Alternatively, a hardware version of the binary morphology functions is implemented to assist in the segmentation process. The hardware binary morphology functions have negligible hardware footprint and power consumption while achieving speed up of 200 times compared to the original software functions.

This paper proposes a steganographic scheme utilizing within and/or after fractal encoding procedures on images for data security. Fractal generation exploits the concepts of iterated function systems (IFS) consisting of a collection of contractive transformations. Fractal images make use of partitioned iterated function systems (PIFS) to determine self similarity within images along with approximating the original uncompressed image. The transformed coefficients are stored in a fractal code table in order to decode the image as an alternative to storing or transmitting image pixel values directly. The proposed steganographic algorithm conceals secret information in the contrast/scaling and brightness/shifting coefficients in the code table, resulting in a stego fractal code table. Using fractal transform as a means of steganography provides a new embedding domain other than the existing steganography tools. The advantages of using fractal compression for securing information within images are: no current fractal detection methods exist, the hidden information is disseminated throughout the image in the spatial domain, the capacity of the image can be increased, and the decoding of the stego table results in a visually undistorted image.