Feature-Based Affine Motion Estimation for Superresolution of a Region of Interest

This chapter presents an interpolation method of low-computation for a Region Of Interest (ROI) using multiple low-resolution images of the same scene. Interpolation methods using multiple images require the accurate motion information between the reference image of interpolation and the other images. Sometimes complex local motions applied to the entire images are estimated incorrectly, yielding seriously degraded interpolation results. The authors apply the proposed Superresolution (SR) method, which employs a simple global motion model, only to the ROI that contains important information of the scene. The ROIs extracted from multiple images are assumed to have simple global motions. At first, using a mean absolute difference measure, they extract the regions from the multiple images that are similar to the selected ROI in the reference image of interpolation and use feature points to estimate the affine motion parameters. The authors apply the Projection Onto Convex Sets (POCS)-based method to the ROI using the estimated motion, simplify the iterative computation of the whole system, and use an edge-preserving smoothing filter to reduce the distortion caused by additive noise. In experiments, they acquire test image sets with a hand-held digital camera and use a Gaussian noise model. Experimental results show that the feature-based Motion Estimation (ME) is accurate and reducing the computational load of the ME step is efficient in terms of the computational complexity. It is also shown that the SR results using the proposed method are remarkable even when input images contain complex motions and a large amount of noise. The proposed POCS-based SR algorithm can be applied to digital cameras, portable camcorders, and so on.

Thermal Human Face Recognition for Biometric Security System

Automatic face recognition has been comprehensively studied for more than four decades, since face recognition of individuals has many applications, particularly in human-machine interaction and security. Although face recognition systems have achieved a significant level of maturity with some realistic achievement, face recognition still remains a challenging problem due to large variation in face images. Face recognition techniques can be generally divided into three categories based on the face image acquisition methodology: methods that work on intensity images, those that deal with video sequences, and those that require other sensory (like 3D sensory or infra-red imagery) data. Researchers are using thermal infrared images for face recognition. Since thermal infrared images have some advantages over 2D images. In this chapter, an overview of some of the well-known techniques of face recognition using thermal infrared faces are discussed, and some of the drawbacks and benefits of each of these methods mentioned therein are discussed. This chapter talks about some of the most recent algorithms developed for this purpose, and tries to give a brief idea of the state of the art of face recognition technology. The authors propose one approach for evaluating the performance of face recognition algorithms using thermal infrared images. They also note the results of several classifiers on a benchmark dataset (Terravic Facial Infrared Database).

Daubechies Complex Wavelet-Based Computer Vision Applications

Many methods for computer vision applications have been developed using wavelet theory. Almost all of them are based on real-valued discrete wavelet transform. This chapter introduces two computer vision applications, namely moving object segmentation and moving shadow detection and removal, using Daubechies complex wavelet transform. Daubechies complex wavelet transform has advantages over discrete wavelet transform as it is approximately shift-invariant, has a better edge detection, and provides true phase information. Results after applying Daubechies complex wavelet transform on these two applications demonstrate that Daubechies complex wavelet transform-based methods provide better results than other real-valued wavelet transform-based methods, and it also demonstrates that Daubechies complex wavelet transform has the potential to be applied to other computer vision applications.

Application of Computer Vision Techniques for Exploiting New Video Coding Mechanisms

One of the main concerns for current multimedia platforms is the provisioning of content that provides a good Quality of Experience to end-users. This can be achieved through new interactive, personalized content applications, as well by improving the image quality delivered to the end-user. This chapter addresses these issues by describing mechanisms for changing content consumption. The aim is to give Application Service Providers (ASPs) new ways to allow users to configure contents according to their personal tastes while also improving their Quality of Experience, and to possibly charge users for such functionalities. The authors propose to employ computer vision techniques to produce extra object information, which further expands the range of video personalization possibilities on the presence of new video coding mechanisms.

Stereo Vision-Based Object Matching, Detection, and Tracking

Computer vision has become very important in recent years. It is no longer restricted to a single camera that is only capable of capturing a single image at any given time. In its place, stereo vision systems have been introduced that not only make use of dual cameras to capture multiple images at once, but they also simulate the exact same nature of the human eye vision. Stereo vision has turned out to be an important research component in the subdivision of computer vision and image processing that deals with the extraction of information from images for the purpose of video surveillance systems, mimicking the human vision for the visually impaired, for robotics, to control unmanned vehicles, for security purposes, virtual reality and 3 Dimensional (3D) televisions, etc. In this chapter, a comprehensive review of all recent algorithms such as stereo matching, object detection, tracking techniques for stereo vision are presented.

Securing Biometrics Using Watermarking

This chapter focuses on the role of watermarking techniques in biometric systems. Biometric systems are automated systems of verifying or recognizing the identity of a living person based on a physiological or behavioral characteristic. While biometric-based techniques have inherent advantages over other authentication techniques, ensuring the security and integrity of data is a major concern. Data hiding techniques are thus used in biometric systems for securing biometric data itself. Amongst all the biometric techniques, fingerprint-based identification is the oldest and the most well established method used in numerous applications because fingerprints are unique and they remain unchanged during the human life span. However, fingerprint images should be watermarked without affecting their quality and their minutia matching ability. Moreover, if the watermark embedded in the fingerprint image is the face image of the same individual, the watermarking scheme will have two levels of security such that it will not only protect the cover fingerprint but also provides a more secure system of personal recognition and authentication at the receiver’s end. This work finds application in a number of security implementations based on multimodal biometric authentication. Computationally intelligent techniques can be employed to develop efficient watermarking algorithms in terms of watermarked image quality and distortion tolerance ability.

Video Authentication

With the growing innovations and emerging developments in sophisticated video editing technology, it is becoming highly desirable to assure the credibility and integrity of video information. Today digital videos are also increasingly transmitted over non-secure channels such as the Internet. Therefore, in surveillance, medical, and various other fields, video contents must be protected against attempts to manipulate them. Video authentication has gained much attention in recent years. However, many existing authentication techniques have their own advantages and obvious drawbacks. The authors propose a novel authentication technique that uses an intelligent approach for video authentication. This chapter presents an intelligent video authentication algorithm for raw videos using a support vector machine, which is a non-linear classifier, and its applications. It covers both kinds of tampering attacks, spatial and temporal. It uses a database of more than 2000 tampered and non-tampered videos and gives excellent results with 98.38% classification accuracy. The authors also discuss a vast diversity of tampering attacks, which can be possible for video sequences. Their algorithm gives good results for almost all kinds of tampering attacks.

Biometric Authentication Based on Hand Vein Pattern

Recently, a new biometric technology based on human hand vein patterns has attracted the attention of many researchers. This chapter discusses vein pattern authentication, which uses the vascular patterns of the back of the hand as personal authentication data. Vein information is hard to duplicate because veins are internal to the human body. Vein authentication is one of the most accurate and reliable biometric technologies, which is widely employed in mission-critical applications such as banking, etc. A dynamic ROI extraction algorithm was presented through which more features can be extracted when compared to the fixed ROI. The extracted ROI was enhanced, and then the noise content was removed. The key features that represent the geometric information of the vein pattern were extracted; they are the bifurcation and ending points. This chapter presents a new vein pattern recognition system by assigning different weights to bifurcation and ending points. The approach is tested on a vein pattern database of 60 different hands. Experimental results show the approach achieves 2.5% of Equal Error Rate (EER) and recognition accuracy of 98.24%.

Recognition of Humans and Their Activities for Video Surveillance

Human Activity Recognition is an active area of research in computer vision with wide-scale applications in video surveillance, motion analysis, virtual reality interfaces, robot navigation and recognition, video indexing, browsing, HCI, choreography, sports video analysis, etc. The analysis of vision-based human activities in videos is an area with increasingly important consequences from security and surveillance to public place and personal archiving. Several challenges at various levels of processing-robustness against errors in low-level processing, view and rate-invariant representations at mid-level processing, and semantic representation of human activities at higher-level processing make this problem hard to solve. The task is challenging due to variations in motion performance, recording settings, and inter-personal differences. In this chapter, the authors explicitly address these challenges. They present a survey of existing work and describe some of the more well-known methods in these areas. They also describe their own research and outline future possibilities. Detailed overviews of current advances in the field are provided. Image representations and the subsequent classification processes are discussed separately to focus on the novelties of recent research. Moreover, the authors discuss the limitations of the state of the art and outline promising directions of research.

Multimodal Biometrics Recognition for Newborns

Identification of newborns at birth is a critical issue for hospitals, birthing centers, and other institutions where multiple births occur. With approximately 300,000 infants born worldwide each day, a large hospital may experience over one hundred new births every day. Correct identification of infants is essential to ensure that each mother travels home with her own child. Mixing, abduction, and illegal adoption of newborns is a global problem, and the research done to solve this problem is minimal. In this chapter, the authors present a multimodal biometric framework for the recognition of newborns.