M. Hassaballah

A Review of SIMD Multimedia Extensions and their Usage in Scientific and Engineering Applications

The volume and complexity of data processed by todays personal computers are increasing exponentially, placing incredible demands on the microprocessors. In the meantime, computing performance that can be achieved by increasing the clock speed of a microprocessor is reaching to physical limits thus making the architectural solutions more prominent. Due to this an important architectural feature is added to recent microprocessors, single instruction multiple data (SIMD), which is a set of instructions that can speed up an application performance by allowing basic operation to be performed on multiple data elements in parallel with fewer instructions. The SIMD computational technique was introduced in the IA-32 Intel® architecture with MMX technology and then further enhanced with Intels introduction of streaming SIMD extensions (SSE), SSE 2 (SSE2) and SSE 3 (SSE3). Although programming using these SIMD extensions enables software to achieve higher performance, several exiting scientific applications are not affected. This paper gives an overview of SIMD multimedia extensions. The features of these extensions are introduced. Available methods for programming with multimedia instruction sets are discussed. It also reviews recent trends to use multimedia extensions to accelerate many applications such as multimedia, scientific and engineering applications, and argues for further use in other significant computationally intensive applications.

Image Features Detection, Description and Matching

Feature detection, description and matching are essential components of various computer vision applications, thus they have received a considerable attention in the last decades. Several feature detectors and descriptors have been proposed in the literature with a variety of definitions for what kind of points in an image is potentially interesting (i.e., a distinctive attribute). This chapter introduces basic notation and mathematical concepts for detecting and describing image features. Then, it discusses properties of perfect features and gives an overview of various existing detection and description methods. Furthermore, it explains some approaches to feature matching. Finally, the chapter discusses the most used techniques for performance evaluation of detection and description algorithms.

Face recognition: challenges, achievements and future directions

Face recognition has received significant attention because of its numerous applications in access control, law enforcement, security, surveillance, Internet communication and computer entertainment. Although significant progress has been made, the state-of-the-art face recognition systems yield satisfactory performance only under controlled scenarios and they degrade significantly when confronted with real-world scenarios. The real-world scenarios have unconstrained conditions such as illumination and pose variations, occlusion and expressions. Thus, there remain plenty of challenges and opportunities ahead. Latterly, some researchers have begun to examine face recognition under unconstrained conditions. Instead of providing a detailed experimental evaluation, which has been already presented in the referenced works, this study serves more as a guide for readers. Thus, the goal of this study is to discuss the significant challenges involved in the adaptation of existing face recognition algorithms to build successful systems that can be employed in the real world. Then, it discusses what has been achieved so far, focusing specifically on the most successful algorithms, and overviews the successes and failures of these algorithms to the subject. It also proposes several possible future directions for face recognition. Thus, it will be a good starting point for research projects on face recognition as useful techniques can be isolated and past errors can be avoided.

Face detection evaluation: a new approach based on the golden ratio {\Phi}

Face detection is a fundamental research area in computer vision field. Most of the face-related applications such as face recognition and face tracking assume that the face region is perfectly detected. To adopt a certain face detection algorithm in these applications, evaluation of its performance is needed. Unfortunately, it is difficult to evaluate the performance of face detection algorithms due to the lack of universal criteria in the literature. In this paper, we propose a new evaluation measure for face detection algorithms by exploiting a biological property called Golden Ratio of the perfect human face. The new evaluation measure is more realistic and accurate compared to the existing one. Using the proposed measure, five haar-cascade classifiers provided by Intel©OpenCV have been quantitatively evaluated on three common databases to show their robustness and weakness as these classifiers have never been compared among each other on same databases under a specific evaluation measure. A thoughtful comparison between the best haar-classifier and two other face detection algorithms is presented. Moreover, we introduce a new challenging dataset, where the subjects wear the headscarf. The new dataset is used as a testbed for evaluating the current state of face detection algorithms under the headscarf occlusion.

Operations on the Similarity Measures of Fuzzy Sets

Measuring the similarity between fuzzy sets plays a vital role in several fields. However, none of all well-known similarity measure methods is all-powerful, and all have the localization of its usage. This paper defines some operations on the similarity measures of fuzzy sets such as summation and multiplication of two similarity measures. Also, these operations will be generalized to any number of similarity measures. These operations will be very useful especially in the field of computer vision, and data retrieval because these fields need to combine and find some relations between similarity measures.

A framework for objective image quality measures based on intuitionistic fuzzy sets

Abstract Measuring the distance or similarity objectively between images is an essential and a challenging problem in various image processing and pattern recognition applications. As it is very difficult to find a certain measure that can be successfully applied to all kinds of images comparisons-related problems in the same time, it is appropriate to look for new approaches for measuring the similarity. Several similarity measures tested on numerical cases are developed in the literature based on intuitionistic fuzzy sets (IFSs) without evaluation on real data. This paper introduces a framework for using the similarity measures on IFSs in image processing field, specifically for image comparison. First, some existing similarity measures are discussed and highlighted their properties. Then, modeling digital images using IFSs is explained. Moreover, the paper introduces an intuitionistic fuzzy based image quality index measure. Second, for improving the perceived visual quality of these IFS-based similarity measures, construction of neighborhood-based similarity is proposed, which takes into consideration homogeneity of images. Finally, the proposed framework is verified on real world natural images under various types of image distortions. Experimental results confirm the effectiveness of the proposed framework in measuring the similarity between images.

Detection and Description of Image Features: An Introduction

Detection and description of image features play a vital role in various application domains such as image processing, computer vision, pattern recognition, and machine learning. There are two type of features that can be extracted from an image content; namely global and local features. Global features describe the image as a whole and can be interpreted as a particular property of the image involving all pixels; while, the local features aim to detect keypoints within the image and describe regions around these keypoints. After extracting the features and their descriptors from images, matching of common structures between images (i.e., features matching) is the next step for these applications. This chapter presents a general and brief introduction to topics of feature extraction for a variety of application domains. Its main aim is to provide short descriptions of the chapters included in this book volume.

A New Class of Similarity Measures for Fuzzy Sets

Fuzzy techniques can be applied in many domains of computer vision community. The definition of an adequate similarity measure for measuring the similarity between fuzzy sets is of great importance in the field of image processing, image retrieval and pattern recognition. This paper proposes a new class of the similarity measures. The properties, sensitivity and effectiveness of the proposed measures are investigated and tested on real data. Experimental results show that these similarity measures can provide a useful way for measuring the similarity between fuzzy sets.

An efficient data hiding method based on adaptive directional pixel value differencing (ADPVD)

Steganography is an important secret information communication technology in which one may send messages without others having knowledge of their existence. This paper proposes a new adaptive steganography method for color images using adaptive directional pixel-value differencing (ADPVD). The proposed method increases the capacity of the hidden secret data and improves the security of the stego-color image as well. The hiding capacity of the original PVD method is investigated by considering three directional edges: horizontal, vertical and diagonal directions using color cover image. The cover image is partitioned into 2-pixel blocks in a non-overlapping fashion and scanned in raster-scan order in all three directions. The proposed method adaptively selects the appropriate embedding directions for each color channel according to the largest embedding capacity. The security is improved since different pixel directions are employed adaptively to embed different number of message bits in each color channel. The experimental results show that the proposed method provides both larger embedding capacity and better visual quality of the stego color image compared with other PVD-based algorithms.

A Random Decision Forests Approach to Face Detection

Face detection has been considered one of the most important areas of research in computer vision due to its wide range of use in human face-related applications. This paper addresses the problem of face detection using Hough transform employed within the random forests framework. The proposed Hough forests-based method is a task-adapted codebooks of local facial appearance with a randomized selection of features at each split that allow fast supervised training and fast matching at test time, where the codebooks are built upon a pool of heterogeneous local appearance features and the codebook is learned for the face appearance features that models the spatial distribution and appearance of facial parts of the human face. Experimental results are included to verify the effectiveness and feasibility of the proposed method.