Ashraf M. A. Ahmad

Robust object detection using cascade filter in MPEG videos

We propose a novel approach for motion vector (MV) based object detection in MPEG-2 video streams. Rather than processing the extracted MV fields that are directly extracted from MPEG-2 video streams in the compressed domain, we perform MV smoothing, perform MV noise reduction, obtain more robust object information, and refine this information through a cascaded filter composed of a Gaussian filter and a median filter. As a result, the object detection algorithm is more capable of accurately detecting objects. We compare the performance of our proposed system with the popular and commonly used spatial filter processing techniques: median filter, mean filter, Gaussian filter, and no filter. Based on experimental results performed over the MPEG7 testing dataset and measuring performance using the standard recall and precision metrics, object detection using the cascade filter is remarkably superior to the alternative filtering techniques. In addition to these results, we describe a user system interface that we developed, where users can maintain the filter parameters interactively.

Multimedia Transcoding in Mobile and Wireless Networks

In recent years, the field of multimedia communication in wireless and mobile video has grown in attraction. As the number of networks, types of devices, and content representation formats increase, the need for multimedia transcoding in wireless and mobile networks intensifies in order to provide a seamless interaction between content creation and usage. Multimedia Transcoding in Mobile and Wireless Networks addresses the transcoding of all multimedia elements including video, audio and text, while addressing both practical and theoretical problems in the field. A must have resource for all stakeholders in mobile and wireless technologies, this book embodies state-of-the-art knowledge of transcoding as it is embedded in today s pioneering technologies and leading research.

Fast and robust object detection framework in compressed domain

In this paper we propose a novel approach for fast and robust motion vector (MV) based object detection in MPEG-1 video streams. By processing the extracted MV fields that are directly extracted from MPEG-1/2 video streams in the compressed domain, through post processing operations, in order to reduce the noise within the MV content, obtain more robust object information, and refine this information through our proposed system which composed of a spatial filter component, a temporal filter component and a texture filter component. As a result, the object detection algorithm is more capable of accurately detecting objects with more efficient performance in terms of runtime. We compare the performance of our proposed system with other popular and commonly related work and techniques. Based on the experimental results performed over the MPEG7 testing dataset and measuring performance by using the standard recall and precision metrics, object detection using our proposed system is remarkably superior to the alternative techniques. In addition to these results, we describe a user system interface that we developed, where users can maintain the parameters interactively.

Fast and robust object-extraction framework for object-based streaming system

Video streaming poses significant technical challenges in the quality of service guarantee and efficient resource management. In this paper, we propose an efficient moving object-extraction algorithm suitable for real-time content-based multimedia streaming systems. A motion vector based object extraction is used to dynamically detect the objects. To utilise the bandwidth efficiently, the video objects can be detected in real time, encoded, and transmitted with higher quality and a higher frame rate than those in the background. To meet the real-time requirement, no computationally intensive operation is included in this framework. Moreover, to guarantee the highest speed, the entire implementation is operating in the compressed domain without any need for decompression.

Multimedia content and the semantic web: Methods, standards and tools.

Process-Aware Information Systems (PAISs) are among today’s hottest topics in the science and practice of information systems. Business processes, workflow management systems, groupware, project management products, enterprise application integration, and businessto-business integration attract a lot of attention from R&D professionals in information systems, management sciences, software engineering, and business-oriented computer science. The book provides an integrated introduction to these areas that have evolved somewhat independently of one another in an overarching framework. Unlike other titles in this area, the book focuses on technological aspects rather than managerial or strategic aspects, presenting a set of common underlying principles to effectively model, design, and implement process-aware information systems. As stated in the Preface, the purpose of the book is to “provide a unifying and comprehensive overview of the technological underpinnings of the emerging field of process-aware information systems engineering.” It is apparently a formidable task. Despite that the book is not exhaustive in terms of coverage of specific techniques, the editors are successful in establishing a broad framework and systematically organizing intelligent works that span the technological spectrum of PAISs, including major concepts, modeling languages, techniques, standards, and tools. Graduate and advanced undergraduate students, teachers, and researchers in computer science and information systems will find the book an excellent textbook for a topic-oriented course. There are 15 chapters in the book, which can be conveniently covered in a typical university semester. Each chapter begins with a short, general description of the problem domain and then progresses to relevant concepts and specific techniques. I especially like the numerous examples and illustrations that clarify and simplify complex topics. At the end of each chapter, there are thought-provoking exercises ranging from simple questions to projects and possible assignment subjects. This is very useful for classroom settings. Practitioners working on workflow and business process management, groupware and teamwork, enterprise application integration, and/or business-to-business integration projects will find the book a valuable reference book as well. Each chapter is self-contained and directly accessible, and the book, on a whole, represents different technical aspects of PAISs in a logically coherent way. Besides, the placement of references at the end of each chapter instead of in a long exhaustive list at the end of the book is extremely valuable to those who want to learn about a particular subtopic or get a quick reference. Also worth mentioning is an up-to-date reading and resource list in the Appendix. It consists of extra suggested readings (complementary to those at the end of each chapter) and URL pointers to relevant portals, standardization bodies, initiatives, and consortia. Brief descriptions and comments are presented for each resource, directing users to further information regarding various aspects of PAISs. The first four chapters form a conceptual basis for understanding of process-aware information systems and a general framework to organize the whole book. It not only discusses major concepts such as process, information systems, and PAIS, but also provides an overview of languages, techniques, and standards at a broad level. After illustrating some of the ongoing trends in information systems as a context of the emergence of an increasing number of PAISs, Chapter 1 then proceeds to an excellent discussion of the definition of a PAIS and different dimensions to classify PAISs. It is according to one of the classification schemes, “according to the nature of the participants” (i.e., P2A, P2P, A2A), that the following three chapters are organized. Chapter 2 gives a survey of P2A processes as embodied in WFMSs, in which workflow management is introduced as a generic concept. However, concrete commercial workflow management systems are not presented and hence, there are no proof-of-concept systems to exemplify the terminology, models, or functions of WFMSs. Chapter 3 explains and characterizes relatively more human-centric P2P processes and systems to support collective group interactions. Chapter 4, on the other hand, introduces more system-centric A2A processes with an innovative view that EAI and B2B integration are really two sides of the same coin, i.e., both are relevant and require each others’ cooperation to integrate application systems and businesses. Although only the basic concepts and approaches are introduced, references to comprehensive, technical, and historical discussions are provided for further study. Chapters 5 through 8 are dedicated to process modeling languages. Chapter 5 demonstrates clearly how UML, a visual and object-oriented modeling standard, though primarily designed for modeling software systems, can be applied to business process modeling. Activity, class, object, sequence, and structure diagrams in UML are examined to describe five major perspectives of process models. Another widely used standard, eEPC, together with its supportive platform, ARIS, are introduced in Chapter 6. A concise section, “How to Correctly Model EPCs” adds particular value to its usefulness in practice, which consists of straightforward rules to avoid common pitfalls and easy-to-follow guidelines to model EPCs. Chapter 7 delivers an effective description of Petri nets from multiperspectives: as a visual tool, as a formal language, and as mathematical structures. Both elementary and high-level Petri nets are analyzed with an eye on modeling single processes, multiple processes, and resources. Chapter 8 closes this part with a more abstract discussion of patterns that can be used to evaluate and adapt modeling languages. Twenty control-flow patterns ranging from very simple patterns such as sequential routing to complex patterns involving complex synchronizations such as discriminator patterns are presented and grouped into six categories. However, these categories are empirical and no explicit criteria to classify patterns are presented. Part III, “Techniques,” includes three chapters: process design and redesign (Chapter 9), process mining (Chapter 10), and transactional business processes (Chapter 11). One of the strengths of Chapter 9 is that it delves into two concrete process redesign methods that can be exploited according to different redesign goals. One is based on heuristic redesign best practices; the other exploits an information processing perspective on business processes. Chapter 10 presents an overview of process mining with event logs, followed by an in-depth discussion of a specific algorithm: the a-algorithm. Some possible solutions are proposed to solve the problems while using the a-algorithm. In Chapter 11, an old technique, transaction management, is investigated in the context of modern processing languages. Atomicity, from the most widely used ACID transaction model, is identified as the transactional property

FAST AND ROBUST OBJECT SEGMENTATION APPROACH FOR MPEG VIDEOS

In this paper we propose an efficient approach for Motion Vector (MV) based object detection in MPEG-1 video streams. Based on experimental results performed over the MPEG testing dataset and measuring performance using the standard recall and precision metrics for perceptual performance, and run time metric for efficiency performance. Our approach is remarkably superior to the alternative techniques. In addition to these results, we describe a user system interface that we developed, where users can maintain the configurations interactively.