An Improvement to Image Segment Based on Human Visual System for Object-based Coding

Article type: Research Article

Authors: Tsai, Chwei-Shyong | Chang, Chin-Chen

Affiliations: Department of Information Management, National Taichung Department of Information Management, National Taichung Institute of Technology, Taichung, Taiwan 404, R.O.C. | Department of Computer Science and Information Engineering Department of Computer Science and Information Engineering National Chung Cheng University

Abstract: A new segmentation technique that is suitable for object-based coding schemes is proposed in this paper. Focusing on DCT frequency coefficients, the proposed technique assigns each frequency coefficient a corresponding rate value according to the importance of edge information. On the other hand, it adopts frequency sensitivity and edge contrast information as well as the properties of the human visual system to weight each frequency coefficient. As a result, the segmentation technique proposed in this paper can effectively reveal the necessary significant edges that make up the shape of objects so as to further facilitate obtaining meaningful objects.

Keywords: MPEG, object-based coding, human visual system, DCT

Journal: Fundamenta Informaticae, vol. 58, no. 2, pp. 167-178, 2003