Title: Fast block prediction-based coding of compound images by exploiting edge orientation

Authors: S. Ebenezer Juliet; V. Sadasivam; D. Jemi Florinabel

Addresses: Department of Computer Science and Engineering, Manonmanium Sundaranar University, Tirunelveli, India ' Department of Computer Science and Engineering, Manonmanium Sundaranar University, Tirunelveli, India ' Department of Computer Science and Engineering, Manonmanium Sundaranar University, Tirunelveli, India

Abstract: Compound images are a combination of text, graphics and natural images. This paper presents a new fast block prediction algorithm for coding of text/graphics parts which exploits the direction of edges in the text/graphics blocks. The proposed scheme classifies the text/graphics blocks into four kinds of blocks using edge characteristics of transformed coefficients of blocks. This edge-based block classification improves the speed of intra-block prediction mode decisions significantly. Experimental results show that the proposed method improves the value of PSNR more significantly than standard JPEG, JPEG-2000 and H.264/AVC-I, while keeping competitive compression ratio and visually lossless quality of text information.

Keywords: compound images; DWT; discrete wavelet transform; block edge classification; block prediction; edge orientation; compression ratio; visually lossless quality; text information; image compression.

DOI: 10.1504/IJSISE.2013.053417

International Journal of Signal and Imaging Systems Engineering, 2013 Vol.6 No.2, pp.100 - 110

Received: 18 Feb 2011
Accepted: 15 Oct 2011

Published online: 21 Apr 2013 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article