A New Efficient Context-Based Relative-Directional Chain Coding
- 1 August 2006
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 21627843,p. 787-790
Abstract
Chain coding is widely used in a variety of image processing applications. In this paper, we present a new chain coding scheme: context-based relative-directional chain coding (CRCC). It applies a novel context modeling with adaptive arithmetic coding to encode contour images. The proposed context modeling is based on the relative-directional chain representation. It provides a favorable conditional probability for the next encoding pixel so that the arithmetic coding is efficiently performed. The experimental results shows that the CRCC coding scheme overall outperforms the chain coding (CC), differential chain coding (DCC) and Differential chain coding mode-8 (DCC-8). More importantly, the proposed CRCC coding scheme completely avoids the cost in computation of eight-directional chain codes and its associated sequence of differencesKeywords
This publication has 5 references indexed in Scilit:
- Contour detection based on nonclassical receptive field inhibitionIEEE Transactions on Image Processing, 2003
- Highly efficient coding schemes for contour line drawings2002
- Coding efficiency of multi-ring and single-ring differential chain coding for telewriting applicationIEE Proceedings - Vision, Image, and Signal Processing, 2001
- Highly efficient coding schemes for contour lines based on chain code representationsIEEE Transactions on Communications, 1991
- On the Encoding of Arbitrary Geometric ConfigurationsIEEE Transactions on Electronic Computers, 1961