Rate-distortion based image segmentation using recursive merging

• Published in: IEEE transactions on circuits and systems for video technology Vol. 10; no. 7; pp. 1121 - 1134
• Main Authors: Lim, Chae Whan, Kim, Nam Chul, Jun, Sung Chul, Jung, Choon Sik
• Format: Journal Article
• Language: English
• Published: 01.10.2000
• ISSN: 1051-8215

In this paper, a rate-distortion based image segmentation algorithm for segmentation-based coding is presented using a recursive merging with region adjacency graph (RAG). In the method, the dissimilarity between a pair of adjacent regions is represented as a Lagrangian cost function considered in a rate-distortion sense. Lagrange multiplier is estimated in each merging step, a pair of adjacent regions whose cost is minimal is searched, and then the pair of regions are merged into a new region. The merging step is recursively performed until some termination criterion is reached. The proposed method thus is suitable for region- or segmentation-based coding. Experimental results for 256x256 Lena show that segmentation-based coding using the proposed method yields PSNR improvement of about 3.5 to approximately 4.5 dB, 1.8 to approximately 2.0 dB, and 1.2 to approximately 1.5 dB over mean-difference based method, distortion-based method, and JPEG, respectively.