Motion picture coding based on region segmentation using genetic algorithm
✍ Scribed by Koh'ichi Takagi; Atsushi Koike; Shuichi Matsumoto; Hideo Yamamoto
- Publisher
- John Wiley and Sons
- Year
- 2002
- Tongue
- English
- Weight
- 464 KB
- Volume
- 33
- Category
- Article
- ISSN
- 0882-1666
- DOI
- 10.1002/scj.1126
No coin nor oath required. For personal study only.
✦ Synopsis
Abstract
The purpose of this paper is to improve the coding efficiency of motion pictures by proposing a region segmentation method using the genetic algorithm (GA). In motion‐compensated predictive coding, region segmentation and motion estimation are closely related to each other, and they should be performed so that the coding complexity and S/N are optimized. Their determination can be reformulated as the optimization problem of assigning the motion vectors to each block of the image. Consequently, this paper examines the data type and the processing method of GA so that the optimal region segmentation is obtained. A computer simulation using a test image was performed as an experiment, and the effectiveness of the proposed method was demonstrated. © 2002 Wiley Periodicals, Inc. Syst Comp Jpn, 33(5): 41–50, 2002; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/scj.1126