GENETIC ALGORITHM IN PATTERN MATCHING PROBLEMS OF VIDEO CODING

概要

論文の詳細を見る
Genetic algorithms have been adopted to solve many optimization and search problems and shown to be very efficient. It is expected beneficial to apply the genetic evolution in solving pattern matching problems of video coding because they can be treated as search problems. However, the tight time-constraint of pattern matching problems of video coding are much different from the characteristics of the other search problems. To cope with the real time constraint of the pattern matching problems of video coding, in this paper, different policies of genetic evolution are investigated. The computational complexity of the proposed genetic search algorithm is very low while comparing with the conventional ones. The proposed genetic algorithm has been applied to solve two kinds of pattern matching problems: the motion estimation and the automatic facial feature extraction. It is shown by simulations that a lot of computations can be saved by applying the proposed algorithm while the searched results are well approved.
社団法人電子情報通信学会の論文
1998-01-21