Multi-pose Face Detection Research Based on Adaboost
Wei Liuliu, Liu Mingyang · 2016
Adaboost is a kind of traversal image searching method, which wastes testing time seriously and should be improved. Skin color detection is used to optimize front-end detection that is relatively stable and has fast speed. In skin color detection, UCS space is used to segment image quickly. Then Adaboost algorithm is used to detect human face based on segmented skin color area. The experiment results show that proposed face detection algorithm has faster speed. Besides, it has lower false detection rate and missing detection rate than Adaboost algorithm. The proposed face detection algorithm can provide reference for design of actual face detection system to some extent.