A face detection algorithm based on Adaboost and new Haar-Like feature

摘要

This paper uses a new face detection method based on Haar-Like feature. New Haar-Like feature is an extension of the Haar-Like feature basis. This article use four new Haar-Like feature, and these features with existing Haar-Like feature are input Adaboost classifier together to select feature, finally constructed classification performance and powerful cascade classifier for face detection. After detection experiments we can see, the algorithm can get better results compared with other traditional face detection classifiers like Haar-Like.