Face Detection Technology Based on Combining Skin Color Model with Improved Adaboost Algorithm
Zhizhen Li, Qinghua Zeng, Xiangdong Li, Yu Fen Yang · 2019
In order to improve the accuracy of face detection, this paper proposes a method based on combination of skin color model and improved Adaboost algorithm. Using skin color model with YCbCr color space to detect the skin color, obtain the area to be detected, and then we can locate accurately the face position with the Adaboost face detection algorithm. Because the traditional Adaboost face detection algorithm needs a long time training and can not effectively distinguish feature values which are aggregated distribution, we have improved Adaboost algorithm based on two-threshold and a new method about suppression of weights updating which which prevents excessive weight gain and avoids the phenomenon of degradation during training in Adaboost.