Face Detection Under Complicated Background Research And Application Base On Adaboost Algorithm

To be able to in a complex background of human face detection rate reached a higher level, this article in order to make use of the pre-test the premise that the detection of pre-processing work. That it can better detect the face area, while removing the face with people similar to part of the test results more accurate, more efficient, the use of time is even shorter.For complex pattern detection problem, based on dominant characteristics of the method to use if only a small number of features is very reliable, in turn related to the use of more features of the image to understand this very difficult problem, and based on the implicit characteristic method is more Lu Rod and ease of scalability, etc., Therefore, this article uses a cascade strong classifier adaboost integral image method to achieve a complex background face detection. Finally, the skin color model, pretreatment technology and adaboost integral Mapping effective combination to achieve complementarity.