人脸识别最难得在于特征量的提取，作者在老师的指导下通过对人脸图像打上网格，对区域块图像做二值分析，通过像素比例来做处理。进而得到人脸区域，识别准确性在一般的应用中已经可以满足要求。-Face recognition is the most difficult to extract the feature extraction, the author in the teacher s guidance through the face image on the grid, the regional block image do two value analysis, Through the pixel ratio to do processing. And then get the face region, the recognition accuracy in the general application has been able to meet the requirements.