Research And Implementation Of Face Recognition Algorithm Based On OpenCV

As a high security biometric technology, face recognition has developed quite fast and has gradually become the main method of biometrics in recent years.OpenCV is an open source cross-platform computer vision library, which contains a lot of common image processing algorithms, and open source OpenCV has good portability, whose code has been optimized. Therefore it can also be used for both research and commerce.Based on OpenCV, the thesis conducts tests of face detection and face recognition respectively. For the face detection, the thesis verifies the effectiveness of image preprocessing methods through several tests. Face detection algorithm is based on the method of AdaBoost that introduces the concept of "integration map", which can effectively improve the efficiency of the algorithm. For the face recognition, the thesis introduces several traditional face recognition methods, and discusses face recognition algorithm which is based on Eigenface. In this way, human face is described in a two-dimensional matrix,and is projected into a feature space which has been trained. At this time we only need to compare the weights of known human face. This algorithm runs fast and is very easy to develop for commercial technologies. Finally, the thesis implements a face recognition system based on the method of Eigenface and the experimental result is fast and accurate, thus verifying the effectiveness of the algorithm.