A Face Recognition Approach Based on Computer Vision

Most face recognition algorithms are easily affected by external factors, or require the recognizer to stand in a fixed position, so the recognition is slow and the accuracy is not very high. This paper implements an application platform of face recognition based on Open CV. Firstly, the face image is collected and preprocessed by optimizing the algorithms of Eigenfaces, Fisherfaces and LBP platform, which are used for face recognition by cooperation with each other. In addition to the restrictions on the recognition of the target to be detected by the face recognition system, eye location is added to the traditional AdaBoost based face detection algorithm to realize the function of face module training and face recognition. After testing, the system is proved to have a friendly interface, stable operation, good robustness to face location and environmental light changes, and it can detect and recognize face in real time quickly and accurately.