Design Of Real- Time Human Face Recognition System Based On OpenCV

In recent years, with the progress of technology and the development of economy, identity security is increasingly valued, the demand for automated identification is increasingly strong in the whole society. As a kind of high security identification technology, face recognition technology has been widely used in many fields and a research hotspot in today’s multiple interdisciplinary.This article implemented a real-time face recognition system based on Open CV, which is an open visual library of cross-platform computer software. The face recognition system consists of three parts: face image acquisition, face image preprocessing and face recognition. The video acquisition device provides real-time videos to a computer, which can capture face image and then deposite to the face sample library. After training the sample library, face recognizer is called to identify the face image from real-time video, then display the recognition result.This system has joined the eye location on the basis of traditional face detection algorithm based on Ada Boost. Firstly, the Ada Boost algorithm was adopted to the real-time face detection, and join the eye localization to screen secondly in order to get better detection results. In section of face recognition, this system adopts PCA face recognition algorithm combining gamma transform and wavelet transform. The algorithm with high recognition rate and recognition speed and robustness to illumination has made full use of the advantages of wavelet transform in image compression and the inhibition of gamma transform in illumination change on the basis of traditional face recognition algorithm based on PCA.This paper has completed the research on face recognition and system design. After testing, the system runs stably with high accuracy, good real-time performance.