Research On Mobile Phone Face Recognition System Based On Wavelet Transform

With the improvement of mobile phone hardware technology and popularization of network, mobile phone is becoming more and more intelligent. Because of the large amount of personal information and important data in mobile phone, many people set a password lock to prevent their information from being stolen by others. Face recognition is a new way to protect mobile phone password which is based on biological characteristic and more fast and stable. Combing face recognition technology with smart phone for authentication is gradually becoming a hot research topic.The methods of face detection, face feature extraction and feature matching are analyzed and researched in this thesis. A face recognition system based on i OS platform is designed and achieved. This thesis mainly includes the following contents:(1) Considering the complex backgrounds and changeable light conditions, the ways to enhance images, such as gray-scale transformation, geometric normalization and histogram equalization are applied in original image. The simulation results show that the enhanced image is better than the original image. Face detection method based on skin color is researched in this thesis. Face region is segmented in YCb Cr color space based on skin color information. The correct detection rate of 87.5% is obtained through experiments.(2) The classical feature extraction method based on Gabor wavelet transform is researched. Considering the limited memory space and computing performance of mobile devices, sophisticated recognition algorithms will affect recognition efficiency, therefore a improved method is adopted based on Gabor wavelet. In this thesis, a feature extraction method combining Gabor wavelet transform and principal component analysis is applied and this method effectively reduces matrix operations in the process of facial feature extraction. This algorithm is applied to the face recognition system and improves operational performance.(3) The i OS system architecture, development language, application components and application life cycle are introduced and analyzed. Using the above techniques and algorithms, a mobile phone face recognition system based on i OS platform is designed and tested on i Phone simulation. Test results show that this system can be applied in mobile devices for personal authentication.