The Method Of Automatic Sign Based On Face Detection And Recognition

At present there are three main methods of Attendance System, Paper-mediated punch card machines, fingerprint machines and IC card photo machines. These three methods have disadvantages: Paper punch card machine can refer to cosmetologists by others, easy to cheat; Fingerprint punch card machines were hunters of the fingerprint, due to manpower in the course of friction would be affected by the clarity of fingerprints, so recognition rate is low; IC card photo punch is needed for everyone to use their own IC card activation punch card machines, and then photographed IC card into the photo folder that corresponds to the end of the month when the statistical picture by hand, the number of sheets, so wastage of manpower and when IC card is lost or damaged, then brought to the attendance of trouble.In this thesis, the face recognition algorithm has been presented to face recognition technology; in attendance systems we use Gabor filters and singular value decomposition, neural networks, BP neural network into face recognition technology. With the establishment of the training face image database to replace the former paper presents the IC card, IC card solved the case of loss or damage to the problems and achieve a dynamic face recognition performance appraisal system. The whole process of implementation steps of sign method is based on the development of ASP and SQL. In this thesis, Facial feature extraction and Face Classification which is based on Neural network also use matlab to make code. At the end when the System is implemented, the matlab code is called by JAVA, and then the component is called by ASP. Finally the automatic sign based on Face detection and recognition is implemented, the interface is friendly and the operation is easy.