Research On Entrance Guard Management System Based On Face Recognition In Dangerous Location

The rapid development of computing technology and the advancement in imageprocessing technology as well as in biometrics feature identification technology in recentlyyears has truly accelerated the development of the human face recognition technology. Theface information cannot be lost; it is not in the form of a password which is not easy to beborrowed, stolen or faked and is harmless to the user. These features are well received by theusers and thus become the ideal way of identity verification. This thesis considers theregional security of flammable and combustible places in petrochemical productions. Wedesigned a face recognition entrance guard management system using face recognitiontechnology and solved problems encountered in traditional technologies such as IC cards,RFID cards, passwords and fingerprint locks like fragile chip, forgotten or stolen passwords,faked fingerprints, etc. This provides a strong support to achieve a more secure, convenient,intelligent and autonomous entrance guard management system.We described the backgrounds of human face detection algorithm, human facerecognition algorithm and image pre-processing technique and the suitable occasions forusing them. Then we chose PCA classical human face recognition algorithm as the theoreticalfoundation of the thesis. The2DPCA algorithm was introduced to address the problem oflarge computation load associated with the PCA algorithm. And the2DPCA algorithm canextract image characteristics more easily, directly and accurately, thus improved the overallresponsive speed of the system. We integrated into the traditional PCA classical algorithmpre-processing image enhancement technology based on local standard deviation and average,effectively reducing the effect of uneven illumination during the process of face recognitionbefore the improved algorithm performing the characteristics extraction. This techniqueeffectively enhanced the robustness for illumination of the traditional PCA algorithm. Toaddress the sensitivity issue for changing in the face expressions of the traditional PCAalgorithm, we proposed the MPCA algorithm. This improvement effectively reduces poorperformance due to the changing in the face expressions. We combined the above threealgorithms and designed a face recognition algorithm that is robust to illumination variation,insensitive to face expressions and has high speed.Simulations were performed on the improved algorithms in the MATLAB environment.And the results show the feasibility of the proposed algorithm. We also combined theimproved algorithm and the entrance guard management system, successfully building asystem based on face recognition, achieving effective management of dangerous locations. Ithas a proper degree of practical use and value for extensions.