Research And Implementation Of Face Detection Technology In The Access Control System

Face detection is a central part to the face identification system. An efficient algorithm of face detection is particularly important to the access control system in high level security or in special occasion. The face has a natural structure of the complex details of the changes.The challenge lies in the fact that Face due to the different appearance, expression, color, etc., have a mode of variability; The person’s face may exist such as glasses, beard appendages; As a three-dimensional objects, the face image is inevitably affected by shadow. Existing face detection methods made a lot of work to address the problem and provided an important inspiration to solve such a complex pattern. But they still inevitably exists about the accuracy, timeliness, efficiency can not meet the practical applications (especially the problem of complex harsh environment).To solve the above problems, this paper designed and implemented a set of system for the special environment of the access control.In order to improve the security of the access control system, this paper used a multi-feature face detection process and combined the background separation, complexion detection as well as profile analysis and people tracking together and resolved some problems caused by single-feature.The results showed that the algorithm in complex background and illumination conditions as well as shelter can quickly and accurately detect the human face. Its false detection rate is low. Compared with other algorithms, the results of the proposed algorithm has a higher robustness.