Font Size: a A A

The Design And Development Of Face Attendance System

Posted on:2017-04-21Degree:MasterType:Thesis
Country:ChinaCandidate:X L MengFull Text:PDF
GTID:2308330482997506Subject:Computer technology
Abstract/Summary:PDF Full Text Request
Because of the stability and particularity of face image, it is widely used in many systems. Complete face work attendance system can make full use of the image information in the database to carry out the management of personnel attendance effectively. The face attendance is related to a large number of face recognition technology, traditional face recognition methods used in the real scene have poor robustness and some other defects. In this paper, according to the actual needs of human face register, the attendance of the light problem, occlusion and other technical difficulties are studied, the face attendance system is designed in several aspects.According to the face detection problem, a face detection method based on Adaboost cascade classifier is designed. Firstly, the input image is finish the denoising process with Gaussian filter, to reduce the influence of noise on the face detection, and the use of Laplacian algorithm for image further sharpen, to enhance the edge information of the image. In order to solve the displacement changes of the image in the real scene, the geometric normalization of the image is carried out. After the pretreatment of the image, and the influence of the illumination, the background and the pixels of the image is reduced. Finally, by using the Haar feature and the Adaboost based cascade classifier, a high detection efficiency of the cascade classifier will be obtained through the combination of iterative training. Experiments show whether the accuracy or speed of the algorithm is improved.In the face recognition stage, a local face recognition algorithm based on feature set matching is proposed. The algorithm has strong robustness to the face information loss and illumination changes. Firstly, SIFT feature is used to detect and extract the key points, SURF feature is as an augmented feature. After the combination of SIFT features and SURF features, the sample images are automatically aligned with the gallery image using extended point matching method. Finally, the face recognition problem is transformed into the distance between the sample image and the image database. Experiments show that the algorithm has good robustness to illumination and occlusion, and can solve the problem of face recognition in real scene.Finally, according to the application requirements, the design and implementation of the human face attendance system based on MVC model. The system can meet the actual needs of human face attendance through the test.
Keywords/Search Tags:Haar features, SIFT features, SURF features, local feature matching, attendance system
PDF Full Text Request
Related items