Realization Of Multimodal Face Recognition Technology Based On Gabor Feature

In this thesis, the use of face recognition technology, according to a company employees to work on the sign of requirements, design of a possible automatic face recognition attendance system. System to store employees face data, including facial image, name and some other key information. System memory can already identify the identity of employees and illegal landing personnel prompts. The purpose is to achieve recognition system correctly recognize human faces, generally divided into two phases, the first is for image acquisition and entry, thus set up a face image of basic data platform portion, the second stage is the training samples and identification part of the test sample is the main focus of research and testing part of the system.In this thesis, the research history and the basic method of face recognition are described, and the pretreatment process of the face image is introduced. Then the principle of Gabor filter and the process of how to select features in face recognition are introduced. Then, the paper introduces the main tools, Gabor features, multi mode partition and SVM., it has a strong operational structure features, can be a great improvement in the speed and accuracy of face recognition. In this paper, we use the Gabor feature and multi mode partition to identify the common role of face recognition.