Font Size: a A A

Research Of The Standard Front Face Recognition System

Posted on:2005-09-18Degree:MasterType:Thesis
Country:ChinaCandidate:H B YanFull Text:PDF
GTID:2168360125470697Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
Human face recognition is a very difficult and significant research task in the field of pattern recognition. It can be applied to many different fields such as security departments, I. D. Card management, teleconference, and so on, and the technique of human face recognition has wide application prospect.In this paper, we study and implement a standard frontal face image recognition system. The process of recognition can be divided into three stages: preprocessing of face image, feature extraction and recognition.In the process of preprocessing of face image, first, the system accomplishes the detection and location of face region and the organs, and then, according to the relation between the assigned distance of two eyes and the distance of two eyes in the real image, the geometrical dimension of face image is normalized. In order to reduce the illumination effect on the performance of recognition, the gray normalization is carried out to the image .At last, the standard face is outputted.In the face feature extraction step, the method of extracting geometrical features and statistic features is elaborated in this paper. Geometrical features extraction is done during the normalization of the image. The purpose of standardization is to get the statistic features of the image. Statistic features extraction uses the method of eigenface recognition based on K-L transformation.Finally, in the step of image recognition, the classic methods (special distance) are combined with artificial neural algorithm to form a hybrid classifier; as a result, the recognition rate is increased evidently.
Keywords/Search Tags:Pattern recognition, Face recognition, Face detection, Features extraction, Neural network
PDF Full Text Request
Related items