Design Of An Image Sensor Based On Conformal Mapping For Face Recognition System

Norazlina, Abd Razak and Shahrudin , Zakaria and Farah Shahnaz, Feroz and Hanissah, Mohamad and Yosza, Dasril (2009) Design Of An Image Sensor Based On Conformal Mapping For Face Recognition System. Project Report. UTeM, Melaka, Malaysia. (Submitted)

PDF (Full Text) Design_Of_An_Image_Sensor_Based_On_Conformal_Mapping_For_Face_Recognition_System_Norazlina_bt_Abd_Razak.pdf - Submitted Version Restricted to Registered users only Download (4MB)

Abstract

The rate of recognition process is a fundamental issue in computer vision. In simple way to say that it is quite useless if it takes few days to recognize a single thing. So researchers try to work out this vision problem. The eye vision system has become an immediate motivation towards improving the machine vision system, focuses more on how the images have been mapped on the retina. It is believed that the mapping has contributed towards less memory used and processing time. This work deals with a face recognition technique based on conformal mapping. For face recognition purpose, neural network has been used and it is found that by applying conformal mapping the recognition process would be much faster with considerable low error. In this research, we have also obtained the result that show the recognition performance is 100% for all images which undergo training phase, while the images which are not in training phase has achieved recognition performance above 80%.

Actions (login required)

View Item