The Feasibility of Implementing a Face Recognition System Based on a Gabor Filter and Nearest Neighbor Techniques in an FPGA Device for Door Control Systems

Door access control systems based on face recognition are geared towards simplifying difficult face recognition problems in uncontrolled environments. Such systems are able to control illumination, offer neutral pose and improve the poor performance of many face recognition algorithms. Door access control systems control illumination and pose in order to overcome face recognition problems. While there have been significant improvements in the algorithms with increasing recognition accuracy, very little research has been conducted on implementing these in hardware devices. Most of the previous studies focused on implementing a simple principal component analysis in hardware with low recognition accuracy. In contrast, the use of a Gabor filter for feature extraction and the nearest neighbour method for classification were found to be better alternatives. Dramatic developments in field programmable gate arrays (FPGAs) have allowed designers to select various resources and functions to implement many complex designs. The aim of this paper is to present the feasibility of implementing Gabor filter and nearest neighbour face recognition algorithms in an FPGA device for face recognition. Our simulation using Xilinx FPGA platforms verified the feasibility of such a system with minimum hardware requirements.