FPGA and Optimization Technique Based Human Gender Classification

Gender identification from face features plays an important role in security, surveillance systems. The
research is carried out with different methods in feature detection , feature extraction, and different
classification algorithms. And also different processing platforms are used. In this paper the identification
of gender from face has been proposed using MATLAB simulation platform and FPGA as processing
platform. Firstly, the image to be examined is taken from pre-recorded video. Then the control points are
detected in MATLAB and the data is transferred to FPGA platform for feature extraction. To make the
system optimized, here optimization algorithm that is Ant Colony Optimization (ACO) is used. With the
help of ACO, optimized features are extracted from the detected control points. Again the extracted data
is send to MATLAB. Finally, with the help of Artificial Neural Network, the extracted data is tested and
gender is identified