基于FPGA的人脸检测系统设计

A face detection system design based on FPGA

选取一种由粗及精的方法,综合采用基于彩色图像的肤色检测方法和基于几何特征的灰度检测方法,在传统肤色检测算法的基础上,加入图像预处理、色彩平衡、二值化、形态学滤波、中值滤波、边缘检测等技术进行人脸检测。搭建了由CMOS摄像头到FPGA核心图像处理单元,再到LCM显示这样一个完整的人脸图像处理及检测系统。

The design combines the detection method of skin color based on color image and that of gray scale based on geometric features in a Coarse-to-fine fashion. On the base of traditional skin color detection method, algorithms including the image preprocessing, color balance, binary module, morphological filtering,median filter and edge detection technology are added to achieve face detection. The system will eventually build a complete face image processing and detecting system, including the CMOS camera, FPGA core image processing unit and LCM display.