基于几何图像滤波技术的人脸3D识别算法研究  

A Study on 3D Face Recognition Algorithm Based on Geometric Image Filter Technology

作　　者：陈初侠[1] 袁宗文[1] Chen Chuxia,Yuan Zongwen(Chaohu University, Chaohu 238000, China)

机构地区：[1]巢湖学院,安徽巢湖238000

出　　处：《黑河学院学报》2018年第10期213-214,共2页Journal of Heihe University

基　　金：巢湖学院自然科学研究资助项目"图像滤波及边缘检测与增强技术研究"(XLY-201110)

摘　　要：几何图像滤波技术的人脸3D识别算法是将人脸网格映射至平面区域中,采用线性插值采样获得几何图像,并对几何图像进行切割,然后采用差分进化算法来优化滤波器,并通过训练对分块图像局部特征进行提取、对相似度计算与融合,最后获得识别结果。实验验证结果表明,基于几何图像滤波技术人脸3D识别算法的识别精度更高、鲁棒性更强,能提高相关部门的工作效率。3 D face recognition algorithm based on geometric image filtering technology refers to the procedures of projecting the face mesh to the planar region, obtaining the geometric image by using linear interpolation sampling, cutting the geometric image, then optimizing the filter by using the differential evolution algorithm, extracting the local features of the block image, calculating and merging the similarity, and finally obtaining the recognition result. The verified results of the experiment shows that 3 D face recognition algorithm based on geometric image filtering technology has higher recognition accuracy and stronger robustness, and can improve the work efficiency of relevant departments.

关 键 词：几何图像 滤波技术 人脸识别 3D识别算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]