一种基于松弛特征关联的3D人脸多特征匹配算法

Based on the Characteristics of Slack Associated 3D Face Many Features Matching Algorithms

针对人脸特征表情变化极为复杂,3D模型形变中的程度较为细微。一些特征点的位置会出现一定程度的飘逸,给准确的三维特征匹配带来了难度。提出基于松弛特征关联的3D人脸多特征匹配算法。通过提取特征点,计算三维特征的松弛特征对应程度,最大程度的完成在误差较大程度下的三维特征匹配。避免传统算法中用直线距离特征匹配带来的弊端。实验仿真程序表明:该方法在误差较大的匹配环境下,优化效果较为明显。

According to face feature expression change is very complex,3 D model of the deformation degree more subtle.Some of the feature point position will appear a certain degree of elegant,give accurate three dimensional feature matching brought the difficulty.Based on the characteristics of the 3 D flabby associated face many features matching algorithms.By extracting feature point,the calculation of 3 d characteristics corresponding characteristics of the relaxation degree,the greatest degree of finish in the error of the larger degree 3 d feature matching.Avoid traditional algorithm with straight distance the malpractices feature matching.Experimental simulation program shows that this method in the error of the larger matching environment,the optimization effect is more apparent.