基于模板匹配的人脸识别跟踪方法研究

在对运动物体的实际追踪过程中,由于目标运动方向不确定、背景环境复杂多变等因素,增加了动态目标跟踪的难度,所以用普通识别方法较难提取出目标特征,从而无法准确跟踪动态目标。提出了基于模板匹配的人脸识别算法,分析了基于方向梯度直方图(HOG)的人脸特征提取算法和基于欧氏距离的人脸识别算法,设计出了人脸识别跟踪系统。实验结果表明通过动态目标视觉检测以及PID控制的角度输出,该系统实现了对特定目标的人脸识别和跟踪。

Research on Face Recognition and Tracking Based on Template Matching

In the actual tracking process of moving objects in daily life,due to the uncertain movement direction of the target and the complex and changeable background environment,it is more difficult to track the dynamic target.Therefore,it is difficult to extract the characteristics of the target with the common recognition method,so the dynamic target cannot be tracked accurately.Face recognition algorithm research based on template matching is presented,face feature extraction algorithm based on HOG and face matching algorithm based on Euclidean distance are analyzed,and face recognition and tracking system is designed.The experimental results show that through dynamic target vision detection and PID control Angle output,face recognition tracking of specific targets is realized.