基于轮廓特征点的高精度识别目标方法

为了实现高成功率、高精度和快速识别跟踪目标，提出对基于轮廓特征点的目标精确识别方法。在识别过程中采用了轮廓提取和多边形拟合算法自动搜寻到图像中要识别和跟踪的目标，同时对目标物轮廓的多边形角点进行亚像素分辨率的定位，从而可以利用目标轮廓角点的精确定位来实现对多边形目标的识别与跟踪。试验结果以及特征点亚像素算法分辨率的分析表明，采用这种自动识别与跟踪目标的方法，其精度可以达到0．02像素。

An Approach of Recognition in High Precision Based on the Contour Feature Points

In order to achieve a high success rate, high precision and fast track the identification of objectives, accurate identification method is put forward. In the process of identification, the extraction and the outline of the polygon fitting algorithm are used to automatically search for images to identify and track targets. The comers of the polygon target are positioned in sub-pixel resolution of the position. The polygons can be replaced with these comers which describe the goals in high precision. The experiment result shows that the identification and tracking method is robust and resolution can be reached 0.02 pixels.