基于模板抽样的快速图像匹配算法

为了提高图像匹配速度,满足某些领域的实时性要求,提出了一种快速图像匹配算法.该算法利用Sobel边缘算子得到模板的灰度边缘图像,并对该边缘图像进行抽样以提取匹配点,从而显著减少匹配过程的计算量.利用遗传算法的非遍历搜索机制,迅速收敛到全局近似最优解,进一步减少了匹配过程的计算量.在此基础上引入精确匹配环节,找出了目标子图像的精确位移及旋转角度.将该算法应用于全自动金丝球焊机的图像识别系统,在主频为1GHz的工控机上实现该算法,匹配时间平均约为37ms,小于系统在60ms内进行匹配的要求,连续多次实验算法均能精确匹配目标的概率为93.8%,满足该系统的实时性与精度要求,取得了理想的效果.

Fast image matching algorithm based on template sampling

A fast image matching algorithm was proposed to increase the running efficiency of the algorithm, and meet the real-time requirement for some applications. A gray edge image of the template was extracted using Sobel marginal operator. Matching points were sampled from the gray edge image and utilized to reduce the time complexity of the matching process. At the same time, the global searching mechanism of the genetic algorithm was exploited to expedite the convergence of the algorithm to the sub-optimal solution to the problem. Based on it, a precisely locating phase was introduced to fix the location and rotating degree of the optimal target sub-image in the scene image. Practical application of the algorithm to the image matching system of a fully automatic gold wire bonder demonstrated that the algorithm could meet the real-time requirement and the locating precision requirement of the system, and they worked as expected.