基于改进Harris角点提取算法的网格图像破损检测

摘要

针对周界网格状围栏破损预警问题,提出一种基于改进Harris角点的网格图像破损检测算法.传统Harris角点提取算法需要对图像中每个像素点计算横纵方向上的一阶导数以及角点响应函数值,算法复杂度高,通过引入灰度“相似度”的参数来计算像素点与其周围像素灰度值的相似程度,从而滤除伪角点,减少Harris角点提取时间,最后通过分析角点分布信息来界定破损区域.对移动机器人采集的典型围栏破损图像进行了检测试验,由实验结果可看出,Harris角点提取时间大大减少,表明该算法有效且满足围栏破损检测实际应用要求.%Concerning the breakage warning problem of grid fence, a grid breakage detection algorithm based on improved Harris comer was proposed. As for traditional Harris corner extraction algorithm, the first derivative in vertical and horizontal direction and the corner response function value need to be calculated, and it must be done for each image pixel, which makes this method time-consuming. Therefore, the parameter of gray "similarity" was introduced to demonstrate the gray similarity between pixel and its ambient pixels, through which to filter pseudo corners, reducing the Harris corner extraction time. Then after analyzing the corners' distribution, the breakage areas can be locked. Breakage detection experiments were taken to various fence images taken by robot, and the results indicate that the Harris corner extraction time decreases largely, and the proposed algorithm is effective, meeting the practical application requirements of fence breakage detection.