样本方差与相似点判断相结合的图像裂缝检测

摘要

为解决传统渗流裂缝检测方法速度较慢的问题,提出一种样本方差与相似点判断相结合的图像裂缝检测算法.采用样本方差提取裂缝区域,提出一种偏差处理方式提取裂缝像素点,减少后续像素处理个数;进行初步裂缝渗流检测;提出一种裂缝断裂区间相似点判断算法并进一步渗流检测,避免裂缝断裂现象,提高检测的精确率.实验结果表明,该算法与传统渗流算法的精确率基本保持一致,降低了裂缝连接算法导致的误检测率,检测速度明显提高.%To solve the problem of slow detection speed of the traditional percolation crack detection method,an algorithm of image crack detection based on the combination of sample variance and similarity judgment was proposed.The sample variance algorithm was used to extract the crack area,and a deviation processing method was proposed to extract the crack pixels to reduce the number of pixels.Percolation detection was carried out preliminarily.The similarity judgment algorithm of fracture zone was proposed,and the percolation process was executed to avoid the phenomenon of fracture to improve the accuracy of crack detection.Experimental results show that the accuracy of the proposed algorithm is consistent with that of the traditional percolation algorithm,and the proposed method not only reduces the false detection rate resulted from crack connection algorithm,but also improves the detection speed obviously.