基于机器视觉的耐火砖裂纹检测

Refractory Brick Crack detection Based on Machine Vision

现代化生产中,对耐火砖裂纹检测具有重要的实际意义。基于机器视觉技术,本文提出了一种有效的检测耐火砖裂纹的方法。该方法先对裂纹感兴趣区域进行定位,再用基于Hessian矩阵的方法提取裂纹中心线的亚像素坐标,最后完成裂纹长度和宽度亚像素计算。实验证明,该方法检测精度高,满足实际检测需要。

In the modernization production,it has the vital practical significance to detect the crack of refractory brick.Based on the machine vision technology,a method that detect the cracks of refractory brick effectively was proposed in this paper.First of all,crack target areas were located with the method.Then the subpixel center line of the crack was detected based on the Hessian matrix method.Finally the length and the width of the crack were culculated.The experimental results show that the detection accuracy is high and the method is able to meet the actual needs.