Abstract: This paper puts forward an effective method of detecting irregularity surface crack. By acquiring the CCD camera image of cracks, MATLAB image processing toolbox is used in grayscale and binarization processing. Thinning can be performed on binary image for acquiring the crack length, crack boundary can be obtained by carrying on the gray contour extraction. The approximation of the real width of a crack is calculated by the correction algorithm proposed in this paper. Compared with other known methods, the accuracy of  calculation results is enhanced by this method. The program can run automatically to reduce the labor intensity during data processing, also, the efficiency has been improved.

Key words: MATLAB image processing toolbox, irregularity crack, correction algorithm

摘要： 提出了一种表面不规则裂纹特征的检测方法，通过获取CCD相机拍照得到裂纹图像，使用MATLAB图像处理工具箱对图像进行灰度化、二值化处理。对二值图像可以进行细化处理以求得裂纹长度，提取灰度等值线可以获得裂纹边界，在提出的修正算法下可以得到裂纹真实宽度的近似值，相比于其他已知方法提高了计算结果的精度，并且实现了自动处理，降低了数据处理过程中的人工劳动强度，提高了数据处理效率。

关键词: MATLAB图像处理工具箱, 不规则裂纹, 修正算法