关键词: 隧道工程；地铁隧道；自适应灰度拉伸；Otsu法；特征提取

Abstract: Aiming at the problems of uneven illumination and complex noise of the crack images collected in the subway tunnel environment, a method combining Mask uniform light with adaptive grayscale stretching was used to improve the contrast of the crack image. By detecting the noise points of the pixels and only operating on the effective pixels in the filtering window, it was possible to filter out a lot of noise while protecting the details of the crack edges. Canny edge detection and Otsu method were combined to segment the filtered image, remove isolated noise and perform morphological connection to obtain a complete binary image of the crack area. Combined with the crack skeleton diagram, the crack length, width and area were calculated. The research results show that compared with the commonly used methods, the proposed algorithm can effectively enhance image contrast and filter noise, and can accurately identify the crack area.

Key words: tunnel engineering; subway tunnel; adaptive grayscale stretching; Otsu method; feature extraction