Research of the algorithm calculating the length of bridge crack based on stereo vision

Bridge crack is a common bridge disease. The usual crack detection focuses on the pavement and bottom of bridges, while the intermediate support structures where cracks exist, such as piers, are seldom considered. As well as, when the camera plane is not parallel to the bridge plane, the crack in the image is only the projection of the true crack in the camera plane, leading to errors of the calculated crack size. To this end, this paper proposes a bridge crack detection method based on stereo vision, which calculates the crack length by combining depth information and crack image. Compared with the existing measurement methods, the method proposed in this paper has good experiment effect in the detection process of crack characteristics.