| With the continuous improvement of the national traffic system,bridges have reached an unprecedented scale.Cable-stayed bridges over 400 meters have reached more than half of the total number of cable-stayed bridges in the world.Bridges have been developed from the construction stage to the operation and maintenance stage,and the bridge safety detection has been more and more important.The surface defect of cable is an important part of safety hazard of cable stayed bridge.The timely discovery of the defects of the PE protective sleeve on the cable surface can grasp the damage of the wire inside the cable.In recent years,the rapid development of digital image processing technology provides a feasible technical means for the detection of bridge cable surface,and it’s also a developing direction for non-destructive testing of bridge cable surface.Due to the limitation of objective conditions,the following phenomena are presented in the cable surface defect detection:uneven illumination,impurity on the cable surface,spiral wound surface of cable and uneven surface of concave and convex surface,which increase the difficulty of automatic video processing of cable surface defect detection.This paper aims at the automatic video processing of bridge cable surface defect detection.First,digital video is converted to digital image.On the basis of this,the key technologies of video data automation processing and software development are carried out.The main research contents and results are as follows:(1)The image background segmentation of the cable surface detection.According to the statistical characteristics of the image sequence,the gray and gradient information of the digital image,combined with the existing image background segmentation algorithm,a background segmentation method based on the gray level gradient of the cable surface defect detection is proposed,and the statistical principle is used to carry out the sequence image to complete the image background segmentation of the cable surface defect detection.(2)The defect area extraction and classification.The surface defect area is extracted from the background segmentation image,and processed by morphology.Then describe the defect area with the area and the ratio of length to width,and separate the serious defects from general defects by discriminant function classification.(3)The development of video automatic processing software for bridge cable surface defect detection.According to the target and requirements of the cable surface defect detection,the software function module of the video data processing tool for creeping rope robot is analyzed and designed,and the tool software is developed based on the MATLAB/GUI platform.The video data of a cable stayed bridge’s surface defect are experimented regularly on the tool software.(4)The innovation of the paper is to determine the processing scheme of the defect extraction after the background segmentation according to the characteristics of the cable surface defect detection video data.In the process of the image background segmentation of the cable surface defect detection,the gray gradient information of the image is used to segment the image,and the interference information and the background mask are eliminated by the statistical characteristics of the image sequence,and obtain the background mask.Finally the background segmentation is completed.The results show that the background segmentation algorithm based on gray-scale gradient is stable,not affected by the uneven illumination and the interference factors of the cable surface.It can quickly and accurately complete the image background segmentation of the cable surface defect detection.The algorithm can correctly distinguish whether there is a defect on the cable surface,and the discriminant function classification method can identify serious defects and general defects effectively.Through the key technology of the video data processing of the creeping rope robot,the developed tool software can complete the automatic processing of the video data of the bridge cable surface detection.It has a positive effect on the field of bridge cable detection,and also has a certain push effect in other surface detection fields. |