The Detection System For Bridge Cable Surface Damage Based On Image Recognition

Nowadays, the cable-stayed bridge and suspension bridge, which have the beautiful structure and better span ability, are applied abroad to the construction of bridge. As the bearing component, the cable is an important part of bridge. So it should resist fatigue reliably, and it can be anticorrosive. Whereas, some factors, such as the rain, fog, gas and so on, could erode the cable in the environment. As a result, the detection of cable is an important part in health detection of bridge.In this thesis, the design of the embedded surface damage detection system bridge cable based on image recognition technology is introduced, after analyzing the currently technology of the bridge cable surface damage detection. It make a target of fast intellective damage detection. Depended on concurrent sensor array of 4 CCD, the system collects the surface image data of the bridge cable, and sends the data to DSPs. This DSPs based on TMS320DM642, includes video signal transform circuit, SDRAM, FLASH, ATA interface and so on. The image of bridge cable could be cropped, pretreated, rectified projection, detected and saved.By the characteristic of bridge cable surface image, we design area cropping, gray transforming to preprocess image. The image's noise can be filtered by the advanced median filter. The image could be emendated by the inserting value arithmetic. The damage could be separated from the background by the dynamic value division.The system provides an effective way to exactly detect the surface damages of bridge cable in real-time, with low costly, and it is another good method in the surface damage detection of bridge cable .