A detection system for surface defects of bridge cable based on machine vision

作者：Li Xinke^*; Gao Chao; Guo Yongcai; He Fuliang

来源：ICIC Express Letters, 2012, 6(9): 2225-2230.

• Core processors
• Damage inspection
• Defect detection
• Defect images
• Detection accuracy
• Detection efficiency
• Detection system
• Edge operator
• Embedded DSP
• Processing speed
• REmove noise
• Traffic safety

Cables are a vital part of cable-stayed bridges. Obviously, the damage inspection and maintenance for bridge cables are indispensable. However, the current methods for cable damage detection have their own disadvantages, which could not meet the requirements of traffic safety properly. It is very important to improve the detection accuracy and detection efficiency. In this work, we propose a detection system for bridge cable surface defect. This detection system, employing embedded DSP (Digital Signal processor) of TI's TMS320DM642 as core processor, uses the improved median filter to remove noise and adopts Sobel edge operator and morphology to extract the defect areas from the background. We use the system to detect the bridge cables. It costs about 0.437 seconds to process and recognize a defect image. The system can detect a defect area larger than 1cm2. Experimental results demonstrate that the cable defect areas can be recognized rapidly and automatically. This system also achieves good performance in the processing speed and stabilization.

• 出版日期2012