| Surface defects detection is one of the most important aspects of cable-stayed bridges’ inspection. Surface defects impact the safety of cable-stayed bridges. The paper adopts the computer vision techniques to detect the defects automatically. However, it is a challenge to detect such defects in a vision system because of illumination inequality and the variation of reflection property of cable-stayed bridges’cables.This paper designs an intelligent vision detection system for surface defects. The vision inspection system first acquires the cable images by the image acquisition system, and then, cuts the sub-image of cable manually. Subsequently, vision detection system enhances the contrast of the cable image using the local contrast method and segmented the enhanced image by improved maximum correlation method. At last, visual inspection system detects defects using the defect identification based on cable images’gradient. And the paper focuses on two key issues:image enhancement and image segmentation. The improved local contrast measure method is used to enhance cable images. It is nonlinear, multiplied noise independent and utilizing a model of contrast perception of human being; it enhances the contrast by mapping relatively low gray range to wider contrast range; therefore, it notably improves the distinction between defects and background. In addition, the promoted automatic thresholding method is applied. It selects a threshold by optimizing the product of object correlation and the weight term that expresses the proportion of thresholded defects.The experimental results demonstrate that the vision detection system detects the Type-Ⅰ defects with a recall of 80.4% and Type-II defects with a recall of 85.2%, and the proposed vision detection system outperforms the related well-established approaches for the application of surface defects detection of cable-stayed bridges’ cables. Furthermore, the proposed vision system is fast with a linear computational time complexity. |
PDF Full Text Request