Damage Identification of Steel Members Based on Infrared Image Processing Technology

Abstract

: In view of the existing infrared image processing algorithms in dealing with damage images of bridge steel members, the signal-to-noise ratio is low, the contrast is low, the resolution is low, the image details are lost, the edges are blurred, and the accuracy of damage recognition is poor. We propose a spatial and time domain filtering combined infrared image enhancement algorithm to improve existing algorithms, suppressing image background noise from multiple directions, enhancing image detail information, strengthening damage edge contours, accurately identifying and extracting damaged parts of steel components. The processing results are quantitatively evaluated by clarity, contrast, peak signal-to-noise ratio (PSNR) and mean square error. The evaluation results show that the infrared image enhancement algorithm based on high-frequency emphasis filtering and nonlinear gray-scale conversion is feasible. The damage recognition effect is remarkable for the steel components under infrared image detection.