科研之友
微信
新浪微博
Facebook
分享链接
摘要
The degree of crack growth in asphalt pavement is an important decision-making factor in road maintenance management. Automatic crack detection is based mainly on digital images; this factor makes effective detection of the degree of crack growth difficult. Infrared thermography was used, and a detection method for the degree of crack growth on the basis of infrared imaging was proposed. Infrared images included gray-level information on cracks and temperature information; the latter provided one additional dimension of information over ordinary images. Temperature information was used to detect the degree of crack growth. Atmospheric temperature was found to be the main factor that affected the temperature difference between a crack and the road surface. This temperature difference varied significantly for different extents of crack growth, and therefore this difference can be used to detect the degree of crack growth. Two classification functions that divided the degree of crack growth into three grades were obtained by classifying data through the use of a support vector machine. A suitable environmental condition for using the detection model was proposed. The experimental results showed that the average model error was 15.4%, which indicated a good application prospect and an improvement in economic benefit for pavement maintenance.
