Machine Vision Aided Locating for Microwave Heating Control of the Asphalt Pavement Maintenance Vehicle

摘要

At present asphalt pavement defects are distinguished subjectively. The recognition depends greatly on human vision, which can bring large variations for micro-wave heating control. The paper adopts machine vision aided locating to create accurate heating control data that is objective and independent of human to overcome the deficiency. Firstly, the micro-wave heating control system (MWHCS) based on machine vision is presented. Then, the algorithms of cubic polynomial warping and reprojection transformation are applied to rectify the image distortion. Thirdly, the automatic and manual methods to create heating control data are discussed. At last, an ecnomical driving circuit design matching with the heating data is demonstrated. Product testing proves that the proposed heating control method with machine vision is prompt and accurate in locating defects, also quite effective in energy saving.