Method of surface defect detection for agricultural machinery parts based on image recognition technology

Abstract

At present, the large use of agricultural machinery for agricultural production has increased the efficiency of agricultural production, thus greatly improving the speed of agricultural production, and also greatly increasing the demand for agricultural machinery. Under the current market scenario, the surface quality of agricultural machinery will have a certain degree of impact on its grade pricing and competitiveness. To find product surface defects in the process of product quality control, product surface fault detection is very important in the process of product quality control. Traditional surface defects are mainly detected manually, which will lead to low efficiency and error prone. In order to solve and find the problems of surface defects and low accuracy of agricultural machinery, this paper proposes a method of inspection of surface micro defects based on image recognition technology. Canny operator is used to extract the weakness of the fault edge, and the algorithm of corrosion function is used to remove the relatively small structural elements in the image, so that the boundary becomes part of the internal contraction, complete the filling of the specified holes, connect the two points that meet the set value, on the other side, the integrity of the second image edge is defective, and complete the identification and detection of the defect section according to the defect type, Complete defect edge images can be obtained by adjusting characteristic defect behavior parameters, so as to complete defect category identification and inspection. In the simulation experiment, the proposed method can accurately detect various small defects of parts, and produce appropriate detection results.