A Pre-sorting Method for Iron Pan Surface Inspection Based on Image Recognition

摘要

In this paper, we propose a method to evaluate the polishing quality of iron pan by using image recognition technology. The method is designed to enable the automatic sorting for iron pot. The inner surface image of iron pan is acquired by a CCD camera, then the image information entropy is selected as the texture feature of these images. 40 groups of information entropy from pre-qualified the iron pan samples are used as the quality detection standard. Then the test samples are chosen, the information entropy of which are calculated, and being compared with the standard. The comparison result shows that it is practicable to pick out the failed sample by the experimental threshold. The method is used in pan inspection line and considerable reduced the load of human inspectors.