Assessment of Pressed Panel Products using Camera Image Processing

Abstract

The defect (necking, crack etc.) detection on press panels during the press forming process is an important step for ensuring the quality of the products. Traditional defect detection has been performed by human inspectors, which is subjective and expensive. It is also very difficult to apply traditional NDE/SHM approaches for such applications. Therefore, it is necessary to develop a new approach to efficiently detect such defects during the press forming process. In this study, we develop an automated crack detection technique for quality assessment of press panels with two camera image processing procedures. The first technique is evaluating the panel edge lines for crack detection, which are extracted from a percolated panel image object. This technique does not require a reference image for defect detection. The second technique is based on the comparison between a reference and a test image using the local image amplitude map. This process also contains compensation for misalignment of the images, which is typical in press lines. Defects are efficiently detected using two image processing techniques. For demonstration, experiments were performed in a laboratory and in actual manufacturing lines. Experimental results show that the proposed techniques could efficiently improve the defect detection rate with improved speed