DAMAGE INSPECTION OF WIPER BLADES USING IMAGE PROCESSING TECHNIQUES

Abstract

The working part of Wiper Blades (WBs), made from PU material, are more vulnerable to contactbased solutions than other methods. In conventional approaches, sharp styluses employed as probes for tracking damaged edges are usually forced to wear out after a long time of the scanning process, thereby producing unreliable results. As a result, a machine vision-based system is proposed in this paper for WB damage inspection. Double threshold and morphological operations are used for defect verification. Defects are then derived using the background subtraction method before applying region-based algorithm for assessment. Results of experiments have demonstrated the reliability and flexibility of the new approach. Differences between the novel approach and the conventional method, which uses piezoelectric sensor, are also presented in this paper.