Wiper Arm Surface Defect Detection

Gan, Sze Liu (2020) Wiper Arm Surface Defect Detection. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text Gan Sze Liu.pdf Restricted to Registered users only Download (2MB)

Abstract

Defect detection on wiper arm surface with the approach of image processing will be presented in this proposal. The detection of the defects is to prevent the presence of appearance or functional faulty on the surface of the wiper arm. The algorithm will process the image of the wiper arm to detect any defects on the surface of it. The basic steps for the image processing system includes grayscale conversion, filtering, edge detection and thresholding. False alarm often occurs due to the complex structure of the wiper, an image classifier are added to differentiate between defects and wiper structure. The average processing time for each image is 0.00025809999999992783 seconds. The final accuracy obtained is 87.9% which includes both good and defected samples.