Bottom-Hat Filtering for Defect Detection With CNN Classification on Car Wiper Arm

Ooi, JiWei (2019) Bottom-Hat Filtering for Defect Detection With CNN Classification on Car Wiper Arm. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text Ooi JiWei.pdf Restricted to Registered users only Download (3MB)

Abstract

Quality control is an essential process for production as it ensures that the product quality is maintained or improved throughout the manufacturing processes. In car wiper industry, the quality of the wipers produced is the main factor of competitiveness given that defects on the wiper may bring negative effect on its market value. As the volume of the wiper produced is very high, hence it is very difficult to monitor the quality of all the car wipers manually. Not only that, additional labors are required to conduct manual inspection, which leads to high production cost and other quality issues. Therefore, this project developed an automated defect inspection system which can be implemented to the manufacturing process of car wiper arm using various techniques of image processing and classification algorithms based mainly on bottom-hat filtering and convolutional neural network (CNN) during the production of car wiper. This research project is in collaboration with Nippon Wiper Blade (M) Sdn. Bhd. The goal of this system is to detect and classify the car wiper arm defects such as bumps, dents, scratches and orange skins within a very short period of time. As result, this system is able to achieve the mean accuracy of 90.5% on defect detection and classification on car wiper arm. With this system, it improves the quality of the wiper produced, reduced its cost of manufacturing and bring growth to the company in car wiper industry