| Summary: | 碩士 === 國立雲林科技大學 === 電機工程系 === 103 === With the growing popularity of golf sport all over the world, the production of relevant products is increasing year by year. In golf utilities, club is the remunerative goods requiring high quality, and Taiwan is No.2 in the worldwide field of golf club production. Within mass production, errors occur easily and drain on manpower hugely in traditional manual examination. Nowadays using automatic detect systems for heads, logo of clubs or balls is common but no similar system for club to reduce manpower consumption. To be more competitive in golf industry, it's urgent in developing related utilities.
On the trend of golf automatic production, this study uses the AOI (Automated Optical Inspection) technology to detect defects on the steel club and classify them at the same time. To get good image of club’s surface in favor of image processing together with computer vision for detection, this study proposes using diffused dark-field side light illumination with mask to light, and area camera matrix on the dynamic rotating club to ignore picturing difficulty.
Experiment results show this system can detect twenty types of flaw and achieve 87% of detection rate with lower cost. The system is close to visual inspection though we know there's still a room to grow. It takes 750ms/club for the system to detect so far which exceeds in the demand for 1ms/club practically. Progress widely even takes manpower place, have a competitive performance.
|
|---|
