Desirability Function Analysis Based Optimization of the On-Machine Diameter Measurement Using Machine Vision Under RGB Light

IoT-based machine vision system for dimensional inspection pays more attention to dimensional measurements of the components in a manufacturing system with the aim of a swift, reliable, error-proof measuring system with minimum human intervention. The use of an IoT system integrated with a machine vision system has a great advantage which improves the overall performance of the system. Such a system can work under a complex system environment. This research focuses on developing an IoT-based machine vision system for diameter measurement. Diameter measurements are performed under RGB light sources to investigate the errors in measurements. The approach of desirability function analysis is applied in order to reduce measurement errors which are further confirmed by a Taguchi method. Experimental results show that red light gives the lowest percentage error in diameter measurement with a value of ±0.2014% while blue light shows the highest percentage error in diameter measurement as ±0.6320%. However, the green light measurement method has a moderate percentage error in diameter measurement with a percentage error of ±0.5528%.