Abstract:

Sorting is a necessary process in industrial mass production applications. Considering the increasing population and demand for consumer products, simpler means to perform image detection and analysis through machine vision applications are crucial. This study demonstrates an effective machine vision application by analysing the correct faces of chain links. In the chain fabrication process, the high throughput production line requires proper alignment of each chain link due to the tolerance difference of each linked face due to the machining procedure. The manual labour currently applied in correcting the face orientation of the chain links in the production flow line increases the fabrication cost and human error. Through a simple machine vision application and an industrial-grade global shutter camera, detection of continuously flowing chain links is achieved by using a marker. The procedure works by detecting the marker on the related face of the chain links through image thresholding and analysis. The study is offering 100% accuracy for sorting single and multi-line chain links in real-time applications. The demonstrated application can be coupled by a sorting mechanism adapted to various quality control and sorting requirements in industrial manufacturing.

Keywords: