Image-based chip detection during turning

Abstract This study proposes a method to analyze chip formation using camera surveillance to enhance safety and efficiency in machine operations. The process involved the face and straight turning of a workpiece under the observation of a camera strategically placed within the workspace. The suggested algorithm carries out initial image preprocessing and edge detection, followed by background subtraction to isolate dynamic elements and filtering based on the size of the objects. Pre-determined masks are applied to eliminate overlaps with existing workspace objects, based on the tool’s trajectory. The research validates that the applied technique effectively recognizes chips in both face and straight turning. Specific filtering techniques improve the algorithm’s capability to detect even smaller chips, and it substantially reduces false alarms, laying the groundwork for long, continuous chip detection systems.