Stereo vision for the measurement of turning tool size

Abstract

During computer numerical control (CNC) machining, interference and collisions can occur due to discrepancies between simulations and actual operating conditions, which can result in time loss and increased costs. The aim of this study was to measure the size of the tools used in CNC machines using automatic sensors in order to determine whether discrepancies exist between simulations and actual machining resulting from human negligence. Using a 3D stereo vision system, we obtained images of turning tools and extracted the required information related to tool features through image processing. Epipolar geometry was applied to simplify the dimensions of the feature search, and the theory and methods of object reconstruction in stereo vision systems were used to obtain the size of the turning tool.