Detection of workpieces on a CNC machining table for measurement and automatic tool positioning

Abstract

This article deals with the localization and measurement of workpieces on top of a CNC machine tool's table. This paper builds upon one of our previous work and covers the same idea, however, the algorithms that had been used were not robust to misalignment. Indeed they required inflexible system settings. The plane of the depth sensor's frame had to be accurately set parallel to the machining table's plane using bubble levels. This process was not convenient since it had to be repeated whenever the camera was moved. In this paper, a Random Sample Consensus (RANSAC) based plane fitting method is applied for the segmentation of workpieces and for determining their sizes and locations for automatic tool positioning.