A simple algorithm for geometric parameters and feature points detection

An algorithm has direct influence on computation efficiency and detection accuracy in digital image. We present a simple algorithm for computing geometric parameters of cutters used in CNC machine centers. The algorithm is a contour-based approach, which the differential segments divided from contour pixel arrays are used for detecting feature points and then the reversed calculations are taken to obtain the geometric parameters of cutter edges. In order to verify the performances of the approach, experiments were done on the tests of real cutters. The experiment results show that there is at most an error of 0.015° of the angle detected compared to the standard, and also the detection accuracy is immune almost from noises in images. The algorithm could improve practical application in cutting tool detection systems and other similar systems.