Research on tool wear detection based on machine vision in end milling process

This paper suggests a novel technique for the tool wear measurement based on machine vision. Tool images are captured between cutting operations using a machine vision system. The gray value difference threshold is determined from the tool image itself and the reference line is found to locate the tool in the image. The edges of the tool wear region are extracted by column scanning. A method of continuity testing is used to find the correct edge position in each wear column. To achieve a more accurate result, the sub-pixel edge detection technology is adopted to extract the edges. Finaly, the tool wear parameters can be obtained after rebuilding the top edge of the wear region and determining the bottom edge of the wear region. The measurement results gotten by the proposed method are compared with those gotten by measuring directly with a microscope. The proposed scheme is shown to be reliable and effective for the automated tool wear measurement.