Cutting Tool Wear Region Extraction Based on L0 Gradientminimization Model and Improved CV Model

Affiliation of Author(s):电子信息工程学院

Journal:2017 2ND INTERNATIONAL CONFERENCE ON MULTIMEDIA AND IMAGE PROCESSING (ICMIP)

Key Words:cutting tool wear region extraction L-0 gradient minimization model CV model

Abstract:To accomplish the wear region extraction in the cutting tool wear detection system exactly, an image segmentation method based on L-0 gradient minimization model and improved CV model is proposed in this paper. Firstly, the original image is smoothed by L-0 gradient minimization model. Then the Otsu method is used to segment the smoothed image to achieve the initial location of the wear region in the cutting tool surface, after which the one-pixel edges are obtained by using the edge following algorithm. Finally, by taking the edges as initial contour, the CV model is applied to segment the original image to extract the wear region exactly. Meanwhile, to reduce the long runtime caused by the low convergence speed, an acceleration factor based on the gradient is introduced to optimize the CV model. A lot of experimental results show the proposed method has a good performance in cutting tool wear region extraction, with high speed and high precision.

Translation or Not:no

Date of Publication:2017-01-01

Co-author:龙云淋,zy

Correspondence Author:wyq

Date of Publication:2017-01-01