Research on Generating Step Value Algorithm for Gray Level Co-occurrence Matrix and Its Application in Tool Monitoring

摘要

The monitoring of tool wear condition can be processed by means of the analysis of workpiece surface texture images. In order to analyze the workpiece surface texture images accurately, the generating step value of Gray Level Co-occurrence Matrix is studied extensively. A new algorithm about generating step value is proposed in this paper. Through analysis and testified by experiments, the optimum value of the generating step value is the step value which makes the textural feature parameters obtain the extreme value (the maximum or minimum value) in the first period. The optimum value of the generating step value is only related with the feed rate. It is unrelated with the other machining parameters such as the tool wear and machining speed. Compared with the other step value, the optimum value of the generating step value makes GLCM more different and diverse for different texture images. Therefore the difference of textural feature parameter values based on the above GLCM is great. Consequently the feature parameters obtained by the reasonable generating step value are sensitive to the tool wear which are beneficial to monitor the tool wear.