Author(s)

A. Zawada-Tomkiewicz and B. Storch

Abstract

The application of image processing techniques in the tool wear estimation A. Zawada-Tomkiewicz and B. S torch Technical University of Koszalin, Poland Abstract The assumptions for a monitoring system of a tool flank wear on the basis of a machined surface digital image acquired from a computer vision system are presented. The method of a VB parameter estimation is based on a texture analysis of a digital image of a machined surface. This method is an indirect method. One directional three-layer artificial neural network with an optimized number of neurons in each layer and time of training acts as an estimator. The settlement of the estimator parameters is based on error minimization between the reference values and the output values from an estimator. Input values for an estimator are the mathematical indexes of a machined surface digital image. The method presented solves the problem of applying a machined surface image for tool flank wear monitoring during cutting. 1. Introduction Turnin

Keywords