Flank wear width is generally recognized as the key indicator for tool life.In the experiments of this study, nine tools made of HSS and stainless steel 316L as work piecewiththreespindlespeeds(550,930 and 1100)rpm and three redial depth of cut (1.5,2 and 2.25)mm were used. The cutting tool wear was measured using optical microscope and vision system based on a proposed algorithm.Maximum and minimum percentage errors in the flank wear width were (8.250% and 0.645 %) respectively. The numerical method used was by a multiple linear and polynomial regression model and developed a polynomial model, especially to predict the flank wear using MATLAB software.Maximum and minimum percentage errors werefound (14% and 0.322 %) respectively.