Measurement and evaluation of surface roughness based on optic system using image processing and artificial neural network

摘要: The traditional devices, used to measure the surface roughness, are very sensitive, and they obtained by scratching of materials. Therefore, optic systems as alternatives these devices avoid unwanted processes that damage surface. In this study, face milling process was applied American Iron Steel Institute (AISI) 1040 carbon steel aluminium alloy 5083 materials using different tools, cutting speeds depth cuts. After processes, roughness values were tester, machined images taken a polarise microscope. converted into binary images, input data train network MATLAB neural toolbox. For training networks, log-sigmoid function selected transfer function, scaled conjugate gradient (SCG) algorithm algorithm, performance trained networks achieved an average 99.926 % for (AA) 99.932 AISI steel. At end prediction programme optical m-file GUI programming developed. Then, tested trial experiments. experiments, with stylus technique compared developed values. When experimental results, results confirmed each other at rate 99.999 %.