CORRELATION OF VISUAL ROUGHNESS PARAMETERS WITH THE AVERAGE VALUE OBTAINED FROM AISI 347 STAINLESS STEEL

This article presents a method that allows to correlate descriptors obtained through the artificial vision techniques of the co-occurrence matrix and Laws, in order to find the best descriptor that allows a measurement of roughness in AISI 347 stainless steel surfaces, similar to the value obtained using a contact profilometer. It was found that by combining the Laws descriptors they allow a better correlation by changing the kernel core that allows to measure curls with an accuracy of more than 92%, this will allow it to be used to quantify the roughness without having contact with the surface, increasing the manufacturing process of parts in inline production systems.