Small Surface Defects Detection on Ceramic Tiles Using Digital Holography

It is hard to detect small surface defects on ceramic tiles during process of visual detection which includes computer vision. This paper proposes new method that includes digital holography to amplify efect of small defects which leads to more efficient detection. To evaluate this new method of defect detection we have conduct experiment both with ceramic tiles with artifficialy made defects and with ceramic tiles with defects coused by manufacturing process. Detection of defects is made by extractions of texture properites from images of ceramic tiles after the effect of amplifying by digital holography. To evaluate properties of ceramic tiles we use well known methods like second order statistical coeficients as well as lokal binary operators.