Automatic Surface Defect Detection for Ceramic Tiles Using Digital Image Processing: A Literature Review

Abstract

The term quality is the most important factor in many industries. Whole business and the brand name are dependent upon the quality of produced goods. Defect detection is generally done manually while manufacturing may be fully automatic. Therefore, our aim is to enhance quality control by integrating various image-processing techniques, before dispatching of the final product to make no rejection rate. An automated system is expensive at the installation but it is a one-time investment, compared to wages paid to labor for a long time it provides high productivity and economical as well. The automatic inspection helps us to know a pattern of defects in a very short time and to determine that defected piece can be reworked or rejecter by avoiding them to mix with fresh tiles. In which Matlab software is used to develop a different algorithm for a different defect. By applying this method, we can obtain accuracy up to 97%. This paper provides the literature review for analysis of techniques, which are used to find defects in tiles that do not match with a reference pattern of tiles stored in the computer. This highly efficient result reflects that this automated system can effetely replace manual tiles inspection with better accuracy and efficiency. In the future, an automatic inspection setup can be designed in which from raw material to final product, one can carry out the 100% inspection of tiles on to the on-line conveyer under the guidance of camera and processor using image processing.