Efficient surface crack detection in ceramic tiles using MATLAB image processing

The significance of quality control in the ceramic tile industry cannot be overstated. Traditional approaches to quality control in this sector have relied heavily on labor-intensive manual inspection methods, which are not only costly but also less efficient. Moreover, these methods often fall short in terms of accuracy, primarily due to the challenging industrial environment and the potential for human error. To address these shortcomings, this research proposed an innovative automated inspection system utilizing advanced image processing techniques specifically designed for the ceramic tile industry. This system detects defects such as corner damage, edge damage and center cracks on tile surfaces and provides information for quality assurance processes by classifying tile quality through rigorous analysis and comparison of various quality parameters. The performance of the system is tested with a total of 120 synthetic data, including those with cracks, damaged corners and discoloration. As a result of the testing process, a 97.5% accuracy rate is obtained. Furthermore, the system operates with a processing time of approximately 0.8 seconds per piece. This study, which offers both high accuracy and efficiency, promises significant improvements in quality control processes by offering an alternative to manual inspection methods.