Visual Inspection of Ceramic Tiles Surfaces Using Statistical Features and LVQ of Artificial Neural Networks

Document Type : Original Article


1 Faculty of Computer and Information Sciences, Zagazig University, Zagazig, Egypt

2 Dept. of Computer Science, College of Computer and Information Sciences, King Saud Univ., Riyadh 11543, Saudi Arabia

3 Dept. of Computer Science, Misr for Engineering and Technology(MET) Academy, Mansoura, Egypt


A product inspection is an important stage in the product manufacturing process before packing process. Visual inspection is an application of computer vision and image processing. Using Automatic Visual Inspection (AVI) in ceramic tile inspection gives more reliable, fast, robust, and less human intervention, increases processing stability, and improves overall production performance. Many approaches had been applied for this purpose starting from primitive pixel by pixel comparison to advanced techniques such as statistical, structural, and signal processing techniques. In this paper, we use statistical approach as a feature extraction approach for visual inspection of ceramic tiles surfaces defects with neural network classification techniques. Our results show that statistical techniques used for extracting ceramic tile features with LVQ neural networks classifiers give subtle results especially LBP and GLCM which give 96% and 90% of correct classification respectively.