Visual Diagnostics Based on Image Wavelet Transform

The image processing described in this paper is used for visual quality control in ceramic tile production. The tiles surface quality is described by the surface defects. The described image processing is based on the discrete wavelet transform method. The diagnostic algorithm is described. It is based on comparing of the wavelet coefficients of the original image without surface defects and the real images of ceramic tiles. The method is verified by using the artificial defects on the image and sensitivity testing on failure contrast and size is done. The algorithm is evaluated experimentaly using the real tile images. The analysis of the detection capabilities and sensitivity expressed in nondetected failures and false proclaimed defect is done also. Optimal connection between the segment size and DSL for each type of surface failure could be used to make efficient system for quality control and failure classification in automated production process.