Automatic classification of travertine stones based on sum and difference histograms algorithm

Many industrial machines are used in travertine stone industry. But classification of this type of stone in terms of quality and appearance is generally done by human experts. Using of human experts for classification has a lot of errors, times, and costs. The reason of choosing of travertine stone is the large variety and increasing use of this stone in the building industry. Also, the classification problem in this stone has not been studied yet. Therefore, the need of automatic and computerized methods for classification of travertine stone is very useful. In this study, a set of travertine stones including 124 images from four groups was prepared. The feature extraction approach called sum and difference histogram (SDH) is used and several classification algorithms are applied. The simulation results show that the proposed approach has acceptable classification performance in travertine stones.