Crack detection in tiles based on hybrid technology of GA and MLP using Image Processing

Abstract: The problem of the crack can be avoided by detecting the cracks on the surface of tiles. Crack detection using individual Genetic Algorithm and Multilayer Perceptron can achieve the efficiency less than the hybrid or combination of these two techniques in tiles industries. The efficient use of image processing with advanced technologies Generic Algorithm and Multilayer Perceptron with self organizing maps which would be required in the real time applications of tiles surface crack detection. Till now it was possible with individual above mentioned methods. This paper shows the possibility of hybrid method combining GA and MLP. Crack/Fault detection using hybrid of GA and MLP will be helpful for deduction in the rejection ratio of the tiles at the time of packaging

Keywords: Genetic Algorithm, Multilayer Perceptron, Self Organizing Maps, Hybrid, Crack Detection.