Abstarct: The process of quality control is made for assurance of designing high quality flawless products. In previous years, some sections of quality control that need visual inspection were done by humans. Nowadays, these works are done by machines and robots. The purpose of surface control is recognition and revelation of surfaces in products and industrial goods that those surfaces do not have ideal quality according to specified criteria. Today this procedure is used for identification of defects in many surfaces such as ceramic tiles, fabric, wood, paper, leather and so on. The revelation of defects of surfaces is possible through texture analysis methods, which explain the characteristic of surfaces. For this purpose we have used, Contourlet transform for analyzing of tile's images. Contourlet is 2D directional transform, which is used to describe curves and delicate details in images. Contourlet expansion is made from basic functions that have distinctive shapes and scales in tendency towards different directions (Anisotropic). Then we used a neural network for training data and finding defective tiles. The results are excellent and detect faults with more than 96.87% accuracy. Compared to the previous works, this is a remarkable primacy.