Pavement cracks classification algorithm based on geometry feature analysis

The cracks of the road is a very important index to assess the quality of a road, since the different classes of the cracks are closely related to evaluate the dangerous degree of the road and the design of the repair strategy. Focusing on the common cracks including horizontal, vertical, block shape and chap cracks, a cracks classification method based on orientation and density characters was described. The main idea of the proposed method was to adopt the differences of the orientation and density distribution of the cracks to distinguish the different cracks. The orientation feature was first utilized to classify the direction cracks and non-direction cracks, and then by using the density feature to distinguish the block and the chap cracks. In order to check the effectiveness of the proposed method in this paper, using a large number of practical test data, by comparing with other algorithms, the results show that this proposed method has higher classification accuracy.