Inteligent crack detecting algorithm on the concrete crack image using neural network

Safety inspection of concrete structures should be strictly carried out since it is closely related with the structural health and reliability. However, it is difficult to find cracks by a visual check for the extremely large structures. So, the development of crack detecting systems has been a significant issue. Final objective of this research is to develop an automatic crack detection system that can analyze the concrete surface and visualize the cracks efficiently. The algorithm is composed of two parts; image processing and image classification. In the first step, cracks are distinguished from background image easily using the filtering, the improved subtraction method, and the morphological operation. The particular data such as the number of pixel and the ratio of the major axis to minor axis for connected pixels area are also extracted. In the second step, the existence of cracks are identified. Backpropagation neural network is used to automate the image classification. Target data values in the training process were generated by inspector's manual classification. In order to verify the first and second step of the proposed algorithm, the algorithm was tested using real surface images of concrete bridge. Backpropagation neural network was trained using 105 images of concrete structure, and the trained network was tested for new 120 new images. The recognition rate of the crack image was 90% and non-crack image was 92%. This method is useful for non expert inspectors, enabling them to perform crack monitoring tasks effectively.