Road crack detection using the particle filter

摘要

Road crack identification is a prerequisite for both road health monitoring and reduction in reconstruction outlay. This paper proposes a Bayesian approach for pavement crack detection. The task is challenging because there is a small difference between a crack and noise besides the cracks have a strong variance in intensity throughout. The devised solution comprises of three steps. First the image data was acquired from the road imaging system. Second image was preprocessed with the erosion technique. Finally application of a particle filter based geometric model is done. The algorithm has the ability to consider any type of the crack. The method proved to be efficient enough to detect low contrast cracks. Experiments show that the proposed method has better outcomes than most of the prevalent algorithms.