Pavement crack detection method based on multilevel denoising model

In order to wipe off all kinds of noisy influence in the detection of pavement surface cracks,the advantages and disadvantages of the existing denoising models were analyzed.Combined with the advantages of the existing denoising algorithms and characteristics of the noise and crack information of pavement image,a novel multilevel denoising model based on the detection of pavement crack was presented.The whole denoising model was composed of gray denoising model,spatial filtering denoising model,crack characteristics denoising model and geometrical features denoising model.The experimental results indicate that the proposed model is more advanced in the pavement image denoising and the extraction of crack.