Research On Pavement Crack Detection Algorithm

With the increasing of automation for the pavement distress detection, the upgrade of hardware requires more highly on real-time of the software algorithm. How to improve its efficiency and applicability has always been the focus of the study.On the basis of analyzing and summarizing research results at home and abroad, this paper researches and uses the high efficiency and applicability detection algorithm to identify the crack ,according to the principle of practicality and the matching with the hardware system. The main research contents are as follows:(1)The uneven illumination in crack image is the topic which is been working to solve by researchers in recent years. This paper analyzes the crack image correction algorithm based on bilinear interpolation. We propose a new crack image correction algorithm for its shortcomings .The experiment proves that the improved algorithm has a good correction effect on image.(2)Classic filtering algorithm always leads to crack edge indistinct, while they can filter out most of the noises. So in order to inherit the advantages and make up the shortcomings, the median filter is improved. At the same time, the correction of crack image resolves the defects of median filter on uneven illumination.(3)On the critical step of segmentation, segmentation algorithm based on histogram of the estimated threshold is improved. The new algorithm which insists on the algorithm for real-time principle, not only refers to improve the segmentation results of the crack, but also simplifies the computation of the algorithm by taking advantage of the characteristics of crack images.Proved by a large number of experiments, this paper better identifies cracks and completes characterization of the crack by applying the above improved algorithms. The recognition result error is samll.