Study On The Crack Extraction Method Based On Pavement Detection Vehicle Image

Highway maintenance work is an important part of ensuring the normal use of the highway Innatural conditions,the external factors and the vehicle loads, the highway will produce a variety of diseases.As an initial crack disease,have adevastating effect on the highway,the pavementcrack detection and recognition is particularly important.Pavement crack detection technology was originally developed by artificial detection, it is time-consuming, dangerous, low efficiency, traffic etc.. With the development of computer and digital image processing, computer aided detection of pavement crack image has become a hot spot of research based on.Pavement crack automatic detection is the process of image acquisition, image transmission,computer processing, the result. Because of the complexity of the road, due to the interference of external factors, the road image acquisition detection vehicle are affected by noise, shadow and light effect is not uniform,so that the existing algorithms can not be better extracted from images of crack, prone to error, omission, the recognition effect can’t meet the requirements.Therefore, in does not change the external factors, how to find a more can adapt to the complex crack extraction algorithm of image is the focus of this paper. This paper mainly do the following several aspects of research work:①In the process of image preprocessing, through theoretical analysis and test of several common filtering operator, the median filtering method is an improved adaptive, and combined with the gray level correction of image preprocessing, the main purpose is to remove the image noise of image gray and Yu Guangzhao caused by uneven. Experimental results show that the removal of the non-uniform illumination and filtering effect is more ideal.②Compared to a variety of image segmentation method, aiming at the cracks in the pavementimage signal is weak, the problem of uneven width, threshold segmentation method is used to determine according to the histogram of solitary valley point, the value of the two principle, on the crack image segmentation.③The segmented image by mathematical morphology theory,geometric connectivity of gray change, structural connectivity in the region, to determine the crack information through thegeometrical characteristics of the image, and then get the crack image,finally the crack image of geometric connectivity features obtained after measurement, crack length, width and reticulatecrack area information.④Finally, the cracks of the extraction algorithm is adopted in this paper, were selected fordifferent types of cracks, cracks, get the information of the crack image. Experimental results show that this method can meet the requirements of the actual detection, and the computational complexity is small, easy to implement.