Research On Key Techniques Of Pavement Crack Video

According to the analysis of the 2018 statistical bulletin on the development of the transportation industry,the highway industry will be gradually transited from a large-scale construction cycle to a large-scale maintenance cycle.Pavement crack detection is the basis of highway maintenance treatment.This article has carried out theoretical and experimental research on several problems of pavement crack detection.UAV aerial photography is applied to video image acquisition of pavement conditions.The image of the pavement condition video is sequenced and frame-sampled to obtain the image to be identified.The main purpose of image processing is crack detection.The process includes denoising of the pavement crack image and segmentation of the image,pavement crack detection and splicing,and extraction of pavement crack feature parameters,and also the classification.This processing method improves the automation level and efficiency of pavement crack energy detection.Aiming at the problems of serious noise,weak crack information,and unclear crack edges in pavement crack images,an improved Otsu pavement crack segmentation algorithm is proposed.The differences between the threshold and the target wave crest and the number of pixels in the overall population are proposed.As weights,the accuracy of target extraction is improved by modifying the gray histogram probability weight factor.The experimental results show that the method can effectively segment the pavement crack images and has good consistency in detecting different pavement cracks.In terms of image acquisition and pre-processing,UAV are applied to the detection of road cracks,that is,images of road cracks are obtained using an image acquisition platform equipped with the UAV.The composition of each module of the UAV pavement acquisition system and the basic functions of the pavement crack image acquisition platform are given.The characteristics of the pavement crack image are analyzed.The pavement crack image enhancement technology is studied,including image graying,equalization,and for road image noise reduction,median filtering and gradient reciprocal weighted filtering are applied to the image to reduce noise.Compared with the experimental results,the latter is more ideal for noise reduction.In image stitching,the relationship between pavement crack splicing and pavement crack detection is analyzed.Through comparison and analysis of a series of pavement crack splicing algorithms,a pavement crack splicing method that is superior to other methods—local area comparison stitching algorithm is proposed.The stitching process of pavement crack images is given,and the algorithm is tested with satisfactory results.Finally,the causes of pavement cracks and the classification of cracks were analyzed from different dimensions.Based on the differences in the geometrical shapes of different types of cracks on the pavement,two-dimensional projection features and damage density characteristics were proposed to determine the types of cracks.The calculation method of the length and width of cracks is given here.The measurement of these cracks can reflect the damage of the road surface to a certain extent,and provides a scientific and technical basis for the further highway maintenance.