Research On Key Techniques For Extracting Crack Information On Airport Pavement

With the rapid development of the civil aviation transportation industry,the traffic flow at the airport is increasing day by day.The airport runway is an important area for carrying aircraft to take off and landing.Under the dual effects of aircraft load and natural climate,various diseases are prone to appear on the airport pavement,and cracks are the manifestations of early disease on the airport pavement.At present,the detection of cracks on the airport pavement is mainly manual,and manual detection is subject to human subjective influence and takes a long time.Therefore,under the premise of ensuring airport safety and operational efficiency,how to quickly detect cracks,extract crack information,and monitor and evaluate the development of cracks has become an urgent problem in airport pavement maintenance and management.Based on the above background,this paper takes the asphalt concrete pavement cracks as the research object,researches crack detection and crack geometric feature extraction technologies,and combines image processing technology to construct a set of airport pavement crack information extraction system.The specific content is as follows:(1)Make an airport road surface crack image data set,use histogram equalization,Laplacian,gamma change and other algorithms to adjust the contrast of the pre-collected road surface crack image,and then use image translation The crack image data set is expanded by the method of image rotation and the image rotation,and the data set is provided for the training of the crack extraction model later.(2)The U-net model is used to extract road surface cracks,the U-net model is trained with the produced data set,and the crack information is extracted with the trained U-net model,and the length,width,and area geometry of the subsequent cracks are classified The extraction of characteristic information provides good input.(3)Fracture geometric feature information extraction.For the cracks that have been extracted,the cracks are roughly classified by calculating the number of connected domains,and the projection method combined with the method of curve fitting is used to classify the cracks finely.Extract the skeleton of the crack image,count the total number of pixels in the crack area,and calculate the length,width,and area of the crack respectively.(4)Build a set of airport pavement crack information extraction system,which realizes the functions of airport pavement crack extraction,crack geometric feature extraction,crack damage early warning and other functions.Experimental simulation proves that the U-net model used in this paper can effectively extract crack information.The developed airport pavement crack information extraction system has the functions of crack image data set processing,crack extraction,crack geometric feature information extraction,and crack damage warning.The efficiency and accuracy of the detection of cracks on the airport pavement facilitates airport managers to formulate road surface maintenance plans more scientifically,and provides guidance for the study of other diseases on the road surface.