Design And Implementation Of Pavement Inspection System Based On Image Analysis

With the rapid development of industries such as automobiles and roads in China,the traffic volume and vehicle load on highways are increasing,which causing road surface cracks and seriously affect the bearing capacity and traffic safety of highways.Therefore,early detection of road surface cracks and timely maintenance of road surface cracks can effectively avoid the further development of cracks and maintain the safety of the car driving on the road.However,the road conditions in China are complex,affecting the collection of road surface images.At the same time,pavement image acquisition and processing technologies are not advanced;there are many deficiencies in the process of image acquisition and processing,leading to a low detection accuracy of cracks on the road surface.Therefore,the study of the pavement detection system based on image analysis is of great significance for ensuring the bearing capacity of the road surface and improving the driving safety of the vehicle.This paper studies the pavement crack detection system,and applies image enhancement processing,image segmentation and edge detection,image feature recognition and feature calculation to the detection of pavement crack images.the crack image will produce noise during the acquisition process,so the crack image needs to be smoothed and denoised.In this paper,a direction-based median filter algorithm is used.The experimental image comparison shows that the direction-based median filter algorithm has the crack image noise.Very good inhibitory effect;the filtering algorithm has good effect on the noise of crack image.In order to further identify and count the crack region information,the crack image needs to be segmented and an adaptive Canny edge detection algorithm is used.Then for the disadvantage that it can only manually set the threshold,adaptively combining global gradient generated dynamic threshold,finally,the experiment finds that the adaptive Canny algorithm not only removes the most of the interference,but also improves the accuracy of the segmentation.After the image segmentation is completed,the morphological expansion and corrosion processes are performed on the phenomena of cracks;then the cracks are classified and identified by the crack cavity method and projection method,and the identified cracks are extracted.The computer image processing results are compared with the actual on-site detection results.The comparison results show that the method can obtain satisfactory results,and then the extracted data results are written into the database.Finally,based on the above data,the road surface breakage rate and pavement condition index(PCI)are calculated.Testing the pavement detection system based on image analysis,it is found that the algorithm used in this system can detect and identify image cracks better,obtain crack parameter information,calculate road surface damage rate,and finally provide maintenance advice based on relevant criteria.This system has achieved the desired results.