Detection And Evaluation Of Cement Concrete Pavement Pothole In Farm And Pastoral Areas

As the main distress of cement concrete pavement,potholes are a great threat to safe driving.Therefore,timely and accurate pothole detection is the key to formulating a reasonable maintenance strategy and ensuring driving safety.At the present stage,the inspection of potholes in farm and pastoral areas is still dominated by manual surveys,but there are many shortcomings in manual inspections,and it can no longer meet the current maintenance work.To overcome the limitations of manual detection,computer vision-based automatic evaluation of pavement distress has gradually become the focus of road maintenance due to its high operating efficiency and good implementation safety.Therefore,based on image processing technology,this paper conducts the following research on the detection of potholes in cement concrete pavement:(1)A pothole segmentation method based on image texture and gray features was established.This method mainly adopts four image processing techniques:texture filtering,gray level,morphological operation and maximum connected domain to achieve complete extraction of the pothole area.According to the overlap rate of the extracted potholes,images with an overlap rate of more than 90%account for 76.8%of the total image,and images with an overlap ratio of more than 80%account for 94%of the total image.Compared with other existing methods,this method has significant advantages in segmentation effect and processing efficiency.(2)Aiming at a data set containing multiple distress types,a pit recognition algorithm based on LIBSVM was proposed.The comprehensive experiments for recognition of potholes show that the recall,precision and F1-Score achieved are 100%,97.4%and 98.7%,respectively.Compared with other existing SVM-based recognition methods,this method has higher accuracy.(3)A method of evaluating pothole damage using fractal dimension D and damage rate DR was proposed.The results showed that when 1.26?D?1.57&1.39<DR?9.71,the pothole is slightly damaged;when 1.57?D?1.77&9.83?DR?18.09,the pothole is moderately damaged;when D?1.77&DR?18.43,the pothole is severely damaged.The evaluation method can be used to easily,quickly and accurately discrimination of pothole damage degree of cement concrete pavement.(4)In order to provide an image containing a complete pothole area,a splicing method based on image features was proposed.The detection of the stitched images shows that it is possible to perform pothole detection and damage determination directly on long images,so as to accurately detect the pothole and eliminate the interference of redundant information.(5)According to the different forms of the existing potholes,the angular second moment and entropy are used to divide the potholes into two types:ordinary and special.Through analysis of the causes of these two types of potholes,it was found that there were several factors including large aggregate mud content,irrational water-cement ratio design,water spraying during construction,and repeated impacts of driving loads.On this basis,corresponding preventive measures and treatment measures are proposed,which provide a reference for the decision of pit maintenance.