Research Of Road Damage Identification And Degree Assessment Technology Based On Unmanned Vehicle

With the leaping development of road transportation in our country, the demands of high level of road maintenance capacity and high grade road from public have become more and more insistent. The fundamental reason for this phenomenon lies in the technologies, which are still traditional in the field of road damage detection and identification. And, the road maintenance strategy which is not scientific enough is also one of the main reasons. So, in order to realize automatic identification of crack of road pavement, reduce labor intensity and danger coefficient of artificial cognition, save maintenance cost and provide scientific basis for the road maintenance department, the technology of damage identification based on unmanned vehicle is researched in this paper. What’s more, the quantitative evaluation of damage degree is also be done.In this paper, unmanned vehicle is used to get a lot of road damage images. According to the characteristics of the crack image, the related algorithms about segmentation and extraction of crack, damage identification and measurement of width, length and area are researched based on the image analysis. First of all, image preprocessing, including image graying, gray transformation, image smoothing and image sharpening, should be done to the images which are gathered by the unmanned vehicle. Then, the algorithm of crack segmentation and extraction based on adaptive sliding window is researched to get binary image of crack. After this, crack can be identified according to geometrical information which should be combined with template matching principle. Subsequently, the related algorithms of morphological image processing are used to do quantitative analysis to crack via calculating the length,width and area of crack. With these results, the algorithm of road damage degree assessment based on fuzzy decision is researched. The effectiveness and stabilization of identification algorithm and degree assessment algorithm can be proved by lots of test result of crack images. At last, the system of road damage identification and degree assessment based on unmanned vehicle is designed and the experimental results could provide reference for road hierarchies and road maintenance decision-making.