Research And Implementation Of Road Markings Extraction And Damaged Markings Detection Based On UAV Aerial Image

In recent years,with the rapid development of road infrastructure construction in our country,the number of road maintenance tasks and the difficulty of maintenance are also increasing.Road marking,an important part of road infrastructure,plays an irreplaceable role in driving directions,regulating traffic travel.So it's very important to maintain the road markings.The traditional method of road maintenance mainly relies on manual inspection.This method need a huge amount of engineering and can not save the test results.Although domestic road test vehicles assist the road workers in testing,their high cost and certain requirements on testing roads make them do not get large-scale use.UAV has good flexibility and is not limited by the space on the ground.It also has low cost and easy to use.With the development of UAV aerial photography technology,UAV has good application prospects.In this paper,we propose a method of road markings detection based on unmanned aerial vehicle images.This algorithm can extract road marking based on UAV aerial images and detect the damaged road markings automatically.The main work of this paper are as follows:(1)This paper divides the roads into two types: the road with incomplete lane dividing line on both sides of the road and the road with complete lane dividing line on both sides,Two kinds of road positioning methods of aerial images are proposed respectively for the two types of roads: one is a road positioning method based on the color and texture features of the road and the other is a highway positioning method based on the lane line of the highway(2)After road positioning,this paper uses the contours to extract the candidate area of the road markings,then,uses the deep learning method and the characteristics of the road markings to filter out the non-marking objects,such as vehicles,pedestrians,etc.Eventually we extracted all the road markings.Experiments show that this algorithm has a high accuracy of reticle extraction,(3)In this paper,a variety of damage detection algorithm for road markings are designed to detect damage road markings.For the most of the road markings: lane line,we designed a detection algorithm to detect lane line damage to improve the accuracy of marking damage detection.Experiments show that the proposed algorithm has good accuracy and practical value(4)According to this design algorithm,we design a road markings inspection and maintenance system.The system has been put into use and runs steady.