Research On The Application Of Computer Vision Technology In Automatic Detection Of Rural Highway

With the continuous development of China’s road traffic construction and the improvement of rural residents’ requirements for road driving conditions,China has put forward higher requirements for road construction quality.Cement concrete pavement is the main form of highway pavement structure in rural areas of China.It has the characteristics of low cost and convenient laying.It plays an important role in the construction of rural roads in China and is the main mode of transportation in rural areas.Due to the rural highway construction fund to carry out the hard times,the technical force is weak,the management system is not effective to form many factors such as construction quality is generally poor,often appear pavement diseases,therefore,affects the driving comfort and service life,thus to strengthen the understanding of rural highway diseases and treatment strategy of research is particularly important.Paper analyzes the rural highway pavement structure and the disease situation,determine the automated testing object of study for rural highway,this paper proposes the methods of image acquisition,the use of computer vision related treatment methods on the rural highway cement concrete pavement were tested and developed is suitable for automated testing platform of rural highway cement concrete pavement,and the practical application of the project.Papers to stone county in chongqing on the basis of the field survey of rural highway,in view of the common pavement structure of rural roads and common diseases,analyzes the cause of the disease and treatment,determines the test object for pavement width of rural road,fracture,crack,crushing plate and the voids and its image characteristics are introduced,the pavement diseases and specific characteristics of the road width on the digital image,for the correct image feature extraction and analysis laid a foundation.Analyses the characteristics of rural road image acquisition,image is introduced the basic principle of perspective transformation,based on the reality of rural road maintenance funds lack of conditions,on the basis of testing environment of rural road scene detection,using image perspective transformation to a wide range of rural cement concrete pavement image acquisition,achieved the goal that reduce costs and improve efficiency.According to the image characteristics of the common diseases of rural roads,the methods of image enhancement,image edge detection,straight line detection and skeleton extraction are used to realize the automatic detection of the common diseases of cement concrete pavement.The method of image edge detection and the pre-calibration data are used to realize the fast detection of the pavement width of rural highway.According to the image perspective transformation based on the rural road surface image acquisition method proposed in this paper to build a detection platform,designed for the rural cement concrete pavement automation detection platform,including software equipment and hardware equipment.The test platform was applied to test the effectiveness of the method and equipment.The image of rural highway pavement is collected by using the detection platform.According to the image characteristics,the width and crack of the pavement,as well as the fracture of broken plate and plate Angle are detected.The results are compared with those of manual detection,and the effectiveness of the method is verified.