The Rural Road Information Extraction And Safety Guidance Optimization Design Based On UAV Image

With the continuous development of the social economy,the number of various type of motor vehicle in the countryside has increased.However,due to the mix of various type of motor vehicle,the lack of necessary planning for the transportation network,the insufficient guidance for traffic safety,which has caused all kinds of traffic accident frequently.Especially in the western country,various type of traffic accident have a great impact on consolidating the result of poverty alleviation and the implementation of rural revitalization strategy.How to optimize the setting of the guide signs for rural road and ensure the travel safety of residents play a very important role in achieving poverty alleviation in the western China.Therefore,combined with the dexterity,convenience and speed of Unmanned aerial vehicle(UAV),a village in western China is selected as the research object to study the method of rural road information extraction and safety guidance optimization design.The main research work are as follows:(1)Research on UAV image data acquisition and road feature knowledge base construction methodSelect Liu Chuying Village in Yuzhong of Gansu province as the research object,a UAV image acquisition plan is designed and UAV is used to obtain the image information of the Village.Combined with the remote sensing image feature analysis method,the characteristics of the road sample in the image map ground object is mainly analyzed,the shape parameter factor and spectral parameter factor of the road are used as the main parameters to construct the road feature knowledge base,which is used as the main criterion for road information extraction.(2)Research on road information extraction method based on UAV image dataBased on the comprehensive analysis of common image data segmentation algorithms,a multi-scale segmentation method is selected and the optimal segmentation scale is determined for UAV image segmentation.The geometry,spectrum,texture and other characteristics of the target sample(rural road)are analyzed in terms of parameters such as uniform value,standard deviation,aspect ratio,and compactness.The object-oriented image classification method is studied and the nearest-neighbor classification method is used to develop the object-oriented rural road information extraction process based on the UAV image.(3)Research on road information extraction process and precision impact analysis based on UAV image dataUsing Ecognition to perform image smoothing,image enhancement,image fusion and other preprocessing on the collected image data of Liu Chuying Village,then,segmentation experiments with different scale parameters are used to find the optimal combination of scale parameters for multi-scale segmentation of the image.The nearest neighbor classification based on rule set is used to extract the road information of Liu Chuying Village through multiple optimization process,and used visual interpretation to establish a confusion matrix to obtain the precision index and Kappa coefficient of the classification.(4)Study on the optimal design method of rural road safety guidanceAccording to the driving characteristics of rural road and the investigation analysis of traffic accident,and combined with relevant design standard and specification,the paper researched the optimized design method of traffic safety guidance for rural road and completed the layout of overall road safety facility and warning signs of Liu Chuying Village,achieved 3D visual analysis of key road section and provided important guarantee for the safety of the residents travel.