Study On 3D Scanning And Modeling Method Of Forest Road

In recent years,with the development of "Intelligent Forestry",the requirement of fast and accurate perception technology of forestry operation environment information has become increasingly urgent.Forest roads,such as fire-proof roads and skidding roads,are important components of forest environment,and their detailed 3D information acquisition,perception and model processing have important research value.In forest area operations such as emergency rescue,operations vehicles such as forest fire trucks need detailed,real-time 3D information of paved and non-paved roads.In addition,intelligent equipments in forest area also need road environment information to complete the functions of positioning and path planning in order to improve production efficiency and safety in the future.At present,the domestic and foreign research on the construction of road surface model mainly focuses on urban roads and expressways,while the research on the construction of 3D model of forest roads is less.Vehicle-mounted mobile mapping system has become an important means to collect 3D data of surface features such as expressways,urban roads and buildings on both sides of the road because of its advantages such as high efficiency,fast speed and non-contact.In this paper,a vehicle-mounted forest road information acquisition system was built composed of laser scanner,IMU,GNSS receiver and other sensors,and the process of external parameter calibration and road surface information acquisition was completed.After fusing the collected data,the fast processing method of laser point cloud data was studied,and a forest road model reflecting 3D information of actual road surface was constructed.Firstly,the parameter setting method of road information acquisition system was studied,the positioning model of the system was confirmed,and the external calibration and accuracy verification of the system was completed in this paper.Secondly,the road data of Jiufeng National Forest Park were collected,and the multi-sensor data were fused in MATLAB.The transformation from two-dimensional laser data to three-dimensional point cloud data was realized by coordinate transformation.Finally,the display,filtering,hole repairing,compression and surface reconstruction of 3D road point cloud were realized through authoring software.Especially for holes formed from filtering and occlusion,the method of repairing model was designed based on the feature of scanning line point cloud,which achieved good repairing effect.The results of this study show that this vehicle-mounted road information acquisition system is low cost,and could be used for data acquisition and modeling of forest roads.The final forest road model could reflect the actual road conditions in a way,and can provide the road information for current forest fire truck to realize performance analysis and intelligent Forest vehicles to complete the functions of positioning,path planning and virtual testing in the future.