Research On Automatic Leveling Method Of Grain Surface Based On 3d Laser Scanning

The people are the foundation of the country,and grain is the livelihood of the people.Food issues are related to the livelihood of the country,and the security of food supply is an important institutional basis for realizing the national security strategy.Food safety storage management is an important part of the whole grain administrative department and granary macro-control system.In the process management of modern grain long-term storage,grain surface cleaning and leveling is the last important management link in the whole process of grain self-produced harvest processing to normal production and reserve management.The grain surface formed after leveling and cleaning is more conducive to the long-term efficient storage of modern grain resources and timely monitoring of grain situation information.This is also the scientific standardization of granary management must be required.At the present stage,most grain surface leveling is mainly manual work,which has problems such as high labor intensity,low efficiency and serious dust pollution.The existing grain leveling devices have high cost and poor portability,and the identification accuracy rate of grain surface leveling area needs to be improved.Based on this,this paper focuses on the problems such as the accurate recognition and positioning of grain heap topography and the selection of grain leveling path during the research and development of automatic grain leveling device,and establishes an automatic grain leveling method based on three-dimensional laser scanning.Threedimensional point cloud data on the surface of wheat grain pile in Pingfang warehouse was obtained by three-dimensional laser scanning equipment.Three dimensional point cloud data were preprocessed by different filtering algorithms to filter outliers,noise and other interference on the grain heap surface.Furthermore,the three-dimensional point cloud data was regionally segmented based on "grain loading height".The convex and convex topography area of grain heap surface was obtained by European clustering algorithm and the grain heap apex was accurately located in each protruding area.Finally,the surface reconstruction of each topography area after segmentation was carried out to calculate the volume of concave and convex bulk grain heap.Then,the optimal path planning was carried out for the characteristic information such as the location of grain piles in the plain grain area,and the results of the three path planning algorithms under different grain piles were compared.The analysis and comparison showed that when the number of grain piles was less than 10,the dynamic programming algorithm could be used;when the number of grain piles was 10-20,the particle swarm optimization algorithm was better;when the grain pile data continued to increase,The genetic algorithm has better performance,and different programming algorithms are selected according to the quantity of grain pile to ensure the maximum utility of flat grain.The results show that the three-dimensional laser scanning technology combined with multivariate data analysis method is feasible for the accurate location,recognition and volume calculation of the concave-convex topography characteristics of grain pile surface.It can provide an important theoretical basis and technical support for the warehouse operation of automatic grain leveling device,and facilitate the follow-up grain condition detection and measurement inventory.The design of grain routing scheme also provides more options for the subsequent actual grain routing process.