| Vegetable cultivation has a long history in China.In the long-term natural selection and artificial cultivation,many rich species have been formed,which occupy an important position in China’s crops.The traditional artificial seeding method is inefficient,timeconsuming and laborious,and causes a large amount of seed waste,which greatly affects the development of the vegetable industry.With the introduction of modern agriculture and the urgent need for economic growth,mechanized sowing technique in vegetable production is becoming more and more important.In this paper,a method for obtaining vegetable seed point cloud data based on laser scanning technology was designed.Combined with the developed software system,the three-dimensional feature calculation of vegetable seeds was realized.It can be used not only for the design of rotational the metering tray type holes,but also for vegetable selection and variety identification.A seed test platform was set up.According to the three-dimensional data distribution law of the obtained pakchoi,eggplant and radish seeds,the nozzles of the metering tray suitable for different seeds were designed to obtain the optimal seeding conditions to improve the seeding performance.The main work and results of this paper are as follows:1.A set of vegetable seed point cloud data acquisition system based on laser scanning technology was constructed.The whole system included 3D laser scanner,mechanical platform,program control software,computer,power supply,clamping device and other peripheral equipment,which could completely collect all the point cloud data on the surface of the vegetable seed.2.The filtering preprocessing and three-dimensional feature calculation method of vegetable seed point cloud data were studied.Filter preprocessing included point cloud denoising,smoothing and simplification,wherein the statistical outlier detection algorithm was used for noise filtering;the moving least squares method was used for point cloud smoothing and data resampling;the voxel raster filtering method was used for point cloud reduction.The 3D feature calculation mainly obtained the length,width,thickness,surface area and volume of the seed.The principal coordinate analysis system was used to establish the standard coordinate system.The OOB bounding box was used to obtain the three-axis size of the seed,and then the sphericity was calculated.The surface area calculation adopted the mapping method,and the Delaunay triangulation algorithm was needed to triangulate the model.The calculation method based on grid model projection was used for volume calculation.Ten vegetables,eggplant and radish seeds were randomly selected,and the triaxial dimensions were compared with the manual measurements using the above algorithm to verify the effectiveness of the algorithm.3.The vegetable seed point cloud data processing and 3D feature calculation software were designed on the Windows 7 system,combined with the MFC base class library and the PCL open source point cloud library,with Visual Studio 2010 as the development platform,the point cloud data processing and 3D feature calculation software were developed.The software mainly included point cloud reading and writing module,point cloud filtering module,3D feature computing module and surface reconstruction module.The whole system can realize the adjustability and visualization of point cloud processing process parameters,and the interface is simple and clear and easy to operate.4.The entirety scheme and working principle of the pneumatic-sheave combined vegetable precision metering device was put forward.The replaceable nozzle of the metering tray was designed,and the material of the nozzle was compared,and finally the alloy was used.According to the triaxial size and sphericity distribution of three representative seeds of pakchoi,eggplant and radish,various nozzle-shaped hole design schemes for each type of seeds were proposed.Conical hole,cylindrical hole and through hole were designed for pakchoi and eggplant seeds.Conical hole,cylindrical hole,kidney ellipsoid hole and through hole were designed for radish seeds.Since the shapes of the three seeds are representative,they provide a reference for the design of the metering tray holes of similarly shaped seeds.Due to the interchangeability of the metering suction nozzle,the scope of application is more extensive.5.The vegetable precision seeding test platform was built.The platform mainly included the pneumatic-sheave combined vegetable precision metering device,the high pressure vortex fan,the motor,the frequency converter and the conveyor belt.Adjusting the conveyor speed and the rotational speed of the metering tray can change the seed spacing to simulate field seeding of different seeds.The test factors were the seed type,the rotational speed of the metering tray,the vacuum degree and the type hole,and the test indicators were the qualified rate,the missing rate and the re-broadcast rate,and the single factor and complete combination tests were carried out respectively.By applying the single factor text for vacuum degree,optimal result showed that most suitable vacuum degree for pakchoi,eggplant and radish were 4kPa,3kPa,5kPa,respectively.The optimal metering tray speed and suction nozzle type hole were obtained by the complete combination test.The result showed that when the rotational speed of the metering tray was 17.5～22.5r/min,the metering device qualified rate was higher,especially at 20r/min.The suction nozzle type holes suitable for pakchoi,eggplant and radish were:conical holes,through hole and kidney ellipsoid hole.Under the optimal suction nozzle,vacuum and rotational speed conditions,the qualified rates of pakchoi,eggplant and radish were 97.0%,93.7%and 95.4%respectively,the re-broadcast rates were 2.4%,4.4%and 2.6%respectively,the missing rates were 0.6%,1.9%and 2.0%,respectively,which could completely meet the requirement of seeding index. |
PDF Full Text Request