Three-dimensional Point Cloud Processing Technology And Calculation Method Of Plant Type Parameters Of Corn Canopy

In the process of visual expression of maize morphological structure,the problems such as the redundancy,noise,loss and traditional measurement of the 3D point cloud data acquisition are complicated and the error is large.In this research,the FastSCAN 3D scanner was used to obtain the 3D point cloud data of corn,and the coronal solid dynamic reconstruction was realized through the point cloud simplification and denoising.Furthermore,the morphological parameters were calculated.Firstly,the obtained three-dimensional point cloud data is simplified by using the three-dimensional point cloud adaptive curvature method based on grille method.Secondly,it USES the method of direct pass filtering,large scale de-noising and bilateral filtering to enrich and reduce the noise in a single region.Again,Geomagic Studio was used to reconstruct the packaged corn model.Finally,the characteristic parameters such as height of maize crown,width of plant and thick stem were obtained by fitting ball and fitting cylinder.The main research contents are as follows:(1)In view of the complexity of 3D point cloud of maize canopy,the curvature is relatively high.In this paper,we propose a method to simplify the three-dimensional point cloud adaptive curvature based on grille method.The method of grating method and random method were used to simplify the point cloud of maize stalks and soil.The results showed that in the early stage of maintaining the original point cloud morphological features,maize leaves,stalks and soil compactness were 32.34% respectively.49.81%;84.98%,the total reduction rate is 48.30%.(2)The effect of different scale noise on the reconstruction of 3D model is discussed.In this paper,based on improved maize morphological characteristics information of 3D point cloud denoising method,based on the 3D point cloud model of corn two-step denoising and adopt different methods to filtering experiments and verification,the average error,maximum error,the standard error of 0.0254,0.1395,0.0238,respectively.The results show that the method denoising is the best,and the de-noising model can maintain the rich details of maize and reduce the surface smoothness of maize leaves caused by bilateral filtering.(3)In the traditional method,the coronal structure and the actual measurement error of corn plant type parameters were calculated.In this research,the characteristic parameters of corncanopy are calculated by fitting ball and fitting cylinder.The average height of maize was0.0700~0.7164 m,and the average plant width was 0.0936~0.6033 m,and the mean stem diameter was 0.3027~2.4413 cm.The regression analysis of maize plant height,plant width and stem thickness was carried out by direct measurement and scanning.The results show that this method can accurately and rapidly obtain the plant parameters of corn canopy,and provides a good theoretical foundation and technical support for studying the light distribution and production practice of maize canopy.