The Method Of Constructing Tree 3d Model Based On Sequence Image

The research of tree and plant modeling has been widely used in agriculture and forestry,ecology,virtual reality,3DGIS and other fields.Due to the complex morphological and structural characteristics of tree branches,there are large differences among different tree species,and the branches of trees are generally blocked by leaves,which makes it very difficult to establish a three-dimensional model of trees.Nowadays,popular modeling software,such as 3DMax or Maya,can be used to model trees.However,for real trees in the real world,such software requires good art skills of tree model designers,and the models are too monotonous to conform to the objective laws of real scenes.Therefore,based on computer graphics,this paper proposes the concept of visual 3d panoramic model.Then,in order to better display the topological structure information of trees,this paper converts the sequence images into point cloud data,and on this basis,carries out point cloud separation and classification,and extracts the skeleton line information of tree branches.Finally,based on the obtained skeleton lines,this paper conducts the solid modeling and artistic rendering of tree branches.The research contents and conclusions of this paper mainly include:(1)The establishment method of the visual 3d panoramic model of trees.Aiming at the problems of existing tree modeling methods,this paper proposes a tree visualization method for visual 3d panoramic model.However,due to the special growing environment of trees,there will be background interference factors in the collected sequence images.Therefore,this paper proposes a background subtraction algorithm based on the tree image,and designs a three-dimensional panoramic model of trees based on this,and the storage space of the model is also small.(2)Skeleton line extraction method of tree branches.Recovery algorithm based on image sequences,the use of motion structure tree three-dimensional point cloud data obtained,and the trunk on point cloud point cloud from the leaves of the point cloud,and then to classify point cloud to extract the skeleton of the separated points,the use of the principle of the shortest distance,the skeleton points to connect,reduction of trees topology information.Compared with the existing methods,the tree skeleton line extraction method in this paper has better topological connectivity,and the quality of the obtained skeleton line is also relatively high.(3)Art rendering method of tree branches.Using the ideas of lofting,the trees based on the classification of point cloud,this paper extracted the biggest plane in each kind of point cloud bounding box,and for each section of the cross section at the ends of the branches,and then based on skeleton line lofting path,lofting processing,and the physical model of trees,and then to apply colours to a drawing of entity model,and add artistic effects,makes some artistic value of the tree didn’t show a better effect.Research results show that the method not only can use ordinary camera series images to achieve 3 d visual effect,trees and extraction of trees skeleton line information can show tree structure more complete information,but also to artistic rendering of trees,to some realistic art value of trees to improve more,be able to add the 3 d scene reality colorful tree model,it has extensive and practical application value.