A method and system for constructing a homogeneous forest scene

Abstract

The invention discloses a homogeneous and heteromorphic forest scene construction method and a homogeneous and heteromorphic forest scene construction system. The method comprises the steps of collecting a 3D (three-dimensional model) of a single tree; skeletonizing the model by use of a circle cutting method to obtain skeletonization data; layering the model; calculating intrastratal morphological description parameters and transmission rule parameters between adjacent layers; calculating a Gaussian distribution value and a variance of each intrastratal morphological description parameter and each transmission rule parameter between the adjacent layers; carrying out stochastic disturbance on the intrastratal morphological description parameters and the transmission rule parameters between the adjacent layers according to the Gaussian distribution values and the variances of the intrastratal morphological description parameters and the transmission rule parameters between the adjacent layers to obtain disturbed parameter values; generating tree trunks in different postures according to the disturbed parameter values, and forming a module set of the model by use of the tree trunks in different postures; and selecting branch and trunk combinations in the module set to generate trees in a forest scene. According to the homogeneous and heteromorphic forest scene construction method and the homogeneous and heteromorphic forest scene construction system, automatic and rapid modeling can be carried out on a large-scale forest, and the light weight degree and the rendering efficiency are greatly improved.

Description

technical field [0001] The present invention relates to the field of scene construction, in particular to a method and system for constructing a homogeneous forest scene. Background technique [0002] Forests are common scenes in nature and play an indispensable role in various outdoor large-scale scenes. In the computer field, with the continuous development of technologies such as virtual reality and 3D games, users' demand for the realism of virtual scenes is also increasing day by day, which leads to an increasing demand for building large-scale virtual visualization forest scenes. However, due to the variety of shapes and complex structures of trees, the difficulty and cost of modeling are greatly increased; at the same time, due to the large number of trees in the forest scene, the amount of data input by the system is very large, resulting in a significant drop in rendering efficiency. It is difficult to get fluent expression in the application. This is contrary to ...

AI Technical Summary

Problems solved by technology

[0004] (1) Long modeling cycle: Due to the complexity of the structure of the tree itself, it often takes a long time for modelers to build a batch of large-scale forest scenes with various shapes

[0005] (2) High cost: Modelers need a long time to learn and understand the meaning of each parameter in the tree modeling rules and the operation technology of tree modeling software, that is, the learning cost is high

It even requires a lot of time-consuming and labor-intensive artists to complete the task of forest modeling, which will inevitably lead to higher modeling (human) costs

[0006] (3) Poor diversity: At present, both games and VR use the reuse method of the whole tree model to generate large-scale forests. Although it is light enough, the diversity of tree forms is seriously insufficient, so its beauty and reality Gender is very lacking

[0007] (4) Rendering calculation is slow: Although the lightweight tree modeling based on reuse reduces the overhead of network transmission, when rendering on the web client, each tree model can only be rendered with an independent dense polygon patch. said that the amount of data and calculation required for rendering has not been reduced, so the efficiency of online rendering of Web3D large-scale forests is still very low, and the experience for users is also poor.

Images