Dynamic LOD optimization for Unity models


foreword

  The previous section mainly described the idea of ​​dynamically generating the sphere model and the generation of the icosahedron. In this section, we will talk about how to realize it. Because if a 6371km sphere with a certain resolution is generated in the future, if each vertex is subdivided and calculated, the number of such vertices will undoubtedly be an astronomical figure. Such a huge overhead is on the current mainstream PC equipment. It's almost impossible to run smoothly. And the number of vertices of a single model in unity cannot exceed 65535, so multiple models must be used to form a sphere, and the management of these models is also a relatively large performance overhead. Therefore, LOD technology is used here. Without affecting the overall effect, the higher the degree of detail, the denser the vertices will be for the part close to the player; otherwise, the lower the degree of detail, the sparser the vertices will be.


About LOD technology

  LOD technology (Level Of Detail) refers to the technology of representing the same object with several models of different complexity. This technology mainly calls models of different complexity according to the change of the position of the object from the viewpoint, that is, the low-complexity model is called when it is far away, and the high-complexity model is called when it is close. Although the use of this technology will lose a certain degree of fidelity, it can reduce the number of polygons drawn in the scene and save system resources. Compared with its improved performance

Guess you like

Origin blog.csdn.net/weixin_44286745/article/details/126669045
Recommended