High efficient texture optimization method for large scene rendering

TL;DR: In this paper, a high efficient texture optimization method for large scene rendering is proposed, in which data reorganization is performed for large-scale models via texture technology, data of rendering objects are significantly reduced, times for submitting the data to a GPU are reduced, GPU processing pressure is effectively alleviated, and scene rendering efficiency is improved.

Abstract: The invention relates to the technology field of geographic space information systems, and discloses a high efficient texture optimization method for a large scene rendering. The method comprises the follow steps of: (1) performing combining processing of textures corresponding to models to be combined, to from a corresponding texture set; (2) establishing a mapping relation of model data associated with small textures for each small texture in the texture set; (3) sequencing the textures in the texture set; (4) combining several small textures in the texture set according to a texture of a given size, to form several large textures of the given size; (5) interchanging identified model data texture coordinates in the step (3; ) and (6) combining models corresponding to the large textures, to obtain a combined model. The high efficient texture optimization method is advantageous in that, data reorganization is performed for large-scale models via texture technology, data of rendering objects are significantly reduced, times for submitting the data to a GPU are reduced, GPU processing pressure is effectively alleviated, and scene rendering efficiency is improved.