In recent years, the comprehensive advancement of new basic surveying and mapping, digital twins and other major projects has put forward further requirements for 3D data formats, such as higher performance, more realistic rendering, and multi-level semantic storage.
In response to these requirements, SuperMap upgraded the Spatial 3D Model (S3M) data format from 2.0 to 3.0, which has breakthroughs in loading performance, high-fidelity rendering, semantic information storage, extensibility and compatibility.
High performance
In order to improve data loading and reading performance, S3M 3.0 supports geometric compression (Meshopt) and texture compression (KTX 2.0), which drastically reduces the amount of data and improves data reading performance, so as to meet the needs of desktop, web, mobile and game engine and other multi-terminal applications; it also supports OBB bounding boxes to achieve more accurate cutting and scheduling switching, improving data download and loading performance; and a new LOD switching mode Geometric_Error, easy to adjust LOD level switching to adapt to different display environments demand, is added.
Tests have shown that it can effectively reduce data capacity and memory usage, and improve the loading performance by about 1 time.
High-fidelity rendering
In terms of high-fidelity rendering, the physics-based material (PBR) in S3M 3.0 adds a new mask texture, which supports data mapping game engine material functions (specify different material functions for the four channels of the mask texture), so as to realize three-dimensional high-fidelity rendering of geographic scenes. As shown in the figure below, S3M 3.0 specifies the self-illuminating material function for the glass material in the game engine, to realize the expression of the night scene effect of real-world 3D data.
PBR materials also add semantic textures, which support the storage of various semantic classification information such as walls, structural columns, windows, doors, and bases, to realize the expression of data pixel-level semantic information.
S3M 3.0 also supports skeletal animation to provide richer scene animation effects.
▲ Specify a self-illuminating material function for the glass material in the game engine based on the mask texture to exhibit the night scene effect
Richer and more levels of semantic information
S3M 3.0 supports the storage of more dimensional vertex attributes and the storage of texture pixel-level semantic information. As shown in the figure below, the results of submersion analysis are stored in multi-dimensional vertex attributes, and the flooding effect that changes over time is dynamically shown.
▲ Store the results of flooding analysis in the multi-dimensional vertex attributes, and dynamically exhibit the flooding effect
More extensible and compatible
S3M 3.0 provides a more extensible and compatible format, supports extended attributes, semantics, etc., and can be conveniently attached to geographic entity semantic information.
In the future, SuperMap will continue to improve S3M 3D data format according to the development of information technology and market demand, to better meet the transmission, exchange, efficient sharing and interoperability of massive 3D spatial data in different terminals in the online and offline environments.
