Loading [a11y]/accessibility-menu.js
3D Asset Size Reduction using Mesh Retopology and Normal Texture Mapping | IEEE Conference Publication | IEEE Xplore

3D Asset Size Reduction using Mesh Retopology and Normal Texture Mapping


Abstract:

There has been a discerning trend over the past few years that increasingly, games have become more and more graphically immersive. A considerable part of this is due to ...Show More

Abstract:

There has been a discerning trend over the past few years that increasingly, games have become more and more graphically immersive. A considerable part of this is due to the shift in trend towards photo-realism. This high detail look is often the main selling point of many video games. Such highly detailed work draws in the player’s attention and makes the game more immersive than ever. However, this also brings in some associated problems like increased asset size, loading and rendering time. Many games load in low detailed models and, based on the player’s field of view and rendering distance, change the level of detail. These games do this to overcome the increased render time and GPU load large size assets have. This paper aims to present a new method to generate a singular 3D model with high enough details and small asset size, eliminating the requirement of constantly switching between the low and high detailed versions of 3D models. The presented approach addresses switching between objects by retopologising the meshes and baking the textures directly onto a singular 3D model, thus reducing the geometry while maintaining the details. When applied to photo scanned 3D models or designer made high poly models, this technique not only preserves the original details of the models but also drastically reduces the asset size. This, in turn, reduces the rendering and loading time. The technique applies to large scale game projects and competitive online games, where loading times and rendering objects at higher distances (without compromising the LOD) are of utmost importance.
Date of Conference: 17-18 December 2021
Date Added to IEEE Xplore: 09 March 2022
ISBN Information:
Conference Location: Greater Noida, India

Contact IEEE to Subscribe

References

References is not available for this document.