A compact random-access representation for urban modeling and rendering

Abstract

Figure 1: Left: a city scene with 5,061 distinct buildings represented using our proposed Non-Uniform Textures (NUTs). The size of this digital model is 42 MB, compressible to 10.6MB PNG files. This model is rendered at 49 fps @ 1280 × 720 on an NVidia GTX580 GPU. An equivalent mesh-based model has 671 million triangles (37GB). Right: a much larger European-style city with 40,400 distinct buildings. This model consists of NUTs (229MB) and a complementary mesh (300MB) for roof structures. The total PNG-compressed NUT file size is 60.5MB. It is rendered at 30 fps. An equivalent mesh model has 1.36 billion triangles, or 61GB. AbstractWe propose a highly memory-efficient representation for modeling and rendering urban buildings composed predominantly of rectangular block structures, which can be used to completely or partially represent most modern buildings. With the proposed representation, the data size required for modeling most buildings is more than two orders of magnitude less than using the conventional mesh representation. In addition, it substantially reduces the dependency on conventional texture maps, which are not space-efficient for defining visual details of building facades. The proposed representation can be stored and transmitted as images and can be rendered directly without any mesh reconstruction. A ray-casting based shader has been developed to render buildings thus represented on the GPU with a high frame rate to support interactive fly-by as well as streetlevel walk-through. Comparisons with standard geometric representations and recent urban modeling techniques indicate the proposed representation performs well when viewed from a short and long distance.

A Survey of Algorithmic Shapes

“…Parallel generation is especially useful in an urban context, with scenes with high complexity and detail. The work of LARS KRECKLAU et al used GPU-accelerated generation in the context of generating and rendering highly detailed building facades [45]; the work of ZHENGZHENG KUANG et al proposes a memory-efficient procedural representation of urban buildings for real-time visualization [46].…”

Semantic volume texture for virtual city building model visualisation

“…For example, gradient calculations, environmental lighting and soft shadows can be used to significantly enhance the border effect (Engel et al, 2006); bounding volume hierarchy (BVH), adaptive sparse voxel octree (SVO), out-of-core rendering techniques (Richter & Döllner, 2014) and texture compression can be used to improve memory efficiency (Kämpe et al, 2013;Novák & Dachsbacher, 2012). Recently, a nonuniform texture (NUT) has been reported for modelling and visualisation at an urban scale (Kuang, Chan, Yu, & Wang, 2013). In contrast to their top-to-bottom level division, we are more concerned with semantic integrity in a small-scale environment.…”

Built by Algorithms – State of the Art Report on Procedural Modeling –

“…Parallel generation is especially useful in an urban context, with scenes with high complexity and detail. The work of LARS KRECKLAU et al (2013) used gpu accelerated generation in the context of generating and rendering high detailed building façades; the work of ZHENGZHENG KUANG et al (2013) proposes a memory-efficient procedural representation of urban buildings for real-time visualization.…”