Journal
COMPUTERS & GEOSCIENCES
Volume 83, Issue -, Pages 37-45Publisher
PERGAMON-ELSEVIER SCIENCE LTD
DOI: 10.1016/j.cageo.2015.07.005
Keywords
3D GIS; Model simplification; Adaptive partitioning; Vertex clustering; Texture-related
Funding
- National Natural Science Foundation of China [40801163, 41023001]
- Fundamental Research Funds for the Central Universities [2014619020203]
Ask authors/readers for more resources
As an important data source in 3D GIS, 3D landmark models are built to simulate the real-world scenario. However, due to the enormous volume and complexity of 3D models, the data transmission under limited bandwidth and the real-time rendering have always been an open problem. In order to improve the visualization and the efficiency, this paper proposes a novel model simplification algorithm in consideration of texture after the analysis of the existing model simplification approaches. Differing from the previous research, our approach defines a new error metric related to the model texture, which extends the vertex clustering scheme in 3D geometry space and 2D texture space independently. Since the uneven distribution of vertices is taken into account, the clustering unit is divided adaptively in consideration of both geometry and texture information. In view of reducing the memory overhead and improving the algorithm efficiency, we don't create new vertices by iterative calculations, but use the inherent vertices in the initial meshes as the characteristic vertices. To demonstrate the feasibility and effectiveness of our strategy, a series of simplification experiments have been carried out on the platform of DirectX 3D, a widely used 3D application programming interface. The results show that the simplified models in consideration of texture preserve more texture details than those traditional ones. It apparently makes a good balance between the reduction rate and visual effects. (C) 2015 Elsevier Ltd. All rights reserved.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available