JISE


  [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20]


Journal of Information Science and Engineering, Vol. 24 No. 1, pp. 83-98


Physically-Based Analysis and Rendering of Bidirectional Texture Functions Data


Ying-Chieh Chen, Sin-Jhen Chiu, Hsiang-Ting Chen and Chun-Fa Chang* 
Department of Computer Science 
National Tsing Hua University 
Hsinchu, 300 Taiwan 
*Department of Computer Science and Information Engineering 
National Taiwan Normal University 
Taipei, 116 Taiwan


    In pursuit of photorealistic rendering of surface material, Bidirectional Texture Function (BTF) has been used frequently in recent years. Its main drawback is its massive data size. To solve this, the Spatial Bidirectional Reflectance Function (SBRDF) techniques compress BTFs into reflectance model parameters. However, SBRDF cannot produce the self-shadowing and self-occlusion mesostructure effects in many real-world surfaces. Aiming at this drawback, we investigate how self-shadowing and self-occlusion affect the surface appearance by additional physically-based analysis. We mainly rely on two physical phenomena to separate self-shadowing and self-occlusion into two independent effects. First, the self-shadowing is view independent. Second, the self-occlusion is independent of changes in lighting direction. After those analyses, we are able to add self-shadowing and self-occlusion effects to SBRDF to achieve rendering quality that is much closer to the original uncompressed BTF data than SBRDF.


Keywords: real-time rendering, GPU, bidirectional texture functions, self-shadow, selfocclusion

  Retrieve PDF document (JISE_200801_06.pdf)