Using Dimensionality Reduction to Create New Materials from Tabular BRDFs

Nunes, Mislene da Silva; Miranda, Gastão Florêncio; Andrade, Beatriz Trinchão

doi:10.1109/sibgrapi.2017.37

Cited by 4 publications

(2 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Aiming to create new materials based on tabular BRDFs, Nunes et al [20] presented an approach that uses PCA to generate a low-dimensional space of BRDFs. Besides, they used Delaunay triangulation to provide a mesh of points, in which these points represent the materials of the low-dimensional BRDF space.…”

Section: Related Workmentioning

confidence: 99%

“…Thus, there are few tabular BRDF databases estimated from real-world materials, and they usually contain a relatively small amount of materials (e.g., 100 [5], 150 [16], and 62 [17] materials). To avoid long acquisition processes, approaches that take advantage of existing databases to create new materials based on tabular data are emerging [5], [18]- [20]. Besides, it is possible to personalize the new materials by choosing desired features about the appearance of materials or editing them [5], [19], [21].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Creation of Materials from Tabular BRDFs

Nunes¹,

Miranda²,

Andrade³

2022

Anais Estendidos Do XXXV Conference on Graphics, Patterns and Images (SIBGRAPI Estendido 2022)

View full text Add to dashboard Cite

Rendering materials with a realistic appearance requires considering how they interact with the light. Bidirectional Reflectance Distribution Functions (BRDFs) are often used to achieve this goal. There are different ways to represent materials from BRDFs, which include tabular BRDFs, analytical models, and linear combinations of a BRDF database. In the last decade, the search for more realism in rendering increased the interest in using tabular BRDFs. However, this approach requires a long acquisition process and high storage space. This master dissertation proposes a pipeline to create new materials from a tabular BRDF database. During this process, we also explored two related topics: we compiled and proposed techniques to evaluate BRDFs, and developed an approach to preprocess and cluster a BRDF database. These researches presented insights and contributions that are useful for contexts other than ours and provided analysis that reinforced our choice of techniques to reach our goal. As a final result, our method creates new materials with realism and consistency.

show abstract

Section: Related Workmentioning

confidence: 99%