High-Performance By-Example Noise using a Histogram-Preserving Blending Operator

Heitz, Eric; Neyret, Fabrice

doi:10.1145/3233304

Cited by 37 publications

(65 citation statements)

References 29 publications

(48 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, as highlighted by 10 (a) to (d), our spot noise can also be used as a primitive in a shader-graph to drive other procedural methods, yielding composite textures. Here, the spot noise is used as a texture space partition in conjunction with two high performance noises [26] to achieve a complex appearance relying on our normal map generation for shading purpose. We linearly blend the two procedural textures by using the amplitude of our spot noise as a blending mask and simulate geometric de- tails during shading using the partial derivatives of our model.…”

Section: Resultsmentioning

confidence: 99%

“…To achieve a high level of realism, virtual appearances should exhibit fine non-stationary details at multiple scales. As shown in our work, a promising venue of future research resides in the conjoint use of procedural stationary texture synthesis [26,23] with spatially varying structured patterns generation to create elaborate surface textures, though filtering issues should be carefully considered.…”

Section: Discussionmentioning

confidence: 99%

“…As stated above, by-example noise textures synthesis are a well suited method for fast texture synthesis without heavy memory consumption, though these methods, even Texton Noise [23], are too time consuming during evaluation in order to be used in critical real-time applications. Recently, Heitz and Neyret [26] introduced an efficient method for computing a by-example texture on-the-fly. By combining Patch-Based texture synthesis and triangle lattice evaluation, this noise requires computing and blending of only three contributions per pixels.…”

Section: Procedural Noisementioning

confidence: 99%

See 2 more Smart Citations

Local spot noise for procedural surface details synthesis

Cavalier¹,

Gilet²,

Ghazanfarpour³

2019

Computers & Graphics

View full text Add to dashboard Cite

To deal with the increasing demand for complex visual details in virtual worlds, procedural methods for content authoring are an expanding field in Computer Graphics. Focusing on on-the-fly texture generation, we present in this paper a content authoring process based on Locally Controlled Spot Noise. Through the control of both the impulses distribution and the spatially-defined kernel, this process can cover a wide range of appearances. In this context, we introduce a new kernel formulation that provides an efficient anisotropic filtering of the generated texture. Furthermore, our method allows users to interactively create the desired appearance by controlling both albedo and meso-geometry of the underlying surface, tackling on-the-fly normal map generation. Our method can be used as an artist friendly tool to model high-quality surface details with direct control over the final appearance in real-time.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Procedural Noisementioning

confidence: 99%

See 1 more Smart Citation

Local spot noise for procedural surface details synthesis

Cavalier¹,

Gilet²,

Ghazanfarpour³

2019

Computers & Graphics

View full text Add to dashboard Cite

show abstract

“…To further reduce the repetition effect and strengthen the sense of reality, pre-processing of textures is implemented using the Heitz's method [39] to preserve the color distribution. This process undoubtedly improves the visual effect.…”

Section: Histogram Color Preservationmentioning

confidence: 99%

An Intuitive and Realistic Visualization Method for Seabed Terrain

Xiao¹,

Tan²,

Liu³

et al. 2020

Preprint

View full text Add to dashboard Cite

Visualization of the seabed terrain is one of the key functions of marine geographic information system. The major challenge here is that the obtained data for seabed terrain usually consists of elevation and sediment only but does not include remote sensing images, which is important for ground terrain visualization, as they can be used as textures to reveal detailed information and achieve realistic visual results. Existing seabed terrain visualizing methods (including annotations, color-mapping, and texture-mapping) have limitations in reality and intuition, as they are inadequate to express the physical characteristics of sediment accurately. This paper presents a novel and advanced 3D visualization method of seabed terrain, which introduces the Physically Based Rendering (PBR) theory into the field of geographic visualization. We analyze the main categories of seabed sediments and their optical features respectively, then develop a procedural method to generate physically based rendering materials for different sediments. Then an enhanced bidirectional reflectance distribution function is employed to achieve accurate and intuitive terrain rendering. We also refine the texture sampling method and propose a procedural seabed objects generation method to construct a more natural and realistic undersea environment. Experimental results reveal our method can make good use of the limited seabed terrain data and get significant improvements in visual effect, which can help users to cognize and analyze the seabed geographic environment more accurately and intuitively.

show abstract

“…Unfortunately, existing approaches have not convincingly solved this problem, still presenting temporal coherence problems and traces of the 'shower door' effect. With the development of coherent noise [Kass and Pesare, 2011] and high performant noise synthesis based on textures [Heitz and Neyret, 2018], there might be a possibility to solve this issue in the near future.…”

Section: Future Workmentioning

confidence: 99%

Real-time watercolor rendering of 3D objects and animation with enhanced control

Montesdeoca¹

View full text Add to dashboard Cite

First, I would like to acknowledge everyone whom I have had the pleasure to share quality time with. Without every single contribution that each one of you has made, I would probably not be writing this. I would like to thank both of my supervisors, Prof. Seah Hock Soon and Assoc. Prof. Hans-Martin Rall for their support and guidance. An immense gratitude goes also to my mentor, Asst. Prof. Davide Benvenuti, for insisting on bringing me back to Singapore. Without him, I would probably be freezing in a Nordic country doing something completely unrelated to what I am doing now. I was lucky to count on the support from all three of you, motivating me and giving me the liberty to do exactly what I have always wanted to do. My endless gratitude goes to my 'adoptive' French supervisors Joëlle Thollot, Pierre Bénard and Romain Vergne. I have looked up to you since I started this endeavour and it has been a true pleasure and an unreal learning experience working together with you. I would like to also thank my partners-in-research, Amir Semmo and Alexandre Blerón, whom I have learned a lot from. A special thanks goes to Henriette Peter, for her invaluable company, support, encouragement and care. Without her affection, I would probably be in a very different place right now. To my family, I literally owe you my life. You have molded my personality and removed all boundaries for me to make it this far. Thank you very much for your limitless support and love. I would also like to express my gratitude to the Singapore International Graduate Award (SINGA), the Interdisciplinary Graduate School (IGS) at the Nanyang Technological University (NTU), Singapore, the Multi-pAtform Game Innovation Centre (MAGIC) and the Maverick team at Inria Grenoble Rhône-Alpes for supporting me financially and academically during my studies.

show abstract

High-Performance By-Example Noise using a Histogram-Preserving Blending Operator

Cited by 37 publications

References 29 publications

Local spot noise for procedural surface details synthesis

Local spot noise for procedural surface details synthesis

An Intuitive and Realistic Visualization Method for Seabed Terrain

Real-time watercolor rendering of 3D objects and animation with enhanced control

Contact Info

Product

Resources

About