A Procedural Model for Snake Skin Texture Generation

A Procedural Model for Snake Skin Texture Generation

A Procedural Model for Snake Skin Texture Generation Jefferson Magalhaes˜ Pinheiro1,2 and Marcelo Walter1 1Instituto de Informatica,´ Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil 2Faculdade de Informatica,´ Centro Universitario´ Ritter dos Reis - UniRitter, Porto Alegre, Brazil Keywords: Texturing, Procedural Generation, Texture Synthesis, Snakes, Serpents, Mathematical Biology. Abstract: There are thousands of snake species in the world, many with intricate and distinct skin patterns. This diversity becomes a problem for users who need to create snake skin textures to apply on 3D models, as the difficulty for creating such complex patterns is considerable. We present a procedural model capable of synthesizing a wide range of texture skin patterns from snakes. Our model was derived from a visual assessment of a large number of snakes, and uses image processing as well as cellular automata to generate textures. Our results show good visual similarity with real skin found in snakes. The resulting textures can be used not only for computer graphics texturing, but also in education about snakes and their visual characteristics. We have also performed a user study to assess the usability of our tool. The score from the System Usability Scale was 85.8, suggesting a highly effective texturing tool. 1 INTRODUCTION generate snake skin textures, with the main purpose of using them in computer graphics applications, such Snakes (reptiles of the Serpentes suborder) are numer- as animations and games. We also discuss the advan- ous in species. They are present on every continent, tages of this technique over traditional texturing meth- except Antarctica and some islands (notably Ireland, ods. We have implemented a tool for our model and Iceland, Greenland, Hawaii and New Zealand) (Bau- we have also assessed the usability of this tool. chot, 2006). Some aquatic species live in the Indian and Pacific oceans as well. Being so widespread, it is natural that snakes are typically depicted in digi- tal media – such as computer animations and video 2 RELATED WORK games – where these animals must be digitalized, which means creating a 3D model and their textures. We were unable to find any work in the related litera- When we consider textures, a complicating factor ture that focuses on snakes (or any reptilian) skin tex- comes into play: there are over 3600 snake species, ture generation, to apply on a 3D model and use on according to The Reptile Database (Uetz et al., 2016), a modern renderer. However, this field was already which means a large variety of colors and patterns. studied in the Mathematical Biology field by (Murray Creating such textures manually, from scratch, may and Myerscough, 1991) and (Cocho et al., 1987a). prove to be a very time-consuming task. Extract- These authors discuss how snake skin patterns can ing textures from photos will be difficult because un- be generated using a mathematical model, which is wrapped snake skin textures are not easily found, and precisely the basis for procedural texture generation. collecting a specimen for photography could also be Cocho uses cellular automata, which are further ex- a complicating factor. Procedural texture generation plained on section 4.1.4, and Murray uses reaction- comes as a possible solution to this problem. Another diffusion, which is further explained on section 4.1.5. advantage of procedural generation is that, since the One drawback of these models for our goals is end result is driven by a mathematical formula, cer- that they can generate textures that are visually too tain parameters (such as colors and random seed) can simple to use on many modern applications. For be modified to generate different patterns, and a col- instance, typically only black-and-white images are lection of visually diverse population. In addition, it synthesized, and in the case of Cocho’s work, the res- demands no art skills from the user. olution is extremely low (about a dozen pixels wide). This paper introduces a model to procedurally Regardless, these works serve as a good basis for a 133 Pinheiro, J. and Walter, M. A Procedural Model for Snake Skin Texture Generation. DOI: 10.5220/0006626401330144 In Proceedings of the 13th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2018) - Volume 1: GRAPP, pages 133-144 ISBN: 978-989-758-287-5 Copyright © 2018 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved GRAPP 2018 - International Conference on Computer Graphics Theory and Applications more elaborate (detailed) texture generation method. The first step in our study was therefore to create Other works focus on snakes in computer graph- a categorization of snakes’ visual pattern, and assign ics, but not specifically on texture generation. One each snake species to one of these categories. This of these focuses on snakes (and other materials) iri- will be useful to help determining the pattern synthe- descence (Dhillon et al., 2014). Another two studies, sis method, and also to determine what pattern types (Panagiotakis and Tziritas, 2006) and (Miller, 1988), are mostly present in nature. Since it is very hard to focus on their movement patterns. find images of a snake’s ventral side (the underside, Regarding procedural texturing, there are many or “belly”), and it is typically not visible, we decided works available in literature. Some works worth men- to consider only the visible dorsal side (“back” and tioning include (Turk, 1991) and (Witkin and Kass, sides). Also, some species have a significantly differ- 1991), who also used reaction-diffusion. However, ent pigmentation color or pattern on their head or tail these methods are too generic, and it is unclear how (such as the Apostolepis assimilis, which has no pat- to apply them to our domain (snakes) such that it cov- tern on its body, but its head is black); again, we did ers all, or most of, snake skin patterns. In addition, not take this into consideration. Both topics will be another goal of our procedural model is to be very explored in future work. fast, allowing for an iterative texture creation process, With the help of a herpetologist, we defined the and these methods are too computationally intensive. following six snake pattern categories. Please note Other texture synthesis methods such as (Efros that when we say “body”, we mean only the dorsal and Leung, 1999) and (Gilet et al., 2014) (the latter side of a snake. We show two figures for each cate- even included a snake as an example) rely on a good gory, as examples. amount and quality of sample input images, therefore, 1. No Pattern. The snake shows no pigmentation would be impractical in the case of snakes (due to the pattern, consisting only of a single solid color. difficulty to obtain good photographies for texturing). Note that we do not consider head or tail pig- (Hendrikx et al., 2013) offers a recent survey on gen- mentation, thus the Apostolepis assimilis is cat- erating procedural content for games, including tex- egorized as having no pattern. tures. 3 SNAKES BIOLOGY Snakes are numerous in species and in skin pattern variation. Some have only a single plain color, while Figure 1: Left: Philodryas aestiva (source: (Nogueira, others have patterns composed of two or more colors. ndb)). Right: Apostolepis assimilis (source: (Nogueira, The pattern can be as simple as stripes or spots, or it nda)). can be so complex that it is even difficult to describe. The Serpentes suborder is split into 26 fam- 2. Longitudinal Stripes. The snake shows one or ilies, the most important (numerous) being the more stripes aligned along its body, that go from Boidae (Boas), Colubridae (Colubrids), Dipsadi- neck to tail or from head to tail. dae, Pythonidae (Pythons), Elapidae, and Viperidae (Vipers and Pit Vipers). 3.1 Skin Pattern We are mostly interested in the snake’s visual appear- ance, that is, pattern and colors. When taxonomists Figure 2: Left: Phalotris lemniscatus (source: (Borges- discover a new snake species, they describe, among Martins, 2007b)). Right: Philodryas olfersii (source: other things, their appearance. However, there is not (Sawaya, nd)). an international standard for categorization of snake skin pattern, such as, “if it looks like this then it 3. Transversal Stripes. The snake shows many should be categorized as Spots; if it looks like that stripes aligned perpendicularly to its body axis, then it should be categorized as Horizontal Stripes”. also known as “rings”. These rings may be in two Instead, taxonomists verbally describe the species’ alternating colors, or may be in multiple colors. pattern. Typical examples include corals and false corals. 134 A Procedural Model for Snake Skin Texture Generation Our goal is to assign each species to one category, so that we can determine the relative proportions be- tween different categories. This was done by means of a visual assessment of each species’ photos. 3.2 Scope Limitation Figure 3: Left: Micrurus altirostris (coral, source: (Borges- Naturally, it would be tremendous work to analyze the Martins, 2007a)). Right: Rhinobothryum lentiginosum (false coral, source: (de Albuquerque, nd)). skin pattern of all 3600+ snake species. Scope had to be limited somehow. 4. Spots. The snake shows regular (similar) shapes For this study, we decided to limit the scope ge- on its body, often elliptical. ographically. In the state of Rio Grande do Sul, Brazil, there are dozens of snake species, at least 81 of which have been identified and catalogued in the book (Abegg and Entiauspe Neto, 2012). In this book, these 81 species have been described as accurately as possible, however, in text format (instead of a well- structured data table). It is important to note that although we limited the Figure 4: Left: Liophis miliaris (source: (de Aguiar Pas- scope to the snakes in this specific region, all major sos, 2013)).

View Full Text

Details

  • File Type
    pdf
  • Upload Time
    -
  • Content Languages
    English
  • Upload User
    Anonymous/Not logged-in
  • File Pages
    12 Page
  • File Size
    -

Download

Channel Download Status
Express Download Enable

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

  • Not to be reproduced or distributed without explicit permission.
  • Not used for commercial purposes outside of approved use cases.
  • Not used to infringe on the rights of the original creators.
  • If you believe any content infringes your copyright, please contact us immediately.

Support

For help with questions, suggestions, or problems, please contact us