Bidirectional Texture Function Modeling

Michal Haindl

Jiří Filip

Institute of Information Theory and Automation of the AS CR

Half-day tutorial at CVPR 2010
Sunday, June 13, 2010 PM (details).


     Robust visual classification, segmentation, retrieval or view/illumination invariant methods dealing with images of textured natural materials, as well as augmented reality applications creating virtual objects in rendered scenes with real material surface optical properties, require realistic physically correct textures. Such texture representation considerably depends on the view and illumination directions and can be efficiently and the most accurately obtained in the form of rough surface textures represented by Bidirectional Texture Function (BTF).

BTF Measurement

Data Modelling

(light circling in front of the object)

     BTF data are the most advanced and accurate digital representation of a real-world material visual properties to date, and their analysis provides abundant information about the measured material that cannot, for the majority, be attained using any alternative visual measurements or representations, e.g., image based relighting, bump/displacement mapping, spatially varying BRDFs, etc. The nature of BTF data allows their straightforward exploitation for design and testing of illumination and view-invariant features and algorithms in numerous robust texture classification, segmentation and retrieval applications. Other image processing problems, such as image restoration, aging modeling, face recognition, security, 3D object recognition, content-based image retrieval and many other tasks can and should benefit from BTF comprehensive information. Additionally, applications of this advanced texture representation allow accurate photo-realistic material appearance approximation for such complex tasks as visual safety simulations or interior design in automotive/airspace industry, architecture or dermatology among others.
     Multispectral BTF is a seven-dimensional function that depends on view and illumination directions as well as on planar texture coordinates. BTF is typically obtained by measurement of thousands of images covering many combinations of illumination and viewing angles. However, the large size of such measurements has prohibited their practical exploitation in any sensible application until recently. During the last few years the first BTF measurement, compression, modeling and rendering methods have emerged.
     In this tutorial we show benefits of using BTF in common applications and explain principles of BTF measurement, modelling and visualization methods.

Detailed Outline

  • Introduction, motivation, texture definition (smooth and rough - BTF) and categorization, texture modeling, evaluation criteria.

  • Basic texture modeling approaches - sampling, intelligent sampling, procedural models, adaptive models - principal differences.

  • Bidirectional Texture Function: principle, measurement methods.

  • BTF analysis and compression.

  • BTF modeling - Markovian models, mixture models, reflectance and sampling approaches.

  • Modeling quality verification, visual perception.

  • BTF visualization.

  • Applications in computer vision and computer graphics.

    Relevant Literature

  • Filip J., Haindl M.:Bidirectional Texture Function Modeling: A State of the Art Survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 31, no. 11, pp. 1921-1940, Oct. 2009

  • Filip J., Chantler M.J., Haindl M.:On Uniform Resampling and Gaze Analysis of Bidirectional Texture Functions. ACM Transactions on Applied Perception, vol. 6, no. 3, Article 18, August 2009, 15 pp.

  • Filip J., Chantler M.J., Green P.R., Haindl M.:A Psychophysically Validated Metric for Bidirectional Texture Data Reduction. ACM Transactions on Graphics 27(5) (proceedings of SIGGRAPH Asia 2008), Article 138, December 2008, 11 pp.

  • Haindl, M., Filip J.: Extreme Compression and Modeling of Bidirectional Texture Function. IEEE Transactions on Pattern Analysis and Machine Intelligence (IEEE TPAMI), IEEE Press, Volume 29, Issue 10, October 2007, pp.1859-1865.

  • Haindl, M., Hatka, M.:BTF Roller. Proceedings of TEXTURE 2005, Heriot-Watt University, October 2005, pp.63-66.

  • Haindl, M., Hatka, M.:Near-Regular Texture Synthesis. Proceedings of CAIP, LNCS Volume 5702, Springer, 2009, pp. 1138-1145.

  • Haindl, M., Havlicek, V., Grim, J.:Probabilistic Discrete Mixtures Colour Texture Models. LNCS Volume 5197, Springer, 2008, pp. 675-682.

    Target Audience

    The tutorial will start from the basic principles and will build on the fundamentals introduced to discuss the latest techniques for texture modeling in the literature. It will, therefore, be suitable for newcomers to the field of computer graphics and computer vision, as well as practitioners who wish to be brought up to date on the state-of-the-art methodology of texture modeling.


    Michal Haindl

    graduated in control engineering from the Czech Technical University (1979), Prague, received PhD in technical cybernetics from the Czechoslovak Academy of Sciences (1983) and the ScD (DrSc) degree from the Czech Technical University (2001). He is a fellow of the IAPR and professor. From 1983 to 1990 he worked in the Institute of Information Theory and Automation of the Czechoslovak Academy of Sciences, Prague on different adaptive control, image processing and pattern recognition problems. From 1990 to 1995, he was with the University of Newcastle, Newcastle; Rutherford Appleton Laboratory, Didcot; Centre for Mathematics and Computer Science, Amsterdam and Institute National de Recherche en Informatique et en Automatique, Rocquencourt working on several image analysis and pattern recognition projects. In 1995 he rejoined the Institute of Information Theory and Automation where he is head of the Pattern Recognition department. His current research interests are random fields applications in pattern recognition and image processing and automatic acquisition of virtual reality models. He is the author of about 250 research papers published in books, journals and conference proceedings.

    Jiří Filip

    received the MSc degree in technical cybernetics and the PhD degree in artificial intelligence and biocybernetics from the Czech Technical University in Prague in 2002 and 2006, respectively. He is currently with the Pattern Recognition Department at the Institute of Information Theory and Automation of the AS CR, Praha, Czech Republic. Between 2002 and 2007, he was a researcher at the Institute of Information Theory and Automation, Academy of Sciences of the Czech Republic. Between 2007-2008, he was a postdoctoral Marie Curie research fellow in the Texture Lab at the School of Mathematical and Computer Sciences, Heriot-Watt University. His current research is focused on analysis, modeling, and human perception of high-dimensional texture data and video sequences.

    Last update 02/02/2010