Fast texture synthesis using tree-structured vector quantization
Proceedings of the 27th annual conference on Computer graphics and interactive techniques
Method of direct texture synthesis on arbitrary surfaces
Journal of Computer Science and Technology - Special issue on computer graphics and computer-aided design
ACM SIGGRAPH 2007 courses
Image restoration with broken curve prediction
International Journal of Computational Vision and Robotics
Hi-index | 0.00 |
Texture analysis and synthesis is very important for computer graphics, vision, and image processing. This paper describes an algorithm which can produce new textures with a matching visual appearance from a given example image. Our algorithm is based on a model that characterizes textures using a non-linear deterministic function. During analysis an example texture is summarized into third function using tree structure vector quantization. An output texture, initially random noise, is then sinthesized from this estimated function. Compared to existing approaches, our algorithm can efficiently generate a wide variety of textures. The effectiveness of our approach is demonstrated using standard test images from the Brodatz texture album.