 |
Chuan Li is a research officer in the Department of Computer Science, University of Bath. He holds a Bachelor Degree in Software Engineering (Zhejiang University, 2005) and a PhD Degree in Computer Science (University of Bath, 2010). He works at the Media Technology Research Center and the Center of Digital Entertainment, both at University of Bath.
His research has mainly focused on developing new techniques for modeling nature scenarios, for example trees and water, using consumer level input devices such a single video camera. He is also interested in procedural modeling of architecture.
Dr. Chuan Li Departmen of Computer Science University of Bath, Bath, UK BA2 7AY cl.chuanli@gmail.com |
 |
||||
Modeling and Generating Moving Trees from Video Chuan Li, Oliver Deussen, Yizhe Song, Phil Willis, Peter Hall
We present a probabilistic approach for the automatic production of tree models with convincing 3D appearance and motion. The only input is a video of a moving tree that provides us an initial dynamic tree model, which is used to generate new individual trees of the same type. We use an approach that fuses global and local constraints to construct a dynamic 3D tree model from a 2D skeleton. Our modeling takes into account factors such as the shape of branches, the overall shape of the tree, and physically plausible motion. Furthermore, we provide a generative model that creates multiple trees in 3D, given a single example model. This means that users no longer have to make each tree individually, or specify rules to make new trees. Results with different species are presented and compared to both reference input data and state of the art alternatives. |
 |
Single Video Based Water Surface Modeling Under review
This project concerns the problem of fully automating the modeling of water surfaces from single viewpoint video ofnatural scenes. Recent advances in this field require dedicated capturing devices and only work in limited conditions. Usingonly a single video, our method produces a model possessing good physical properties. The method is a three-step model fittingprocess. First, shape from shading is used to acquire an initial surface geometry. Second, a model is then fitted to the shape fromshading data. The model’s geometry and velocity conform with the shallow water equation, producing a physically constrainedsurface. Finally, a multi-layer framework is used to further optimize the fit between the model and the data. Comprehensivequantitative and qualitative experiments suggest our method outperforms alternative single video based approaches and isrobust across a wide range of scenes. |
| Publications: | |||
Probabilistic Tree Modeling and Generation from Video. Chuan Li, Oliver Deussen, Yizhe Song, Phil Willis and Peter Hall.
|
|||
Realtime Video Based Water Surface Approximation. Chuan Li, Martin Shaw, David Pickup, Darren Cosker, Peter Hall and Phil Willis. |
|||
Reconstructing Mass-Conserved Water Surfaces using Shape from Shading and Optical Flow. David Pickup, Chuan Li, Darren Cosker, Peter Hall and Phil Willis. Asia Conference on Computer Vision 2010 |
|||
Finding Semantic Structure in Image Hierarchies and Laplacian Graph Energy. Yizhe Song, Pablo Arbalaez, Peter Hall, Chuan Li and Anupriya Balikai. |
|||
ARTcams: Attributed Rational Tensor Cameras. Chuan Li, Peter Hall and Phil Willis. |
|||
Environment Matting into Photographs using Coloured RTcams. Chuan Li, Peter Hall and Phil Willis. |
|||
Learning a Stable Structure to Describe Dynamic Texture. Chuan Li and Peter Hall. |
|||
Colour Constancy based on Model Selection. Chuan Li and Peter Hall. |
|||
RTcams: A New Perspective on Non-photorealistic Rendering from Photographcs. Peter Hall, John Collomosse, Yizhe Song, Peiyi Shen and Chuan Li. |
|||