Zilong Dong 董子龙, Sixth year Ph.D. student State Key Lab of CAD&CG, Zijingang Campus, Zhejiang University, Hangzhou, China, 310058 Phone: +86-571-88206681 Fax: +86-571-88206680 Email: zldong (AT) cad.zju.edu.cn Curriculum Vitae: pdf(English),pdf(Chinese)

This work demonstrates that various visually appealing montage effects can be online created from a live video captured by a rotating camera, by accurately retrieving the camera state and segmenting out the dynamic foreground. The key contribution is that a novel fast bilayer segmentation method is proposed which can effectively extract the dynamic foreground under rotational camera configuration, and is robust to imperfect background estimation and complex background colors. Our system can create a variety of live visual effects, including but not limited to, realistic virtual object insertion, background substitution and blurring, non-photorealistic rendering and camouflage effect.

In this work, we are trying to track the camera pose in the natural scenes at realtime framerate. We plan to achieve this by two steps. At the first stage, we will recover the sparse 3D structure of the targeted scenes, and track the camera based on the preprocessed information. At the second stage, we strive to recover the 3D structure of the scene and the camera pos simutaneousely. At present, the first goal has been completed to some extent, and we are moving towards the direction of realtime 3D reconstruction. [project page]

Augmented reality is a technique to fuse virtual and real worlds. The proposed augmented reality system aims to use an off-the-shelf web camera, and augments related virtual media information onto the Chinese-Character-Markers surrounded by black rectangles, which provides a new learning experience for Chinese character. The proposed method enhances the marker detection method with edge extraction, and performs better under complicated shading condition. To reliably solve the camera parameters, the proposed system first uses the homography to estimate the initial values, and then incorporate adaptive smoothness constraint to reduce the image noise. This constraint makes the solved camera parameters more accurate and stable.

ACTS is a automatic camera tracking system which can efficiently and stably recover camera motion as well as 3D scene strucutre from videos and film sequences. It can track all kinds of the camera motion efficiently and stably, which can be rotational or free-move. Especially, the variable focal length can be handled by the system in a robust way. Guofeng Zhang designed the general framework and implemented the kernel. I design the software UI, which iteracts with the tracking core and show the results, and complete most of them except the timeline and play control, which are Yunlong Shao's work. [project page]

In this work, trees are synthesized from samples of scanned range data or normal photographs, based on a statistical model which is two-level Markovian, and a new concept Planton, describing a set of similar samples, is put forth by Rui Wang. I finished the main programming work of the learning and synthesizing algorithm, integrated with the existing multi-view geometry and range data visuliazation applications.

 Since 7/22/2009. Last update on 14/09/2010.