 |
 |
 |
 |
 |
 |
 |
 |
English |
| |
!NOTICE! ~Click HERE to visit my homepage v2.0~
About me
I am a second year Master Candidate in the Broadband Network & Digital Media Lab at Dept. of Automation of Tsinghua University.I primarily work on computer vision and video processing for 3D application. Currently I'm busy with project ofcamera self-calibration and image-based modelling.
By the way, as a typical southerner I am, job chances from the Yangtze Delta Area really appeal to me.
Education
Master Candidate
- Sep. 2008 ~ Jul. 2011 (expected)
- My Supervisor: Associate Professor Guihua Er
- Broadband Network & Digital Media Lab, Tsinghua University
Bachelor Degree
- Sep. 2004 ~ Jul. 2008
- Department of Automation, Tsinghua University
Research Interests
- Camera self-calibration and Image-based modeling
- 3DTV System
- 3D for mobile platform
Research Projects
Semi-automatic 2D to 3D conversion using bidrectional tracking
This project presents a novel semi-automatic method for the generation of stereoscopic video from monocular video. Firstly, one key frame is chosen from the monocular video, meanwhile, objects segmentation and depth assignment are done by interactive operation. Then, the objects are tracked based on improved KLT(Kanade-Lucas-Tomashi) feature points tracking algorithm with feature point refining and bidirectional tracking method. The depth image of these non-key frames can be produced using corresponding depth information in key frame. At last, stereoscopic video is generated by the DIBR(depth image based rendering)and 3-D combination with the original monocular frame and depth image. The semi-automatic method improves the quality and efficiency by interactive operation and tracking algorithm, respectively.[pdf] <!--My bachelor thesis (in chinese) -->
Objective evaluation of 3D image
In this project, a novel objective evaluation of depth image based rendering (DIBR) is proposed for the 3D video in format of a monocular video augmented by the gray-scale depth image. The metric is composed of Color and Sharpness of Edge Distortion (CSED) measure. Color distortion measures the luminance loss of the rendered image compared with the reference, and sharpness of edge distortion calculates a depth weighted proportion of remaining edge to the original edge. Comparing to the conventional quality metrics such as MSE and PSNR, our metric represents not only the color artifact but also the synthesis error with above two aspects. Subjective assessment of the different rendering methods is done as well, and the obtained results show significant agreement with our objective metric.[pdf]
Multi-cameras settings for 3D video conference
In this project, I focus on the real-size and no-aliasing performance in 3D video conference system using the multi-camera capture and autostereoscopic multiview dispaly. Firstly I build up two separate systems for capture and display process. The captured disparity and the displayed depth are analysed in detail for corresponding system respectively. Then, I combine two systems together under their relationship and fomula the unified expession for size-scaling and depth-compressing. With the guidance of the two expressions, we can give advice on both settings of camera array and the perfect-peformance area for viewers. [More details...]
Service standard of 3D content on mobile phone
In this project, we try to give a draft proposal of service standard for mobile network operators. Several 3D content services based on current techonology are discussed in the draft. Most modules in the 3DTV system are considered and optimized to adapt for mobile platform.
Camera self-calibration and 3D reconstruction
The project focus on the 3D model reconstruction from a sequence of images taken with comsumer cameras. Neither the camera motion nor the camera settings have to be awared, while after structure from motion and self-calbration process, 3D structure of scene and the motion of camera will be obtained in the euclidean space. In the project I introduce a self-accelerate features matching algorithm which guarantee both precision and computation complexity. Furthermore, I use kruppa equations to calculate the focus length, then nonliear optimization is taken using focuses above as initialization values.
Publications
Conference Papers
- Hang Shao, Xun Cao, Guihua Er, "Objective quality assessment of depth image based rendering in 3DTV system",3DTV Conference: The True Vision Capture, Transmission and Display of 3D Video, 2009
Patents
- 基于双向跟踪和特征点修正的平面视频转立体视频的方法，戴琼海;尤志翔;邵航，公开号：CN101287142
- 一种视频帧序列中物体轮廓跟踪的方法和装置，戴琼海;邵航;曹汛;尔桂花，公开号：CN101477690
Develops
Tools & Platform
- Mostly used: C/C++, Matlab
- Glanced at: Java, Python, PHP, Javascript, HTML, CSS, Batch Scipt.
- OS: Widnows, Linux.
What's been keeping me busy
- Self-accelerate mathcing algorithm for SIFT features.
- Focal length recovery for uncalibrated video sequences.
What's been already Finished
- Objective quality assessment for 3D image generated by DIBR.
- Semi-automatic 2D to 3D conversion using bidirectional tracking.
Honors and Awards
- Tsinghua-Guanghua Fist Prize Scholarship, 2009
- Excellent Graduate of Tsinghua University, 2008
- First scholarship to study progress, 2007
- Second scholarship to study progress, 2006
- Third scholarship to study progress, 2005
- Second scholarship to Freshmen, 2004
Social Skills
- Monitor of AG21(my graduate class), 2009.9~now
- Party branch committee member of AG21, 2009.2~now
- Monitor of DA44(my undergraduate class), 2007.9~2008.7
- Member of Student Union in Department of Automation, 2005~2006
Hobbis
Digging in various network applicatios
I find myself a great interest in tasting what is just released in network applications. My prime motive is to improve my working efficiency and make my life as convenient as possible.
Soccer
I'm a five-year Arsenal F.C. fan and I play soccer as well. Whether you are fond of talking about Arsenal or playing soccer (Elevens, Sevens , Free), please feel free to contact me.
Links
Refer
Broadband Network and Digital Media Lab
Blogroll
Contacts
Address: Room 724, Central Main Building, Tsinghua University,Beijing,China
Zip Code: 100084
Tel: +8610 62788613 EXT.813
E-mail: thushao(at)gmail.com
Wave: thushao(at)googlewave.com ~new~
HomeTown
Jiangshan City locates in the southwest of Zhejiang Province and neighbors to Jiangxi and Fujian Province. I was living in the city for 18 years and left for Beijing in 2004 to attend Tsingha University.
In recent years, I only have two chances every year to return back to my family in Jiangshan. Oh... , I miss it. By the way, once you had a plan to visit to the city, please let me know, I would feel a great honour to paly the host.