Construction of 2D Multi-Resolution Cylindrical Panoramas from a Hand-Held Camera
Back to Zhigang's Homepage | Computer Science | School of Engineering | CCNY
 
We have developed a system that can automatically construct a multi-resolution 360-degree panoramic representation from image sequences taken by a hand-held camera undergoing rotation and zoom. . Although a simple 2D rigid motion model is used to estimate inter-frame motion parameters, our mosaicing methodology enables precise cylindrical panoramic mosaic. Moving objects are detected and separated from images based on motion information, and their accurate contours are extracted using a modified active contour algorithm. A multi-resolution representation is built for the more interesting areas by means of camera zooming. The mosaic construction method is fast, robust and automatic, taking about 1 second for a pair of 384*288 color images for image registration and mosaicing on a 266MHz PC. No camera calibration, feature extraction, image segmentation or complicated nonlinear optimization processing is required in our algorithms. The construction of the panoramic representation can be used in virtual reality and very low bit-rate video coding.

    Examples

From a video sequence (44.7MB BMP zip file)
To    Conic Mosaic (JPEG , 268KB)
Then To Cylindrical Mosaic (JPEG,  163KB)

Works in both  Indoors (JPEG,  621KB )

and   Outdoors (JPEG, 798KB)

Related Publications:

  • H. Luo, Z. Zhu, G. Xu, C. Xiong, "Image-based Panoramic VR Modeling Techniques", Proc. Of the 9th National Conference on Image and Graphics, Xi'an, China, April 1998 (Best conference paper) (in Chinese).
  • Zhigang Zhu, Guangyou Xu, Heng Luo, Qiang Wang, Automating the Construction of Dynamic and Multi-Resolution 360° Panorama for Natural Scenes with Moving Objects, Proceedings of the IEEE Virtual Reality Conference, Houston, Texas, March 13-17, 1999, pp. 85.
  • Zhigang Zhu, Guangyou Xu, Edward M. Riseman, Allen R. Hanson, Fast Generation of Dynamic and Multi-Resolution 360( Panorama from Video Sequences, Proceedings of the IEEE International Conference on Multimedia Computing and Systems, Florence, Italy, June 7-11, 1999. (icmcs99.ps.gz: 10.7 MB)
  • Zhigang Zhu, Guangyou Xu, Allen R. Hanson and Edward M. Riseman, Fast construction of dynamic and multi-resolution 360-degree panorama from video sequences. Technical Report TR #99-10, Computer Science Department, University of Massachusetts at Amherst, February, 1999. (UMASSTR99-10.ps.gz,  6.9MB)<>
  • <>Z. Zhu. G. Xu, E. Riseman and A. Hanson, Fast Construction of Dynamic and Multi-Resolution 360° Panoramas from Video Sequences. Image & Vision Computing Journal, 2005, in press. 

    • Collaborators:
    Guangyou Xu, Professor, Department of Computer Science and Technology, Tsinghua University, Beijing
    Xueyin Lin, Professor, Department of Computer Science and Technology, Tsinghua University, Beijing
    Heng Luo, MS student, Department of Computer Science and Technology, Tsinghua University, Beijing
    Lin Song, MS student, Department of Computer Science and Technology, Tsinghua University, Beijing
    Qiang Wang, Ph.D. student, Department of Computer Science and Technology, Tsinghua University, Beijing

    Allen R. Hanson, Professor, Department of Computer Science, UMass-Amherst
    Edward M. Riseman, Professor, Department of Computer Science, UMass-Amherst

    Supported by
  • China National Science Foundation Project (No. 69805003), Evolution of Vision Algorithms for the Constructing of Layered Scene Models, 01/99-12/01, Principal Investigator (Zhu).
  • China National High-Tech Project (863 Plan), Image-based VR Scene Modeling and Image Retrieval for Internet/Intranet, 7/1997-6/1998, PI (Zhu)
  • National Science Foundation Project (Grant Number EIA- 9726401), Automatic Interpretation of High-Altitude Image Data for Eco-System Modeling, $1,800,000, 02/01/98 - 01/31/01, PI (Riseman), Co-PIs (Hanson)

    •