International Journal of Computer Vision


Special Issue on Modeling and Representations of Large-Scale 3D Scenes





Guest Editors

Zhigang Zhu (zzhu@ccny.cuny.edu), CUNY City College and Graduate Center

Takeo Kanade (Takeo.Kanade@cs.cmu.edu), Carnegie Mellon University





Manuscript Submission Deadline (Extended): March 19, 2007


All papers will be peer reviewed by following the IJCV reviewing procedures. The final manuscript of each paper will be published online first as soon as it is accepted to this special issue. The final printed volume will appear when all the papers are sent to the Publisher.





Scene modeling and representations of large-scale 3D urban and historical scenes, both dynamic and static, indoor and outdoors, are critical in many military and civilian applications, such as mapping, surveillance, transportation, development planning, archeology and architecture. Research in this area has attracted increasing attention of both academia and industry. Quite a few research groups have major research projects in large-scale 3D scene modeling.  Organizations like Microsoft and Google are trying to include 3D city models in their products (Microsoft Virtual Earth and Google Earth).


In addition to using traditional aerial imagery, there is a growing interest in using both closer-range airborne and ground video/lidar sensors for rapid, accurate and/or realistic 3D modeling of large-scale scenes. Representations are also critical for modeling large-scale 3D man-made urban or historical scenes with fine structures, textureless regions, sharp depth changes, and occlusions. Sometimes, dynamic aspects also need to be modeled and represented for applications such as airborne surveillance, urban transportation, and event analysis. Papers that address these topics are scattered throughout various conferences, workshops, and vision/graphic journals. The proposed special issue will foster focused attention on the latest research results in this interesting and challenging area.


The special issue will focus on the modeling and representation of large-scale 3D scenes, particularly cities, cultural sites, historical scenes, and large-scale dynamic scenes. Researchers are invited to submit papers on methods and systems that lead to solutions to such large-scale problems. Topics of interest include (but not limited to):


q       Scene modeling and visualization using ground and/or aerial video and imagery

q       3D mosaics, layered representations, volumetric representations, and other visual representations

q       3D building detection and modeling from aerial LIDAR and/or imagery

q       3D and moving objects extraction from video of large-scale dynamic scenes

q       Automatic alignment of a large number of large-scale range images

q       View planning and representations for large-scale site modeling

q       2D-3D registration and texture mapping for large-scale 3D scenes

q       Multi-camera approaches for 3D and dynamic scene modeling




Submission Procedure


Papers should be of full journal length with detailed experimental results of 3D modeling and representations of large-scale real scenes, as well as analysis and/or evaluation appropriate for journal publication. Submissions should follow the guidelines set out by IJCV. Authors should submit their papers via the IJCV website




Please choose the Article Type "Special Issue - 3D Scenes".