BUildings Detection and Description System

Chung-An Lin and Ramakant Nevatia
Institute for Robotics and Intelligent Systems
Department of Computer Science
University of Southern California
Los Angeles, California




Previous Work:



  1. Hypothesize roof candidates.
  2. Select good hypotheses based on image evidence.
  3. Verify hypotheses based on wall and shadow evidence.

Most Recent Results:

Modelboard Results: (09/15/95)

Demo Movies of Modelboard Results: (12/14/95)

Fort Hood Results: (12/06/95)

Evaluation & Analysis:

Interactive System: (USC-BUDDIES)


"Buildings Detection and Description from Monocular Aerial Images," Chungan Lin and Ramakant Nevatia, ARPA Image Understanding Workshop 1996.
"3-D Descriptions of Buildings from an Oblique View Aerial Image," Chungan Lin and Ramakant Nevatia, IEEE International Symposium on Computer Vision, 1995, pp377-382.
"Detection of Buildings from Monocular Images," Chungan Lin, Andres Huertas and Ramakant Nevatia, Ascona Workshop, 1995.
"Detection of Buildings Using Perceptual Grouping and Shadows," C. Lin, A. Huertas and R. Nevatia, IEEE Proceedings of Computer Vision and Pattern Recognition 1994, pp62-69.
Detection of Buildings from Monocular Views of Aerial Scenes using Perceptual Grouping and Shadows, A. Huertas, C. Lin and R. Nevatia, ARPA Image Understanding Workshop 1993, pp 253-260.