Research in knowledge-based automatic feature extraction

Abstract : This report provides a description of USC's activities as part of the DARPA Automated Population of Geospatial Data-bases (APOD) project. Focus of this effort is on detection, delineation, and 3-D description of buildings from aerial images. Buildings are important in all environments and accurate models of them are needed for tasks such as mission planning and rehearsal, tactical training, damage assessment, and change detection. An overview of a system for modeling buildings using multiple views is given and its performance is quantitatively evaluated. The automatic system is limited to rectilinear-shaped buildings with flat or symmetric gabled roofs, and assumes that ground height, camera models, and illumination parameters are known. Use cues from IFSAR data is proposed. It is shown that such cues can not only greatly improve the efficiency of the automatic system but also improve the quality of the results. The resolution of such data, however, is typically lower than that of electro-optical images and the data may have missing or erroneous elements. An approach that allows a user to assist the automatic system in modeling buildings is described. The approach is designed to be efficient in user time and effort while preserving the quality of the models created. The assisted system also is able to handle multicomponent buildings, buildings having several layers, and buildings having arbitrary shapes. Simple models can be created by only one or two clicks in many cases. Efficient editing of automatically derived models also is possible. Results on images of Ft. Hood, TX, Ft. Benning, GA, and Washington D.C. sites are presented.

[1]  Ramakant Nevatia,et al.  3-D descriptions of buildings from an oblique view aerial image , 1995, Proceedings of International Symposium on Computer Vision - ISCV.

[2]  Ramakant Nevatia,et al.  Detection and description of buildings from multiple aerial images , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[3]  Ramakant Nevatia,et al.  Model Registration and Validation , 1995 .

[4]  Gérard G. Medioni,et al.  Fast Convolution with Laplacian-of-Gaussian Masks , 1987, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[5]  E. Baltsavias,et al.  Automatic Extraction of Man-Made Objects from Aerial and Space Images (II) , 1995 .

[6]  Ramakant Nevatia,et al.  Detection of Buildings from Monocular Images , 1995 .

[7]  Ramakant Nevatia,et al.  Matching Images Using Linear Features , 1984, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[8]  Ramakant Nevatia,et al.  Segmentation and description based on perceptual organization , 1989, Proceedings CVPR '89: IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[9]  Ramakant Nevatia,et al.  Perceptual organization for segmentation and description , 1989 .

[10]  Akira Ishii,et al.  Three-View Stereo Analysis , 1986, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[11]  Ramakant Nevatia,et al.  Detecting changes in aerial views of man-made structures , 2000, Image Vis. Comput..

[12]  Ramakant Nevatia,et al.  Using Perceptual Organization to Extract 3-D Structures , 1989, IEEE Trans. Pattern Anal. Mach. Intell..

[13]  Ramakant Nevatia,et al.  A System for Building Detection from Aerial Images , 1997 .