Map-Guided Interpretation Of Remotely-Sensed Imagery
by Tenenbaum, J. M., Barrow, H. G., Bolles, R. C., Fischler, M. A. and Wolf, H. C.
Technical Note 196
Institution: AI Center, SRI International
Address: 333 Ravenswood Ave., Menlo Park, CA 94025
Note: In IEEE Pattern Recognition and Image Processing 1979.
A map-guided approach to interpretation of remotely sensed imagery is described, with emphasis on applications involving continuous monitoring of predetermined ground sites. Geometric correspondence between a sensed image and a symbolic reference map is established in an initial stage of processing by adjusting parameters of a sensor model so that image features predicted from the map optimally match corresponding features extracted from the sensed image. Information in the map is then used to constrain where to look in an image and what to look for. With such constraints, previously intractable remote sensing tasks can become feasible, even easy, to automate. Four illustrative examples are given, involving the monitoring of reservoirs, roads, railroad yards, and harbors.
|Bolles, Robert C||Program Director|
|Fischler, Martin A||Principal Scientist|
|Tenenbaum, J. M.||Alumnus|
|Wolf, Helen Chan||Alumnus|