Automatic Processing of Airborne Remote Sensing Data for Pattern Discrimination of Jungle and Other Vegetation Areas.
Abstract
Aerochrome Infrared aerial photography taken over Barro Colorado Island in the Panama Canal Zone is the main source for simulating airborne multi-spectral line-scan data of tropical forest areas. With these new data, the variations of forest canopy spatial statistics as a function of angle of view and angle of sun have been studied. Using 50x50 m (on the ground) microdensitometer scan ensembles generated from the aerial photography, and frequency distribution and power spectrum snalyses of the resulting data, the behavior of a young-mature forest at five different angles of view, and a mature and secondary forest at different times of day were studied. Time and angular variations were applied to various tropical vegetation types seen at Tocache, Peru, and illustrated in a 2x12 matrix form, and a preliminary qualitative as well as quantitative analysis has been presented. The authors conclude that the discussed techniques do have considerable merit for automatic vegetation mapping by airborne means, but that extensive further work would be required to produce the data catalogue and algorithms necessary to develop an on-line, operating system. (Modified author abstract)
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1973
- Accession Number
- AD0762583
Entities
People
- Dale E. Mcivor
- Leonard A. Leschack
- William R. Brinley Jr.