Task 2: Yellowstone National Park Data, 1398
There are no author-identified significant results in this report.
Engineering topics
Publications and source records attributed to Thomson, F. J..
There are no author-identified significant results in this report.
The author has identified the following significant results. Results of crop mensuration and recognition studies in the Sacramento Valley indicate that the measurement of acreages of rice fields from ERTS-1 data can be dramatically improved by utilizing a processing technique which estimates the proportion of rice and bare soil in field boundary elements classified as corn by conventional pattern recognition techniques.
Crop species recognition and mensuration in Sacramento Valley from ERTS-1 data
Terrain classification maps of Yellowstone National Park
There are no author-identified significant results in this report.
There are no author-identified significant results in this report.
The goal of the second recognition map was to delineate various crop species in a portion of the Sacramento Valley, and at the same time to determine how accurately each could be classified and measured from ERTS-1 data. The new recognition map, a new and concise display of the old map, and classification and mensuration accuracy data are presented and discussed. The mensuration accuracy, in particular, is affected by the presence of an edge effect one resolution wide surrounding nearly all fields. Points on the edge are misclassified because they contain a mixture of, crop and bare soil. Using a processing technique to estimate the proportions of unresolved objects in this edge region, a much improved mensuration capability will be demonstrated.
A cooperative ERTS-1 investigation involving U. S. Geological Survey, National Park Service, and Environmental Research Institure of Michigan (ERIM) personnel has as its goal the preparation of terrain classification maps for the entire Yellowstone National Park. Excellent coverage of the park was obtained on 6 August 1972 (frame 1015-17404). Preliminary terrain classification maps have been prepared at ERIM by applying multispectral pattern recognition techniques to ERTS-MSS digital taped data. The color coded terrain maps are presented and discussed. The discussion includes qualitative and quantitative accuracy estimates and discussion of processing techniques.
There are no author-identified significant results in this report. Remotely sensed multispectral scanner and return beam vidicon imagery from ERTS-1 is being used for: (1) water depth measurements in the Virgin Islands and Upper Lake Michigan areas; (2) mapping of the Yellowstone National Park; (3) assessment of atmospheric effects in Colorado; (4) lake ice surveillance in Canada and Great Lakes areas; (5) recreational land use in Southeast Michigan; (6) International Field Year on the Great Lakes investigations of Lake Ontario; (7) image enhancement of multispectral scanner data using existing techniques; (8) water quality monitoring of the New York Bight, Tampa Bay, Lake Michigan, Santa Barbara Channel, and Lake Erie; (9) oil pollution detection in the Chesapeake Bay, Gulf of Mexico southwest of New Orleans, and Santa Barbara Channel; and (10) mapping iron compounds in the Wind River Mountains.
There are no author-identified significant results in this report.
There are no author-identified significant results in this report.
There are no author-identified significant results in this report.
There are no author-identified significant results in this report.
There are no author-identified significant results in this report.
The radiances in two thermal infrared channels of an airborne scanner system were ratioed to produce images that recorded compositionally diagnostic emittance variations for several silicate rock types near Pisgah Crater, California. The images demonstrate that the ratio method is capable of enhancing emittance variations in the presence of temperature extremes that differ by no more than 25 C, with no temperature corrections.
The feasibility of solving practical multispectral data reduction problems by available processing techniques and the identification of areas for more technical development requirements are considered. Parameter mapping and ratio mapping techniques prove their usefulness in a corn blight watch experiment. A hybrid computer processing system is advocated for complete realization of aircraft and spacecraft sensor potentials.
A method is presented which can be used to map silicate rock-type from aerial infrared data. The method has been partially tested over a sand quarry at Mill Creek, Oklahoma, in which highly siliceous targets were discriminated from nonsilicates in the scene. The technique is currently being tested experimentally on basic silicates. On the basis of the Mill Creek results and theoretical considerations, percent SiO2 differences as small as 14% should be detectable with the University of Michigan's currently available detectors.
Remote sensing techniques for water resources evaluation and hydrobiological features mapping with multiband scanner imagery