Search NASA⌕ Search

Engineering topics

Mccloy, K. R.

Publications and source records attributed to Mccloy, K. R..

6 records
All sourcesNASA NTRSDOE OSTI
NASA NTRS1977

Water utilization, evapotranspiration and soil moisture monitoring in the south east region of south Australia

The author has identified the following significant results. It was established that reliable estimates of sand and coastal scrub areas can be determined from LANDSAT image classification by the Vec classifier more economically than by conventional means from a map of the coastal zone produced by photointerpretation using 1:10,000 aerial photography. Current LANDSAT imagery is also suitable for monitoring for large scale storm damage to the zone, but the normal change in sand areas extent due to man's activity or other reasons, is about 5 to 10 m per year, occasionally being as great as 30 m per year, so that it is considered that LANDSAT D will have the resolution necessary to monitor these changes but not current imagery.

Mccloy, K. R.↗
NASA NTRS1977

The vector classifier

A linear classifier is developed. The classifier algorithm derives, from the picture element (pixel) response values, the response values for a point (called footpoint) consisting of determined proportions of distinct and operator defined land surfaces. The algorithm decides whether the pixel contains the content of the defined surfaces by testing the footpoint to pixelpoint distance against an operator specified threshold value. The test can be either of statistical or geometric form. Those pixels which give a positive result to the test are subclassified according to the proportions of the distinct land surfaces. The practical characteristics of the classifier are discussed and the classification results achieved, in a number of test areas, are described. The artificial nature of the classifier assumptions relate to certain land surface conditions. Testing the suitability of the classifier for other conditions is discussed.

Mccloy, K. R.↗