A Method for Automatic Water Mass Classification.
Abstract
A method for automatic water mass classification from satellite radiometric data uses Bayesian decision theory. The mean, standard deviation and gradient of corrected satellite infrared temperature data is computer with a window and assigned to the center pixel. The probability for each statistic is computed for each class of ocean front/water mass based upon historical statistical values. The class having the maximum probability is then tested to reduce the chance of the misclassification. The result is a class value assigned to each pixel position representative of a particular oceanic front/water mass. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 06, 1984
- Accession Number
- ADD011405
Entities
People
- C. W. Horton
- R. E. Coulter
Organizations
- United States Department of the Navy