Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

A Genetic Algorithm Based Tool for Analysis and Modeling of Multi-Spectral Data

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan

Citation:  Paper number  033126,  2003 ASAE Annual Meeting . (doi: 10.13031/2013.14066) @2003
Authors:   Tong Wang, Chi Ngoc Thai
Keywords:   Spectral Imaging, Vegetation Indices, Genetic Algorithm, Statistical Modeling

This report describes the development of statistical models to determine the relationships between vegetation indices and bush bean plant nitrogen stress levels using 2 approaches, statistical analysis with exhaustive search by SAS procedures and optimal search using genetic algorithm (GA). With SAS, polynomial regression was used to fit data under all possible twowaveband combinations. The best wavebands identified were 700nm, 710 nm, 720 nm, and 750 nm. The best model selected by both SAS and GA was a 2nd degree polynomial for RVI (= R 2 / R 1) with an adjusted R2 value of 0.9144. Comparison of these two approaches showed that GA integrated both statistical analysis and model selection and was as accurate but more efficient than SAS.

(Download PDF)    (Export to EndNotes)