Share Email Print

Proceedings Paper

An automatic method for producing robust regression models from hyperspectral data using multiple simple genetic algorithms
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

This paper presents a new method for automatically determining the optimum regression model, which enable the estimation of a parameter. The concept lies on the combination of k spectral pre-processing algorithms (SPPAs) that enhance spectral features correlated to the desired parameter. Initially a pre-processing algorithm uses as input a single spectral signature and transforms it according to the SPPA function. A k-step combination of SPPAs uses k preprocessing algorithms serially. The result of each SPPA is used as input to the next SPPA, and so on until the k desired pre-processed signatures are reached. These signatures are then used as input to three different regression methods: the Normalized band Difference Regression (NDR), the Multiple Linear Regression (MLR) and the Partial Least Squares Regression (PLSR). Three Simple Genetic Algorithms (SGAs) are used, one for each regression method, for the selection of the optimum combination of k SPPAs. The performance of the SGAs is evaluated based on the RMS error of the regression models. The evaluation not only indicates the selection of the optimum SPPA combination but also the regression method that produces the optimum prediction model. The proposed method was applied on soil spectral measurements in order to predict Soil Organic Matter (SOM). In this study, the maximum value assigned to k was 3. PLSR yielded the highest accuracy while NDR’s accuracy was satisfactory compared to its complexity. MLR method showed severe drawbacks due to the presence of noise in terms of collinearity at the spectral bands. Most of the regression methods required a 3-step combination of SPPAs for achieving the highest performance. The selected preprocessing algorithms were different for each regression method since each regression method handles with a different way the explanatory variables.

Paper Details

Date Published: 19 June 2015
PDF: 10 pages
Proc. SPIE 9535, Third International Conference on Remote Sensing and Geoinformation of the Environment (RSCy2015), 953502 (19 June 2015); doi: 10.1117/12.2192320
Show Author Affiliations
Dimitris Sykas, National Technical Univ. of Athens (Greece)
Vassilia Karathanassi, National Technical Univ. of Athens (Greece)

Published in SPIE Proceedings Vol. 9535:
Third International Conference on Remote Sensing and Geoinformation of the Environment (RSCy2015)
Diofantos G. Hadjimitsis; Kyriacos Themistocleous; Silas Michaelides; Giorgos Papadavid, Editor(s)

© SPIE. Terms of Use
Back to Top