24 February 2004 Ecosystem classification using artificial intelligence neural networks and very high spatial resolution satellite imagery
Author Affiliations +
Abstract
This study investigates the potential of classifying complex ecosystems by applying the radial basis function (RBF) neural network architecture, with an innovative training method, on multispectral very high spatial resolution satellite images. The performance of the classifier has been tested with different input parameters, window sizes and neural network complexities. The maximum accuracy achieved by the proposed classifier was 78%, outperforming maximum likelihood classification by 17%. Analysis showed that the selection of input parameters is vital for the success of the classifiers. On the other hand, the incorporation of textural analysis and/or modification of the window size do not affect the performance substantially. The new technique was applied to the area of Lake Kerkini (Greece), a wetland of great ecological value, included in the NATURA 2000 list of ecosystems.
© (2004) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Iphigenia Keramitsoglou, Iphigenia Keramitsoglou, Haralambos Sarimveis, Haralambos Sarimveis, Chris T. Kiranoudis, Chris T. Kiranoudis, Nicolaos Sifakis, Nicolaos Sifakis, } "Ecosystem classification using artificial intelligence neural networks and very high spatial resolution satellite imagery", Proc. SPIE 5232, Remote Sensing for Agriculture, Ecosystems, and Hydrology V, (24 February 2004); doi: 10.1117/12.511041; https://doi.org/10.1117/12.511041
PROCEEDINGS
9 PAGES


SHARE
Back to Top