HR: 1340h
AN: H13C-0438    [Abstracts]
TI: Comparison between Neural Network and Fuzzy Logic system for Soil Moisture Estimation using Microwave Remote Sensing Data
AU: * Lakhankar, T Y
EM: tarendra@ce.ccny.cuny.edu
AF: Civil Engineering Department, City University of New York, Steinman Hall, Convent Avenue at 140th Street, New York, NY 10031 United States
AU: Ghedira, H
EM: ghedira@ce.ccny.cuny.edu
AF: Civil Engineering Department, City University of New York, Steinman Hall, Convent Avenue at 140th Street, New York, NY 10031 United States
AU: Khanbilvardi, R M
EM: khanbilvardi@ccny.cuny.edu
AF: Civil Engineering Department, City University of New York, Steinman Hall, Convent Avenue at 140th Street, New York, NY 10031 United States
AB: Artificial neural networks and Fuzzy logic have been applied to a wide range of problems in several disciplines. They have been successfully applied to image processing, and have shown a great potential in the classification of remote sensing data. However, a successful application of these methods in remote sensing data classification requires a good comprehension of the effect of their internal parameters and especially those that are related to the algorithm structure and to the training process. In this work we report the application of backpropagation neural network and fuzzy logic in estimating the soil moisture level using Synthetic Aperture Radar (SAR) data. The potential of SAR images in spatial soil moisture estimation depends on the ability of these algorithms to define the complex relationship that exists between the backscattered energy and the moisture content of the soil. A study area located in Oklahoma (97d35'W, 36d15'N) has been chosen for this project. Several textural measures derived from Radarsat-1 images acquired in Scansar Mode during the summer of 1997 were used as input for two algorithms. The soil moisture data measured by ESTAR Instrument (Electronically Scanned Thinned Array Radiometer) during the SGP97 campaign (operated by NASA) were used as truth data in the training process. The effect of some parameters related to the training process on classification performance was investigated for both methods. The preliminary results showed that for neural networks, the variations of the number of hidden layers and the number of nodes by layer have no significant effect on classification accuracy. However, the retained threshold value used in the output layer affects significantly the overall classification. Concerning, the fuzzy logic algorithm, the preliminary results showed that the cluster radius selection have a significant effect on classification accuracy.
DE: 3360 Remote sensing
DE: 1866 Soil moisture
DE: 1894 Instruments and techniques
SC: Hydrology [H]
MN: 2004 AGU Fall Meeting