Water-Stressed Crops Detection Using Multispectral WorldView-2 Satellite Imagery

Abstract

The paper presents a method for automatic detection and monitoring of small waterlogged areas in farmland, using multispectral satellite images and diverse classifiers. In the waterlogged areas, excess water significantly damages or completely destroys the plants, thus reducing the average crop yield. Automatic detection of (waterlogged) crops damaged by the combined effect of rainfall and rising underground water is an important tool for government agencies dealing with yield assessment and disaster control. The paper describes the application of two different machine learning algorithms to the problem of identifying crops that have been affected by rising underground water levels inWorldView-2 satellite imagery. Satellite images of central European region (Northern Serbia), taken in May and July 2010, with spatial resolution of 0:5m and 8 spectral bands were used to train the classifiers and test their performance when it comes to identifying the water-stressed crops. WorldView-2 satellite provides 4 new bands potentially useful in agricultural applications: coastal-blue, red-edge, yellow and near-infrared 2. We propose a methodology based on Multilayer Perceptron neural networks and Genetic Programming to achieve per-pixel classification. The classifiers constructed are able to achieve 99.4% accuracy when trained and evaluated on a single image and 97.8% accuracy when the testing is done on an image taken under different atmospheric and solar geometry conditions.