You are here

Optimization of spectral indices and long-term separability analysis for classification of cereal crops using multi-spectral RapidEye imagery

TitleOptimization of spectral indices and long-term separability analysis for classification of cereal crops using multi-spectral RapidEye imagery
Publication TypeJournal Article
Year of Publication2016
AuthorsGerstmann, H
Secondary AuthorsMöller, M
Corporate AuthorsGläßer, C
JournalInternational Journal of Applied Earth Observation and Geoinformation
Volume52
Start Page115
Pagination11
Date Published06/2016
Type of ArticleResearch Article
Keywordscrop classification, RapidEye, spectral separability, time series, vegetation index
Abstract

Crop monitoring using remotely sensed image data provides valuable input for a large variety of applications in environmental and agricultural research. However, method development for discrimination between spectrally highly similar crop species remains a challenge in remote sensing. Calculation of vegetation indices is a frequently applied option to amplify the most distinctive parts of a spectrum. Since no vegetation index exist, that is universally best-performing, a method is presented that finds an index that  is optimized for the classification of a specific satellite data set to separate two cereal crop types. The (eta-squared) measure of association – presented as novel spectral separability indicator – was used for the evaluation of the numerous tested indices. The approach is first applied on a RapidEye satellite image for the separation of winter wheat and winter barley in a Central German test site. The determined optimized index allows a more accurate classification (97%) than several well-established vegetation indices  like NDVI and EVI (<87%). Furthermore, the approach was applied on a RapidEye multi-spectral image time series covering the years 2010–2014. The optimized index for the spectral separation of winter barley and winter wheat for each acquisition date was calculated and its ability to distinct the two classes was assessed. The results indicate that the calculated optimized indices perform better than the standard  indices for most seasonal parts of the time series. The red edge spectral region proved to be of high significance for crop classification. Additionally, a time frame of best spectral separability of wheat and barley could be detected in early to mid-summer. 

DOI10.1016/j.jag.2016.06.001