IMAGE PROCESSING, PATTERN RECOGNITION
Recognition of abandoned agricultural lands using seasonal NDVI values
E. A. Terekhin
Belgorod State University, Federal and Regional Centre for Aerospace and Ground Monitoring of Objects and Natural Resources, Scientific and Technological Equipment Common Use Centre, Belgorod, Russia
This paper explores the potentialities of discriminant analysis for the identification of abandoned agricultural lands using their reflectance spectrum properties. A method of automated detection of fallow lands is proposed. Using experimental data received from the agricultural lands of the Belgorod Region, we propose equations that allow an agrarian land to be classified as an arable or fallow land in an automated mode. The accuracy of fallow land recognition is 71%. It is found that seasonal normalized difference vegetation index (NDVI) values computed from MODIS data in the period of late September - early October are most informative in terms of abandoned agricultural land identification. It is shown that the use of the minimal NDVI values is much more efficient for the identification of fallow land when compared with the mean NDVI values.
abandoned agricultural lands, stepwise discriminant analysis, remote sensing, NDVI, MODIS, reflectance spectrum properties.
PDF file (335 kB)
E. A. Terekhin, “Recognition of abandoned agricultural lands using seasonal NDVI values”, Computer Optics, 41:5 (2017), 719–725
Citation in format AMSBIB
\paper Recognition of abandoned agricultural lands using seasonal NDVI values
\jour Computer Optics
Citing articles on Google Scholar:
Related articles on Google Scholar:
|Number of views:|