A multi-temporal masking classification method for field-based agricultural crop mapping

Arıkan, Mahmut
This study describes the field-based classification of agricultural crops using multi-date Landsat 7 ETM+ images acquired in May, July, and August 2000. The study area is located in north-west of Turkey with a size of about 1 5 km x 1 1.3 km and grows a variety of crops. The objective was to identify the summer (August) crops within the agricultural fields. The classification methodology is based on a multi- temporal masking of Landsat 7 ETM+ images. First, a supervised per-pixel classification of the three images (May, July, and August 2000) was performed using a maximum likelihood classifier algorithm. The accuracy of classified outputs was computed by comparing them with the ground truth information. Those classes with a high classification accuracy were masked out and the August image was re-classified using the unmasked classes only excluding the masked areas from the classification. The masking technique was applied to overcome the problems caused by the spectral overlaps between the classes. After completing the classification process, the multi- temporal classified output of the August image was analyzed in a field specific manner in the integration of remote sensing and geographic information system (GIS). In each field, the percentages of classified pixels were computed and the field was assigned a class label based on the highest percentage value. The resulting classification accuracy of the multi-temporal masking technique was 81%, which was 10% more accurate than the classification of the August image only. An immediate updating of a GIS database was provided by means of directly entering the analysis results into the database.
Citation Formats
M. Arıkan, “A multi-temporal masking classification method for field-based agricultural crop mapping ,” M.S. - Master of Science, Middle East Technical University, 2003.