Temporal Signature Matching for Land Cover Classification

Objects on the surface of earth describe unique patterns in time and this is exploited in the proposed land cover classification method enumerated in this paper. The applicability of various distance measures to reasonably find similarity or dissimilarity between time series vegetation index patterns for land cover classification is demonstrated. These distance metrics have inherent advantages for satellite image classification as land use practices such as field crops/vegetation exhibit temporal shifts from pixel to pixel based on the geographic location. Experimental results on time series of MODIS EVI (250m) and AWiFS derived NDVI (56m) demonstrate the applicability and validate the performance of the proposed method. The classification method is also compared with k nearest neighbor (kNN) and support vector machine (SVM) classifier to understand its utility. * Corresponding author.