Vol. 10, Issue 2 (2021)

Author(s):
Sugavaneshwaran Kannan, Ragunath Kaliaperumal, Pazhanivelan S, Kumaraperumal R and Sivakumar K

Abstract:
A Research study was conducted to estimate the paddy area in the Nagapattinam district of Tamil Nadu using Sentinel 1A SAR data through multi-temporal feature extraction. Multi-temporal Sentinel 1A data of GRD data at VH polarizations were obtained for the study area at 12 days intervals. Collected data were processed using MAP scape-RICE software. Sentinel 1A is an active SAR microwave data, which captures the crop characteristics irrespective of the weather condition and illumination. Ground truth observations collected during cropping period were used to derive the paddy signature from the processed satellite images. The dB values extracted as signature then subjected to the multi-temporal feature extraction method for delineating the paddy growing areas. Around 99, 00, 700 ha was mapped as the paddy growing area in Nagapattinam district. Accuracy assessment was done with 40 percent of the ground truth data. The overall classification accuracy was 94 percent with a kappa score of 0.88.