Comparison of fusion and NDVI deference's methods in extraction of rice-planted area using multi-temporal Landsat TM and IRS-1C panchromatic data (Case study some-Sara district - Rasht)

Optimal use of water and soil resources has always been the aim in agricultural development in order to achieve maximum productivity by using of minimum land and water resource consumption. Nowadays، remote-sensing technology represents a powerful tool for researchers and policy makers. In particular، the availability of an ever increasing number of remote sensing satellites makes it possible to consider new advanced applications of remote sensing، ranging from monitoring and management of natural resources، mapping of land-cover/land-use (e. g.، monitoring of agricultural areas، etc.) to economical investigation of production. In this study Landsat (TM7) and panchromatic band of IRS-1C data were used for extraction of rice planted area of Some-Sara district، Iran and its villages using RGB-HIS fusion method and NDVI deference’s employed in this research. Evaluation was performed by field observation and by analyzing the classification results using a maximum likelihood classifier for both fused image resulted from fusion method and NDVI deference’s images. The accuracy of these methods were compared. The results was shown that the NDVI deference’s of two landsat images acquired with 45 days interval، yields higher accuracies (Overall accuracy=92. 4%، Kappa =0. 83%) than the fusion method using RGB-HIS transformation (Overall accuracy= 87. 47%، kappa=72. 76%) method. Meanwhile، total estimated rice planted area is about 26675. 30 ha which is corresponded to the area reported by agricultural organization of Gilan province.