A Novel Water Index for Extraction of Water Features using Landsat-8 Images in Prayagraj District, India

The satellite images can be categorized in three classes on the basis of the spectral bands, viz., panchromatic, multispectral and hyperspectral. The multispectral imageries are very useful in the feature extraction and enhancement, the combination of bands creates magical effect not only to delineate some features but also to eliminate the others. Several water indices on the basis of band combinations have been proposed by the researchers for the open surface water detection. In the present work a new index has been proposed for open surface water detection from landsat-8 satellite imagery. The proposed index can detect the water class more efficiently as compared to the existing indices which has been supported by the quantitative data. The results are validated in terms of the number of pixels of water class with an accuracy of 99.77%.