The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Publications Copernicus
Articles | Volume XL-3/W3
19 Aug 2015
 | 19 Aug 2015


Kuldeep and P. K. Garg

Keywords: Remote Sensing, Image Classification, Object Extraction, Flood Mapping

Abstract. River flooding in planar region is the most significant type of natural disaster that modern society is exposed to, affecting several thousand people each year. Recent flood events, population growth concerns have augmented the call for global methods which utilise both spatial and temporal dynamics. Object oriented classification approaches based on the segmentation are being adopted for extraction of variety of thematic information from high resolution satellite images. Generation of landuse/cover map which is one of the important inputs to the model for flood inundation mapping and for accurate assessment of damage due to floods requires advanced methods of image classification. The Cartosat-1 (PAN) satellite data has been fused with the LISS-III (MX) to obtain the color image containing both high spatial and spectral information. The fused image is further classified to obtain the landuse/cover map using object based classification approach. The classification results are assessed by calculating overall accuracy and kappa index with the help of error matrix. The overall accuracy of classification has been obtained 86.00% with kappa index 0.7815. The objective of this study is to develop a GIS aided model for flood inundation mapping of the surrounding of the part of the Yamuna River which flows through the two districts i.e. Yamuna Nagar and Saharanpur in states of Haryana and Uttar Pradesh respectively. The model considers the five parameters viz. topography (slope, elevation) information, landuse/cover, time series data of surface water elevation, river geometry and location of the rain gauge station. Field survey has been conducted to validate the positional accuracy of the DEM and landuse/cover classes using DGPS. A final flood inundation map has been prepared by combining all weighted layers with in GIS environment. The flood inundation maps can further be used for quick identification of areas of potential flood hazard to minimize the flood losses.