An efficient framework for segmentation and classification of water bodies from aerial images

Abstract

There is a growing need for the classification of water bodies using newlineremote sensing images. Water-body analysis is significant in extracting new newlinewater-bodies from aerial images and also helps in preserving existing water-bodies by continuous monitoring. Extraction of water bodies from high-resolution aerial images is seen as a necessary activity for both civilian and military purposes. Water-body segmentation helps in extracting water bodies like lakes, ponds, rivers and reservoirs from high-resolution aerial images. This also helps in discovering new water bodies.The state-of-the-art methods include rule based, thresholding based, region based, edge based, gradient based, clustering based and neural network based methods. To address the water-body segmentation problem, new methods have to be proposed. The challenges that need to be tackled in the study are boundary detection and shadow removal newline newline