Embedded Files
Academic Projects
Academic Projects
The main objective of the research is to implement an object-oriented deep CNN-based model for the detection of selective features in hyperspectral imagery. The future research work aims at object-level feature extraction considering the spectro-spatial aspect of hyperspectral images. A high-resolution hyperspectral dataset is required consisting of features for object detection.
Detection of mars craters using Faster-RCNN Resnet-50 (2022 GeoAI Martian challenge)
The focus of our 2022 GeoAI challenge is Mars crater detection. In this object detection challenge, Mars crater benchmark datasets that contain over 100,000 images and 300,000 craters have been prepared located in nearly every corner of the Mars surface.
In this research work, non-linearity in the data has been handled by incorporating kernel with the MPCM algorithm. Nine different kernel functions have been proposed to classify nine different classes with non-linearity among them. The Landsat-8 dataset has been used for classification and the Formosat-2 dataset has been used as a reference image (resampled at 10m resolution). The mean Membership Difference method was used as an accuracy assessment for the optimization and identification of the best kernel for Formosat-2.
In this research work, paddy stubble burnt fields were identified by using fuzzy-based machine learning algorithms for a test site in Patiala, Punjab, India. A Euclidean Distance (ED) and Gaussian Kernel-based Modified Possibilistic c-Means (MPCM) algorithms were used for this purpose. The main objective of this research work was to test the effectiveness of the machine learning approach for the identification of paddy stubble burnt field and also to find out the better classifier amongst Euclidean Distance (ED) and Gaussian Kernel MPCM classifier.
Surging glaciers that often originate practically from the same feeder zones it is seen that one glacier will show rapid advance in one valley; yet another, in the adjacent valley, will exhibit degeneration and retreat. This study will try to analyze and monitor the nature of fast glacier flow (surging phenomena) by estimating the surface area change of these surge-type glaciers. The prime focus of this study is to visually and digitally analyze the multi-date, high-resolution satellite data pertaining to each year from 1992 to 2018.
Real-time data monitoring is an effective way of maintaining data with a visual map from which it is possible to interpret data in an effective way. It enables data administrators to review the overall processes and functions performed on the data in real-time, or as it happens, through graphical charts and bars on a central interface/dashboard. The goal of this work is to take live data from the website data.gov.in and visualize the data dynamically as well as the map.
The objective of the work is to detect land use/land cover changes in Kuttanad taluk, Alappuzha, Kerala due to flood using GIS techniques in classes such as: Water bodies, Agriculture, Urban areas. The pre-flood and post-flood image of the entire taluk was digitized and different classes were assigned. The percentage areal coverage of different classes were calculated and change in percentage areal coverage before and after the flood was calculated .
Page updated
Google Sites
Report abuse