International Journal Of Imaging

Abstract

PREDICTING LANDCOVER CHANGES IN HIGH-RESOLUTION SATELLITE IMAGERY THROUGH DEEPLEARNING TECHNIQUES

Author: Rajesh. G, Navaneeth.D, Bharath.M,Venkat Tiru Gopal Reddy.K

Abstract:

Accurate land cover classification is of paramount importance for a wide range of applications, including urban planning, environmental monitoring, and natural resource management. Traditional land cover classification systems often rely on manual feature engineering, which can be time-consuming and subject to human bias. In addition, these systems struggle to handle complex landscapes and may not adapt well to changing environmental conditions. To overcome these limitations, our proposed Artificial intelligence based system leverages deep learning techniques, to automate the classification process. The dataset utilized for this research pairs each satellite image with a corresponding mask image that encodes land cover annotations using an RGB colour scheme We train the model to learn the intricate spatial patterns and spectral signatures associated with each land cover class, making it highly adaptable to diverse and dynamic environments. Land cover change detection is c