Identifying Land features from satellite images using deep learning

Land use and Land Cover changes have been subject to study for longer time due to its impact on climate and environment. Identification of Land features helps in keeping track of changes occur over time. Feature extraction or identification of remote sensing images can be done using deep learning techniques such as CNN. Feature extraction on low resolution satellite images is difficult and we propose to use various deep learning techniques to effectively identify land features using CNN. In this Study we experiment with parameterize multi filter CNN which can be used as the training depth. This will help in further research towards remote sensing image analysis on feature identification using novel CNN techniques.