A Critical Review of Methods for Automated Detection of Mangrove Deforestation

Nowadays, there has been significant interest in the dependable and efficient automated anticipation and evaluation of alterations in mangrove landcover. This interest has been driven by the application of advanced technologies such as deep learning networks and machine learning techniques. Automated mangrove landcover change detection is highly valuable, providing timely, accurate insights into ecosystem shifts using advanced technologies like remote sensing, machine learning, etc. It offers benefits such as frequent monitoring for rapid change detection, efficient analysis of extensive datasets, accurate alteration identification, early conservation responses, data-driven decision-making, improved ecological understanding, standardized long-term monitoring, informed resource allocation, community empowerment, and effective policy advocacy, making it a powerful tool for mangrove ecosystem protection and management. While a few researchers have introduced techniques for automated detection of changes in mangrove landcover, this realm of study still possesses untapped potential. Several aspects within this domain remain inadequately explored. In response, our study undertakes a comprehensive evaluation of current methods, spanning diverse aspects including data origins, methodologies, and beyond. Through an extensive exploration, our research seeks to furnish valuable insights to fellow researchers and propose remedies for the unexplored domains within this field.