Efficacy of small-scale RPAS for monitoring mangrove species distribution under urbanisation threats
摘要
Increasing urbanisation threats faced by mangrove ecosystems necessitate accurate monitoring methods. This study evaluated efficacy of small-scale remotely piloted aircraft systems (RPAS) combined with segment anything model (SAM) for mapping mangroves in urban areas of Central Vembanad lake. High-resolution imagery (4:3 pixels) was collected using DJI drone with FC7303 RGB sensor. RPAS achieved ground sampling distance of 0.84 inches/pixel. SAM model achieved 92% accuracy in land-cover classification. Mangroves and associated vegetation cover 56.34% of study area, while built-up areas occupy 19.51%. RPAS–SAM approach proved to be powerful tool for detailed monitoring, indicating complex urban-mangrove interface and subsequent impact on mangrove distribution.