Mapping and monitoring of mangrove species based on remote sensing technology play a crucial role in biodiversity conservation and management. This paper employs CiteSpace to visualize the literature and presents a comprehensive review of the researches conducted in this domain, focusing primarily on bibliometric characteristics, diverse sensors, and classification algorithms. Since the publication of the first remote sensing-based study on mangrove species classification in 2004, the number of publications in this field has exhibited a general upward trend up to 2023. China, the United States, and India lead in publishing research on mangrove species mapping, with researchers in the United States being particularly active in international collaborations. Mapping of mangrove species is predominantly concentrated on single time points and across 53 small regions, with the majority of research sites located in India and China. Existing studies have utilized various remote sensing image for mangrove species classification, including airborne hyperspectral, spaceborne visible, infrared, multispectral, hyperspectral, synthetic aperture radar (SAR), and drone-borne visible, infrared, multispectral, hyperspectral, light detection and ranging (LiDAR) data. Classification algorithm development has evolved four stages, from pixel-based methods to object-oriented approaches, progressing to approaches incorporating machine learning algorithms, and currently advancing towards ensemble learning and deep learning. Research in this field still faces several challenges in data fusion, classification algorithm enhancement, increased number of classification species, and large-scale long-term mapping. The studys findings would provide valuable guidance to researchers and practitioners in advancing and enhancing the management and conservation of mangroves
- Autor/es: Yuqi Wu, Chunyan Lu, Kexin Wu, Wenna Gao, Nuocheng Yao, Jingwen Lin
- Año de publicación: 2025
- País: China
- Idioma: Inglés
- Fuente de indexación: Scopus