193 / 2023-09-05 19:49:38

Advancing Environmental Monitoring with UAV-Based Remote Sensing Technologies

Hyperspectral imaging,LiDAR technology,Environmental monitoring,Sensitive ecosystems

Uncrewed aerial vehicles (UAVs), commonly referred to as drones, have emerged as a promising method for high-resolution environmental monitoring in proximity to the Earth's surface. These UAV systems offer a cost-effective and replicable approach to capturing remote sensing data with exceptional spatial precision and a favourable signal-to-noise ratio. The integration of advanced lightweight hyperspectral and Light Detection and Ranging (LiDAR) sensors into UAV platforms has expanded the potential to glean valuable physiochemical and structural insights from the environment. While the adoption of these systems has gained traction in fields like agriculture and forestry research, their utilization in monitoring intricate mine environments remains largely unexplored. The monitoring of sensitive mine environments, such as the swamp vegetation within longwall mining regions, presents a critical yet challenging task due to the inherent complexities involved. Presently, monitoring these remote and demanding environments primarily relies on ground-based methods. This reliance on terrestrial approaches is partly attributable to the absence of a comprehensive framework and the formidable obstacles associated with employing UAV-based sensor systems for monitoring such sensitive ecosystems.



This research endeavors to confront the pertinent challenges by developing a suite of UAV-based hyperspectral and LiDAR systems, complete with an integrated workflow tailored for mapping and potentially monitoring highly heterogeneous and intricate environments. The development of these systems encompasses several key components, including sensor hardware integration, calibration procedures, mission planning strategies, and the design of a processing pipeline aimed at generating actionable datasets. Novel algorithms and processing routines have been devised to establish an efficient data retrieval process and facilitate the generation of functional products. Throughout the development process, an array of data processing analyses has been conducted at distinct intermediate stages to ensure the consistent quality of data throughout the workflow. These designed systems and methods have been put into practice within a peat swamp environment to produce a precise geo-spatialized hyperspectral raster and LiDAR point cloud. The performance of the UAV-based hyperspectral and LiDAR data has been rigorously evaluated against ground-based measurements, encompassing fine-scale mapping assessments, the derivation of crucial parameters such as vegetation indices and LiDAR matrices maps, leaf area index calculations, canopy height modeling, and classification tasks.



In summary, this study furnishes a quantifiable foundation for evaluating the condition of vegetation in sensitive environments through the development of integrated UAV-based hyperspectral and LiDAR data acquisition systems, coupled with their associated processing protocols.