UAV imaging for shallow-water habitat mapping: Comparing performance between hyperspectral, multispectral and RGB imaging

Key Points

• To compare the performance of different imaging modalities for UAV imaging in shallow-water habitat mapping.
• Employed hyperspectral imaging to simulate multispectral and RGB images.
• Utilized spectral response functions from two existing cameras for image simulation.
• Applied machine learning techniques to evaluate imaging performance.
• Performance differences were noted among hyperspectral, multispectral, and RGB imaging modalities.
• Hyperspectral imaging showed the highest accuracy for habitat mapping, followed by multispectral and RGB.
• Specific performance metrics were derived from simulated data, influencing future research directions.