Adaptive Grouping Distributed Compressive Sensing Reconstruction of Plant Hyperspectral Data

Area of Science:

• Remote Sensing
• Data Science
• Spectroscopy

Background:

• Hyperspectral technology is crucial for vegetation analysis.
• Effective compressive reconstruction is needed for efficient data handling.
• Existing methods face challenges in maintaining spectral information.

Purpose of the Study:

• To develop an effective spectral compressive reconstruction method for plant hyperspectral data.
• To improve data storage, transmission, and spectral information fidelity.
• To enhance quantitative remote sensing research and applications.

Main Methods:

• Proposed the spectral adaptive grouping distributed compressive sensing (AGDCS) algorithm.
• Analyzed spectral characteristics to construct a joint sparse model.
• Adaptively grouped spectral bands for compressed sensing reconstruction.

Main Results:

• AGDCS significantly improved spatial domain image reconstruction visual effects.
• Achieved higher peak signal-to-noise ratio (PSNR) at low sampling rates compared to OMP and GPSR.
• Demonstrated lower spectral domain errors (RMSE, MAPE, MAE) than GPSR.
• Showcased relatively high reconstructed efficiency.

Conclusions:

• AGDCS offers superior performance for plant hyperspectral data reconstruction.
• The adaptive grouping strategy enhances reconstruction accuracy and efficiency.
• This method is promising for advancing quantitative remote sensing of vegetation.