Simulation of EO-1 Hyperion Data from ALI Multispectral Data Based on the Spectral Reconstruction Approach

Area of Science:

• Remote Sensing
• Geospatial Data Analysis
• Image Processing

Background:

• Hyperspectral data offers richer information than multispectral data for advanced remote sensing applications.
• Acquiring hyperspectral data is challenging and expensive, limiting its widespread use.
• Abundant multispectral data exists, presenting an opportunity for simulation.

Purpose of the Study:

• To develop and evaluate a method for simulating hyperspectral data from multispectral Advanced Land Imager (ALI) data.
• To enable the creation of hyperspectral datasets where direct acquisition is not feasible.
• To enhance the utility of existing multispectral archives for research and development.

Main Methods:

• A spectral reconstruction approach was employed to simulate hyperspectral data.
• Inherent spectral information was extracted from source ALI data and reassigned.
• 106 bands of Hyperion data were simulated from corresponding ALI data.

Main Results:

• The simulation method demonstrated good performance in reconstructing hyperspectral data.
• Visual interpretation and statistical comparisons confirmed the quality of the simulated bands.
• Classification results indicated that the simulated data preserved essential information effectively.

Conclusions:

• The proposed method successfully simulates hyperspectral data from multispectral sources.
• This technique facilitates algorithm testing, extends multispectral data utility, and aids virtual sensor design.
• Spectral reconstruction offers a viable solution for generating hyperspectral datasets for research and application development.