Implementation of the directly-georeferenced hyperspectral point cloud.

Area of Science:

• Remote Sensing
• Geospatial Data Analysis
• Computer Vision

Background:

• Hyperspectral imaging (HSI) data preprocessing typically involves radiometric, atmospheric, geometric correction, and resampling, resulting in raster images.
• Conventional raster formats compromise the spatial-spectral integrity of HSI data, impacting application performance.
• A new data format, Directly-Georeferenced Hyperspectral Point Cloud (DHPC), was developed to preserve spatial-spectral integrity more effectively.

Purpose of the Study:

• To describe the DHPC data fusion workflow for improved HSI data processing.
• To provide tools for integrating the DHPC workflow into existing processing pipelines.
• To enhance the accessibility and implementation of the DHPC format for data providers and end-users.

Main Methods:

• The DHPC workflow modifies the digital surface model (DSM) used in geometric correction.
• A MATLAB script is provided to derive the point spread function (PSF) of HSI data and convolve it with the DSM.
• This process ensures better spatial consistency between elevation data and HSI data.

Main Results:

• The DHPC format preserves spatial-spectral integrity more effectively than traditional raster images.
• DHPC files are up to 13 times smaller than conventional rasters, facilitating data distribution.
• The derived PSF aids in characterizing HSI data and quantifying spectral contributions.

Conclusions:

• The DHPC data fusion workflow offers a transparent and approachable method for enhancing HSI data processing.
• Implementation of DHPC leads to improved spatial-spectral integrity, reduced storage needs, and better application results.
• The DHPC format is ideal for advanced applications like virtual and augmented reality site exploration.