Spacecube: A fast inverse hyperspectral georectification system
By: Thomas P. Watson, Eddie L. Jacobs
Potential Business Impact:
Makes special cameras map land faster and better.
Hyperspectral cameras provide numerous advantages in terms of the utility of the data captured. They capture hundreds of data points per sample (pixel) instead of only the few of RGB or multispectral camera systems. Aerial systems sense such data remotely, but the data must be georectified to produce consistent images before analysis. We find the traditional direct georectification method to be slow, and it is prone to artifacts. To address its downsides, we propose Spacecube, a program that implements a complete hyperspectral georectification pipeline, including our own fast inverse georectification technique, using OpenGL graphics programming technologies. Spacecube operates substantially faster than real-time and eliminates pixel coverage artifacts. It facilitates high quality interactive viewing, data exploration, and export of final products. We release Spacecube's source code publicly for the community to use.
Similar Papers
From Orbit to Ground: Generative City Photogrammetry from Extreme Off-Nadir Satellite Images
CV and Pattern Recognition
Creates 3D city views from space photos.
Improved 3D Gaussian Splatting of Unknown Spacecraft Structure Using Space Environment Illumination Knowledge
CV and Pattern Recognition
Builds 3D space junk models from photos.
From Orbit to Ground: Generative City Photogrammetry from Extreme Off-Nadir Satellite Images
CV and Pattern Recognition
Creates 3D city views from satellite pictures.