BRDF modelling using linear kernel models on satellite data. BRDF is an important part of optical remote sensing. Usually, the change in view/illumination geometry is considered an annoyance, and is treated as noise. Pity the fools! There are a very simple ways of accounting for these angular effects, among them using simple linear kernel models. This repository has some exploration on how to use them.
There's some software that you can play around. It is designed for ahem educational purposes, so you've been warned.
If this is useful to you and you end up using it, please consider citing it. It's on Zenodo. Somewhere. You find this, I'm sure you can find that ;)
You can run the notebooks directly online (takes a while, be patient!)
Introduction to linear BRDF models
- More stuff coming whenever I feel like it ;)
