I am working with hyperspectral airborne data and would like to know if anyone could help me to do this kind of correlation chart between spectral bands? What software to perform? What would the procedure look like? Thank you.
I took an Electro-Optical Infrared analysis course in which our final project was an airborne LIDAR topographical mapping project. For problems like this, I've always been a fan of linear algebraic techniques such as principal components analysis (PCA).
https://en.wikipedia.org/wiki/Principal_component_analysis
The process itself is rather mechanical if you have enough matrix algebra experience and a copy of MatLab on hand. Charting techniques associated with PCA should not be hard to track down and implement. You can do it in ExCel using pivot charts, for example. The real challenge is in the interpretation of the PCA results - something only a human can do as far as I know. And with hyperspectral data you have far too many dimensions to chart and correlate all at once, so you have to pick and choose certain spectra if you want your charts to be intelligible. That part is sort of like running queries against a relational database, which in this case is your multidimensional data set.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Software