I am reading about the working principles and applications of hyperspectral imaging (HSI). In this link, the uses of HSI over the range of the wavelength, from visible to longwave infrared have been summarized by the picture below.

My question is, is there an equation that relates the chemical composition / color / etc of a material to its hyperspectral signature ?

For example, I read that water is transparent to visible light, but it absorbs strongly at wavelengths of 1450 and 1900 nm so I know if I want to detect water in my project then i obviously should not use those wavelengths. Is there an equation that tells me the absorption strength of certain chemical compounds at different wavelengths ?

enter image description here

  • 1
    Welcome to GIS SE RuiQi. Could you please narrow this question down to one answerable question? As you are new to this community, please consider taking the tour: gis.stackexchange.com/tour – Aaron May 26 '16 at 1:57
  • 1
    Hi, I have edited my question. I hope it is much clearer now – RuiQi May 26 '16 at 2:21
  • If such a table exists it's a very closely guarded trade secret. Vast amounts of study in this area has been done with particular interest to a specific sensor/software as they have access to and are keeping their findings secret or not in readable form. In my experience hyperspectal imagery is classified (supervised) to find areas of interest using 'training areas', for example you indicate where you know to be pine trees and the software finds everything similar and calls it pine trees. Disclaimer: I have only had occasional exposure to this imagery and have not used all available software. – Michael Stimson May 26 '16 at 4:29

The usual library is ASTER.

For application of that, and background, I suggest Baldridge, A. M., S.J. Hook, C.I. Grove and G. Rivera, 2009.. The ASTER Spectral Library Version 2.0. Remote Sensing of Environment, vol 113, pp. 711-715.

For water, do you want liquid or frozen water? Sea water or fresh? Distilled or tap? In any case, start at http://speclib.jpl.nasa.gov/search-1/water

Don't forget that what you are sensing the signature through contributes to the spectral response.

  • Thanks for the database ! Am i right to assume that there is no known or general formula that relates the chemical composition / color / etc of a material to its spectral signature ? I realized that this was actually what my question was. – RuiQi May 27 '16 at 1:51
  • I don't think the question makes sense. The spectrum of something is a set of measurements (typically thousands of points), not a single value that could come out of an equation. – BradHards May 27 '16 at 3:25
  • I mean the input to the formula could be the chemical composition / etc. and a certain wavelength and the output would be the approximated reflectance at that wavelength. I don't know though its just a thought. Im a computer engineering student doing a simple project on image processing so I always thought the guys over at material science (whatever this field is in) would have a general formula. – RuiQi May 27 '16 at 6:26
  • @BradHards I've a question that's tangentially related to the original question: is the ASTER library available on the internet, or is it necessary to fill out the request and receive a CD containing the data via snail? – blazs Oct 28 '16 at 13:33
  • @blazs that is a long way from GIS.SE scope, but you can get individual files via the search engine (e.g. Conifers is at speclib.jpl.nasa.gov/speclibdata/…). It doesn't seem to be possible to download everything at once though. – BradHards Oct 28 '16 at 23:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.