Hyperspectral Thermal Image Unmixing
The most common images contain the three bands of the visible to the human eye spectrum: Red, Green and Blue, visualized through the RGB color model.
Satellite images usually are obtained with multispectral sensors. This means that the captured images have not only these 3 common bands, but also bands beyond the visible light range in the electromagnetic spectrum, e.g. in the Near Infrared (NIR), Short Wave Infrared (SWIR) or in Thermal Infrared (TIR). This allows the determination and extraction of more information from the image that the human eye cannot normally perceive.
On the other hand, hyperspectral images contain even more bands, having shorter wavelength range when compared to simple RGB or Multispectral images, and thus, even more information can be acquired throughout the E/M spectrum.
One of the uses of image spectroscopy - or hyperspectral imaging, is the identification of the materials that are present in the scene. Some of the fields that this can be applied are in agriculture, geology, astronomy, molecular biology and physics.
And this becomes feasible, as generaly, each pixel value of a hyperspectral image can be compared to a material database, regarding the spectral signature that it generates. However, when it comes to low spatial resolution depending to the application, each pixel may contain more than one material. That means that in each pixel, a mixture of different materials exists.
The process of the ‘unmixing’ of these mixed pixels is called hyperspectral image unmixing.
In order to better and quicker understand the theoretical background and the procedure of the hyperspectral image unmixing, GEO University provides a selection of video notes crafted to present you the use of the spectral unmixing process, with focus on thermal images.
Hyperspectral Thermal Image Unmixing is a free course with which you can quickly understand spectral unmixing on thermal images.
Initially, some thermal imaging concepts are presented along with the concepts of spectral mixing and unmixing.
A set of notes is also included that act as a guide to apply spectral unmixing on MODIS thermal images.
And all of the above, with the use of the ENVI software. But of course, it is up to you the choice of the software for the implementation of the showcased algorithms.
What will you learn?
- Basic thermal theory concepts
- Spectral mixing/unmixing
- Application of spectral unmixing on thermal data
Need more information?
E-mail here: info@geo.university
