One of the most common purposes of hyperspectral image data is to provide information on vegetation in an agricultural context. This information, such as health status, chlorophyll content and nitrogen demand, water content, structural information, etc. is hidden within the reflectance data.
In order to extract this information from the data one way is to use spectral indices, which belong to empirical-statistical methods. By this, specific wavelengths of the electromagnetic spectrum, which are known to be sensitive to the searched vegetation parameter, are used and set into relation.
A well-known example for this is the NDVI, which uses two different wavelength regions. In the visible red region chlorophyll content causes an absorption of the incoming radiation, while in the near infrared the cell structure of leaves causes a high reflectance value. The healthier a leaf is, the higher is the discrepancy between both values.