What are multispectral and hyperspectral images?

Multispectral and hyperspectral images are types of images obtained using special sensors that record reflected light in different spectral ranges.

• Multispectral images contain data in a limited number (usually from 3 to 10) of broad spectral channels, such as red, green, blue, near-infrared, etc. They are widely used in remote sensing for monitoring vegetation, water bodies, soils, etc.

• Hyperspectral images have many more channels (hundreds or more), each covering a narrow spectral range. This allows for detailed spectral information about materials and objects, which is useful for precise classification, mineral detection, composition analysis, and other tasks.

Thus, hyperspectral data provide a deeper understanding of surface characteristics but require more computational resources for processing.