Multispectral imaging records spatial images in several selected wavelength bands. An optical filter array, separate filtered optical paths or a tunable selection mechanism lets different detector channels sample different portions of the optical spectrum.
Comparing red and near-infrared spectral reflectance can distinguish vegetation-like responses from some other surfaces, but measured channel values also depend on illumination, atmosphere and sensor response. The bands must sample corresponding locations: motion between sequential exposures can create apparent spectral differences at edges. More bands add information only when their calibration, spatial sampling and signal-to-noise ratio support the intended comparison.
An image formed by dividing corresponding calibrated pixels in two spectral bands. If their signals have a common multiplicative factor, and , their ratio is and does not depend on . This explains suppression of some common illumination effects and enhancement of spectral contrasts. Additive atmospheric radiance and small denominators prevent exact cancellation and can amplify errors. Normalized difference vegetation index is a related normalized spectral contrast.
The normalized contrast of red and near-infrared reflectances. Large positive values indicate a vegetation-like spectrum, but soil, atmosphere, shadows and canopy saturation affect its interpretation.
A displayed image assigning measured bands to colour channels that differ from ordinary visual perception. Assigning near-infrared to red is a common way to highlight vegetation.
The steep increase of healthy green vegetation reflectance between red absorption and the high near-infrared scattering plateau, typically around 0.68–0.75 micrometres.
Articles by others on the same topic
Multispectral imaging is a technique that captures image data at specific frequency ranges across the electromagnetic spectrum. Unlike traditional imaging that typically uses only visible light, multispectral imaging collects data across multiple wavelengths, including ultraviolet, visible, and infrared light. The key features of multispectral imaging include: 1. **Multiple Wavelengths**: Multispectral cameras capture data from several discrete bands, usually ranging from 3 to 10 different wavelengths, though some systems may capture more.