NDVI Camera

An NDVI (Normalized Difference Vegetation Index) camera is a specialized multispectral imaging device used primarily to assess plant health and monitor vegetative density. Deployed on platforms ranging from agricultural drones and tractors to low-Earth orbit satellites, these cameras are critical tools in precision agriculture, forestry management, and environmental science.

Scientific Principles

The operation of an NDVI camera relies on the way vegetative cellular structures interact with the electromagnetic spectrum.

Healthy plants contain chlorophyll, which strongly absorbs visible red light (typically around 660nm) to drive photosynthesis. Simultaneously, the spongy mesophyll leaf structure strongly reflects near-infrared (NIR) light (typically around 850nm) to prevent the plant from overheating. When a plant experiences stress—such as disease, nutrient deficiency, or dehydration—chlorophyll production drops, causing it to absorb less red light and reflect less NIR light. By capturing and comparing these two specific wavelengths, an NDVI camera can detect stress before it becomes visible to the human eye.

Optical Design and Engineering

Standard commercial digital cameras are equipped with an optical shortpass filter (an IR-cut filter) over the sensor to block infrared light, mimicking human vision. To function as an NDVI camera, this internal filter is removed.

While some entry-level NDVI cameras use a single modified sensor, professional and scientific-grade setups utilize precision multi-sensor rigs to eliminate spectral crosstalk. The optical path in a multi-sensor rig generally involves:

  1. Broadband Light Collection: Incident light enters the primary objective lens.
  2. Dichroic Beamsplitter: The light strikes a dichroic mirror mounted at a precise angle. This optical component is engineered to transmit visible light while reflecting infrared wavelengths, splitting the image into two identical, simultaneous paths.
  3. Optical Bandpass Filters: The transmitted visible light travels toward a dedicated sensor, passing through an optical bandpass filter strictly centered at 660nm (often with a narrow Full Width at Half Maximum, or FWHM, of 10nm to 20nm) to isolate the Red channel.
    • The reflected NIR light travels toward a second sensor, passing through a different bandpass filter centered at 850nm (or 800nm, depending on the target vegetation) to isolate the NIR channel.

Optical Challenges: AOI and Blue Shift

The physical layout of optics is critical for accurate data collection. The Angle of Incidence (AOI) of the light striking the dichroic beamsplitter and the bandpass filters must be strictly controlled. If the objective lens creates a highly convergent light cone, light hits the edges of the thin-film filters at a steeper angle than at the center.

This steep angle induces a blue shift in the filter's transmission spectrum, altering its center wavelength. In multispectral imaging, a pronounced blue shift can allow unwanted background light to bleed into the sensors, fundamentally skewing the NDVI calculation for pixels located near the edges of the sensor.

The NDVI Calculation

Once the isolated Red and NIR light values are captured for a given area, the data is processed pixel-by-pixel using the normalized difference formula:

NDVI = (NIR - Red)/(NIR + Red)

The resulting values always fall on a scale between -1.0 and +1.0:

  • 0.6 to 1.0: Dense, healthy canopy with high chlorophyll absorption and high structural reflectance.
  • 0.2 to 0.5: Sparse vegetation, crops in early growth stages, or plants experiencing physiological stress.
  • 0 to 0.1: Bare soil, rock, or dead flora.
  • -1.0 to 0 (Negative values): Typically indicative of non-vegetative surfaces such as water, snow, or dense clouds.

Applications

  • Precision Agriculture: Farmers use NDVI cameras on drones to generate color-coded maps of their fields. This allows them to spot pest infestations, water shortages, or nutrient deficiencies days or weeks before the damage becomes visible to the naked eye.
  • Forestry Management: Monitoring the health of large forest canopies and tracking the impact of wildfires or deforestation.
  • Environmental Monitoring: Tracking drought conditions, mapping wetlands, and studying ecological changes over time.

Practical Application: Precision Agriculture

A common application of an NDVI camera is monitoring large-scale agricultural operations, such as a commercial vineyard, for water stress.

A multi-sensor NDVI rig is mounted to a drone and flown over the vineyard at solar noon to minimize shadows. As it flies, the dual sensors simultaneously trigger, capturing perfectly aligned pairs of 660nm and 850nm images. These raw image pairs are fed into photogrammetry software, which aligns the pixels and applies the NDVI formula to generate a false-color orthomosaic map.

 

Back to blog