LogoLogo
Home PageCreate a Ticket
Double 4K Sensor
Double 4K Sensor
  • 📷Double 4K
    • Double 4K Sensor Image Capture
    • Double 4K - Firmware Release Notes
    • Double 4K - Updating Firmware
    • Double 4K - Sensor Troubleshooting
    • Double 4K - Hardware Overview
    • Double 4K - Using Pix4D
    • Double 4K Multispectral - NDVI and NDRE Calculations
    • Double 4K NDVI/NDRE Sensor - NDVI and NDRE Calculations
    • Double 4K - Accessing the Diagnostics Web Pages
    • ❓FAQS
      • FAQS: How do I use a new MicroSD Card with my Sentera Sensor?
Powered by GitBook
On this page
Export as PDF
  1. Double 4K

Double 4K Multispectral - NDVI and NDRE Calculations

This document describes the NDVI and NDRE calculations for the Double 4K (D4K) Multispectral camera.

PreviousDouble 4K - Using Pix4DNextDouble 4K NDVI/NDRE Sensor - NDVI and NDRE Calculations

Multispectral

The D4K Multispectral camera is a bit different than other cameras, since each band must be normalized to each other before indices that use both cameras are calculated.

For the RGB side of the camera, the stacked filters result in the following spectral response, as shown in Figure 1:

In Figure 1, red/green/blue correspond to channels of the resulting jpeg image. Note that there is some cross talk between the channels. That is, the blue channel has primarily blue wavelengths, but also has small contributions from the green and red wavelengths as well.

If it is necessary to separate the channels further, we can use a system of equations to subtract out the effect of the out of band channels on each band. This calculation assumes that the incoming light is approximately uniform (which is the case with daylight). We also add the constraint that the power across each of the filtered bands is equal. This allows us to perform calculations across each of the color bands, even if the filter widths vary.

This results in the following:

DN [color] = digital number of that band (from the jpeg)

[color] = Corrected color after subtracting out of band colors

Blue = 1.377DNblue - 0.182DNGreen - 0.061DNRed

Green = -0.199DNblue + 1.420DNGreen - 0.329DNRed

Red = -0.034DNblue - 0.110DNGreen + 1.150*DNRed

For the NIR/Red Edge camera, we have the following combined spectral response, as shown in Figure 2:

Note that from this imager, we only use the Blue channel for NIR and the Red channel for Red Edge. The green channel is discarded when performing index and band calculations. This is because there is not any additional useful information in the green band that isn't already captured by the blue and red bands.

Like the RGB method above (and subject to the same restrictions), we can subtract the out of band channels in the NIR and Red Edge portions via the following:

RedEdge = -0.956DNblue + 1.000DNred NIR = 2.426DNblue - 0.341DNred

When attempting to calculate indices that use both cameras, both the camera settings and the normalization factors used in the band math equations need to be considered during any index computations you may make.

You will first have to normalize each of your images for the total exposure opportunity. This will allow images with differing ISO and exposure times to be compared properly.

DN = Digital Number

Gain = ISO / 100 [Exif Tag]

Shutter = Shutter time in seconds [ Exif Tag]

ExposureOpportunity = DN / ( Gain * Shutter )

The normalization values for each band separation matrix were chosen independently for each camera to keep 8-bit values in their proper ranges (typically 0-255). A normalization value of 1/750 was used for RGB, and 1/277.7 was used for RedEdge. This means that a step increase in RGB values implies a 2.7x step in the RedEdge values. So, if you want to compute indices that use both cameras you must consider our band math gains. This is done after the bands are separated using the above system of equations.

So, you end up with the following:

NDVI = (2.700 * NIR_2 - Red_1) / (2.700 * NIR_2 + Red_1)

NDRE = (NIR_2 - RedEdge_2) / (NIR_2 + RedEdge_2)

📷
Figure 1: Stacked Filters
Figure 2: Combined Spectral