Question: NDVI: How is NIR value removed/separated from the RED+NIR that enters the red channel when using a single camera?

Anice is asking a question about general
Follow this topic

by Anice | December 13, 2018 02:37 | #17883

Good day! I am new to the concept of NDVI and got interested in it since it allows for easier monitoring of the health of a crop with a use of a camera only. I have at hand the infragram kit and successfully modified a Canon camera (CHDK enabled) using red filter.

I would like to inquire regarding the post-processing of acquired images. I have learned that the red filter allows Red + NIR to pass. How can Red and NIR values be separated in the red channel? I have read in some comments that(a) 80% of the NIR in the blue channel is subtracted from the values in the red, and (b) 40% of the NIR is subtracted. What are the basis for these percentage? What factors/settings may affect this percentage?


You are correct that both red and NIR pass through the red filter (Rosco Fire # 19) and both are captured in the red channel of the photograph. It is not possible to distinguish the colors in a single channel after the photo has been taken.

For some cameras the spectral response of each channel is known, and for very few cameras this spectral response includes the NIR region. In even fewer cases, the spectral response of the camera after it has been converted with a red filter is known.

An example of this type of information and how it can be used to estimate the degree of NIR contamination in the red channel is here. For the PowerShot A2200 described there, the amount of NIR contaminating the red channel should be equal to about twice the brightness of the blue channel.


Thank you so much Chris! This information is very helpful. I plan on researching about the spectral sensitivity of my camera (Canon SX230 HS) and see if the adjustment for the Powershot A2200 may also be applicable for it.

Regards, Anice

Reply to this comment...

Log in to comment