How can I have a value of NDVI starting from an IR picture? I want to find a numeric value of NDV...
Public Lab is an open community which collaboratively develops accessible, open source, Do-It-Yourself technologies for investigating local environmental health and justice issues.
How can I have a value of NDVI starting from an IR picture? I want to find a numeric value of NDVI in pictures taken with a normal camera without IR filter. is it possible?
Computing NDVI requires a value for the "brightness" of both red light and near infrared light for the same part of a scene (or for each pixel). Unmodified consumer cameras do not record near infrared light so their photos cannot produce NDVI. If you remove the IR cut filter from a camera, then all three color channels (R,G,B) will have near infrared light mixed in with the red, green, or blue light. The proportion of NIR:Visible for any pixel will vary greatly depending on several factors, so it is not possible to know how much of the "brightness" is from visible versus NIR light.
If you then add a filter which blocks only blue and green light (a red filter), then the blue channel will capture mostly NIR light and the red filter will capture red light with some NIR mixed in. The brightness of the NIR in the blue channel will not be as great as the relative brightness of NIR in the scene because typical sensors are not as sensitive to NIR as they are to visible light. If you exaggerate the brightness of the blue channel, the values in the blue and red channels can represent NIR and RED, respectively, for computing NDVI. Exaggerating the values in the blue channel can be done with a custom white balance or with a calibration procedure.
Computing NDVI requires a value for the "brightness" of both red light and near infrared light for the same part of a scene (or for each pixel). Unmodified consumer cameras do not record near infrared light so their photos cannot produce NDVI. If you remove the IR cut filter from a camera, then all three color channels (R,G,B) will have near infrared light mixed in with the red, green, or blue light. The proportion of NIR:Visible for any pixel will vary greatly depending on several factors, so it is not possible to know how much of the "brightness" is from visible versus NIR light.
If you then add a filter which blocks only blue and green light (a red filter), then the blue channel will capture mostly NIR light and the red filter will capture red light with some NIR mixed in. The brightness of the NIR in the blue channel will not be as great as the relative brightness of NIR in the scene because typical sensors are not as sensitive to NIR as they are to visible light. If you exaggerate the brightness of the blue channel, the values in the blue and red channels can represent NIR and RED, respectively, for computing NDVI. Exaggerating the values in the blue channel can be done with a custom white balance or with a calibration procedure.
More here: https://publiclab.org/wiki/ndvi-plots-ir-kit
Chris
Reply to this comment...
Log in to comment