Question: Why the color of Infragram picture is different from what I take with my Raspberry Pi using Picamera API

iman is asking a question about ndvi
Follow this topic

by iman | March 30, 2021 19:01 | #26064


When I connect to my Camera using Raspberry pi local network and Infragram, the picture is:

image description

But when I take a picture with my Raspberry Pi using Picamera API with the same settings as Infragram, the picture is:

image description

Why is it different while settings are the same? Is Infragram's preview picture, a processed picture? If yes, how can I process the latter picture to get the former's colours?



4 Comments

@cfastie , wondering if you might be around and able to drop some knowledge on this question?

Is this a question? Click here to post it to the Questions page.

Reply to this comment...


I don't know what your camera's settings were and I don't know how your camera has been modified (does it have a red filter?). I don't know what the Infragram settings were and I don't always understand how Infragram works.

However, the blue photo looks a lot like photos taken by cameras modified to capture NIR (e.g., a Pi NoIR, or camera with the IR cut filter removed) and then modified by adding a red filter that passes only red and NIR. If the white balance settings exaggerate the blue channel (where NIR is captured) then foliage in the photos will have the blue tint as in your blue photo.

The red photo looks like one taken by the same camera without the white balance exaggeration of the blue channel. The red channel will capture more light (for three reasons) and without an artificial white balance setting foliage will look red.

The blue photos can sometimes be used to create NDVI images by using the red channel for red light and the blue channel for NIR light. The red photos generally cannot be used to create meaningful NDVI images unless they are RAW image files apparently because the jpeg processing discards some information.

Chris

Is this a question? Click here to post it to the Questions page.

Thank you @cfastie!


Reply to this comment...


Thanks @cfastie . The problem was with the Auto white balance and changing the awb from "auto" to "greyworld" solved the issue.

Reply to this comment...


Log in to comment