Public Lab Research note

NDVI MAPS for Precision Farming

by Monoxyde | September 11, 2015 18:27 11 Sep 18:27 | #12210 | #12210

What I want to do

I want to get Aereal (UAV) prescription maps for precision farming

My attempt and results

As a first approach , I purchased in PublicLab , a Mobius Point & Shoot filter Wratten # 25red and a custom white balance. Good compromise between lightness and result but slow too extreme for aerial mapping and too many problems in the shadows. Too many aberrations. I recently changed a Canon Helf 130is with Kodak Wratten # 25 Red and made white balance ​​several tests.

SOME SHOOTS and ImageJ Outputs using NDVI_VGYRM Lut :







What do you think? Some advice? In the meantime, I purchased a Canon PowerShot SX260 HS , with GPS for the conversion . In both Powershot , I installed CHDK ( ) to have additional features such as INTERVALLOMETER and many other extra Canon features. If I obtain satisfactory results , I will proceed to the flight test.


These are all very good NDVI images. The NDVI values seem to be close to the correct range, and there are not that many anomalies (most leaves are green and few non-leaves are green). Healthy foliage should probably have NDVI values closer to 5, so maybe there should be more yellow in the images. But the absolute values for NDVI will always be shifted unless a calibration procedure is used. The protocol you are using has great potential to make useful NDVI maps of farm fields.

Reply to this comment...

Thanks Chris, your judgment lifts me up. I will try to make different calibrations. As soon as I can, I share the results and procedures. Thanks again for your time. Luca.

Reply to this comment...

Chris, do you feel that the Photomonitoring plugin is able to work better if I provide two images (from a modified camera and the other by the same camera but not edited), at the same time?

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

Reply to this comment...

I am not sure what your question is. You could use the two-camera feature of the plugin but give it the same Wratten 25 photo for both the VIS and NIR photos. Tell it to use the appropriate channel from each photo, and I assume the plugin would give the same result as if you used the single camera feature and gave it the same photo and chose the appropriate channels.

Reply to this comment...

Thank you, I thought, that by offering to plugin a VIS photo plus a NIR photo, Shooting by two identical cameras with the same parameters ISO etc ... it might help to extract data more faithful. Anyway, I'll find a way to bring in, more yellow channel, As well as from your As well as from your observation. Thank you Chris. Luca.

Reply to this comment...

Login to comment.