I can't figure out what's causing some images to have a different color.???
I'm flying a matrice100 w/ X3 RGB Camera taking wheat images to be stitched together for plant health. With the coloring being erratic, the plant health analysis will not work correctly.

It can come from the weather, cloud passage.  The white balance should not be on auto.

another subject: what method do you use to do crop monitoring in RGB?

LV_Forestry Posted at 12-21 22:39
another subject: what method do you use to do crop monitoring in RGB?

Actually I am experimenting. I have a X3 NIR camera as well. Only in the last few days I started “trying “ to use the RGB , thinking it would be helpful in my crop scouting to have a RGB image . Of course the issue with the random coloring of the images is screwing up any analysis after stitching.

LV_Forestry Posted at 12-21 22:37
It can come from the weather, cloud passage.  The white balance should not be on auto.

Thanks for the info. I’ll see if I can find the white balance setting.

Ky Farmer Posted at 12-22 03:54
Actually I am experimenting. I have a X3 NIR camera as well. Only in the last few days I started “trying “ to use the RGB , thinking it would be helpful in my crop scouting to have a RGB image . Of course the issue with the random coloring of the images is screwing up any analysis after stitching.

As far as method. My main challenge is determining which solution to use . My end goal is to have the ability end up with an image to be loaded to John Deere ops center or to my Farm data  Management software (FieldAlytics). This seems to be a challenge without spending a lot of money for a few different solutions to get from mission planning to a image that can be ingested into my main GIS management platforms. It’s been like “herding cats”.

Ky Farmer Posted at 12-22 04:02
As far as method. My main challenge is determining which solution to use . My end goal is to have the ability end up with an image to be loaded to John Deere ops center or to my Farm data  Management software (FieldAlytics). This seems to be a challenge without spending a lot of money for a few different solutions to get from mission planning to a image that can be ingested into my main GIS management platforms. It’s been like “herding cats”.

With the RGB alone you can already do weed detection if the resolution is satisfactory and the culture is not too advanced. With the Red and Green and NIR Band you can do NDVI and GNDVI, see which best suits your crops.Don't rush to the miracle software that promises you prescription maps and crop forecasts...

With QGIS you can do wonders, and it's open source. Do not hesitate to send me a set of RGB photos and also from X3 to see what it is possible to do with it.