Vegetation indices calculed by Fiji

Hi there,
Did anyone calculated vegetation indices from images with Fiji, or by any other free image analysis software?
Thanks in advance,

1 Like

if you have the appropriate images layers you can simply calculate the indices (e.g., NDVI):

If you have problems to import an georeference image use OpenSouce GIS tool, e.g.:

Quantum GIS:
Saga GIS:

The include tools to calculate vegetation inidices.

1 Like

Thanks for your answer. For example this is the image that i would like to process

And to calculate NDVI index i need NIR channel which i cant get from Fiji, from Fiji i can get only R, B and G channel so that is the problem for me.


Hello Pedja,
There are many ways to accomplish what you want to do without special equiptment. The software used by all the sources is simply a ratio of bands, it is just that they know what they are looking for and how to spot it in the images. Attached is an example of that type of processing using your own image (which is very good quality by the way).

List your objectives in order of importance, such as Good/Bad plant ratio or ground cover percentage etc…etc… and go from there and we can help much further.

Hello again Pedja,
Just in case I may have confused you I have attached an actual NDVI image from this past Sunday (2019-03-11) and the imformation it contains. This is how it is sent to the user before any analysis is performed. That is left up to the user because they are the ones that are familiar with what data is needed for the analysis which can be many different ways. A NVDI image simply uses bands which are more sensitive to vegatation. The analysis is done mathematically or by using algorithms to classify the information needed.

I am also attaching your image with the same processing as is typically done to extract data. I just used guess work as to what I think are healthy plants, what are weak plants and what is just dirt. So see if this is what you had in mind.
Red= Healthy plants, Green= Weak plants and Blue= Dirt.
The classification can be much tighter with exact specs. And can be done in ImageJ.

Hello Bob,
Thank you for answer. I must say that i was confused a little, but if i understood you processed my image and this is how it looks like when you calculate NDVI index from my image in Fiji or m i wrong?
Also if that is true can you write me step by step how you did it in Fiji?

Hello again Pedja,
Yes, that is a NDVI image of your image.
Because of the great quality of your image all I done was to use Trainable Weka Segmentation. You will find it in the Plugins menu under Segmentation.I just choose what I wanted to classify (healthy plants, weak plants, and dirt) and trained the algorithm for them. I caution you because with your images it will be easy to OVER train the algorithm. Overall it is easy to use and understand so it may be all you need to do for what you want.
As for the NDVI, it is only an index or colorchart to go by. The client still has to decide what it is they are looking for.
But just for the fun of it I enhansed the Signal/Noise ratio (or dynamic range) of each channel of your image and recombined them to show you how your image can be enhansed if desired, but I doubt you will need to do that.

If you have any further questions just ask,

1 Like

Thank you very much for your help.
I have tried Trainable Weka Segmentation and it worked, so once more thank you.

You are quite welcome. I am glad to have helped.


Guys, how can I make my own vegetation index?

I want to use just RGB channels.

For example: VARI Index = (Rgreen −Rred)/(Rgreen +Rred −Rblue)

Thanks a lot,

Hello Tiago,
Yes, that would be one formula to use. There are very many variations used to analyze vegetation. The typical NDVI index only looks for differences in water content between areas and plants. The exact data you are looking for can be narrowed greatly many different ways.
A bit of advice, subtract the blue from the dividend (first parenthese) instead of the second. The blue affects both bands both before and after.
If you need more specific advice, feel free to ask.