What is the best Index using Pix4D Fields to identify high vigor spots in a vineyard at harvest?
1 Like
Great question Donald.
I think there could be a few indices that are helpful, but first we need to filter out data that makes the index messy. Here is a NDVI example from Washington. You can see a lot of content here.
First thing first, lets trim the data to focus on the crop only, even a block at a time if necessary (so we compare oranges to oranges/apples to apples). Poor joke.
Then create your index, in this example NDVI:
Now when I use the advanced layer visualization tools on the right I focus the right to mask out the soil by limiting the range on the histogram by moving it to above 0.3 (typical for soil). I tend to place my low and high marks on either side of the major spike in the histogram.
That provides a pretty dynamic image. If you leave the options to transparent then anything below that range will not be visible.
Another option to add is the histogram equalization.
Feel free to change the palette to your preference. I tend to like the thermal/glow one or even grayscale.
Now you can see the variation across this vineyard by row, there are some areas or renovation and areas of more mature vines that are more bright, obviously there is an edge effect on the plants likely from the dust and other pest pressure. I would share to PIX4Dcloud and scout to ground truth and then use the map and annotations to help manage and communicate with team members.
There are many other indices out there, NDVI is just one simple example. I would do a little research on the web and at your local university to see what works best in your area for your crops and conditions. The University Extension is one of the best resources we have in the US.
1 Like
Thank you very much. This is an area I have not become the most skilled. I will put some time on this to get the most out of the platform.
Thanks, Donald
Sorry to open an old post! How did you get on with this Donald?
Is there a way to measure the height and width of the vines with the aim of calculating a potential yield?
Hi Chris,
You could process using Accurate processing to obtain a detailed DSM. The relation with potential yield would be something to investigate. Do you have any papers to consult?