Dear Pix4D Support Team,
I am conducting research related to the creation of prescription (VRA) maps based on NDVI layers in Pix4Dfields.
I need to know the exact interpolation method that Pix4Dfields uses when generating VRA maps (e.g., IDW, kriging, spline, or another algorithm), as well as whether the user can influence the choice of this model.
Could you please confirm:
-
Which interpolation algorithm is used when zoning NDVI values in VRA maps?
-
Can the resolution and interpolation parameters be adjusted by the user?
Thank you in advance for your assistance.
Kind regards,
Rados
Hi Rados,
PIX4Dfields zonation methods are as follows:
Equal intervals
It will split the values inside the visible pixels of the histogram into equal intervals/ranges. The number of intervals refers to the Number of zones set in the left-hand-side menu.
Quantile
This zone distribution will be split into regular intervals, considering the data distribution. Quantiles are values that divide a data set into equal-sized portions. They are used to describe the distribution of data by indicating where specific data points fall relative to the whole set.
This setting is helpful to highlight and boost even slight heterogeneity in your crop development, allowing you to exploit those for application. Keep in mind that the quantile zones will represent the same area in the field, if setting the Zone detail to Fine.
Manual intervals
It helps if the user has a model using a certain vegetation index and knows it relates to a field status. Then, they need to manually set the brackets.
Thank you for your answer Jose.
The NDVI image is already a raster (each pixel has a value from -1 to +1) and here are no gaps between points — every pixel from the camera is already precisely georeferenced so, there is no need to “predict” values between samples as in chemical soil analysis (e.g., IDW, kriging)
Pix4Dfields can immediately take those pixels and crop them according to Equal Intervals or Quantile rules.
Please correct me if I got it wrong.