Statistical analysis on time series pixel by pixel

I would like to know if there are any functions or libraries to simplify the statistical analyzes of image collections pixel by pixel. For example, calculate the mean or variance of an index for each pixel of a collection of images by masking invalid pixels. This operation is very simple with the Google Earth Engine libraries. Is there something similar on Sentinel Hub as well?

Hi Gino,

I’m assuming that you are trying to calculate multi temporal statistics by pixel, for example, calculating the Maximum NDVI over a time period by pixel. With Sentinel Hub APIs you can do this within the custom script (a piece of javascript that is part of your request). For example, in the custom scripts repository we have a script that does this here.

This can easily be adapted to calculate statistics for other indices or individual bands. You can even generate cloudless mosaics using the same tools. You can read more about these functions in our documentation here.

As you are interested in statistics you can also bypass the pixels completely and generate statistics without downloading the underlying imagery using our Statistical API.

If you have any more questions or I have got the wrong end of the stick then let us know :slight_smile: