I am trying to split the Hansen dataset https://developers.google.com/earth-engine/datasets/catalog/UMD_hansen_global_forest_change_2018_v1_6?hl=fr into 250x250 pixel tiles within a region.

For that I think I can use image.clip(geometry), but I am not sure how I can split a region in rectangles (how to map the rectangle coordinates to the pixels) so that each of them have the desired 250x250 pixel size.

  • 1
    Hi Mr. Liu, what have you tried so far? include some code. What is your final goal? Apr 21, 2019 at 20:39
  • gis.stackexchange.com/questions/315541/… . You possibly want to use neighborhoodToArray(KERNEL)
    – Kuik
    Apr 25, 2019 at 13:47
  • @RodrigoE.Principe I'm trying to get a bunch of 250x250 pixel tiles, 30m per pixel, from the Hansen Data within a specific region. I plan to feed those tiles to Tensorflow. I tried to use image.clip(geometry) but that normally requires some sort of Geometry, so I'm not sure how to proceed.
    – Mr. Liu
    May 5, 2019 at 18:27
  • @Kuik could you explain a bit how neighborhoodToArray(KERNEL) works? I'm not sure if I understand it propertly. Does it store in every pixel, a matrix with the values of the neighbors up to the size of the kernel, being that pixel the center of the matrix?
    – Mr. Liu
    May 5, 2019 at 23:13


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.