Using Google Earth Engine, I'm trying to do a supervised classification over a semi-urban area by placing points with different numeric classes on areas of vegetation, impermeable surfaces (roads and buildings), and exposed soil. I have about 50 points of each classification type and have merged them into a new feature collection. When I try to import a landsat image over my area of interest, the image only covers about half of the points. In the tutorial "Hands-on Intermediate Training", at about 12:10, the same thing happens but he is able to correct the issue by moving one point:


I'm using essentially the same code as in the tutorial. I've moved about 60 points around and nothing changes.

Has anyone else had this problem and had any success getting the image to cover all the points?

  • Could you include the code that you are using, so that others can replicate the behavior? Nov 15, 2017 at 19:23

1 Answer 1


Without seeing code it's hard to say, but my recollection is that the roi point was moved to the scene center (NOT the training points). The better option might be to classify one of these instead of a single scene:

var l8raw = ee.ImageCollection('LANDSAT/LC08/C01/T1');
var composite = ee.Algorithms.Landsat.simpleComposite({
  collection: l8raw, 
  asFloat: true
Map.addLayer(composite, {bands: ['B4', 'B3', 'B2'], max: 0.3}, 'composite');

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.