0

I am trying to improve image classification in GEE by using the STM of the mean. I first calculated the mean of the four bands.

Then I added the new column with mean to the image.

Then I used the new image with 5 bands (including the mean) to classify.

However, I keep getting errors in my classification. Here is my code:

Map.addLayer(image, imageVisParam);
Map.centerObject(image,11);

//select the bands 
var bands = ['b1', 'b2', 'b3', 'b4'];

var stm = image.select(bands).reduce(ee.Reducer.mean()).rename('STM')
print(stm)

var classNames = Water.merge(BuiltupArea).merge(Cropland).merge(Vegetation).merge(Wetland).merge(clouds)
print(classNames);

var imageWithSTM = image.addBands(stm)

var classifier = ee.Classifier.smileRandomForest(50)
  .train({
    features: classNames,
    classProperty: 'Class', 
    inputProperties: bands.concat(['STM'])
  });
  
print(classifier, 'classifier')

var classified = imageWithSTM.classify(classifier)


print(classified, 'classified');

//Add the classification to the map view, specify colours for classes
Map.addLayer(classified,
{min: 0, max: 5, palette: ['0000FF', 'FF0000', '008000','FFFF00','98ff00','ffc82d']},
'classification')

I get this error when I run the codes: Image (Error) Property 'b1' of feature '0' is missing. classification: Layer error: Property 'b1' of feature '0' is missing.

How can I change my code to use all five bands or more in the classification without getting an error?

Improve this question
5
  • 2
    What error do you get? Include your asset and training data in the code, so the error can be replicated. Make sure the assets are shared.
    – Daniel Wiell
    Sep 18 at 11:05
  • Please share the particular code snippet rather the whole repository link.
    – Padmanabha
    Sep 18 at 14:19
  • @Padmanabha I have edited the question with the code snippet
    – Ndahambelela
    Sep 18 at 18:48
  • Perhaps feature 0 is outside the bounds of the image? check developers.google.com/earth-engine/guides/…
    – Oliver Lopez
    Sep 25 at 13:32
  • It seems like that's not the case. When I run the classification using the same training data but without the additional STM column, it functions perfectly. The issue seems to have been introduced when I added the new field.
    – Ndahambelela
    Sep 26 at 9:50

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.