I am trying to perform a supervised image classification to detect water surface area change on two different LANDSAT TM 5 Images, separated by several years. How can I add the numeric property, 'class', to store the binary landcover values water or non-water in my ground truth points? After creating a collection of points on my image, I tried splitting the data into the validation and training sets. The points didn't extract the landcover type into the feature collection properties though. I tried editing the feature collection properties from the settings displayed in the imports section at the very top, but I could not specify a numeric property in order to store landcover type.
Here is a link to my code: https://code.earthengine.google.com/9b906a6e605dcdb08c367480f68a7ff7
Best Answer
I'd do something like this:
You can access the feature collections in the following link: https://code.earthengine.google.com/3f4c2908fcf7c469357b80783f3ef26f
You should mask clouds, use another path/row to include the missing part of the lake, and make more training points. You can also try another classifier like svm, randomForest, etc.