Developed by Agrosatélite Geotecnologia Aplicada Ltda.
This folder contains the scripts to classify, separate and post-process the Temporary Crop and Perenial Crop classes.
We recommend that you read the Agriculture Appendix of the Algorithm Theoretical Basis Document (ATBD), since important informations about the classification and separation methodology can be found in there.
You can see the metodology for temporary and perrenial classification in the image bellow:
First, you need to copy the scripts in this folder and in the utils folder to your Google Earth Engine (GEE) account.
Then, in your GEE account, go to the Assets tab and create the following directory structure:
- MAPBIOMAS/C5/AGRICULTURE/TEMPORARY_PERENNIAL/RESULTS
and three Image Collections:
- MAPBIOMAS/C5/AGRICULTURE/TEMPORARY_PERENNIAL/RESULTS/RAW
- MAPBIOMAS/C5/AGRICULTURE/TEMPORARY_PERENNIAL/RESULTS/SEPARATION
- MAPBIOMAS/C5/AGRICULTURE/TEMPORARY_PERENNIAL/REFERENCE_MAP
On the REFERENCE_MAP collection, you must provide reference maps for the sampling, with pixel values of 1 for sugar cane and 0 for other classes, and a property year
with the year of the map.
You also need to provide the training samples for the crops separation in Feature Collection:
- MAPBIOMAS/C5/AGRICULTURE/TEMPORARY_PERENNIAL/SAMPLES
Each sample must have a property class
with value 1 for annual crop sample and 2 for perrenial crop sample.
To run the classification, follow these steps:
-
Open the script agriculture/temporary_perennial/classification.js;
-
On line 2 (variable
api
), set the path to the api.js script you copied to your GEE account; -
On line 7 (variable
outputCollection
), set the output path for the classification results; -
On line 10 (variable
years
), set the years you want to classify; -
On line 17 (variable
tiles
), set the WRS (path and row) you want to classify; -
On line 42 (variable
periods
), set the periods of the region you want to classify (more information about that you can read on the ATBD Agriculture Appendix); -
On line 67 (variable
imageCollection
), set the collection you want to use to create the mosaics; -
On line 70 (variable
referenceCollection
), set the path to your reference map that will be used for sampling; -
Run the script.
To apply the temporal and spatial filters, follow these steps:
-
Open the script agriculture/temporary_perennial/temporal_spatial_filter.js;
-
On line 2 (variable
input
), set the input path to the raw classification result; -
On line 5 (variable
output
), set the path for the filtered result; -
On lines 8 and 9 (variables
startYear
andendYear
), set the year range you want to apply this script; -
Run the script.
To run the separatino os temporaryand perennial crops, follow these steps:
-
Open the script agriculture/temporary_perennial/separation.js;
-
On line 2 (variable
api
), set the path to the api.js script you copied to your GEE account; -
On line 5 (variable
normalization
), set the path to the normalization.js script you copied to your GEE account; -
On line 10 (variable
input
), set the input path to the filtered classification result; -
On line 13 (variable
outputCollection
), set the output path for the classification results; -
On line 16 (variable
years
), set the years you want to classify; -
On line 23 (variable
tiles
), set the WRS (path and row) you want to classify; -
On line 45 (variable
periods
), set the periods of the region you want to classify (more information about that you can read on the ATBD Agriculture Appendix (ATBD)); -
On line 75 (variable
trainingSamples
), set the path to your training samples; -
Run the script.
To run the temporary crops post-processing, follow these steps:
-
Open the script agriculture/temporary_perennial/temporary_temporal_spatial_filter.js;
-
On line 2 (variable
input
), set the input path to the raw separation result; -
On line 5 (variable
output
), set the path for the temporary crop filtered result; -
On line 8 (variable
classOfInterest
), set the pixel value of the temporary crop as noted on the crops separation map; -
On lines 11 and 12 (variables
startYear
andendYear
), set the range of years you want to filter; -
Run the script.
To run the perennial crops post-processing, follow these steps:
-
Open the script agriculture/temporary_perennial/perennial_temporal_spatial_filter.js;
-
On line 2 (variable
input
), set the input path to the raw separation result; -
On line 5 (variable
output
), set the path for the temporary crop filtered result; -
On line 8 (variable
classOfInterest
), set the pixel value of the perennial crop as noted on the crops separation map; -
On lines 11 and 12 (variables
startYear
andendYear
), set the range of years you want to filter; -
Run the script.