Specifying region in export image in GEE Python API

Question

When I export an image in GEE code editor I can pass a geometry to the 'region' parameter without problem. However, it doesn't work for Python API and I get below error for the shown code:

task = ee.batch.Export.image.toCloudStorage(
    image=imageOfSeries.toArray(),
    description=AssetName,
    fileNamePrefix=AssetName,
    bucket=GCbucketName,
    scale=scale,
    region=box,
    fileFormat='TFRecord',
    formatOptions={
        'patchDimensions': [patch_size, patch_size],
        'tensorDepths': [tDepth],
        'compressed': True
    }
)


Traceback (most recent call last):
  File "D:/Shahriar/OneDrive - SUNY ESF/Thesis/Codes/Landsat_Classification/ExtractLandsatArrayImage.py", line 205, in <module>
    'compressed': True
  File "C:\Python36\lib\site-packages\ee\batch.py", line 332, in toCloudStorage
    Task.ExportDestination.GCS)
  File "C:\Python36\lib\site-packages\ee\batch.py", line 854, in _prepare_image_export_config
    _canonicalize_parameters(config, export_destination)
  File "C:\Python36\lib\site-packages\ee\batch.py", line 1418, in _canonicalize_parameters
    config['region'] = _canonicalize_region(config['region'])
  File "C:\Python36\lib\site-packages\ee\batch.py", line 1503, in _canonicalize_region
    return json.dumps(region)
  File "C:\Python36\lib\json\__init__.py", line 231, in dumps
    return _default_encoder.encode(obj)
  File "C:\Python36\lib\json\encoder.py", line 199, in encode
    chunks = self.iterencode(o, _one_shot=True)
  File "C:\Python36\lib\json\encoder.py", line 257, in iterencode
    return _iterencode(o, 0)
  File "C:\Python36\lib\json\encoder.py", line 180, in default
    o.__class__.__name__)
TypeError: Object of type 'Geometry' is not JSON serializable

How can I solve this problem?

Answer 1

I think getting box as a dict makes it JSON serializable, try:

task = ee.batch.Export.image.toCloudStorage(
    image=imageOfSeries.toArray(),
    description=AssetName,
    fileNamePrefix=AssetName,
    bucket=GCbucketName,
    scale=scale,
    region=box.getInfo(),
    fileFormat='TFRecord',
    formatOptions={
        'patchDimensions': [patch_size, patch_size],
        'tensorDepths': [tDepth],
        'compressed': True
    }
)

In case you have another issue (or the same), you can also try a function I included in geetools:

from geetools import tools

region = tools.geometry.getRegion(box)

task = ee.batch.Export.image.toCloudStorage(
    image=imageOfSeries.toArray(),
    description=AssetName,
    fileNamePrefix=AssetName,
    bucket=GCbucketName,
    scale=scale,
    region=region,
    fileFormat='TFRecord',
    formatOptions={
        'patchDimensions': [patch_size, patch_size],
        'tensorDepths': [tDepth],
        'compressed': True
    }
)