eolearn.features.radiometric_normalization

Module for radiometric normalization

class eolearn.features.radiometric_normalization.ReferenceScenesTask(*args, **kwargs)[source]

Bases: eolearn.core.eotask.EOTask

Creates a layer of reference scenes which have the highest fraction of valid pixels.

The number of reference scenes is limited to a definable number.

Contributor: Johannes Schmid, GeoVille Information Systems GmbH, 2018

Parameters
  • feature ((FeatureType, str) or (FeatureType, str, str)) – Name of the eopatch data layer. Needs to be of the FeatureType “DATA”.

  • valid_fraction_feature ((FeatureType, str)) – Name of the layer containing the valid fraction obtained with the EOTask ‘AddValidDataFraction’. Needs to be of the FeatureType “SCALAR”.

  • max_scene_number (int) – Maximum number of reference scenes taken for the creation of the composite. By default, the maximum number of scenes equals the number of time frames

Stores initialization parameters and the order to the instance attribute init_args.

execute(eopatch)[source]

Implement execute function

class eolearn.features.radiometric_normalization.BaseCompositingTask(*args, **kwargs)[source]

Bases: eolearn.core.eotask.EOTask

Base class to create a composite of reference scenes

Contributor: Johannes Schmid, GeoVille Information Systems GmbH, 2018

Parameters
  • feature ((FeatureType, str)) – Feature holding the input time-series. Default type is FeatureType.DATA

  • feature_composite ((FeatureType, str)) – Type and name of output composite image. Default type is FeatureType.DATA_TIMELESS

  • percentile (int or list) – Percentile along the time dimension used for compositing. Methods use different percentiles

  • max_index (int or NaN) – Value used to flag indices with NaNs. Could be integer or NaN. Default is 255

  • interpolation (str) – Method used to compute percentile. Allowed values are {‘geoville’, ‘linear’, ‘lower’, ‘higher’, ‘midpoint’, ‘nearest’}. ‘geoville’ interpolation performs a custom implementation, while the other methods use the numpy percentile function. Default is ‘lower’

  • no_data_value (float or NaN) – Value in the composite assigned to non valid data points. Default is NaN

Stores initialization parameters and the order to the instance attribute init_args.

execute(eopatch)[source]

Compute composite array merging temporal frames according to the compositing method

Parameters

eopatch – eopatch holding time-series

Returns

eopatch with composite image of time-series

class eolearn.features.radiometric_normalization.BlueCompositingTask(*args, **kwargs)[source]

Bases: eolearn.features.radiometric_normalization.BaseCompositingTask

Blue band compositing method

  • blue (25th percentile of the blue band)

Parameters

blue_idx (int) – Index of blue band in feature array

Stores initialization parameters and the order to the instance attribute init_args.

class eolearn.features.radiometric_normalization.HOTCompositingTask(*args, **kwargs)[source]

Bases: eolearn.features.radiometric_normalization.BaseCompositingTask

HOT compositing method

  • HOT (Index using bands blue and red)

The HOT index is defined as per

Zhu, Z., & Woodcock, C. E. (2012). “Object-based cloud and cloud shadow detection in Landsat imagery.” Remote Sensing of Environment, 118, 83-94.

Parameters
  • blue_idx (int) – Index of blue band in feature array

  • red_idx (int) – Index of red band in feature array

Stores initialization parameters and the order to the instance attribute init_args.

class eolearn.features.radiometric_normalization.MaxNDVICompositingTask(*args, **kwargs)[source]

Bases: eolearn.features.radiometric_normalization.BaseCompositingTask

maxNDVI compositing method

  • maxNDVI (temporal maximum of NDVI)

Parameters
  • red_idx (int) – Index of red band in feature array

  • nir_idx (int) – Index of NIR band in feature array

Stores initialization parameters and the order to the instance attribute init_args.

class eolearn.features.radiometric_normalization.MaxNDWICompositingTask(*args, **kwargs)[source]

Bases: eolearn.features.radiometric_normalization.BaseCompositingTask

maxNDWI compositing method

  • maxNDWI (temporal maximum of NDWI)

Parameters
  • nir_idx (int) – Index of NIR band in feature array

  • swir1_idx (int) – Index of SWIR1 band in feature array

Stores initialization parameters and the order to the instance attribute init_args.

class eolearn.features.radiometric_normalization.MaxRatioCompositingTask(*args, **kwargs)[source]

Bases: eolearn.features.radiometric_normalization.BaseCompositingTask

maxRatio compositing method

  • maxRatio (temporal maximum of a ratio using bands blue, NIR and SWIR)

Parameters
  • blue_idx (int) – Index of blue band in feature array

  • nir_idx (int) – Index of NIR band in feature array

  • swir1_idx (int) – Index of SWIR1 band in feature array

Stores initialization parameters and the order to the instance attribute init_args.

class eolearn.features.radiometric_normalization.HistogramMatchingTask(*args, **kwargs)[source]

Bases: eolearn.core.eotask.EOTask

Histogram match of each band of each scene within a time-series with respect to the corresponding band of a reference composite.

Contributor: Johannes Schmid, GeoVille Information Systems GmbH, 2018

Parameters
  • feature ((FeatureType, str) or (FeatureType, str, str)) – Name of the eopatch data layer that will undergo a histogram match. Should be of the FeatureType “DATA”.

  • reference ((FeatureType, str)) – Name of the eopatch data layer that represents the reference for the histogram match. Should be of the FeatureType “DATA_TIMELESS”.

Stores initialization parameters and the order to the instance attribute init_args.

execute(eopatch)[source]

Perform histogram matching of the time-series with respect to a reference scene

Parameters

eopatch (EOPatch) – eopatch holding the time-series and reference data

Returns

The same eopatch instance with the normalised time-series

class eolearn.features.radiometric_normalization.ReferenceScenes(*args, **kwargs)[source]

Bases: eolearn.features.radiometric_normalization.ReferenceScenesTask

A deprecated version of ReferenceScenesTask

Stores initialization parameters and the order to the instance attribute init_args.

class eolearn.features.radiometric_normalization.BaseCompositing(*args, **kwargs)[source]

Bases: eolearn.features.radiometric_normalization.BaseCompositingTask

A deprecated version of BaseCompositingTask

Stores initialization parameters and the order to the instance attribute init_args.

class eolearn.features.radiometric_normalization.BlueCompositing(*args, **kwargs)[source]

Bases: eolearn.features.radiometric_normalization.BlueCompositingTask

A deprecated version of BlueCompositingTask

Stores initialization parameters and the order to the instance attribute init_args.

class eolearn.features.radiometric_normalization.MaxNDVICompositing(*args, **kwargs)[source]

Bases: eolearn.features.radiometric_normalization.MaxNDVICompositingTask

A deprecated version of MaxNDVICompositingTask

Stores initialization parameters and the order to the instance attribute init_args.

class eolearn.features.radiometric_normalization.MaxNDWICompositing(*args, **kwargs)[source]

Bases: eolearn.features.radiometric_normalization.MaxNDWICompositingTask

A deprecated version of MaxNDWICompositingTask

Stores initialization parameters and the order to the instance attribute init_args.

class eolearn.features.radiometric_normalization.HOTCompositing(*args, **kwargs)[source]

Bases: eolearn.features.radiometric_normalization.HOTCompositingTask

A deprecated version of HOTCompositingTask

Stores initialization parameters and the order to the instance attribute init_args.

class eolearn.features.radiometric_normalization.MaxRatioCompositing(*args, **kwargs)[source]

Bases: eolearn.features.radiometric_normalization.MaxRatioCompositingTask

A deprecated version of MaxRatioCompositingTask

Stores initialization parameters and the order to the instance attribute init_args.

class eolearn.features.radiometric_normalization.HistogramMatching(*args, **kwargs)[source]

Bases: eolearn.features.radiometric_normalization.HistogramMatchingTask

A deprecated version of HistogramMatchingTask

Stores initialization parameters and the order to the instance attribute init_args.