Core API

class hugin.engine.core.AverageMerger(*args, **kwargs)

Bases: hugin.engine.core.PredictionMerger

get_prediction()

update(xstart, xend, ystart, yend, prediction)

class hugin.engine.core.CategoricalConverter(num_classes)

Bases: object

class hugin.engine.core.CloneComponentGenerator(base_component)

Bases: hugin.engine.core.RasterGenerator

Generator generating an clone of an existing component

class hugin.engine.core.IdentityModel(*args, num_loops=1, **kwargs)

Bases: hugin.engine.core.RasterModel

fit_generator(train_data, validation_data=None)

predict(batch, batch_size=None)

Runs the predictor on the input tile batch

Parameters:batch – The input batch Returns:returns a prediction according to model configuration

save(destination=None)

class hugin.engine.core.NullMerger(*args, **kwargs)

Bases: hugin.engine.core.PredictionMerger

get_prediction()

update(xstart, xend, ystart, yend, prediction)

class hugin.engine.core.PredictionMerger(height, width, depth, dtype)

Bases: object

get_prediction()

update(xstart, xend, ystart, yend, prediction)

class hugin.engine.core.RasterGenerator

Bases: object

Base class used by handlers generating new data components

class hugin.engine.core.RasterModel(name=None, batch_size=1, swap_axes=True, input_shapes=None, output_shapes=None)

Bases: object

fit_generator(train_data, validation_data=None)

predict(batch)

Runs the predictor on the input tile batch

Parameters:batch – The input batch Returns:returns a prediction according to model configuration

save(destination=None)

class hugin.engine.core.SkLearnStandardizer(model_path, with_gti=True, *args, copy=True, with_mean=True, with_std=True, **kw)

Bases: hugin.engine.core.RasterModel

fit_generator(train_data, validation_data=None)

save(destination)

hugin.engine.core.identity_metric(prediction, gti)

hugin.engine.core.identity_processor(arg)

hugin.engine.core.metric_processor(func)

hugin.engine.core.postprocessor(func)