eda_plugin.examples.analysers.KerasTilingWorker

class eda_plugin.examples.analysers.KerasTilingWorker(*args: Any, **kwargs: Any)

Bases: KerasWorker

KerasWorker with pre and postprocessing.

Add background subtraction, resize, intensity normalization and tiling to preprocessing. Add Stitching to the postprocessing.

__init__(*args, **kwargs)

Call the init function of KerasWorker with the settings supplied.

Methods

__init__(*args, **kwargs)	Call the init function of KerasWorker with the settings supplied.
extract_decision_parameter(network_output)	Return the a value of the ndarray.
post_process_output(network_output, input_data)	Stitch the images recevied from the network to an array with the same size as input.
prepare_images(images)	Background subtraction, resize, intensity normalization and tiling.
run()	Run the model.

extract_decision_parameter(network_output: numpy.ndarray)

Return the a value of the ndarray.

post_process_output(network_output: numpy.ndarray, input_data) → numpy.ndarray

Stitch the images recevied from the network to an array with the same size as input.

prepare_images(images: numpy.ndarray)

Background subtraction, resize, intensity normalization and tiling.

run()

Run the model.

Prepare the images, infer the model, calculate the decision parameter and construct the image that will be displayed in the GUI. Preparation and postprocessing are optional and can be implemented by subclasses as necessary for the specific model. Specific implementations can be found in examples.analysers.keras