continunet.network

continunet.network.unet module

UNet model for image segmentation in keras.

class continunet.network.unet.Unet(input_shape: tuple, filters: int = 16, dropout: float = 0.05, batch_normalisation: bool = True, trained_model: str = None, image: ndarray = None, layers: int = 4, output_activation: str = 'sigmoid', model: Model = None, reconstructed: ndarray = None)

Bases: object

UNet model for image segmentation.

build_model()

Build the UNet model.

compile_model()

Compile the UNet model.

convolutional_block(input_tensor, filters, kernel_size=3)

Convolutional block for UNet.

decode_image()

Returns images decoded by a trained model.

decoding_block(input_tensor, concat_tensor, filters, kernel_size=3)

Decoding block for UNet.

encoding_block(input_tensor, filters, kernel_size=3)

Encoding block for UNet.