datamodules.DivaHisDB.utils package

Submodules

datamodules.DivaHisDB.utils.functional module

gt_to_int_encoding(matrix: Tensor, class_encodings: List[int]) → Tensor[source]

Convert ground truth tensor to integer encoded matrix

Parameters:

  • matrix (torch.Tensor) – Image as a tensor of size [C x H x W] (BGR)

  • class_encodings (List[int]) – class encoding so which class (index) has what value (element)

Returns:

integer encoded matrix

Return type:

torch.Tensor

gt_to_one_hot(matrix: Tensor, class_encodings: List[int])[source]

Convert ground truth tensor or numpy matrix to one-hot encoded matrix

Parameters:

  • matrix (torch.Tensor or np.ndarray) – float tensor from to_tensor() or numpy array shape (C x H x W) in the range [0.0, 1.0] or shape (H x W x C) BGR

  • class_encodings (List[int]) – List of int Blue channel values that encode the different classes

Returns:

Tensor of size [#C x H x W] sparse one-hot encoded multi-class matrix, where #C is the number of classes

Return type:

torch.LongTensor

datamodules.DivaHisDB.utils.image_analytics module

get_analytics(input_path: Path, data_folder_name: str, gt_folder_name: str, get_gt_data_paths_func) → Tuple[Dict[str, Any], Dict[str, Any]][source]

Get the analytics for the dataset. If the analytics file is not present, it will be computed and saved.

Parameters:

  • input_path (Path) – Path to the root of the dataset

  • data_folder_name (str) – Name of the folder containing the data

  • gt_folder_name (str) – Name of the folder containing the ground truth

  • get_gt_data_paths_func (Callable) – Function to get the paths to the data and ground truth

Returns:

Tuple of analytics for the data and ground truth

Return type:

Tuple[Dict[str, Any], Dict[str, Any]]

get_class_weights(input_folder, workers=4) → List[float][source]

Get the weights proportional to the inverse of their class frequencies. The vector sums up to 1

Parameters:

  • input_folder (Path) – Path to the dataset folder (see above for details)

  • workers (int) – Number of workers to use for the mean/std computation

Returns:

The weights vector as a 1D array normalized (sum up to 1)

Return type:

List[float]

datamodules.DivaHisDB.utils.output_tools module

output_to_class_encodings(output, class_encodings, perform_argmax=True)[source]

This function converts the output prediction matrix to an image with the colors of the class encodings.

Parameters:

  • output (np.array of size [#C x H x W]) – output prediction of the network for a full-size image, where #C is the number of classes

  • class_encodings (list) – Contains the range of encoded classes

  • perform_argmax (bool) – perform argmax on input data

Returns:

np.array of size [H x W] (BGR)

save_output_page_image(image_name, output_image, output_folder: Path, class_encoding)[source]

Helper function to save the output during testing in the DIVAHisDB format

Parameters:

  • image_name (str) – name of the image that is saved

  • output_image (np.array of size [#C x H x W]) – output image at full size

  • output_folder (Path) – path to the output folder for the test data

  • class_encoding (list) – list with the class encodings

Returns:

mean iou of this image

Return type:

float

datamodules.DivaHisDB.utils.single_transform module

class IntegerEncoding(class_encodings: List[int])[source]

Bases: object

Convert ground truth tensor to integer encoded matrix.

Parameters:

class_encodings (List[int]) – class encoding so which class (index) has what value (element)

class OneHotEncoding(class_encodings: List[int])[source]

Bases: object

Convert ground truth tensor to one-hot encoded matrix.

Parameters:

class_encodings (List[int]) – class encoding so which class (index) has what value (element)

Module contents