trw.transforms.transforms_random_crop

Module Contents

Classes

TransformRandomCrop	Add padding on a numpy array of samples and random crop to original size
TransformRandomCropJoint	Add random padding & cropping on a numpy or Torch arrays. The arrays are joints and the same padding/cropping applied on all the arrays

Functions

_transform_random_crop(feature_name, feature_value, padding, mode='edge', constant_value=0, size=None)

Add a specified padding to the image and randomly crop it so that we have the same size as the original

trw.transforms.transforms_random_crop._transform_random_crop(feature_name, feature_value, padding, mode='edge', constant_value=0, size=None)

Add a specified padding to the image and randomly crop it so that we have the same size as the original image

This support joint padding & cropping of multiple arrays (e.g., to support segmentation maps)

Parameters

• feature_name – a feature name or a list of feature names

• feature_value – the value of the feature or a list of feature values

• padding – the padding to add to the feature value. If None, no padding added

• constant_value – a constant value, depending on the mode selected

• padding – a sequence of size len(array.shape)-1 indicating the width of the padding to be added at the beginning and at the end of each dimension (except for dimension 0)

• mode – numpy.pad mode. Currently supported are (‘constant’, ‘edge’, ‘symmetric’)

• size – if None, the image will be cropped to the original size, else it must be a list of the size to crop for each dimension except for dimension 0

Returns

a padded and cropped image to original size

class trw.transforms.transforms_random_crop.TransformRandomCrop(padding, criteria_fn=None, mode='edge', constant_value=0, size=None)

Bases: trw.transforms.transforms.TransformBatchWithCriteria

Add padding on a numpy array of samples and random crop to original size

Parameters

• padding – a sequence of size len(array.shape)-1 indicating the width of the padding to be added at the beginning and at the end of each dimension (except for dimension 0). If None, no padding added

• criteria_fn – function applied on each feature. If satisfied, the feature will be transformed, if not the original feature is returned

• mode – numpy.pad mode. Currently supported are (‘constant’, ‘edge’, ‘symmetric’)

• size – the size of the cropped image. If None, same size as input image

Returns

a randomly cropped batch

class trw.transforms.transforms_random_crop.TransformRandomCropJoint(feature_names, padding, mode='edge', constant_value=0, size=None)

Bases: trw.transforms.transforms.TransformBatchJointWithCriteria

Add random padding & cropping on a numpy or Torch arrays. The arrays are joints and the same padding/cropping applied on all the arrays

Parameters

• feature_names – these are the features that will be jointly padded and cropped

• padding – a sequence of size len(array.shape)-1 indicating the width of the padding to be added at the beginning and at the end of each dimension (except for dimension 0). If None, no padding added

• criteria_fn – function applied on each feature. If satisfied, the feature will be transformed, if not the original feature is returned

• mode – numpy.pad mode. Currently supported are (‘constant’, ‘edge’, ‘symmetric’)

• size – the size of the cropped image. If None, same size as input image

Returns

a randomly cropped batch