thermography.classification package

This package contains the implementation of the module classification routines implemented for the thermography package.

Submodules

thermography.classification.inference module

class Inference(checkpoint_dir: str, model_class: type, image_shape: numpy.ndarray, num_classes: int)[source]

Bases: object

Class implementing the inference procedure to classify new images according to a preexisting moduel.

Example: 
classification_model = ThermoNet3x3
inference = Inference(checkpoint_dir, classification_model, img_shape, num_classes)
img_list = [img_1, img_2, ...] # len(img_list) = N
class_probabilities = inference.classify(img_list) # class_probabilities.shape = [N, num_classes]

Initializes the inference object by loading the weights of an already trained model.

Parameters:
  • checkpoint_dir – Directory of the checkpoint where the weights of the model are stored.
  • model_class – Class associated to the stored weights. This class is used to build the tf.graph which will be used for the inference.
  • image_shape – Shape of the images fed to the classifier. The images passed to the self.classify function are resized according to this parameter.
  • num_classes – Number of classes predicted by the model.

Warning

The parameters must all be consistent with the preexisting weights!

classify(image_list: list) → numpy.ndarray[source]

Classifies the image list passed as argument using the model loaded in self.model.

Parameters:image_list – Python list of numpy arrays representing the images to be classified. All images are classified as a mini-batch.
Returns:A numpy array of shape [len(image_list), self.num_classes] containing the class probability for each image passed as argument.

Note

If the images contained in the input parameter are not of the same shape as the one store in self.image_shape, the input images are resized to fit the desired image shape.

model

Returns the model being used for classification.