Skip to main content

algorithm

The algorithm module contains the core algorithm for reconstructing an object from a series of images. The main function is reconstruct, which takes an ImageSeries object and returns the reconstructed object in the spatial domain (as a complex value).

Function: reconstruct​

fpm_py.algorithm.reconstruct(
    image_series: ImageSeries,
    output_scale_factor: int,
    pupil_0: torch.Tensor,
    iteration_terminator: TerminatorType,
    optimizer: OptimizerType
):
    ...

Core algorithm. Reconstructs an object from a series of images.

Args:​

  • image_series (ImageSeries): The series of images to reconstruct.
  • output_scale_factor (int): The scale factor of the output image. If not provided, the minimum needed size will be calculated.
  • pupil_0 (torch.Tensor): The initial calculation of the pupil function. If not provided, a boolean circular pupil will be created.
  • iteration_terminator (TerminatorType): The function that determines when to stop iterating. If not provided, the default is to iterate 10 times.
  • optimizer (OptimizerType): The optimizer function that updates the object and pupil. If not provided, the default is the tomas optimizer.

Returns:​

  • torch.Tensor: The reconstructed object in the spatial domain.