organoid_tracker.neural_network.image_loading module

organoid_tracker.neural_network.image_loading.extract_patch_array(full_images: Dict[TimePoint, Image], start_zyx: Tuple[int, int, int], patch_shape_zyx: Tuple[int, int, int]) → ndarray

Extracts a patch from the full images. Coordinates are assumed to be in position coordinates. Offsets of the images are taken into account.

full_images must contain a continuous set of time points, but doesn’t have to be in order.

organoid_tracker.neural_network.image_loading.fill_none_images_with_copies(full_images: Dict[TimePoint, Image | None])

Fills any None values in the array with the array of the closest time point that is available.

(At the start or end of a movie, some time points may be missing. In that case, it’s best to just copy the nearest available image.)

Raises a ValueError if no images are available at all.