import numpy as np
[docs]def compute_mean_image(frames):
"""
Compute mean image of given frames
Parameters
-------
frames : list of numpy.ndarray
All frame must have same shape
Returns
-------
mean_image : numpy.ndarray
Mean image of given frame.
"""
frame = frames[0]
prev_frame = frame
mean_image = np.zeros(shape=frame.shape, dtype=np.float32)
N = len(frames)
for frame in frames:
if not prev_frame.shape == frame.shape:
raise ValueError("Given frames does not have same shape. {} != {}".format(
prev_frame.shape, frame.shape))
mean_image += frame
prev_frame = frame
return mean_image / N