def phase_detection(masked_spectrum, distance):
shape = masked_spectrum.shape
shape_center = [dim / 2 for dim in shape]
filtered_hologram = get_shifted_idft(masked_spectrum)
focus_mask = get_mask(shape, np.ones(20), shape_center)
focus_feature = filtered_hologram * focus_mask
feature_spectrum = get_shifted_dft(focus_feature)
propagation_array = get_propagation_array(shape, distance)
propagated_spectrum = propagation_array * feature_spectrum
propagated_hologram = get_shifted_idft(propagated_spectrum)
radious = 50
separation = 400
window = np.ones(radious)
spectrum2sensor = np.angle
spectrum2sensor = np.abs
spectrum2sensor = get_intensity
spectrum2sensor = normalize
left_center = shape_center[0], shape_center[1] - separation / 2
left_mask = get_mask(shape, window, left_center)
# left_bundle = left_mask * propagated_spectrum
left_bundle = left_mask * propagated_hologram
left_sensor = spectrum2sensor(get_shifted_dft(left_bundle))
left_peak = get_circles(left_sensor, 1, 50)[0][1]
right_center = shape_center[0], shape_center[1] + separation / 2
right_mask = get_mask(shape, window, right_center)
right_bundle = right_mask * propagated_hologram
right_sensor = spectrum2sensor(get_shifted_dft(right_bundle))
right_peak = get_circles(right_sensor, 1, 50)[0][1]
# showimage(normalize(np.maximum(left_bundle, right_bundle)))
# showimage(normalize(np.maximum(left_sensor, right_sensor)))
distance = get_distance(left_peak, right_peak)
fitness = distance
return fitness
def phase_detection(masked_spectrum, distance):
shape = masked_spectrum.shape
shape_center = [dim / 2 for dim in shape]
filtered_hologram = get_shifted_idft(masked_spectrum)
focus_mask = get_mask(shape, np.ones(20), shape_center)
focus_feature = filtered_hologram * focus_mask
feature_spectrum = get_shifted_dft(focus_feature)
propagation_array = get_propagation_array(shape, distance)
propagated_spectrum = propagation_array * feature_spectrum
propagated_hologram = get_shifted_idft(propagated_spectrum)
radious = 50
separation = 400
window = np.ones(radious)
spectrum2sensor = np.angle
spectrum2sensor = np.abs
spectrum2sensor = get_intensity
spectrum2sensor = normalize
left_center = shape_center[0], shape_center[1] - separation / 2
left_mask = get_mask(shape, window, left_center)
# left_bundle = left_mask * propagated_spectrum
left_bundle = left_mask * propagated_hologram
left_sensor = spectrum2sensor(get_shifted_dft(left_bundle))
left_peak = get_circles(left_sensor, 1, 50)[0][1]
right_center = shape_center[0], shape_center[1] + separation / 2
right_mask = get_mask(shape, window, right_center)
right_bundle = right_mask * propagated_hologram
right_sensor = spectrum2sensor(get_shifted_dft(right_bundle))
right_peak = get_circles(right_sensor, 1, 50)[0][1]
# showimage(normalize(np.maximum(left_bundle, right_bundle)))
# showimage(normalize(np.maximum(left_sensor, right_sensor)))
distance = get_distance(left_peak, right_peak)
fitness = distance
return fitness
def get_peak_coords(spectrum):
# TODO: must select by position not by intensity
shape = spectrum.shape
center = [dim / 2. for dim in shape]
intensity = get_intensity(spectrum)
circles = [(-get_distance(center, circle[1]), circle[0], circle[1])
for circle in get_circles(intensity, 2, 20)]
circles.sort()
peak = circles[0][2]
peaks_row = (peak[0] - center[0]) / float(center[0])
peaks_col = (peak[1] - center[1]) / float(center[1])
return peaks_row, peaks_col
def get_peak_coords(spectrum):
# TODO: must select by position not by intensity
shape = spectrum.shape
center = [dim / 2. for dim in shape]
intensity = get_intensity(spectrum)
circles = [(-get_distance(center, circle[1]), circle[0], circle[1])
for circle in get_circles(intensity, 2, 20)]
circles.sort()
peak = circles[0][2]
peaks_row = (peak[0] - center[0]) / float(center[0])
peaks_col = (peak[1] - center[1]) / float(center[1])
return peaks_row, peaks_col