Add tools to calculate subpixel shifts. Possible improvements still on the table

This commit is contained in:
Abe Levitan committed 2019-04-10 12:20:39 -04:00
1 parent 91f11643e9
commit ace81cc66d
3 files changed
+83 -8

No files matched your search

+57 -4
View File
@@ -71,8 +71,57 @@ def find_subpixel_shift(im1, im2, search_around=(0,0), resolution=10):
search_around (array_like) : Default (0,0), the shift to search in the vicinity of
resolution (int): Default is 10, the resolution to calculate to in units of 1/n
"""
pass
#
# Here's my approach, perhaps it's a little unconventional. I will first
# calculate the phase correlation function as found in ____ (cite a paper
# defining it). This is strongly peaked, so I can take a small window
# of say, 10x10 pixels, and then do a sinc interpolation of that area
# using an FFT with upsampling by a factor of resolution in reciprocal
# space
#
# If last dimension is not 2, then convert to a complex tensor now
if im1.shape[-1] != 2:
im1 = t.stack((im1,t.zeros_like(im1)),dim=-1)
if im2.shape[-1] != 2:
im2 = t.stack((im2,t.zeros_like(im2)),dim=-1)
cor_fft = cmath.cmult(t.fft(im1,2),cmath.cconj(t.fft(im2,2)))
# Not sure if this is more or less stable than just the correlation
# maximum - requires some testing
cor = t.ifft(cor_fft / cmath.cabs(cor_fft)[:,:,None],2)
# Now, I need to shift the array to pull out a contiguous window
# around the correlation maximum
try:
search_around = search_around.cpu()
except:
search_around = t.tensor(search_around)
window_size = 15
shift_zero = tuple(-search_around + t.tensor([window_size,window_size]))
cor_window = t.roll(cor, shift_zero, dims=(0,1))[:2*window_size,:2*window_size]
# Now we upsample this window
cor_window_fft = cmath.fftshift(t.fft(cor_window,2))
upsampled = t.zeros(tuple(t.tensor(cor_window_fft.shape)[:-1] * resolution) + (2,),
dtype=cor.dtype,device=cor.device)
upsampled[:2*window_size,:2*window_size] = cor_window_fft
upsampled = t.roll(upsampled,(-window_size,-window_size),dims=(0,1))
upsampled = t.roll(cmath.cabssq(t.ifft(upsampled, 2)),(-window_size*resolution,-window_size*resolution), dims=(0,1))
# And we extract the shift from the window
sh = t.tensor(upsampled.shape).to(device=upsampled.device)
cormax = t.tensor([t.argmax(upsampled) // sh[1],
t.argmax(upsampled) % sh[1]]).to(device=upsampled.device)
subpixel_shift = ((cormax + sh // 2) % sh - sh//2).to(dtype=upsampled.dtype)
return search_around.to(device=upsampled.device, dtype=upsampled.dtype) + \
subpixel_shift / resolution
def find_pixel_shift(im1, im2):
"""Calculates the integer pixel shift between two images by maximizing the autocorrelation
@@ -94,9 +143,13 @@ def find_pixel_shift(im1, im2):
if im2.shape[-1] != 2:
im2 = t.stack((im2,t.zeros_like(im2)),dim=-1)
cor = cmath.cabs(t.ifft(cmath.cmult(t.fft(im1,2),
cmath.cconj(t.fft(im2,2))),2))
cor_fft = cmath.cmult(t.fft(im1,2),cmath.cconj(t.fft(im2,2)))
# Not sure if this is more or less stable than just the correlation
# maximum - requires some testing
cor = cmath.cabs(t.ifft(cor_fft / cmath.cabs(cor_fft)[:,:,None],2))
#cor = cmath.cabs(t.ifft(cor_fft,2))
sh = t.tensor(cor.shape).to(device=im1.device)
cormax = t.tensor([t.argmax(cor) // sh[1],