You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello,
I have notice what seems to be a drift in the average values of the different images of a given FOV.
For a given FOV f, let GT_f be the average of the 400 noisy images I_f_i the drift in average value is D_f_i = | mean(GT_f) - mean(I_f_i) |
For channel 1, among all FOV: max_f(max_i(D_f_i)) = 17.8 and mean_f(mean_i(D_f_i))=1.14 when the average signal is 138.8 .
This induces a bias in computation of GT and makes the noise non-centered (I provide below the mean values of the noise N_f=I_f_249 - GT_f for each FOV of channel 1 ).
I'm working with the raw images.
In my case this impacts metrics of denoising methods.
Have you observed this issues, and do you have a standard correction for this ?
Thanks in advance.
Jean Ollion
Hello,
I have notice what seems to be a drift in the average values of the different images of a given FOV.
For a given FOV
f,
letGT_f
be the average of the 400 noisy imagesI_f_i
the drift in average value isD_f_i = | mean(GT_f) - mean(I_f_i) |
For channel 1, among all FOV:
max_f(max_i(D_f_i)) = 17.8
andmean_f(mean_i(D_f_i))=1.14
when the average signal is138.8
.This induces a bias in computation of GT and makes the noise non-centered (I provide below the mean values of the noise
N_f=I_f_249 - GT_f
for each FOV of channel 1 ).I'm working with the raw images.
In my case this impacts metrics of denoising methods.
Have you observed this issues, and do you have a standard correction for this ?
Thanks in advance.
Jean Ollion
The text was updated successfully, but these errors were encountered: