How does the imresize function work?

Hi, I was wondering if the resize function changes the size of the pixel? Is the resize function same as CCD binning? I like to change the size of my image but not change the ratio. My original image has a pixelsize of 5,5 um.

J = imresize(I, 0.5);

I tried to see if resizing changed the result (the size of my object and the integrated intensity) and it does comes out differently than if I would calculate with my orginial image which takes a long time (with convolution of my image and my mask aperture which is considerably smaller image than my orginial image). Try to figure out why it coems out differently.

Answer 1

[Jan]

imresize does not have or use any information about the size of the pixels. It combines the values of the pixels only to create a smaller or greater output array.

If the resizing changes your "result", this must be an effect on how the "result" is calculated. So please post the code you are using and explain the differences between the output and your expectations ins detail.

[Happy PhD]

My question is linked to this question: https://se.mathworks.com/matlabcentral/answers/414070-fast-convolution-of-an-large-image-and-small-mask

What I want to calculate is worst alpha (angle in mrad) and Q (energy within said angle) that gives the highest SF-value.

I execpt to get the same alpha, Q and SF, since the image is unchanged other than redued in size. Instead Q is completely different and alpha isnt exactly the same.

original image gives;

SF = 

ans =

      5.6508

alpha = 

ans =

      3.6000

Qm = 

ans =

     24.1148

alpha_x alpha_y

ans =

      2.2000    5.0000

resized image by 0.5 and bilinear gives:

SF = 

ans =

      2.4068

alpha = 

ans =

      3.8333

Qm = 

ans =

     60.2866

alpha_x alpha_y

ans =

      2.6667    5.0000

Edit: Removed code.

[Jan]

Of course reducing the number of pixels to a quarter of the original number changes the contents of the image. imresize does not change the size of the pixels, but their number. I do not know, what the shown parameters are, but it seems to be obvious that their value if affected by the reduction of the pixels. Why do you assume, that the values do not change? I think, that this assumption is the only problem.