# Name

acfnpl - calculate the normalized autocorrelation function of an image using padding with zeroes, multiplication in Fourier space, and normalization of the result by the actual number of pixels used for calculating the ccf coefficients.

# Usage

output = acfnpl(image, center=True)

## Input

image
input image (real)
center
if set to True (default), the origin of the result is at the center; if set to False, the origin is at (0,0), the option is much faster, but the result is difficult to use

## Output

output
autocorrelation function of the input image. Real. The origin of the autocorrelation function (term ccf(0,0,0)) is located at (int[n/2], int[n/2], int[n/2]) in 3D, (int[n/2], int[n/2]) in 2D, and at int[n/2] in 1D.

# Method

• In order to calculate the normalized and lag-normalized autocorrelation function of an image f, first the imagea is normalized by subtracting its average and by dividing it by its standard deviation. Next, Fourier transform is calculated, then the modulus squared in Fourier space as `|hat(f)|^2`, then the inverse Fourier transform, the acfnpl is windowed out using the size of original images, and the resulting autocorrelation function is normalized by the lag, i.e., the actual number of pixels in image that entered the calculation.

• In real space, this corresponds to:
• `a\cfnpl(n)=(1/(nx-|n|)sum_(k=0)^(nx-1)f((k+n)-Ave_f)(f(k)-Ave_g))/(sigma_f^2)`

• `n = -(nx)/2, ..., (nx)/2`

• with the assumption that `f(k)=0 fo\r k<0 or kgenx`

• Note: acfnpl is free from "wrap around" artifacts.

# Reference

Pratt, W. K., 1992. Digital image processing. Wiley, New York.

Pawel A. Penczek

category 1
FUNDAMENTALS
category 2
FOURIER

fundamentals.py

# Maturity

stable
works for most people, has been tested; test cases/examples available.

# Bugs

None. It is perfect.

acfnpl (last edited 2013-07-01 13:12:49 by localhost)