Fractional Fourier Transform Phase for Image Matching

Abstract

Fractional Fourier transform is the generalization of the Fourier transform. In this article the synthesis of phase and magnitude of fractional Fourier transform is demonstrated. The influence of phase and magnitude on the synthesis results is shown. The fractional phase correlation function using fractional Fourier transform is calculated, and it is used for image matching. The use of fractional phase correlation can achieve better results compared to the use of the classical phase correlation. The comparison between phase correlation and fractional phase correlation for biometric iris images is given. The selection of optimal parameter for fractional phase correlation is proposed.

Fractional Wavelet Transform Phase for Iris Image Key Points Matching

“…For the fractional Fourier transform, a different amount of information about the edges can contain both the phase and the magnitude, depending on the value of the parameter 𝑎. It can be shown by FRFT phase and FRFT magnitude synthesis (Protsenko, 2022).…”

“…The fractional operations like fractional convolution (Mustard, 1998), fractional correlation (Mendlovic, 1995) can also be considered as a generalization of classical operations. Some fractional operations can help to obtain better results than classical methods (Protsenko, 2022).…”