signal similarity measurement - cross-correlation

Hi,

I am simulating an ultrasound imaging system. Here I need to compare two large time series using xcorr(). But, the computation time is higher than I expected and cross-correlation is taking up most of the time. I tried things such as phase zero search in analytic signal etc. But, I need something other than cross-correlation. Is there any other algorithm than cross-correlation to measure the similarity between two signals?

Answer 1

[Image Analyst]

Have you tried normalized cross correlation? If it's an image you can use SSIM, which is popular.

[Image Analyst]

There is a function normxcorr2() in the Image Processing Toolbox, if you have it.

[Mohammad Al Nagdawi]

the output of normxcorr2() is matrix. Is there is cross-correlation function output is single value similar to MI