Phase correlation


DrPL
06-15-2006, 04:48 AM
Hi all,
Has anyone implemented a phase correlation technique using the NR rlft3 (albeit in 2 dimensions) and fourn(...) functions as listed in the book? I'd like to use it in image analysis

Basically, depending on which website you read, you either
multiply the FFT of image 1 with the complex conjugate of image 2, or simply take the FFT of both images, and then
do a exp(i*(angle(FFT1)-angle(FFT2)), and finally do an inverse FFT.

This latter approach just looks wrong to me (!). I'm wading through the NR code and am finding it difficult coding the complex conjugate FFT.

Can anyone help - I'd be very grateful.

Thanks

Paul