Evaluation the efficiency and accuracy of retinal images registration algorithm


Abstract in English

In this paper, we describe an algorithm to register the retinal images by employing a relatively cross-correlation function. Pixel to pixel accuracy has been investigated and evaluated among registered images by calculating the local cross-correlation between extracted vessels profiles along tracked vessels after transforming the images into alignment.

References used

Harvey A, Lawlor J, McNaught A, Williams J and Fletcher- Holmes D 2002 Hyperspectral imaging for the detection of retinal diseases, Proc SPIE. 4816.325-335
Mordant DJ, AlAbboud I, Muyo G, Gorman A, Sallam A, Ritchie P, Harvey A, McNaught AI 2011 Spectral imaging of the retina, Eye (Lond). Mar;25(3).309-20
Alabboud I, Muyo G, Gorman A, Mordant D, McNaught A, Petres C, Petillot Y, Harvey A 2007 New spectral imaging techniques for blood oximetry in the retina, Proc. SPIE 6631

Download