A novel method for speckle reduction in optical coherence tomography by image registration

Speckle noise is a key factor that can influence the image quality of optical coherence tomography (OCT). The averaging of multiple B-scans can effectively suppress the speckle noise. Because of the sample motion, the images have to be exactly aligned before averaging. In this paper, we propose a new method for OCT image registration that combines global and local registration. The method is able to align the large global displacements in axial and lateral directions, as well as local displacements caused by non-linear deformation between images. Compared with other OCT image registration methods, our method improves the signal-to-noise ratio and contrast-to-noise ratio