We describe an algorithm based on shrinkage in the curvelet domain to attenuate speckles in optical coherence tomography (OCT) images. The algorithm exploits the curvelet transform’s sparse representation of edge discontinuities that are common in OCT images and its ability to map signals and noise into different areas in the curvelet domain. The speckle attenuation is controlled by a single parameter that determines the threshold in the curvelet domain. Applying the algorithm to OCT images shows significant improvement of image quality.
© 2009 Optical Society of America
Original Manuscript: January 9, 2009
Revised Manuscript: April 6, 2009
Manuscript Accepted: April 13, 2009
Published: May 8, 2009
Vol. 4, Iss. 7 Virtual Journal for Biomedical Optics
Zhongping Jian, Zhaoxia Yu, Lingfeng Yu, Bin Rao, Zhongping Chen, and Bruce J. Tromberg, "Speckle attenuation in optical coherence tomography by curvelet shrinkage," Opt. Lett. 34, 1516-1518 (2009)