In this chapter, we present a two-stage strategy, in which stripe noise removal is addressed by image decomposition and vessel segmentation is solved by an active contours approach. Besides, a joint model is also given in this chapter to improve the speed of the algorithm while retaining the quality of the segmentation and destriping. The experimental results are presented on both simulated and real retinal imaging data, demonstrating the effective performance of the joint model for segmenting vessels from the OCTA images corrupted by stripe noise. Furthermore, the stripe direction influence is tested to prove the flexibility of the joint model.
|Title of host publication
|Photo Acoustic and Optical Coherence Tomography Imaging
|Subtitle of host publication
|Angiography: An Application in Vessel Imaging
|Ayman El-Baz, Jasjit S. Suri
|IOP Publishing Ltd.
|Number of pages
|Published - 1 Dec 2021