Abstract:
In past, retinal diseases were identified by visual filed test or by Fundography. Both of these techniques do not show layered structure of eye and disease can’t be detected at early stage. Optical Coherence tomography (OCT), the most imminent methodology used by ophthalmologist in modern era which is non-invasive, for the detection of retinal disorders, as it shows layered structure of an eye and is very handy for early diagnosing of diseases like glaucoma, Macular Edema, Macular hole, central serous Retinopathy and Pigment Epithelium Detachment. Acquisition of OCT images is based on Low coherence interferometry. Due to the principle of interference, additive and multiplicative noise is generated in images. Speckle (multiplicative noise) reduction or suppression in OCT is not simple as speckle carries not only the noise but structural information of image. While designing algorithms for automated analysis of OCT images, removal of speckle noise is necessary for better diagnosis of the retinal disorders and detection of retinal layers for early detection of disease such as glaucoma, diabetic retinopathy. Researchers around the world have been working on different techniques to obtain better results. In this thesis, different methodologies have been discussed and implemented and an overall review is discussed by comparing certain parameters on a given dataset. Results show how these mathematical models helps compress the multiplicative noise and improve quality of an image.
