Automatic Identification of Visual Impairment Using Optic Disc Segmentation

The extraction of optic disc from the fundus image is key process for automatic recognition of visual impairment. Automatic recognition is the technique which is used to detect the disease occurs in optic disc (OD).The proposed methodprincipal component analysis allows to detect Optic disc contour more accurately. Principal component analysis (PCA) is the preprocessing step along the mathematical morphological operations for optic disc segmentation. This stage is very important to get final result veryprecisely. The principal component analysis applied on the input retinal fundus image to get better representation of gray scale image. The main idea of PCA to reduce the dimensionality and then speed up the automation process. Along with this PCA number of mathematical morphological operations such as Inpainting algorithm, stochastic watershed, and geodesic transformations will be done to extract optic disc. By extracting the required features we can identify the visual impairment. Firstly, Cup to Disc ratio will be calculated to identify the glaucoma by using Hough Transform. Secondly, Micro aneurysms will be detected to identify the Diabetic Retinopathy. Extended minima transform and Blob Algorithm are used to detect Micro aneurysms.