Texture analysis of Choroidal Neovascularization retinal images using Covering Blanket Method

Choroidal Neovascularization (CNV) is characterized by growth of a choroidal vascular construction into the macula go together with enhanced vascular permeability. The CNV is sight risk obstacle that occurs in 7% to 12% patients of high myopia. This paper presents an empirical evaluation of retinal images to calculate approximately the fractal dimension (FD) of textured images using Covering Blanket Method. The classification accurateness of the Blanket method is contrasted with other on hand texture analysis methods mentioned in the literature. The publicly available DRIVE database is used for performance evaluation of manually labeled images along with 56 images including both 25 normal and 31 pathological, supplied by ophthalmologist. Performance of presented method on both sets of images is found to be improved with respect to other existing solutions. The obtained results show high accuracy with discrimination rate of 96% using Blanket method.