Diabetic retinopathy (DR) and glaucoma are common eye diseases that affect a blood\nvessel in the retina and are two of the leading causes of vision loss around the world. Glaucoma is a\ncommon eye condition where the optic nerve that connects the eye to the brain becomes damaged,\nwhereas DR is a complication of diabetes caused by high blood sugar levels damaging the back of\nthe eye. In order to produce an accurate and early diagnosis, an extremely high number of retinal\nimages needs to be processed. Given the required computational complexity of image processing\nalgorithms and the need for high-performance architectures, this paper proposes and demonstrates\nthe use of fully parallel field programmable gate arrays (FPGAs) to overcome the burden of real-time\ncomputing in conventional software architectures. The experimental results achieved through software\nimplementation were validated on an FPGA device. The results showed a remarkable improvement\nin terms of computational speed and power consumption. This paper presents various preprocessing\nmethods to analyse fundus images, which can serve as a diagnostic tool for detection of glaucoma and\ndiabetic retinopathy. In the proposed adaptive thresholding-based preprocessing method, features\nwere selected by calculating the area of the segmented optic disk, which was further classified\nusing a feedforward neural network (NN). The analysis was carried out using feature extraction\nthrough existing methodologies such as adaptive thresholding, histogram and wavelet transform.\nResults obtained through these methods were quantified to obtain optimum performance in terms of\nclassification accuracy. The proposed hardware implementation outperforms existing methods and\noffers a significant improvement in terms of computational speed and power consumption.
Loading....