A Transfer Learning Based Web App for Glaucoma Detection Using Low-Cost Ophthalmoscopic Camera

Glaucoma is the most fundamental cause of irreversible blindness in persons over a few decades. The condition is extremely prevalent in several nations, and in the worst-case situation is that many countries do not have enough ophthalmologists. According to data from a global glaucoma research, roughly 12 million individuals would be symbiotically blind from early-stage glaucoma by 2020. To evaluate the disease in the eye, fundus photography is used. Fundus photos require a lengthy visual inspection. The transfer-learning based web-app for glaucoma detection utilizing low-cost ophthalmoscopic cameras is presented in this paper. Fundus evaluation is an important part of an eye exam that helps the web-app provide vital diagnostic information to both ophthalmologists and non-ophthalmologists. The vision impairment occurs over time because it is not a rapid consequence. Only in the later stages those signs are noticed. The optic nerve was studied using 2-Dimensional retinal fundus pictures, which is a difficult task that can result in a significant cost for detecting regular glaucoma. There are a variety of different tests that can be used to rule out glaucoma, including tonometry which signifies pressure on eye, gonioscopy which examines angle is closed or open, OCT - Optical Coherence Tomography, Ophthalmoscopic Imaging to visualize retina as well as optic nerve and Fundus Copying. In this paper, an effective embedded based eye testing architecture is presented.