The leading cause of blindness in India are Glaucoma and Diabetic retinopathy Existing diagnostic methodologies in the field of ophthalmology, are not proven accurate for early detection of these. This is primarily due to the asymptomatic nature of the diseases in early stages and lack of retinal specialists. It has been noticed that, if left untreated these diseases can cause serious vision loss and blindness. To overcome this problem, we propose a system capable of early detection of the diseases accurately, using deep learning. In addition to that, the diagnostic system alerts the patient about his condition soon from a screening standpoint. This earliest referral could save his potential vision.

The system, which runs on Convolutional neural networks, is trained with dataset containing fundus images of affected and unaffected cases of both diseases. The trained system is capable of predicting the diseases from multiple set of possibilities. The prime focus of the work is to develop a system for early detection of the eye diseases with less intervention of a specialist.

A device which has a Braille pattern recognizer and a Braille keypad through which digital communication is made possible for a blind person or both blind and deaf. The pattern recognizer converts the normal languages to Braille code which are produced letter by letter as projections thus recognizable. Design of a pattern recognizer is the important task. It is implemented by means of 6 solenoids, representing a single Braille character.

Upon electromagnetic actuation, by providing the driving current through a relay board, every solenoid projects out forming combinations of symbols. By using this concept we can improve response time, power efficiency and most importantly the size of the product. The information is downloaded and decoded wirelessly using a wi-fi module and then the code is delivered to the ARDUINO UNO board which has been programmed accordingly to magnetize the particular cells of a symbol.

Communication can be set up with devices like mobile phones through an application ‘BLYNK’ which has its own cloud storage and an application interface for a normal user. Through this cloud storage a person using the braille device can send and receive text messages. In this growing digital world it is necessary that everywhere you see a growth it is not just the common people it is the entire people. There are some existing devices in this domain but nothing is affordable. This project deals with the implementation of a digital communication system that is affordable at the same time being useful and providing the right services for a blind or both blind and deaf person.