abhijit-code / drishti---for-blind

A smart Assistant for Navigation and go-to text Reading for blind

Drishti---For-Blind

A smart Assistant for Navigation and go-to text Reading for blind

Drishti helps to read the text in front of the user, and it also conveys the environment around the visually impaired, by describing the objects and relationship b/w the objects. Examples are a 'person sitting on a bench,' 'a dog is sleeping,' 'a stop sign on the road'. On users command the image is captured with a camera connected to the raspberry pi. If the user wants text description of the image, the image is analysed by Google's Cloud Vision API using Optical Character Recognition to detect the characters, letters in the picture, and if the user wants the scene description of the nearby environment, the image is sent to the Microsoft Cognitive Services to analyse the image using Computer Vision. Generated result is stored in the Dynamo DB (Cloud database). When the user asks Alexa to 'Read the text' or 'Explain the environment', the Alexa Skill Kit triggers AWS Lambda Function to fetch the results from the database (Dynamo Db). The Alexa app then recites the stored result on the user's mobile. This innovative device, designed to serve the visually impaired, is handy, easy to use with high accuracy.