A wearable text recognition device for the visually impaired.

What is it?

A wearable device that recognizes printed text, logos, and object labels in the environment and relays this to the wearer over audio.

Capable of reading common objects like street signs, menus, flyers, mail, and business cards.

Automatically detects language and translates to English if necessary

Made using the Raspberry Pi Zero W, Google Cloud Platform, and GitHub.

How does it work?

Camera built into sunglasses captures an image of the environment

Raspberry Pi Zero W uploads image to Google Cloud Platform through the Google Cloud Vision API

Google Cloud Platform uses machine learning models to quickly detect printed text, logos, and object labels in the image

Device recieves a response from the API request with detected text, logos, and labels in a .json file.

The Google Text-to-speech API is used to convert the detected features into an .mp3 audio file.

The audio file is played back to the user through connected headphones.

The process repeats at a desired interval.