EyeRead

• 

  Webcam setup to read text

We started off wanting to make a program to make inputting equations to online platforms easier for you but quickly realized that our text based idea had much more potential than that. Now it has become a tool that both the visually impaired and children can use to better their lives and combat illiteracy at any age. It converts video to text and then to audio so you can read almost anything on spot!

We built it using opencv to analyze and filter an image. Then we ran tesseract on that image to convert all the words to text. After that we used goggle's text to speech api to convert it to speech and read the words out loud.

Right now we have a rough algorithm for comparing word similarity to sense if a page has been flipped or an image has shifted so we can save time. We also have a rough way to sense a finger and find a word near it but not well enough to isolate a single word and it is definitely a future project for us.

Built With

• google-text-speech
• opencv
• tesseract