We as college students often take photos of the notes on the board in class, due to multiple inconveniences e.g writing speed, erasing speed, handwriting, etc. As such, we've always wanted to be able to easily convert pictures of lecture notes to easily readable and searchable documents, not only for convenience but also to help us study efficiently.
What it does
Using Google Translate's Image to Text Recgonition, the written notes on the whiteboard is translated from image to text. This text is then inputted manually as text file into our application in order to organize into readable notes. The application also pulls important dates that it finds within the image and automatically creates and assigns due dates to Google Calendar.
How we built it
We used Python and the google calendar API.
Challenges we ran into
Very early on we ran into problems pertaining to processing text from images because of how primitive OCR (Optical Character Recgonition) APIs are. We also had trouble implementing Google Calendar's API into our program.
Accomplishments that we're proud of
After long toil, we managed to use the help of an internet forum in order to figure out how to fix our problem with Google's Calendar API, making it work well.
What we learned
Its really hard to identify and process written text through images
What's next for Notr
Currently, the program we have now is only partially done. Without a more powerful image to text processing API that can read handwriting, it can not be implemented into our program to use to convert lecture notes to text documents like we initially wanted. We hope to also include some sort of auto image processing using a special code to identify graphs/charts/etc that professors have deemed important. These pictures are then auto-cropped within the app and integrated into our translated notes. Ideally, this is all done as soon as the student takes a picture of the board.