Picturize - Auto Note Taker

Inspiration

I come back from school every single day and go like, "ugh," when I realize I have to take a whole chapter of notes for Chemistry class each week from the textbook. I end up not wanting to do it so I procrastinate. This made me want to build an app that took notes for me.

What it does

Picturize is an app that can take notes for you on virtually anything! Create short paragraphs, paragraphs, or bullet points from a picture of a text (textbook, newspaper, newsletter, magazine, article, screenshots etc.) all in seconds! Literally, all you need to do is point your camera at a page in a textbook and then Picturize will analyze the page, pinpoint the most important parts of the text in the picture, and will give you the option either to read those notes or to let it dictate the notes for you.

How I built it

I used Android Studio to make the whole Android app. I used the Google Cloud Platform to do Optical Character Recognition on every picture I receive so I know what text is in the picture. I then run this text through a keyword/importance analysis algorithm to see what parts of the text are most important. Finally, I show this to the user and give them multiple options to share, etc.

Challenges I ran into

The biggest challenge I ran into was figuring out how to do Optical Character Recognition. I was first looking at tesseract which was another free but old Google library to do OCR but it never performed up to par with images that weren't preprocessed. This is why I switched to the Google Cloud which provided higher accuracy OCR to many images which were taken under many conditions. It still doesn't work perfectly and you need to take a somewhat solid picture but it does work pretty well.

Accomplishments that I'm proud of

I was able to hack this together in less than a week and make the app look pretty professional. I also do have over 10k active users!

What I learned

I learned a lot about the different techniques I can use to analyze the text. I also did learn about the Google Text to Speech API and a few other APIs as well.