By extracting text information from images, we can utilize the information to improve user experience. We focused on developing understand of Microsoft's cognitive service and implementing an app for using Microsoft's Computer Vision API.
What it does
This app allows to split the total amount of bill and share to friends. It gets the total amount on receipt by taking or loading a picture. The amount is divided evenly, then those amount will be shared to the selected user's friends. The friends get a message to pay the amount on the app.
How we built it
Based on Android environment, we built it using Java, Microsoft's Computer Vision API, and Firebase DB We collaborated android cell phone's camera and gallery along with various API's to process our data.
Challenges we ran into
Accomplishments that we're proud of
As we planned in the beginning of the Hackathon, we managed to almost finish an app that uses the computer vision API of Azure to identify the total cost written on a given print receipt. The app constitutes of multiple parts that were handled using the Firebase Authentication system and its database. It is a possible extension to an app, such as Venmo or PayPal, in which a user can simply take a photo and send a payment request to his or her friends. The process within the app divides up the computed total cost evenly by n, number of friends, hence the name "DutchPay" came.
What we learned
We learned a lot about Microsoft's Vision recognition technology. It would be nice if we could implement image enhancing & size reduction process in our app because latency delayed an app to process the data whenever pictures were too large. Although we could not completely finish the hacking that we initiated, Not only that we made a great progress on what we wished to achieve, but we also learned a better way to do group programming using Github's branches and learned quite a deal about android programming.