Automation of the tedious checkout lines

What it does

INVENGO is a cross platform app that allows the store/warehouse to learn products and users by image recognition, and by doing so allowing us to track how a user interacted with a product

How I built it

We created an Android app for learning products and a UWP app for learning users and tracking their interaction with products. The API for login and internal logic is built with C# on .NET 4.5 and MSSQL with stored procedures

Challenges I ran into

Android OS doesn't allow apps to take images and submit them to our CDN without confirming each shot, and by doing so we couldn't to make a timer on the camera which made the product learning too tedious. So we shifted toward streaming video and analyzing frames which was not possible on mobile devices. We ended up recording a whole video, sending it to CDN and there it's analyzed.

Accomplishments that I'm proud of

Not giving in

What I learned

To document more on the permissions that Android is giving to developers

What's next for Invengo

Develop iOS app, provide other options for product identification: barcode / qr / input

Built With

Share this project: