In light of COVID-19, many on-campus dining locations have had to change the way they function. At RIT, one of these locations, the Commons, is frequented by many students, yet has an inefficient and not-COVID-friendly way of handling in-person orders. Frustrated by these, we decided to revamp the in-place system.

What it does

We take an image of an order slip, we extract the slip from the image (including straightening it and removing the background), we record the food items ordered according to markings on the slip, we display the order as well as total items from unfulfilled orders.

How we built it

Using open-cv, PyCharm, pyimage search, HTML, CSS, and FileZilla, we were able to put our heads together for roughly 22 hours to bring an implemented application with a GUI to life.

Challenges we ran into

We had trouble recognizing all of the boxes on the slips as well as determining the best camera setup.

Accomplishments that we're proud of

We got the boxes to work properly after about eight hours of fiddling with the program and trying different techniques.

What we learned

We learned a lot about computer vision, including contours. We also got to learn more about tkinter and pillow, as well as using FTP to upload a website for the first time.

What's next for Overhaul Ordering

Future steps in the project could include using OCR (optical character recognition) to recognize the names on the slips and using those instead of order numbers, improving the quality and aesthetic of the GUI, increasing the amount of test cases to increase the variability of marks on slips which will result in accurate readings, and possibly creating executables so this application can run on any modern Windows or MacOS system.


Website: link Presentation: link Python project: link

Built With

Share this project: