Download the file 'fingerCursor.py' and simply run the program. Make sure you have the packages installed with pip on your machine! Watch this video to see a demo of AirWrite!
From hours of long calculus homework, to intense days of trying to input a diagram onto google docs, we could do no more. We had difficulty creating alternatives for doing things on the web to easily upload, while doing work through handwriting is less acceptable and doesn't easily upload to certain platforms. With many high school and college students in mind, we had to make a difference in terms of how a teacher could send hand written notes online, as well as how a student can submit work this way as well. COVID has employed around 12x the use of online learning for schools all across the US, and in an effort to create a better and more well off society, we pursued AirWrite!
Our program runs on a computer, and tracks your finger based on location in the frame of the computer's wide angle camera. You are then able to draw and create new diagrams, letters, words, and numbers, simply with the movements of your fingers.
For this project, we used one of our personal favorite programming languages: Python. In terms of python, we chose OpenCV to create our computer vision model, and numpy arrays for our frame by frame, to allow the processing of the data by the computer vision model.
At first, we were unsure on how we were going to start the program and what must we do in order to even recognize movements using OpenCV. However, over countless hours of research and Stack Overflow articles, we were able to create the final product that is currently in the GitHub repository linked below. Following this, we continue to struggle with the frame rate on our machine's camera output, and have a difficult time finding the proper lighting such that our algorithm can run its full course, to the best of its ability.
We are glad we were able to create a functional algorithm that detects writing in the air! We truly believe this is the next revolutionary idea in the education industry and that algorithms like these can make writing in the classroom so much easier. We hope this project can gain tons of outreach, and we can teach kids the power of AI, and Computer Vision and how it can affect almost every aspect of our lives.
We learned a great amount about functionalities of the Python modules, OpenCV and NumPy. The fact that we could create such a complex algorithm with just three imported modules shows how technology has come to progress over the years. We loved the way that the simplicity of different libraries, can come together, and we love how we were able to incorporate them in this way.
We hope to implement this algorithm into an application that can used by teachers and students across the world. This way, we can allow the classroom to once again be revolutionized, and take the world by storm with new technology, following the internet age. The future of writing is truly here!
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent