Slouchn’t

Are you sick of having back and neck pain? We spend our time working 24/7 with our back pain attacking us every second in the present and future. So, how do we save our backs? USE SLOUCHN'T!

Comment

Inspiration

Slouchn’t was inspired by the high rates of back pain induced by bad posture, which is a severe issue that affects people of all ages, especially students and adults. This issue continues to impact people today, and the habit of bad posture has only worsened ever since COVID-19 started. Thus, we’re introducing Slouchn’t, a very accessible and effective way to improve your posture while working on your laptop or computer.

What it does

Slouchn’t first calibrates with the user to determine their correct and incorrect posture. With this collected data, the program will determine the range the user would have to meet in order to maintain a good posture. If the user does not have the right posture - straight back and neck - for a certain amount of time, the program will urge the user to adjust their posture through audio and video effects. The program also includes tips the user can keep in mind for maintaining the perfect posture.

How we built it

We used the opencv-python package, along with mediapipe to build our computer vision program. It prompts the webcam of the computer, which detects the motion of the person using specific body landmarks, including the shoulders, nose and eyes. The distances between these landmarks are then used to determine whether or not the person has incorrect posture. If the distance drops under a certain calculated threshold, the user is told to fix their posture before continuing their work. We tested this on multiple people to raise the accuracy of our program.

Challenges we ran into

One of our main challenges was figuring out how we should detect the bad posture of a user just using a webcam, since the webcam may not distinguish between the user looking down or slouching. We also struggled with initializing git and synchronizing the project files with all computers as well since it was our first time not using an online IDE.

Accomplishments that we're proud of

We are proud of being able to learn and implement computer vision, and then make it a running program that’s realistic and can help many people today with posture problems. Moreover, despite the challenges we were able to turn it into something relatively useful. We are also proud of expanding our learning of python and using different packages in our programs.

What we learned

Throughout this experience, we gained more knowledge about computer vision, how to implement sounds and utilizing motion detection. Some of our members are relatively new to hackathons, so we also learnt how to initialize git and do live collaboration on vscode. We also enhanced our python skills, along with our teamwork and communication skills.

What's next for Slouchn’t

We are hoping to include more features into Slouchn’t, such as customizable alert sounds for when bad posture is detected and also incorporate cleaner visuals. In the future, we are planning to make this program into a more convenient one by hosting it on a web server where anybody can access it, instead of keeping it as a python script. To further our reach, we can also incorporate a separate webcam instead of the computer one, as some people may not have a webcam in their laptop. This separate webcam also allows us to look at the user from different angles, further increasing our accuracy in helping the user with their posture. We also hope to collect our user's emails so that our program can send them their daily or weekly summary of how much they slouched.

Built With

Share this project:

Updates