Patients who just recovered from stroke often get post-stroke muscle paralysis syndrome; a condition where some parts of their body (mainly hands and feet) lost their muscle memory. In a pre-pandemic world, people with post-stroke muscle paralysis would need to go to physiotherapy and do a guided exercise. However, knowing that access to physiotherapy is not as easily available to everyone, our team decided to replicate the experience with the help of a motion detector.
What it does
Muscle Mate prompts users to follow a series of exercise videos. Using a motion-tracker, it tracks whether users are doing the exercises properly.
How I built it
- using python OpenCV library to create a motion tracker
- creating HTML and CSS pages for the website
Challenges I ran into
- Detecting hands using opencv library was a challenge. It was our first time using doing motion detection, so we were not aware of all the built-in functions that were available. Our attempts of trying multiple algorithms of background subtraction and canny edge detector resulted in inconsistent behavior with different lighting and skin color.
Accomplishments that I'm proud of
- Working motion detector. We decided to put a trackbar so user can specify the HSV colorspace of the object they're tracking.
- Learning HTML and CSS in a day & understanding how everything works.
- The idea of helping people with post-stroke muscle paralysis and giving back to the community
What I learned
- using openCV library to track motion
- positioning elements in CSS
What's next for Muscle Mate
- We are taking a step back to redesign the motion detector so that it has better accuracy and precision. For example, differentiating fingers or palm, left hand or right hand
- Working on user authentication and authorization so users can save their progress.
Muscle Mate as an idea has a lot of potentials to help people. We are very excited to explore more on what Muscle Mate can (and will) do!