The inspiration behind the project was a combination of medicine and technology. As a team, we try our best to not miss an opportunity in making healthcare accessible and affordable. If we cannot use our knowledge and skills to make the world a better place, then what’s the use for it? Hence, medicine and technology inspire us, it brings out the best in us and motivates us. The feeling you get when you know that your solution is going to save the lives of millions or make one’s life much easier is definitely the best feeling ever. Also, the fact that no one in India has come up with the solution for this problem using ComputerVision
What it does
Our project uses Computer Vision and Deep Learning algorithms to detect if a patient is in a state of paralysis. If the algorithm detects that a person is in a state of paralysis, we have come up with a solution to alert the nearby person as well as any of the relatives/friends via an email. Our project revolves around making the solution as cheap and feasible for anyone to use.
How we built it
To detect moving objects from the video stream from a static camera, we have used background subtraction to separate the background from the foreground. After much experimentations with MOG, MOG2, and GMG algorithm, we have tuned the MOG algorithm to work perfectly in any conditions. We have used SMTP and socket communication for communicating via email to the required person.
Challenges we ran into
We had to code everything from scratch and shoot a video and tune our code according to the video for the best accuracy. We tried hard to send a text message or a phone call but did not get permission to do so on such short notice and we decided to use email as the alerting method. We had a hard time training our Deep learning in such a way that detects only the human and nothing else with pretty good accuracy
Accomplishments that I'm proud of
We are very proud that we have completed the implementation of your project and works with great accuracy as shown in the demonstration video. We will soon give it out for tryout in the nearest medical facility and finetune it if necessary. We are proud that we could create a finished product and expand on our idea more than what we had originally planned. Additionally, this project worked much better than expected.
What we learned
Working under pressure is something we learned to handle. We learned the importance of dividing work among us so that the work can be completed faster the final product portrays the best of everyone’s ability. The rigorous planning and jamming sessions, and the pressure of the deadline, taught us things that only projects like these can do. Since we are new to Arduino, we learned the basics from the workshops.
What's next for NeuroVision
1)We want to alert the user using SMS or Phone call instead of an email. 2)Fine-tune the algorithm further using reinforcement learning which could further increase the accuracy. 3)Add support for dynamic cameras.