Inspiration
There was a lot of inspiration for this project. One of them was addressing theft crimes which have been increasing in the recent years, so to take initiative, our team worked towards helping lowering these crimes. Additionally, assisting the elderly in their own homes and act as first responders. Looking back in Another inspiration came from Canada’s recent wildfire, the starting fire was caught on CCTV. With our model integrated we can detect fires early on, preventing any catastrophic damages.
What it does
Our project integrates an advanced object detection AI model with CCTV cameras, turning it into a proactive safety solution. This AI model constantly analyzes the video feed, identifying and labeling objects within its view. The system is designed to send alerts to authorities or emergency contacts.
How we built it t
Python was our choice of language and to help us with the computer vision we used openCV and paired that with Yolov10 cnn mode for object detections and trained the cnn on fire detection and fall detection using roboflow.
Challenges we ran into
In the process of training our AI model, we came across many errors that we had to go back to look for. One of them were training time took long more than 6 hours. Also figuring out how to track people were one of the challenges we ran to.
Accomplishments that we're proud of
We were able to do a very well object detection.
What we learned
We learned the do’s and don’ts of how to properly train a model. We learned how to properly classify label images.
What's next for Meelo
Our next plan is working towards integrating this system into police body cams, fact checking them and sending alerts to the higher officials.
Log in or sign up for Devpost to join the conversation.