Inspiration
With the development of Singapore's society and factories, security and safety issues are becoming more critical. With a lack of manpower, people sometimes need to supervise different areas at the same time and intervene in time to stop accidents when they happen.
What it does
To detect objects in a marked area. Can be used for surveillance purposes, baby care, etc.
How we built it
- Color recognition API using KNN classifier To identify objects colour Dataset: Color samples
- Highlight Region Of Interest and detect entrance. With a script to extract ROI region using mouse event for new inputs
- Yolo V4 & Deepsort To identify object & extract object information Dataset: COCO dataset ## Challenges we ran into Limited time ## Accomplishments that we're proud of We made use of KNN classifier, yolo and deepsort, to detect and tract objects, and pass the detected object to color identification. When detected objects move into region of interest area, they will be marked as 'enter'. ## What we learned Deep Learning, Computer Vision, Object Detection using yolo & deepsort, Tensorflow, Python, Surveillance ## What's next for The Guardian AI Can link to IoT solution to build a complete or half auto system to guard and protect specific area
Built With
- cocodataset
- deepsort
- machine-learning
- python
- yolo
Log in or sign up for Devpost to join the conversation.