Welcome to Hackster!
Hackster is a community dedicated to learning hardware, from beginner to pro. Join us, it's free!

Accident Detection From CCTV base in Bangkok.

It's a guardian angel for public spaces, roads, and workplaces – not just an innovation, but the future of accident prevention

20
Accident Detection From CCTV base in Bangkok.

Things used in this project

Hardware components

Raspberry Pi 4 Model B
Raspberry Pi 4 Model B
×1
Camera (generic)
×1

Software apps and online services

OpenCV
OpenCV – Open Source Computer Vision Library OpenCV
The primary asset for detecting traffic accidents is the OpenCV object detection model. The image input from CCTV cameras will be processed through three distinct models: Car Detection Model: This model serves to determine distance and speed. It includes distance estimation, which will subsequently be utilized to calculate speed data relative to historical frames.
Note that There are a lot of efficient pre-trained models for car detection available online, and it will be consider later on the project based on our time limitations and resources. Accident Classification Model: This model operates on a frame-by-frame basis to detect accidents. It classifies the type of accident in each frame. Fire Signal Detection Model: This model detects fire and car signals on a frame-by-frame basis. Prior to detection, the images will undergo preprocessing to enhance high-contrast or light-intensive areas. Each of these models will produce confident score outputs. Based on these scores, the program will detect and trigger alerts for moderators through an application when a traffic accident occurs. The scores from each model will be weighted and utilized to calculate the type and severity of the accident.
Kaggle Dataset for Vehicle Tracking
Roboflow Dataset for Vehicle Tracking
Roboflow Dataset for Accident Detection
Kaggle Dataset for Accident Detection
Github Dataset for Fire and Smoke Detection
Jupyter Notebook
Jupyter Notebook

Story

Read more

Credits

Park Saranunt
4 projects • 3 followers
Contact
Oranus Kotsuwan
0 projects • 3 followers
Contact
Paweena Kanokhong
0 projects • 3 followers
Contact
Chayawat
0 projects • 3 followers
Contact

Comments

Please log in or sign up to comment.