https://github.com/sohitbennett/roadsafe
A Deep learning computer vision system for real-time traffic safety monitoring.
https://github.com/sohitbennett/roadsafe
computer-vision esrgan keras numpy pandas python scikit-learn tensorflow tesseract-ocr yolov5 yolov8
Last synced: 3 months ago
JSON representation
A Deep learning computer vision system for real-time traffic safety monitoring.
- Host: GitHub
- URL: https://github.com/sohitbennett/roadsafe
- Owner: SohitBennett
- Created: 2025-09-16T20:54:28.000Z (9 months ago)
- Default Branch: main
- Last Pushed: 2025-09-16T21:02:58.000Z (9 months ago)
- Last Synced: 2025-09-16T23:48:37.053Z (9 months ago)
- Topics: computer-vision, esrgan, keras, numpy, pandas, python, scikit-learn, tensorflow, tesseract-ocr, yolov5, yolov8
- Language: Jupyter Notebook
- Homepage:
- Size: 2.71 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# 🚦 RoadSafe
RoadSafe is a deep learning–powered computer vision system designed for **real-time traffic safety monitoring**.
The project integrates multiple models for **helmet detection, seatbelt compliance, vehicle occupancy recognition, and license plate extraction**, enabling automated enforcement and improved road safety.
---
## ✨ Features
- 🪖 **Helmet Detection** – Ensures two-wheeler riders follow safety compliance.
- ⛑️ **Seatbelt Detection** – Identifies drivers/passengers not wearing seatbelts.
- 🧍 **Occupancy Detection** – Counts people in vehicles to monitor overcrowding.
- 🔢 **License Plate Extraction** – Detects and extracts number plates using OCR (Tesseract).
- 🖼️ **Image Enhancement** – Integrated **ESRGAN** for super-resolution on low-quality traffic footage.
- ⚡ **Real-Time Analysis** – Built with **YOLOv5/YOLOv8 + OpenCV** for live video stream monitoring.
---
## 🛠️ Tech Stack
- **Languages**: Python
- **Deep Learning Frameworks**: TensorFlow, Keras, PyTorch
- **Models**: YOLOv5, YOLOv8, ESRGAN
- **Libraries**: OpenCV, NumPy, Pandas, scikit-learn
- **OCR**: Tesseract OCR
---
SCREENSHOT -
