README.md

---
title: Train obstruction detection V1
emoji: 🚗
colorFrom: blue
colorTo: green
sdk: streamlit
sdk_version: "1.29.0"
app_file: app.py
pinned: false
---


# Traffic Light Detection with Protection Area

This Streamlit app performs traffic light detection and monitors objects within a user-defined protection area. The app uses:
- ShuffleNetV2 for traffic light detection
- YOLO for object detection
- Interactive point selection for defining protection areas

## Features
- Upload and process images
- Interactive protection area definition
- Traffic light state detection
- Object detection within protection area
- Real-time visualization

## How to Use
1. Upload an image using the file uploader
2. Click on the image to define 4 points for the protection area
3. Use "Reset Points" if you need to start over
4. Click "Process Detection" when ready
5. View the results showing traffic light state and detected objects

## Models
- Traffic Light Detection: ShuffleNetV2
- Object Detection: YOLOv8

## Requirements
See requirements.txt for all dependencies.