# Quick Start Guide - Testing the Model

This guide will help you quickly test the Highway Vehicle Detection model on new data.

## Prerequisites

1. **Install Python dependencies:**
   ```bash
   pip install -r requirements.txt
   ```

2. **Download/clone the repository** from Hugging Face:
   ```bash
   git clone https://huggingface.co/bichuche0705/highway-vehicle-detection-code
   cd highway-vehicle-detection-code
   ```

   Or download manually from: https://huggingface.co/bichuche0705/highway-vehicle-detection-code

## Model Files Location

The trained models are located at:
- **Best model (recommended)**: `training_runs/yolov8m_stage2_improved/weights/best.pt`
- **Alternative location**: `models/yolov8m_stage2_improved_best.pt`
- **Stage 1 model**: `training_runs/yolov8m_stage1_smart/weights/best.pt`

## Method 1: Simple Test Script (Recommended for Quick Testing)

Use the `test_model.py` script for easy testing:

### Test on an Image:
```bash
python test_model.py your_image.jpg
```

The script will:
- Automatically find the model
- Detect vehicles in the image
- Save the result as `your_image_result.jpg`

### Test on a Video:
```bash
python test_model.py your_video.mp4 --video
```

With custom output:
```bash
python test_model.py your_video.mp4 --video --output output_video.mp4
```

## Method 2: Using example_usage.py

```python
from example_usage import detect_vehicles, process_video

# For images
detect_vehicles("test_image.jpg", model_path="models/yolov8m_stage2_improved_best.pt")

# For videos
process_video("test_video.mp4", output_path="output.mp4")
```

## Method 3: Using main.py (Full Application)

The `main.py` script provides advanced features like vehicle counting and tracking:

```bash
# Basic usage (auto-detects first .mp4 in directory)
python main.py --model training_runs/yolov8m_stage2_improved/weights/best.pt

# With specific video
python main.py --model training_runs/yolov8m_stage2_improved/weights/best.pt --video your_video.mp4

# Save output video
python main.py --model training_runs/yolov8m_stage2_improved/weights/best.pt --video input.mp4 --output output.mp4
```

**Important**: You need to specify the `--model` path because the default path in `main.py` might not match the repository structure.

## Method 4: Direct Python Code

```python
from ultralytics import YOLO

# Load model
model = YOLO('training_runs/yolov8m_stage2_improved/weights/best.pt')

# Test on image
results = model('your_image.jpg')
results[0].show()  # Display results
results[0].save('output.jpg')  # Save results

# Test on video
results = model('your_video.mp4', save=True)
# Output saved to: runs/detect/predict/
```

## Detected Vehicle Classes

The model detects 8 vehicle types:
1. **auto** - Three-wheelers
2. **bus** - Public transport vehicles
3. **car** - Passenger cars
4. **lcv** - Light Commercial Vehicles
5. **motorcycle** - Two-wheelers
6. **multiaxle** - Multi-axle heavy vehicles
7. **tractor** - Agricultural/construction vehicles
8. **truck** - Heavy vehicles

## Troubleshooting

### Model Not Found Error

If you get "Model file not found":
1. Make sure you've downloaded/cloned the repository completely
2. Check that the model file exists at one of these paths:
   - `training_runs/yolov8m_stage2_improved/weights/best.pt`
   - `models/yolov8m_stage2_improved_best.pt`

### Import Errors

If you get import errors:
```bash
pip install ultralytics opencv-python numpy torch torchvision
```

Or install all requirements:
```bash
pip install -r requirements.txt
```

### CUDA/GPU Issues

If you have CUDA available, PyTorch will use it automatically. For CPU-only:
- The model will run but will be slower
- No additional setup needed

## Example Output

When running the model, you should see:
- Bounding boxes around detected vehicles
- Class labels (auto, bus, car, etc.)
- Confidence scores
- Output saved to files

## Need Help?

Check the repository README.md for more detailed documentation, or review the code comments in:
- `main.py` - Full application with counting
- `example_usage.py` - Simple examples
- `test_model.py` - Quick test script