Car Detection System

A comprehensive car detection system built on YOLOv5, providing real-time vehicle detection and classification capabilities.

= Overview

This project implements a state-of-the-art car detection system using YOLOv5 (You Only Look Once version 5), a fast and accurate object detection model. The system can detect various types of vehicles including cars, trucks, buses, and motorcycles in images and video streams.

(' Features

• Real-time Detection: Fast inference for live video streams
• Multiple Vehicle Types: Detects cars, trucks, buses, motorcycles, and more
• High Accuracy: State-of-the-art YOLOv5 models for precise detection
• Multiple Input Formats: Supports images, videos, and webcam streams
• Easy Training: Tools for custom dataset training
• Export Options: Convert models to various formats (ONNX, TensorRT, etc.)
• Web Interface: Optional Flask REST API for easy integration

= Installation

Prerequisites

• Python 3.8 or higher
• CUDA-compatible GPU (optional, for faster inference)

Quick Start

1. Clone the repository bash git clone https://github.com/SoObai/Car-Detection-System.git cd Car-Detection-System

2. Install dependencies bash pip install -r requirements.txt

3. Download pre-trained weights (optional) ```bash

   Download YOLOv5s weights

   wget https://github.com/ultralytics/yolov5/releases/download/v7.0/yolov5s.pt ```

= Usage

Basic Detection

Detect cars in an image

bash python detect.py --source path/to/image.jpg --weights yolov5s.pt

Detect cars in a video

bash python detect.py --source path/to/video.mp4 --weights yolov5s.pt

Real-time webcam detection

bash python detect.py --source 0 --weights yolov5s.pt

Advanced Options

Save results

bash python detect.py --source path/to/image.jpg --weights yolov5s.pt --save-txt --save-conf

Adjust confidence threshold

bash python detect.py --source path/to/image.jpg --weights yolov5s.pt --conf 0.5

Use different model sizes

```bash

YOLOv5n (nano) - fastest

python detect.py --weights yolov5n.pt

YOLOv5s (small) - balanced

python detect.py --weights yolov5s.pt

YOLOv5m (medium) - more accurate

python detect.py --weights yolov5m.pt

YOLOv5l (large) - very accurate

python detect.py --weights yolov5l.pt

YOLOv5x (extra large) - most accurate

python detect.py --weights yolov5x.pt ```

< Training

Train on Custom Dataset

1. Prepare your dataset in YOLO format: dataset/ %%% images/ % %%% train/ % %%% val/ %%% labels/ %%% train/ %%% val/

2. Create dataset configuration (data/cars.yaml): yaml path: ../dataset train: images/train val: images/val nc: 1 # number of classes names: ['car'] # class names

3. Start training: bash python train.py --data data/cars.yaml --weights yolov5s.pt --epochs 100

Training Options

Resume training

bash python train.py --resume runs/train/exp/weights/last.pt

Multi-GPU training

bash python train.py --data data/cars.yaml --weights yolov5s.pt --device 0,1

Hyperparameter tuning

bash python train.py --data data/cars.yaml --weights yolov5s.pt --hyp data/hyps/hyp.scratch-high.yaml

= Validation

Validate trained model

bash python val.py --weights runs/train/exp/weights/best.pt --data data/cars.yaml

Validate with custom confidence

bash python val.py --weights runs/train/exp/weights/best.pt --data data/cars.yaml --conf 0.5

=' Model Export

Export to different formats

ONNX format

bash python export.py --weights yolov5s.pt --include onnx

TensorRT format

bash python export.py --weights yolov5s.pt --include engine --device 0

CoreML format

bash python export.py --weights yolov5s.pt --include coreml

TensorFlow formats

bash python export.py --weights yolov5s.pt --include saved_model pb tflite

< Web API

Start Flask REST API

bash python utils/flask_rest_api/restapi.py

API Usage

bash curl -X POST "http://localhost:5000/v1/object-detection/yolov5s" \ -H "Content-Type: image/jpeg" \ --data-binary @path/to/image.jpg

= Project Structure

Car-Detection-System/ %%% detect.py # Detection script %%% train.py # Training script %%% val.py # Validation script %%% export.py # Model export script %%% requirements.txt # Python dependencies %%% models/ # Model configurations % %%% yolo.py # YOLO model definition % %%% common.py # Common model components % %%% hub/ # Pre-configured models %%% data/ # Dataset configurations % %%% coco.yaml # COCO dataset config % %%% hyps/ # Hyperparameter files %%% utils/ # Utility functions % %%% general.py # General utilities % %%% plots.py # Plotting functions % %%% torch_utils.py # PyTorch utilities % %%% flask_rest_api/ # Web API %%% classify/ # Classification module %%% segment/ # Segmentation module

< Supported Datasets

• COCO: Common Objects in Context
• VOC: Pascal VOC
• ImageNet: Large-scale image classification
• Custom: Your own datasets

=

Detection Classes

The system can detect various vehicle types: - Car - Truck - Bus - Motorcycle - Bicycle - And more (depending on training data)

= Performance

| Model | Size | Speed (ms) | mAP@0.5 | mAP@0.5:0.95 | |-------|------|------------|---------|--------------| | YOLOv5n | 1.9M | 6.3 | 28.0 | 45.7 | | YOLOv5s | 7.2M | 6.4 | 37.4 | 56.8 | | YOLOv5m | 21.2M | 8.2 | 45.4 | 64.1 | | YOLOv5l | 46.5M | 10.1 | 49.0 | 67.3 | | YOLOv5x | 87.7M | 12.1 | 50.7 | 68.9 |

Performance metrics on COCO dataset

> Contributing

We welcome contributions! Please see our Contributing Guidelines for details.

How to Contribute

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

= License

This project is licensed under the MIT License - see the LICENSE file for details.

=O Acknowledgments

• Ultralytics for the YOLOv5 implementation
• PyTorch for the deep learning framework
• The open-source community for various tools and libraries

= Support

If you encounter any issues or have questions:

1. Check the Issues page
2. Create a new issue with detailed information
3. Join our community discussions

= Updates

Stay updated with the latest features and improvements by: - Starring this repository - Watching for updates - Following our release notes

---

Made with d' for the computer vision community