Object Detection using OpenCV and YOLO in Rust

Overview

This project implements real-time object detection using the YOLO (You Only Look Once) deep learning model with OpenCV in Rust. The system loads a pre-trained YOLOv4 model, processes video frames, and detects objects with bounding boxes and confidence scores.

Features

• Loads YOLOv4 model weights and configuration.
• Reads class labels from a coco.names file.
• Processes video frames for real-time object detection.
• Applies non-maximum suppression (NMS) to reduce false positives.
• Displays detected objects with class labels and confidence scores.

Installation

Prerequisites

Ensure you have the following installed:

• Rust (latest stable version) with Cargo
• OpenCV (compiled with DNN module support)
• opencv-rust bindings

Setting Up the Project

Clone the repository and navigate to the project directory:

 git clone https://github.com/orbitalisvoid/yolo_tracking.git
 cd object-detection-rust

Install dependencies:

 cargo build

Usage

Prepare YOLO Model Files

Download the necessary YOLOv4 files and place them inside a yolo directory:

• yolov4.weights
• yolov4.cfg
• coco.names
• Video file (0.mp4) for testing

Run the Object Detection

 cargo run

This will process the video and print detected object classes with confidence scores.

Project Structure

.
├── src
│   ├── main.rs        # Main application file 
├── yolo
│   ├── yolov4.weights # YOLO model weights
│   ├── yolov4.cfg     # YOLO model configuration
│   ├── coco.names     # Class labels
│   ├── 0.mp4          # Sample video
├── Cargo.toml         # Rust dependencies and project metadata
└── README.md          # Project documentation

Configuration

Modify the YOLO_FOLDER_PATH constant in main.rs to point to the correct location of the YOLO model files.

const YOLO_FOLDER_PATH: &str = "/path/to/yolo/folder";

Future Enhancements

• Real-time Webcam Support: Implementing live detection using a webcam.
• Improved Performance: Optimize processing speed using GPU acceleration.
• Bounding Box Visualization: Draw detected objects on frames using OpenCV.

License

This project is licensed under the MIT License. Feel free to use and modify it as needed.

Acknowledgments

• OpenCV Team
• YOLO Developers (AlexeyAB, PJ Reddie)
• Rust OpenCV Bindings Community

yolo_tracking