OpenCV with CUDA Docker Setup

This project provides a Docker-based environment for running OpenCV with CUDA support, enabling GPU-accelerated computer vision tasks. The setup includes a complete build of OpenCV 4.10.0 with CUDA 12.6.3 and cuDNN support.

Features

• OpenCV 4.10.0 with CUDA support
• CUDA 12.6.3 with cuDNN
• GPU-accelerated image processing
• Pre-configured Docker environment
• Example code demonstrating CUDA-enabled OpenCV operations

Prerequisites

• Docker
• NVIDIA Container Toolkit
• NVIDIA GPU with CUDA support
• Docker Compose

Installation

1. Clone this repository:

git clone <your-repo-url>
cd <repo-name>

2. Ensure you have the NVIDIA Container Toolkit installed:

Usage

1. Build and run the container using Docker Compose:

docker-compose up --build

2. The example code will:
   • Check for CUDA availability
   • Create a random test image
   • Apply GPU-accelerated Gaussian blur
   • Save both original and processed images

Project Structure

.
├── Code/
│   ├── Dockerfile          # OpenCV with CUDA build configuration
│   └── main.py            # Example code demonstrating CUDA usage
├── docker-compose.yml     # Docker Compose configuration
└── README.md             # This file

Technical Details

• Base Image: NVIDIA CUDA 12.6.3 with cuDNN
• OpenCV Version: 4.10.0
• CUDA Architectures: 6.1, 7.0, 7.5, 8.0, 8.6, 8.9, 9.0
• Python Version: 3.12.3
• Key Features:
  • CUDA acceleration
  • cuDNN support
  • OpenGL support
  • Video codec support
  • TBB threading

Testing CUDA Support

The included main.py script demonstrates CUDA functionality by:

1. Checking CUDA availability
2. Creating a test image
3. Applying GPU-accelerated Gaussian blur
4. Saving the results

Contributing

Feel free to submit issues and enhancement requests!