flutter_cv2\

*** version 0.0.1 ***

A custom Flutter plugin that bridges the power of OpenCV in C++ with real-time camera access in Flutter. Designed to enable real-time image streaming, frame analysis, and snapshot capturing—perfect for AI/ML projects focused on computer vision.

🚀 Currently available on Linux with upcoming support for Web, Android, Windows, macOS, and iOS in weekly updates.

✨ Why I Built This

As a developer passionate about AI and computer vision, I constantly faced a pain point: lack of a cross-platform real-time camera plugin that works seamlessly with OpenCV and Flutter.

Most camera plugins are either platform-limited or lack direct access to frame buffers required for AI/ML models. So, I decided to step up and build my own solution — one that combines the low-level power of C++ OpenCV with the flexibility of Flutter.

If you work with mobile or edge AI, this is the plugin you've been waiting for.

✅ Features

  • 🔧 Real-time camera feed powered by C++ OpenCV.
  • 🎯 Cv2Camera widget to render frames in your Flutter app.
  • 🔄 Frame subscription via onFrame (NumPy-array format).
  • 🖼️ Access raw image bytes via onByte.
  • 📸 Snapshots with onSnap — save or analyze.
  • 🔄 Flip camera feed (horizontal/vertical/both).
  • 🔌 Built with FFI and native C++.
  • 💻 Linux supported now. Cross-platform rollout coming weekly.

🚀 Getting Started

  1. Add to your pubspec.yaml:
dependencies:
  flutter_cv2_camera:
    git:
      url: https://github.com/jamesadewara/flutter_cv2_camera

  1. Linux prerequisites:

    • Make sure you have OpenCV (opencv4) installed.
    • CMake >= 3.10.

  2. Your Flutter app must run on Linux for now:

    flutter run -d linux

Windows, Web, and Android support will be added soon — follow the repo for updates.

🛠️ Usage

final controller = Cv2CameraController();

Cv2Camera(
  controller: controller,
  onFrame: (frame) {
    // Access as NumPy array (in Uint8List form)
    processFrame(frame.bytes);
  },
  onByte: (bytes) {
    // Save or stream
    upload(bytes);
  },
  flipCode: 0, // vertical: 0, horizontal: 1, both: -1
  width: 300,
  height: 250,
)

To capture a snapshot:

final bytes = await controller.takeSnap();
// Save or analyze

📷 Screenshots / Demo

🤝 Contributing

I would love collaborators! If you:

  • ❤️ OpenCV
  • 💡 Know Flutter + C++
  • 🧠 Love AI or mobile edge computing

Let’s build the future of mobile-first AI vision systems together. PRs and issues are highly welcome.

🔮 Roadmap

  • ✅ Linux support
  • 🔜 Web support
  • 🔜 Android support
  • 🔜 Windows support
  • 🔜 macOS & iOS

📩 Stay Updated

Follow me on LinkedIn or GitHub for weekly updates.

License

MIT

Libraries

flutter_cv2_camera