Background Remover for Laravel

A Laravel wrapper for the artisan-build/bg-remover C++ binary, providing fast and efficient background removal using OpenCV's GrabCut algorithm and optional ML-based segmentation.

Features

• 🚀 Fast - Native C++ performance using OpenCV
• 🤖 ML Support - Optional ONNX Runtime for ML-based segmentation (U2-Net, RMBG)
• 🎯 Simple API - Easy-to-use Laravel service class
• ☁️ Cloud Storage - Built-in support for Laravel Storage (S3, local, etc.)
• 📦 Multi-platform - Supports Ubuntu (Forge) and macOS ARM64
• 🔒 Checksum Verification - Ensures binary integrity
• ⚡ Queue Support - Async processing via Laravel queues
• 🔧 Build from Source - Includes full C++ source code and build system
• 🎨 Demo UI - Interactive web interface for testing

Requirements

• PHP 8.2+
• Laravel 11.0+ or 12.0+

Installation

Install the package via Composer:

composer require artisan-build/background-remover

Install the binary for your platform:

php artisan background-removal:install

Platform Auto-Detection

The install command automatically detects your platform:

• Ubuntu/Debian → bg-remover-ubuntu-x86_64 (for Laravel Forge, Vapor, etc.)
• macOS → bg-remover-macos-arm64 (for Apple Silicon development)

Manual Platform Selection

If auto-detection fails or you need a specific platform:

php artisan background-removal:install --platform=ubuntu
php artisan background-removal:install --platform=macos-arm64

Custom Platform Support

The package only maintains binaries for platforms we actively use. For other platforms (Windows, Raspberry Pi, Arch, etc.), see the FORKING.md guide to build your own binary.

Quick Install Binary (Standalone)

If you're not using Laravel or want to quickly install the binary:

# Run the installation script
./install-binary.sh

This will auto-detect your platform and download the appropriate binary to bin/bg-remover.

Building from Source

This package now includes the complete C++ source code. You can build the binary yourself:

Quick Build (macOS)

# Install OpenCV and ONNX Runtime
brew install opencv onnxruntime

# Build with ML support
cd cpp-src
make ML=1

# Or build without ML support
make

# Install
mkdir -p ../bin
cp bg-remover ../bin/

Quick Build (Ubuntu/Debian)

# Install dependencies
sudo apt-get update
sudo apt-get install -y g++ make pkg-config libopencv-dev

# Optional: Install ONNX Runtime for ML support
# Follow: https://github.com/microsoft/onnxruntime/releases

# Build with ML support (if ONNX Runtime installed)
cd cpp-src
make ML=1

# Or build without ML support
make

# Install
mkdir -p ../bin
cp bg-remover ../bin/

Docker Builds (Cross-platform)

cd cpp-src

# Build for Ubuntu (includes ML support)
make ubuntu

# Install
mkdir -p ../bin
cp bg-remover-ubuntu-x86_64 ../bin/bg-remover
chmod +x ../bin/bg-remover

For detailed build instructions, see cpp-src/README.md.

Demo UI

Try out the background remover with a simple web interface:

1. Install the binary (see above)
2. Start the demo:

   cd demo
   php -S localhost:8000

3. Open http://localhost:8000 in your browser
4. Upload an image and see the result

The demo features:

• Drag-and-drop image upload
• Side-by-side comparison
• Download processed images
• Support for JPG and PNG

For more details, see demo/README.md.

Usage

Basic Usage

use ArtisanBuild\BackgroundRemover\Services\BackgroundRemovalService;

$service = app(BackgroundRemovalService::class);

// Remove background from local files
$service->removeBackground(
    inputPath: '/path/to/input.jpg',
    outputPath: '/path/to/output.png'
);

Using Laravel Storage

use ArtisanBuild\BackgroundRemover\Services\BackgroundRemovalService;

$service = app(BackgroundRemovalService::class);

// Remove background using Storage disks
$service->removeBackgroundFromStorage(
    inputDisk: 's3',
    inputPath: 'uploads/photo.jpg',
    outputDisk: 's3',
    outputPath: 'processed/photo-no-bg.png'
);

Using Queues (Recommended for Production)

use ArtisanBuild\BackgroundRemover\Jobs\RemoveBackgroundJob;

// Dispatch to queue
RemoveBackgroundJob::dispatch(
    inputDisk: 's3',
    inputPath: 'uploads/photo.jpg',
    outputDisk: 's3',
    outputPath: 'processed/photo-no-bg.png'
);

// Dispatch to specific queue
RemoveBackgroundJob::dispatch(...)
    ->onQueue('image-processing');

// Dispatch to specific connection
RemoveBackgroundJob::dispatch(...)
    ->onConnection('redis');

Configuration

Publish the config file:

php artisan vendor:publish --tag=background-remover-config

This creates config/background-remover.php:

return [
    // Path to the binary (auto-installed to base_path('bin/bg-remover'))
    'binary_path' => env('BG_REMOVER_BINARY_PATH', base_path('bin/bg-remover')),

    // GitHub release settings
    'github' => [
        'repo' => 'artisan-build/bg-remover',
        'version' => 'latest', // or specific version like 'v1.0.0'
    ],

    // Platform (null = auto-detect)
    'platform' => env('BG_REMOVER_PLATFORM', null),

    // Timeout in seconds
    'timeout' => env('BG_REMOVER_TIMEOUT', 60),

    // Temporary directory for processing
    'temp_dir' => env('BG_REMOVER_TEMP_DIR', sys_get_temp_dir()),

    // Queue configuration
    'queue' => [
        'connection' => env('BG_REMOVER_QUEUE_CONNECTION', null),
        'queue' => env('BG_REMOVER_QUEUE', 'default'),
    ],
];

Environment Variables

# Custom binary path
BG_REMOVER_BINARY_PATH=/usr/local/bin/bg-remover

# Force specific platform
BG_REMOVER_PLATFORM=alpine

# Timeout (seconds)
BG_REMOVER_TIMEOUT=120

# Queue settings
BG_REMOVER_QUEUE_CONNECTION=redis
BG_REMOVER_QUEUE=image-processing

Laravel Vapor Deployment

The binary is automatically downloaded during build. No additional configuration needed!

# Vapor will automatically run:
php artisan background-removal:install

The Ubuntu binary works on AWS Lambda environments (including Vapor) and includes full ML support.

Laravel Forge Deployment

Add to your deployment script:

php artisan background-removal:install

The binary is installed to base_path('bin/bg-remover') and committed (or installed on each deploy).

Production Best Practices

1. Use Queues

Always process images asynchronously in production:

RemoveBackgroundJob::dispatch(...)
    ->onQueue('image-processing');

2. Configure Timeout

For large images, increase the timeout:

BG_REMOVER_TIMEOUT=120

3. Monitor Storage

The service downloads files to temporary storage for processing. Ensure adequate disk space:

// Default uses sys_get_temp_dir()
// Or customize:
BG_REMOVER_TEMP_DIR=/path/to/large/temp/dir

4. Error Handling

Wrap processing in try-catch:

use Illuminate\Process\Exceptions\ProcessFailedException;

try {
    $service->removeBackgroundFromStorage(...);
} catch (ProcessFailedException $e) {
    Log::error('Background removal failed', [
        'error' => $e->getMessage(),
        'input' => $inputPath,
    ]);
}

Testing

# Run package tests
composer test

# Run with coverage
composer test:coverage

How It Works

1. Download: Images are downloaded from Laravel Storage to temporary files
2. Process: The C++ binary processes the image using OpenCV's GrabCut algorithm
3. Upload: Processed images are uploaded back to Laravel Storage
4. Cleanup: Temporary files are automatically deleted

Platform Notes

Ubuntu (Laravel Forge, Vapor, etc.)

• Uses system OpenCV libraries
• Includes ONNX Runtime for ML support
• Standard glibc-based binary
• ~23KB binary size

macOS ARM64 (Development)

• Native Apple Silicon binary
• Uses Homebrew OpenCV and ONNX Runtime
• Full ML support included
• Universal x86_64 support coming soon

ML Mode

All binaries are compiled with ML support via ONNX Runtime. To use ML-based segmentation:

# Download an ONNX model (e.g., U2-Net, RMBG-1.4)
# Then use with --ml flag
bg-remover -i input.jpg -o output.png --ml --model path/to/model.onnx

ML mode provides superior edge detection and works better with complex backgrounds compared to the traditional GrabCut algorithm.

Project Structure

background-remover/
├── src/                      # Laravel package source
│   ├── Commands/             # Artisan commands
│   ├── Jobs/                 # Queue jobs
│   ├── Providers/            # Service providers
│   └── Services/             # Background removal service
├── cpp-src/                  # C++ source code
│   ├── bg-remover.cpp        # Main C++ implementation
│   ├── Makefile              # Build system with ML support
│   ├── Dockerfile.ubuntu     # Ubuntu build with ONNX Runtime
│   └── README.md             # Build documentation
├── demo/                     # Interactive demo UI
│   ├── index.html            # Web interface
│   ├── process.php           # Processing endpoint
│   └── README.md             # Demo documentation
├── config/                   # Laravel config
├── tests/                    # Package tests
├── bin/                      # Binary installation directory
├── install-binary.sh         # Binary installation script
└── README.md                 # This file

Troubleshooting

Binary Not Found

# For Laravel projects
php artisan background-removal:install

# For standalone usage
./install-binary.sh

# Or build from source
cd cpp-src && make && mkdir -p ../bin && cp bg-remover ../bin/

Permission Denied

# Make binary executable
chmod +x bin/bg-remover

Platform Not Supported

Build from source using the included C++ code:

cd cpp-src
# Follow instructions in cpp-src/README.md

Or see FORKING.md for building custom platform binaries.

Credits

• Binary: artisan-build/bg-remover
• Algorithm: OpenCV GrabCut
• Maintained by: Artisan Build

License

MIT License. See LICENSE.md for details.

Related Projects

• artisan-build/bg-remover - The C++ binary
• artisan-build/scalpels - Collection of Laravel utilities