Перейти до основного вмісту

Camera Service Service

wdio-camera-service is a 3rd party package, for more information please see GitHub | npm

A WebdriverIO service that enables camera feed injection for testing applications that use camera/video inputs. This service allows you to mock camera feeds with pre-recorded video files during automated testing.

Features​

  • 🎥 Inject custom video feeds into Chrome browsers during testing
  • 🔄 Dynamically change camera sources during test execution
  • 📁 Automatic video directory management
  • 🔄 Automatic format conversion - Use MP4, WebM, PNG, JPG and more (requires FFmpeg)
  • 💾 Smart caching - Converted files are cached to avoid re-conversion
  • 🧪 Perfect for testing camera-dependent applications like QR code scanners, video conferencing, etc.

Installation​

npm install --save-dev wdio-camera-service

Configuration​

Add the camera service to your WebdriverIO configuration:

// wdio.conf.ts
export const config: WebdriverIO.Config = {
// ... other config
services: [
['camera', {
defaultCameraFeed: './camera/default.mjpeg',
videoDirectory: './camera/video',
}],
],
// ... other config
};

Service Options​

OptionTypeRequiredDefaultDescription
defaultCameraFeedstring✅-Path to the default video file
videoDirectorystring✅-Directory for session-specific video files
ffmpegPathstring❌'ffmpeg'Custom path to FFmpeg executable
cacheEnabledboolean❌trueEnable caching of converted files
outputFormat'mjpeg' | 'y4m'❌'mjpeg'Output format for converted files

Supported Formats​

The service supports multiple input formats with automatic conversion:

Native Formats (No Conversion)​

  • .mjpeg - Motion JPEG
  • .y4m - YUV4MPEG2

Video Formats (Requires FFmpeg)​

  • .mp4 - MPEG-4 Video
  • .webm - WebM Video
  • .avi - AVI Video
  • .mov - QuickTime Video
  • .gif - GIF (animated or static)

Image Formats (Requires FFmpeg)​

Images are converted to a single-frame MJPEG (Chrome loops it automatically):

  • .png - PNG Image
  • .jpg / .jpeg - JPEG Image
  • .bmp - Bitmap Image

FFmpeg Requirement​

FFmpeg is only required when using non-native formats (MP4, WebM, PNG, etc.). If you only use .mjpeg or .y4m files, FFmpeg is not needed.

Installing FFmpeg​

macOS:

brew install ffmpeg

Ubuntu/Debian:

sudo apt-get install ffmpeg

Windows:

winget install FFmpeg
# or
choco install ffmpeg

Or download from: ffmpeg.org

Browser Support​

  • ✅ Chrome/Chromium/Android Chrome - Full support
  • ❌ Firefox - Not supported
  • ❌ Safari - Not supported
  • ❌ Edge - Not supported (unless Chromium-based)

Note: This service only works with Chrome/Chromium browsers as it relies on Chrome-specific command line arguments for camera mocking.

Android SDK Supports (With Chrome)​

SDKVersionSupport?
3112✅
3313✅
3414✅
3515✅
3616✅

Usage​

Basic Usage​

The service automatically injects the default camera feed when the browser starts:

describe('Camera Tests', () => {
it('should use default camera feed', async () => {
await browser.url('https://example.com/camera-app');
// Your default camera feed is now active
});
});

Using Different Formats​

You can use various formats as your default camera feed:

// wdio.conf.ts
export const config: WebdriverIO.Config = {
services: [
['camera', {
// Use a PNG image as the default feed (Chrome loops it automatically)
defaultCameraFeed: './camera/qr-code.png',
videoDirectory: './camera/video',
}],
],
};

Changing Camera Source​

Use the changeCameraSource command to switch video feeds during test execution:

describe('Camera Tests', () => {
it('should change camera source dynamically', async () => {
await browser.url('https://example.com/camera-app');

// Start with default feed, then switch to a different video
await browser.changeCameraSource('path/to/barcode-video.mjpeg');

// Test barcode scanning functionality
await expect($('#barcode-result')).toHaveText('123456789');

// Switch to a PNG image (automatically converted to video)
await browser.changeCameraSource('path/to/qr-code.png');

// Test QR code scanning
await expect($('#qr-result')).toHaveText('QR Content');
});
});

File Structure​

project/
├── camera/
│ ├── default.mjpeg # Default camera feed (MJPEG)
│ ├── qr-code.png # QR code image (auto-converted)
│ ├── barcode-sample.mp4 # Barcode video (auto-converted)
│ └── video/ # Auto-generated session videos
│ ├── 0-0.mjpeg # Session-specific copies
│ └── .cache/ # Cached converted files
│ └── abc123.mjpeg # Hash-based cache
├── test/
│ └── specs/
│ └── camera.e2e.ts # Test files
└── wdio.conf.ts # WebdriverIO configuration

Video File Requirements​

  • Format: MJPEG (Motion JPEG) or Y4M
  • Extension: .mjpeg or .y4m
  • Location: Relative to your project root or absolute paths

Other Formats​

When using non-native formats, the service will automatically convert them using FFmpeg. Converted files are cached in videoDirectory/.cache/ to avoid repeated conversions.

Creating MJPEG Files Manually​

You can convert existing video files to MJPEG format using FFmpeg:

# Convert MP4 to MJPEG
ffmpeg -i input.mp4 -q:v 2 output.mjpeg

# Convert image to single-frame MJPEG
ffmpeg -i input.png -frames:v 1 -q:v 2 output.mjpeg

API Reference​

Browser Commands​

browser.changeCameraSource(videoFilePath: string)​

Changes the active camera source to a different video file.

Parameters:

  • videoFilePath (string): Path to the video/image file (relative to project root). Supports all formats listed above.

Returns: Promise\<void\>

Example:

// Use MJPEG file (native)
await browser.changeCameraSource('camera/new-feed.mjpeg');

// Use MP4 file (auto-converted)
await browser.changeCameraSource('camera/video.mp4');

// Use PNG image (auto-converted to looping video)
await browser.changeCameraSource('camera/qr-code.png');

Error Handling​

The service will throw errors in the following cases:

  • Missing configuration: When defaultCameraFeed or videoDirectory is not specified
  • File not found: When the specified video file doesn't exist
  • FFmpeg not found: When using non-native formats without FFmpeg installed
  • Conversion failed: When FFmpeg fails to convert a file
  • Unsupported format: When using an unrecognized file extension
  • Unsupported browser: When used with non-Chrome browsers (logs warning instead of error)

Error Types​

ErrorDescription
FfmpegNotFoundErrorFFmpeg required but not installed
ConversionErrorFFmpeg conversion failed
UnsupportedFormatErrorUnknown file extension

Example Test Cases​

import { browser } from '@wdio/globals';

describe('Camera Application Tests', () => {
it('should scan QR codes from image', async () => {
await browser.url('https://qr-scanner-app.com');

// Inject QR code image (auto-converted to video)
await browser.changeCameraSource('camera/qr-code.png');

await $('#start-camera').click();
await expect($('#qr-result')).toHaveText('Expected QR Content');
});

it('should detect faces from video', async () => {
await browser.url('https://face-detection-app.com');

// Inject face video (MP4 auto-converted)
await browser.changeCameraSource('camera/face.mp4');

await $('#start-detection').click();
await expect($('#face-count')).toHaveText('1 face detected');
});
});

Troubleshooting​

Common Issues​

  1. "FFmpeg is required but not found"

    • Install FFmpeg using your package manager
    • Or provide a custom path via ffmpegPath option
  2. "Default camera feed does not exist"

    • Ensure the defaultCameraFeed path is correct
    • Verify the file exists and has proper permissions
  3. "New source camera feed does not exist"

    • Check the path passed to changeCameraSource()
    • Verify the file exists
  4. "Unsupported format"

    • Check the file extension is in the supported list
    • Ensure FFmpeg is installed for non-native formats
  5. Camera not working in browser

    • Verify you're using Chrome/Chromium
    • Check browser console for permission errors
    • Ensure the test site allows camera access

Debug Tips​

  • Set logLevel: 'debug' in your WebdriverIO config to see detailed logs
  • Check the videoDirectory/.cache/ for converted files
  • Verify MJPEG files play correctly in media players
  • Run ffmpeg -version to check FFmpeg installation

Contributing​

  1. Fork the repository
  2. Create a feature branch
  3. Add tests for new functionality
  4. Submit a pull request

License​

MIT License—see LICENSE file for details

Welcome! How can I help?

WebdriverIO AI Copilot