flutter_face_detection

flutter_face_detection is a Flutter package designed to streamline face detection and liveness face detection for developers. It combines the powerful functionality of the Flutter camera library and Google ML Kit, enabling seamless integration into camera-based applications. The package is optimized for detecting facial features, such as eye openness and smiling, with configurable thresholds, making it ideal for liveness verification scenarios.

Key Features:

• Face Detection: Accurately detects faces in real-time using Google ML Kit.
• Liveness Detection: Supports advanced liveness detection by analyzing facial movements such as eye openness and smiling.
• Threshold Customization: Developers can easily adjust thresholds for detecting eye openness and smiles to suit various use cases.
• Camera Integration: Combines with the flutter_feature_camera library to handle camera functions like image streaming and camera switching.

Dependencies:

• flutter_feature_camera: Facilitates camera functionality required for real-time face detection.

This library is ideal for applications that require accurate and real-time face and liveness detection, such as biometric verification, security, and user authentication.

Features

Liveness Face Detection

Detect Left Eye Close & Right Eye Open

Detects liveness by checking if the left eye is open and the right eye is closed.

Future<void> onImageStream(CameraImage cameraImage,
    int sensorOrientation,
    DeviceOrientation deviceOrientation,
    CameraLensDirection cameraLensDirection,) async {
  final inputImage = FlutterFaceDetection.inputImageFromCameraImage(
    cameraImage,
    sensorOrientation: sensorOrientation,
    deviceOrientation: deviceOrientation,
    cameraLensDirection: cameraLensDirection,
  );
  if (inputImage != null) {
    final result = await detectLivenessLeftEyeOpenAndRightEyeClose(inputImage);
    // process result
  }
}

Detect Right Eye Close & Left Eye Open

Detects liveness by checking if the right eye is open and the left eye is closed.

Future<void> onImageStream(CameraImage cameraImage,
    int sensorOrientation,
    DeviceOrientation deviceOrientation,
    CameraLensDirection cameraLensDirection,) async {
  final inputImage = FlutterFaceDetection.inputImageFromCameraImage(
    cameraImage,
    sensorOrientation: sensorOrientation,
    deviceOrientation: deviceOrientation,
    cameraLensDirection: cameraLensDirection,
  );
  if (inputImage != null) {
    final result = await detectLivenessRightEyeOpenAndLeftEyeClose(inputImage);
    // process result
  }
}

Detect Smiling

Detects liveness by checking if the person is smiling.

Future<void> onImageStream(CameraImage cameraImage,
    int sensorOrientation,
    DeviceOrientation deviceOrientation,
    CameraLensDirection cameraLensDirection,) async {
  final inputImage = FlutterFaceDetection.inputImageFromCameraImage(
    cameraImage,
    sensorOrientation: sensorOrientation,
    deviceOrientation: deviceOrientation,
    cameraLensDirection: cameraLensDirection,
  );
  if (inputImage != null) {
    final result = await detectSmiling(inputImage);
    // process result
  }
}

Getting started

Change Min SDK Version

in android/app/build.gradle, change minSdkVersion to 21

// ..
android {
    // ..
    defaultConfig {
        // ..
        minSdkVersion 21 // change the sdk version
        // ..
    }
}

Import Flutter Feature Camera

flutter_feature_camera: ^0.0.6-beta

Extend Base Mixin Class

Extend Stateful Widget with BaseMixinFlutterFaceDetectionCamera and BaseMixinFeatureCameraV2

Usage

Initialized Streaming Camera

@override
void initState() {
  super.initState();
  initializeStreamingCamera(
    cameraLensDirection: CameraLensDirection.front,
    onCameraInitialized: onCameraInitialized,
  );
}

void onCameraInitialized(_) {
  setState(() {});

  initializeLivenessFaceDetection();
}

Start Streaming Image

void streamingCameraTap() {
  startImageStream(
    onImageStream: onImageStream,
  );
}

Future<void> onImageStream(
    CameraImage cameraImage,
    int sensorOrientation,
    DeviceOrientation deviceOrientation,
    CameraLensDirection cameraLensDirection,
    ) async {
  final inputImage = FlutterFaceDetection.inputImageFromCameraImage(
    cameraImage,
    sensorOrientation: sensorOrientation,
    deviceOrientation: deviceOrientation,
    cameraLensDirection: cameraLensDirection,
  );
  if (inputImage != null) {
    // process detect liveness here
  }
}