Mendeteksi Wajah dengan ML Kit di iOS

Anda dapat menggunakan ML Kit untuk mendeteksi wajah dalam gambar dan video.

Sebelum memulai

1. Jika belum menambahkan Firebase ke aplikasi, lakukan dengan mengikuti langkah-langkah di panduan memulai.
2. Sertakan library ML Kit di Podfile Anda:

   pod 'Firebase/MLVision', '6.25.0'
   # If you want to detect face contours (landmark detection and classification
   # don't require this additional model):
   pod 'Firebase/MLVisionFaceModel', '6.25.0'
   

   Setelah menginstal atau mengupdate Pod project, pastikan untuk membuka project Xcode menggunakan .xcworkspace-nya.
3. Di aplikasi Anda, impor Firebase:

   Swift

   import Firebase

   Objective-C

   @import Firebase;

Pedoman gambar input

Agar ML Kit dapat secara akurat mengenali wajah, gambar input harus berisi wajah yang diwakili oleh data piksel yang memadai. Secara umum, setiap wajah yang ingin dideteksi dalam suatu gambar harus berukuran minimal 100x100 piksel. Jika Anda ingin mendeteksi kontur wajah, ML Kit membutuhkan input resolusi yang lebih tinggi: setiap wajah harus berukuran minimal 200x200 piksel.

Jika ingin mendeteksi wajah dalam aplikasi real-time, Anda mungkin perlu mempertimbangkan dimensi gambar input secara keseluruhan. Gambar yang lebih kecil dapat diproses lebih cepat. Jadi, untuk mengurangi latensi, ambil gambar dengan resolusi yang lebih rendah (dengan tetap memperhatikan persyaratan akurasi di atas), dan pastikan bahwa wajah subjek menempati gambar seluas mungkin. Baca juga Tips untuk meningkatkan performa real-time.

Fokus gambar yang buruk dapat mengurangi akurasi. Jika tidak memperoleh hasil yang dapat diterima, coba minta pengguna untuk mengambil ulang gambar.

Orientasi wajah terhadap arah kamera juga dapat memengaruhi fitur wajah yang terdeteksi oleh ML Kit. Baca Konsep Deteksi Wajah.

1. Mengonfigurasi detektor wajah

Misalnya, buat objek VisionFaceDetectorOptions seperti salah satu contoh berikut:

// High-accuracy landmark detection and face classification
let options = VisionFaceDetectorOptions()
options.performanceMode = .accurate
options.landmarkMode = .all
options.classificationMode = .all

// Real-time contour detection of multiple faces
let options = VisionFaceDetectorOptions()
options.contourMode = .all

// High-accuracy landmark detection and face classification
FIRVisionFaceDetectorOptions *options = [[FIRVisionFaceDetectorOptions alloc] init];
options.performanceMode = FIRVisionFaceDetectorPerformanceModeAccurate;
options.landmarkMode = FIRVisionFaceDetectorLandmarkModeAll;
options.classificationMode = FIRVisionFaceDetectorClassificationModeAll;

// Real-time contour detection of multiple faces
FIRVisionFaceDetectorOptions *options = [[FIRVisionFaceDetectorOptions alloc] init];
options.contourMode = FIRVisionFaceDetectorContourModeAll;

2. Menjalankan detektor wajah

1. Dapatkan instance VisionFaceDetector:

   Swift

   lazy var vision = Vision.vision()
   
   let faceDetector = vision.faceDetector(options: options)
   

   Objective-C

   FIRVision *vision = [FIRVision vision];
   FIRVisionFaceDetector *faceDetector = [vision faceDetector];
   // Or, to change the default settings:
   // FIRVisionFaceDetector *faceDetector =
   //     [vision faceDetectorWithOptions:options];
   

2. Buat objek VisionImage menggunakan UIImage atau CMSampleBufferRef.

   Untuk menggunakan UIImage:

   1. Jika perlu, putar gambar sehingga properti imageOrientation-nya adalah .up.
   2. Buat objek VisionImage menggunakan UIImage yang sudah diputar dengan benar. Jangan tentukan metadata rotasi apa pun—yang harus digunakan adalah nilai default, yaitu .topLeft.

      Swift

      let image = VisionImage(image: uiImage)

      Objective-C

      FIRVisionImage *image = [[FIRVisionImage alloc] initWithImage:uiImage];

   Untuk menggunakan CMSampleBufferRef:

   1. Buat objek VisionImageMetadata yang menentukan orientasi data gambar yang terdapat dalam buffering CMSampleBufferRef.

      Untuk mendapatkan orientasi gambar:

      Swift

      func imageOrientation(
          deviceOrientation: UIDeviceOrientation,
          cameraPosition: AVCaptureDevice.Position
          ) -> VisionDetectorImageOrientation {
          switch deviceOrientation {
          case .portrait:
              return cameraPosition == .front ? .leftTop : .rightTop
          case .landscapeLeft:
              return cameraPosition == .front ? .bottomLeft : .topLeft
          case .portraitUpsideDown:
              return cameraPosition == .front ? .rightBottom : .leftBottom
          case .landscapeRight:
              return cameraPosition == .front ? .topRight : .bottomRight
          case .faceDown, .faceUp, .unknown:
              return .leftTop
          }
      }

      Objective-C

      - (FIRVisionDetectorImageOrientation)
          imageOrientationFromDeviceOrientation:(UIDeviceOrientation)deviceOrientation
                                 cameraPosition:(AVCaptureDevicePosition)cameraPosition {
        switch (deviceOrientation) {
          case UIDeviceOrientationPortrait:
            if (cameraPosition == AVCaptureDevicePositionFront) {
              return FIRVisionDetectorImageOrientationLeftTop;
            } else {
              return FIRVisionDetectorImageOrientationRightTop;
            }
          case UIDeviceOrientationLandscapeLeft:
            if (cameraPosition == AVCaptureDevicePositionFront) {
              return FIRVisionDetectorImageOrientationBottomLeft;
            } else {
              return FIRVisionDetectorImageOrientationTopLeft;
            }
          case UIDeviceOrientationPortraitUpsideDown:
            if (cameraPosition == AVCaptureDevicePositionFront) {
              return FIRVisionDetectorImageOrientationRightBottom;
            } else {
              return FIRVisionDetectorImageOrientationLeftBottom;
            }
          case UIDeviceOrientationLandscapeRight:
            if (cameraPosition == AVCaptureDevicePositionFront) {
              return FIRVisionDetectorImageOrientationTopRight;
            } else {
              return FIRVisionDetectorImageOrientationBottomRight;
            }
          default:
            return FIRVisionDetectorImageOrientationTopLeft;
        }
      }

      Kemudian, buat objek metadata:

      Swift

      let cameraPosition = AVCaptureDevice.Position.back  // Set to the capture device you used.
      let metadata = VisionImageMetadata()
      metadata.orientation = imageOrientation(
          deviceOrientation: UIDevice.current.orientation,
          cameraPosition: cameraPosition
      )

      Objective-C

      FIRVisionImageMetadata *metadata = [[FIRVisionImageMetadata alloc] init];
      AVCaptureDevicePosition cameraPosition =
          AVCaptureDevicePositionBack;  // Set to the capture device you used.
      metadata.orientation =
          [self imageOrientationFromDeviceOrientation:UIDevice.currentDevice.orientation
                                       cameraPosition:cameraPosition];

   2. Buat objek VisionImage menggunakan objek CMSampleBufferRef dan metadata rotasi:

      Swift

      let image = VisionImage(buffer: sampleBuffer)
      image.metadata = metadata

      Objective-C

      FIRVisionImage *image = [[FIRVisionImage alloc] initWithBuffer:sampleBuffer];
      image.metadata = metadata;

3. Lalu, teruskan gambar ke metode detect(in:):

   Swift

   faceDetector.process(visionImage) { faces, error in
     guard error == nil, let faces = faces, !faces.isEmpty else {
       // ...
       return
     }
   
     // Faces detected
     // ...
   }
   

   Objective-C

   [faceDetector detectInImage:image
                    completion:^(NSArray<FIRVisionFace *> *faces,
                                 NSError *error) {
     if (error != nil) {
       return;
     } else if (faces != nil) {
       // Recognized faces
     }
   }];
   

3. Mendapatkan informasi tentang wajah yang terdeteksi

for face in faces {
  let frame = face.frame
  if face.hasHeadEulerAngleY {
    let rotY = face.headEulerAngleY  // Head is rotated to the right rotY degrees
  }
  if face.hasHeadEulerAngleZ {
    let rotZ = face.headEulerAngleZ  // Head is rotated upward rotZ degrees
  }

  // If landmark detection was enabled (mouth, ears, eyes, cheeks, and
  // nose available):
  if let leftEye = face.landmark(ofType: .leftEye) {
    let leftEyePosition = leftEye.position
  }

  // If contour detection was enabled:
  if let leftEyeContour = face.contour(ofType: .leftEye) {
    let leftEyePoints = leftEyeContour.points
  }
  if let upperLipBottomContour = face.contour(ofType: .upperLipBottom) {
    let upperLipBottomPoints = upperLipBottomContour.points
  }

  // If classification was enabled:
  if face.hasSmilingProbability {
    let smileProb = face.smilingProbability
  }
  if face.hasRightEyeOpenProbability {
    let rightEyeOpenProb = face.rightEyeOpenProbability
  }

  // If face tracking was enabled:
  if face.hasTrackingID {
    let trackingId = face.trackingID
  }
}

for (FIRVisionFace *face in faces) {
  // Boundaries of face in image
  CGRect frame = face.frame;

  if (face.hasHeadEulerAngleY) {
    CGFloat rotY = face.headEulerAngleY;  // Head is rotated to the right rotY degrees
  }
  if (face.hasHeadEulerAngleZ) {
    CGFloat rotZ = face.headEulerAngleZ;  // Head is tilted sideways rotZ degrees
  }

  // If landmark detection was enabled (mouth, ears, eyes, cheeks, and
  // nose available):
  FIRVisionFaceLandmark *leftEar = [face landmarkOfType:FIRFaceLandmarkTypeLeftEar];
  if (leftEar != nil) {
    FIRVisionPoint *leftEarPosition = leftEar.position;
  }

  // If contour detection was enabled:
  FIRVisionFaceContour *upperLipBottomContour = [face contourOfType:FIRFaceContourTypeUpperLipBottom];
  if (upperLipBottomContour != nil) {
    NSArray<FIRVisionPoint *> *upperLipBottomPoints = upperLipBottomContour.points;
    if (upperLipBottomPoints.count > 0) {
      NSLog("Detected the bottom contour of the subject's upper lip.")
    }
  }

  // If classification was enabled:
  if (face.hasSmilingProbability) {
    CGFloat smileProb = face.smilingProbability;
  }
  if (face.hasRightEyeOpenProbability) {
    CGFloat rightEyeOpenProb = face.rightEyeOpenProbability;
  }

  // If face tracking was enabled:
  if (face.hasTrackingID) {
    NSInteger trackingID = face.trackingID;
  }
}

Contoh kontur wajah

Ketika mengaktifkan deteksi kontur wajah, Anda juga akan melihat sekumpulan titik untuk setiap fitur wajah yang terdeteksi Titik-titik ini mengikuti bentuk fitur wajah. Baca Ringkasan Konsep Deteksi Wajah untuk mengetahui informasi detail tentang cara kontur direpresentasikan.

Gambar berikut mengilustrasikan bagaimana titik-titik ini dipetakan ke wajah (klik gambar untuk memperbesar):