FaceMLKit

A powerful Android SDK for face detection and quality assessment using Google ML Kit. FaceMLKit provides easy-to-use APIs for detecting faces, extracting face regions, and assessing face quality with advanced image processing techniques.

Features

• Face Detection: Accurate face detection using Google ML Kit
• Face Extraction: Extract and crop the largest face from images
• Quality Assessment: Advanced face quality checking with blur detection
• Image Processing: Automatic brightness adjustment and image enhancement
• Fallback Processing: Robust handling when no faces are detected
• Coroutine Support: Both callback-based and suspend function APIs

Installation

Add the following dependency to your build.gradle file:

dependencies {
    implementation(files("libs/facemlkit-release.aar"))
    
    // Required ML Kit dependencies
    implementation 'com.google.mlkit:face-detection:16.1.5'
}

API Reference

FaceMLKit Class

getInstance(context: Context): FaceMLKit

Get singleton instance of FaceMLKit.

detectFaces(bitmap: Bitmap, callback: (FaceDetectionResult) -> Unit)

Detect all faces in the provided bitmap and return results via callback.

detectFaces(bitmap: Bitmap): FaceDetectionResult

Suspend version of face detection that returns all detected faces.

extractLargestFace(bitmap: Bitmap, callback: (Bitmap?) -> Unit)

Find the largest face in the image, crop it, resize to 224x224, and return via callback.

extractLargestFace(bitmap: Bitmap): Bitmap?

Suspend version that extracts and returns the largest face as a 224x224 bitmap.

extractLargestFaceWithQuality(bitmap: Bitmap, callback: (Bitmap?, FaceQualityResult) -> Unit)

Extract the largest face and perform quality assessment, returning both the face bitmap and quality results via callback.

extractLargestFaceWithQuality(bitmap: Bitmap): Pair<Bitmap?, FaceQualityResult>

Suspend version that extracts the largest face and returns both the bitmap and quality assessment.

close()

Release ML Kit detector resources and clean up.

Data Models

FaceDetectionResult

data class FaceDetectionResult(
    val success: Boolean,           // Whether detection found any faces
    val faces: List<Face>,          // List of ML Kit Face objects
    val message: String             // Status message about detection result
)

FaceQualityResult

data class FaceQualityResult(
    val isGoodQuality: Boolean,     // Whether face passes quality checks
    val qualityScore: Float,        // Quality score from 0.0 to 1.0
    val laplacianRawScore: Double,  // Raw sharpness measurement
    val issues: List<QualityIssue>, // List of detected problems
    val failureReason: String?      // Description of quality failure
)

QualityIssue

enum class QualityIssue {
    MULTIPLE_FACES,      // More than one face detected
    NO_FACE_DETECTED,    // No face found in image
    UNDEREXPOSED,        // Image is too dark
    BLURRY_FACE         // Face lacks sharpness
}

Usage Examples

Basic Face Detection

val faceMLKit = FaceMLKit.getInstance(context)

// Callback version
faceMLKit.detectFaces(bitmap) { result ->
    if (result.success) {
        Log.d("Faces", "Found ${result.faces.size} faces")
    }
}

// Coroutine version
val result = faceMLKit.detectFaces(bitmap)

Face Extraction

// Extract face only
faceMLKit.extractLargestFace(bitmap) { faceBitmap ->
    faceBitmap?.let { 
        // Use 224x224 face bitmap
    }
}

// Extract with quality check
faceMLKit.extractLargestFaceWithQuality(bitmap) { faceBitmap, quality ->
    if (quality.isGoodQuality) {
        // Use high quality face
    }
}

Requirements

• Android API level 21+
• Google ML Kit Face Detection
• Kotlin Coroutines support