在配置好 OpenCV Android SDK 之後(具體見前一篇文章Android Studio中使用OpenCV Android SDK
)，可以使用 OpenCV 封裝的介面很方便地進行各種影象處理操作。

這裡簡單介紹如何直接使用 OpenCV 訓練的人臉模型直接從攝像頭幀檢測人臉。

1. 新建Android Project

這裡可以直接使用預設的 Android Studio 專案模板， Activity 選擇 Empty Activity 。

2. 配置OpenCV Android SDK

參考前一篇文章: Android Studio中使用OpenCV Android SDK

3. 向 AndroidManifest.xml 中新增 Camera 相關的 Permission

在 AndroidManifest.xml 檔案 <application> 節點前新增如下程式碼：

<uses-permission android:name="android.permission.CAMERA"/>

<uses-feature android:name="android.hardware.camera" android:required="false"/>
<uses-feature android:name="android.hardware.camera.autofocus"
 
 android:required="false"/>
<uses-feature android:name="android.hardware.camera.front" android:required="false"/>
<uses-feature android:name="android.hardware.camera.front.autofocus" android:required="false"/>

4. 新增OpenCV訓練的人臉模型

將 OpenCV Android SDK 中 sdk/etc 目錄下的 lbpcascade_frontalface.xml

5. 修改 MainActivity.java 的程式碼

修改 MainActivity.java 的程式碼為:

public class MainActivity extends AppCompatActivity
        implements CameraBridgeViewBase.CvCameraViewListener {


    private CameraBridgeViewBase openCvCameraView;
    private CascadeClassifier cascadeClassifier;
    private Mat grayscaleImage;
    private int absoluteFaceSize;


    private BaseLoaderCallback mLoaderCallback = new BaseLoaderCallback(this) {
        @Override
        public void onManagerConnected(int status) {
            switch (status) {
                case LoaderCallbackInterface.SUCCESS:
                    initializeOpenCVDependencies();
                    break;
                default:
                    super.onManagerConnected(status);
                    break;
            }
        }
    };


    private void initializeOpenCVDependencies() {


        try {
            // Copy the resource into a temp file so OpenCV can load it
            InputStream is = getResources().openRawResource(R.raw.lbpcascade_frontalface);
            File cascadeDir = getDir("cascade", Context.MODE_PRIVATE);
            File mCascadeFile = new File(cascadeDir, "lbpcascade_frontalface.xml");
            FileOutputStream os = new FileOutputStream(mCascadeFile);


            byte[] buffer = new byte[4096];
            int bytesRead;
            while ((bytesRead = is.read(buffer)) != -1) {
                os.write(buffer, 0, bytesRead);
            }
            is.close();
            os.close();


            // Load the cascade classifier
            cascadeClassifier = new CascadeClassifier(mCascadeFile.getAbsolutePath());
        } catch (Exception e) {
            Log.e("OpenCVActivity", "Error loading cascade", e);
        }


        // And we are ready to go
        openCvCameraView.enableView();
    }


    @Override
    public void onCreate(Bundle savedInstanceState) {
        super.onCreate(savedInstanceState);


        getWindow().addFlags(WindowManager.LayoutParams.FLAG_KEEP_SCREEN_ON);


        openCvCameraView = new JavaCameraView(this, -1);
        setContentView(openCvCameraView);
        openCvCameraView.setCvCameraViewListener(this);
    }


    @Override
    public void onCameraViewStarted(int width, int height) {
        grayscaleImage = new Mat(height, width, CvType.CV_8UC4);


        // The faces will be a 20% of the height of the screen
        absoluteFaceSize = (int) (height * 0.2);
    }


    @Override
    public void onCameraViewStopped() {
    }


    @Override
    public Mat onCameraFrame(Mat aInputFrame) {
        // Create a grayscale image
        Imgproc.cvtColor(aInputFrame, grayscaleImage, Imgproc.COLOR_RGBA2RGB);


        MatOfRect faces = new MatOfRect();


        // Use the classifier to detect faces
        if (cascadeClassifier != null) {
            cascadeClassifier.detectMultiScale(grayscaleImage, faces, 1.1, 2, 2,
                    new Size(absoluteFaceSize, absoluteFaceSize), new Size());
        }


        // If there are any faces found, draw a rectangle around it
        Rect[] facesArray = faces.toArray();
        for (int i = 0; i <facesArray.length; i++)
            Core.rectangle(aInputFrame, facesArray[i].tl(), facesArray[i].br(), new Scalar(0, 255, 0, 255), 3);


        return aInputFrame;
    }


    @Override
    public void onResume() {
        super.onResume();

        if (!OpenCVLoader.initDebug()) {
            Log.e("log_wons", "OpenCV init error");
            // Handle initialization error
        }
        initializeOpenCVDependencies();
        //OpenCVLoader.initAsync(OpenCVLoader.OPENCV_VERSION_2_4_6, this, mLoaderCallback);
    }
}

6. 編譯程式，執行

編譯之後執行程式，程式會獲取手機上的預設相機(一般為後置攝像頭)並進行人臉檢測。

效果：

參考資料

知乎日報: Android相機開發那些坑
Android Developers: Camera
OpenCV Documentation: OpenCV4Android SDK