目录
效果展示
代码解释
本次效果的实现是在以下三篇文章的知识基础上进行的：
●SeetaFace2库的引入请查看：Android NDK开发：SeetaFace2人脸识别算法简介
●人脸检测请查看：Android NDK开发：SeetaFace2实现人脸检测
●人脸特征点检测请查看：Android NDK开发：SeetaFace2实现人脸特征点检测
/**
* 人脸匹配
*/
public class FaceRecognizerActivity extends AppCompatActivity {
private Button mBt;
private ImageView mImg;
private FaceDetector2 faceDetector;
private PointDetector2 pointDetector;
private FaceRecognizer2 faceRecognizer;
private int registerIndex;
private boolean moduleFlag = false;
private Handler handler = new Handler(){
@Override
public void handleMessage(Message msg) {
super.handleMessage(msg);
mImg.setImageBitmap((Bitmap) msg.obj);
}
};
@Override
protected void onCreate(Bundle savedInstanceState) {
super.onCreate(savedInstanceState);
setContentView(R.layout.activity_facerecognizer);
//将模型拷贝到SD卡中
initView();
initFace();
mBt.setOnClickListener(new View.OnClickListener() {
@Override
public void onClick(View v) {
//如果人脸模型初始化成功了再操作
if(moduleFlag){
Toast.makeText(FaceRecognizerActivity.this, "匹配中请稍等", Toast.LENGTH_SHORT).show();
new Thread(new Runnable() {
@Override
public void run() {
//加载进行匹配的图像
Bitmap bitmap = BitmapFactory.decodeResource(getResources(), R.drawable.tang08);
//这里必须进行copy否则修改不了
Bitmap copy = bitmap.copy(Bitmap.Config.ARGB_8888, true);
//利用Bitmap创建Canvas，为了在图像上绘制人脸区域
Canvas canvas = new Canvas(copy);
Paint paint = new Paint(Paint.ANTI_ALIAS_FLAG);
paint.setColor(Color.RED);
paint.setStyle(Paint.Style.STROKE);
paint.setStrokeWidth(3);
SeetaImageData seetaImageData = ConvertUtil.ConvertToSeetaImageData(bitmap);
//人脸检测
SeetaRect[] detects = faceDetector.Detect(seetaImageData);
if(detects.length>0){
//将所有检测到的人脸与注册到数据库的人脸进行匹配
for(int i = 0 ; i < detects.length ; i++){
SeetaRect faceRect = detects[i];
SeetaPointF[] seetaPoints = pointDetector.Detect(seetaImageData, faceRect);//根据检测到的人脸进行特征点检测
float[] similarity = new float[1];//用来存储人脸相似度值
int targetIndex = faceRecognizer.Recognize(seetaImageData, seetaPoints, similarity);//匹配
Log.e("人脸匹配",targetIndex+"======="+registerIndex+"====="+similarity[0]);
//如果匹配值大于0.7说明是同一个人
if(similarity[0]>0.7){
//将匹配出来的人脸区域绘制出来
Rect rect = new Rect(faceRect.x,faceRect.y,faceRect.x+faceRect.width,faceRect.y+faceRect.height);
canvas.drawRect(rect,paint);
}
}
//通知主线程更新UI
Message obtain = Message.obtain();
obtain.obj = copy;
handler.sendMessage(obtain);
}
}
}).start();
}else {
Toast.makeText(FaceRecognizerActivity.this, "人脸模型尚未初始化成功请稍等", Toast.LENGTH_SHORT).show();
}
}
});
}
/**
* 初始化人脸检测器
*/
private void initFace() {
new Thread(new Runnable() {
@Override
public void run() {
//初始化检测器（参数是模型在SD卡的位置）
faceDetector = new FaceDetector2(Environment.getExternalStorageDirectory()+ File.separator+"seetaface"+File.separator+"SeetaFaceDetector2.0.ats");
pointDetector = new PointDetector2(Environment.getExternalStorageDirectory()+ File.separator+"seetaface"+File.separator+"SeetaPointDetector2.0.pts5.ats"); //特征点
faceRecognizer = new FaceRecognizer2(Environment.getExternalStorageDirectory()+ File.separator+"seetaface"+File.separator+"SeetaFaceRecognizer2.0.ats"); //人脸匹配
Bitmap registBitmap = BitmapFactory.decodeResource(getResources(), R.drawable.tang01);
//利用SeetaFace2提供的转换方法获取SeetaRect（人脸识别结果）
SeetaImageData RegistSeetaImageData = ConvertUtil.ConvertToSeetaImageData(registBitmap);
SeetaRect[] faceRects = faceDetector.Detect(RegistSeetaImageData);
if(faceRects.length>0){
//获取人脸区域（这里只有一个所以取0）
SeetaRect faceRect = faceRects[0];
SeetaPointF[] seetaPoints = pointDetector.Detect(RegistSeetaImageData, faceRect);//根据检测到的人脸进行特征点检测
registerIndex = faceRecognizer.Register(RegistSeetaImageData, seetaPoints);//将人脸注册到SeetaFace2数据库
}
runOnUiThread(new Runnable() {
@Override
public void run() {
Toast.makeText(FaceRecognizerActivity.this, "人脸模型初始化成功", Toast.LENGTH_SHORT).show();
}
});
//模型加载标记
moduleFlag = true;
}
}).start();
}
private void initView() {
mBt = findViewById(R.id.bt_face);
mImg = findViewById(R.id.img);
}
@Override
protected void onDestroy() {
super.onDestroy();
//释放
if(faceDetector!=null){
faceDetector.dispose();
}
if(faceRecognizer!=null){
faceRecognizer.Clear();
faceRecognizer.dispose();
}
if(pointDetector!=null){
pointDetector.dispose();
}
}
}
网友评论