Android 使用Face++进行人脸识别

识别精度很好，可获取的信息也很多，比如左右眼睛的位置、鼻子的位置、嘴巴的位置、人脸的大小、年龄、性别、微笑程度等等

首先，要在Face++官网上注册一个账号，获取key和secret



然后创建一个应用



进入管理中查看key和secret



然后下载SDK http://www.faceplusplus.com.cn/dev-tools-sdks/

接下来在Android Studio中创建一个项目

导入SDK jar包

MainActivity.java

import android.app.Activity;
import android.graphics.Bitmap;
import android.graphics.BitmapFactory;
import android.graphics.Canvas;
import android.graphics.Color;
import android.graphics.Paint;
import android.os.Bundle;
import android.util.Log;
import android.view.ViewGroup;
import android.widget.ImageView;

import com.facepp.error.FaceppParseException;
import com.facepp.http.HttpRequests;
import com.facepp.http.PostParameters;

import org.json.JSONArray;
import org.json.JSONException;
import org.json.JSONObject;

import java.io.ByteArrayOutputStream;

public class MainActivity extends Activity {

private ImageView mIv;
private Bitmap mFaceBitmap;

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

//创建一个ImageView，加载到Activity视图上
mIv = new ImageView(this);
setContentView(mIv, new ViewGroup.LayoutParams(ViewGroup.LayoutParams.WRAP_CONTENT, ViewGroup.LayoutParams.WRAP_CONTENT));
//要进行人脸识别的图片
mFaceBitmap = BitmapFactory.decodeResource(getResources(), R.mipmap.face);

mIv.setImageBitmap(mFaceBitmap);

//网络操作，放到另一个线程中执行
new Thread(new Runnable() {
@Override
public void run() {
//创建http请求，参数为key和secret
HttpRequests httpRequests = new HttpRequests(yourKey, yourSecret);
ByteArrayOutputStream out = new ByteArrayOutputStream();
//将位图数据写入到字节流中
mFaceBitmap.compress(Bitmap.CompressFormat.JPEG, 100, out);
//字节流转换为字节数组
byte[] bytes = out.toByteArray();
//发送到服务器的参数
PostParameters parameters = new PostParameters();
//设置图片的字节数组
parameters.setImg(bytes);
try {
//接收服务器返回的JSONObject数据
JSONObject jsonObject = httpRequests.detectionDetect(parameters);
Log.e("tag", jsonObject.toString());
//根据JSONObject的数据在人脸图上标出对应的点
final Bitmap bitmap = parseBitmap(mFaceBitmap, jsonObject);
//把处理后的位图显示出来
mIv.post(new Runnable() {
@Override
public void run() {
mIv.setImageBitmap(bitmap);
}
});
} catch (FaceppParseException e) {
e.printStackTrace();
}
}
}).start();
}

/**
* 在人脸图上画点
*/
private Bitmap parseBitmap(Bitmap mFaceBitmap, JSONObject jsonObject) {
if (mFaceBitmap == null || jsonObject == null) return null;
//创建一个空的大小和mFaceBitmap一样的位图
Bitmap bitmap = Bitmap.createBitmap(mFaceBitmap.getWidth(), mFaceBitmap.getHeight(), Bitmap.Config.ARGB_8888);
//获取bitmap的画布，在画布上进行绘制
Canvas canvas = new Canvas(bitmap);
//创建画笔，抗锯齿
Paint mPaint = new Paint(Paint.ANTI_ALIAS_FLAG);
//画笔为黄色
mPaint.setColor(Color.YELLOW);
//画笔的大小为10
mPaint.setStrokeWidth(10);
//画笔为圆形
mPaint.setStrokeCap(Paint.Cap.ROUND);
//JSONObject数据解析
try {
JSONArray faces = jsonObject.getJSONArray("face");
//人脸数
int faceCount = faces.length();

//首先在画布上绘制mFaceBitmap图片
canvas.drawBitmap(mFaceBitmap, 0, 0, mPaint);

//在人脸上画点
for (int i = 0; i < faceCount; i++){
//拿到每张人脸的信息
JSONObject face = faces.getJSONObject(i);
//拿到人脸的详细位置信息
JSONObject position = face.getJSONObject("position");

//中心点
double centerX = position.getJSONObject("center").getDouble("x");
double centerY = position.getJSONObject("center").getDouble("y");
//拿到的并不是实际的像素值，而是一个相对于图片大小的比例，所以要处理一下
centerX = centerX * bitmap.getWidth() / 100;
centerY = centerY * bitmap.getHeight() / 100;
canvas.drawPoint((float) centerX, (float) centerY, mPaint);

//左眼睛
double eyeLeftX = position.getJSONObject("eye_left").getDouble("x");
double eyeLeftY = position.getJSONObject("eye_left").getDouble("y");
eyeLeftX = eyeLeftX * bitmap.getWidth() / 100;
eyeLeftY = eyeLeftY * bitmap.getHeight() / 100;
canvas.drawPoint((float) eyeLeftX, (float) eyeLeftY, mPaint);

//右眼睛
double eyeRightX = position.getJSONObject("eye_right").getDouble("x");
double eyeRightY = position.getJSONObject("eye_right").getDouble("y");
eyeRightX = eyeRightX * bitmap.getWidth() / 100;
eyeRightY = eyeRightY * bitmap.getHeight() / 100;
canvas.drawPoint((float) eyeRightX, (float) eyeRightY, mPaint);

//鼻子
double noseX = position.getJSONObject("nose").getDouble("x");
double noseY = position.getJSONObject("nose").getDouble("y");
noseX = noseX * bitmap.getWidth() / 100;
noseY = noseY * bitmap.getHeight() / 100;
canvas.drawPoint((float) noseX, (float) noseY, mPaint);

//嘴巴左边
double mouthLeftX = position.getJSONObject("mouth_left").getDouble("x");
double mouthLeftY = position.getJSONObject("mouth_left").getDouble("y");
mouthLeftX = mouthLeftX * bitmap.getWidth() / 100;
mouthLeftY = mouthLeftY * bitmap.getHeight() / 100;
canvas.drawPoint((float) mouthLeftX, (float) mouthLeftY, mPaint);

//嘴巴右边
double mouthRightX = position.getJSONObject("mouth_right").getDouble("x");
double mouthRightY = position.getJSONObject("mouth_right").getDouble("y");
mouthRightX = mouthRightX * bitmap.getWidth() / 100;
mouthRightY = mouthRightY * bitmap.getHeight() / 100;
canvas.drawPoint((float) mouthRightX, (float) mouthRightY, mPaint);

//人脸范围
double width = position.getDouble("width");
double height = position.getDouble("height");
width = width * bitmap.getWidth() / 100;
height = height * bitmap.getHeight() / 100;
mPaint.setStyle(Paint.Style.STROKE);
canvas.drawRect((float) (centerX - width / 2), (float) (centerY - height / 2), (float) (centerX + width / 2), (float) (centerY + height / 2), mPaint);
}

return bitmap;
} catch (JSONException e) {
e.printStackTrace();
return null;
}
}
}

//创建http请求，参数为key和secret
HttpRequests httpRequests = new HttpRequests(yourKey, yourSecret);

在这里传入的就是上面申请的key和secret

这是我的测试图片face.jpg



运行结果:



这个是我写的封装类FacePlus.java