Java使用OpenCV实现人脸识别/人眼检测/图片截取/合成/添加水印

Posted 2023-01-15 零度anngle

官方下载opencv安装文件: http://opencv.org/releases.html，以windows版本为例，下载opencv-3.1.0.exe

安装后，在build目录下 D:\\opencv\\opencv\\build\\java，获取opencv-310.jar，copy至项目opncv目录（需要新建）

同时需要dll文件 与 各识别xml文件，进行不同特征的识别（人脸，侧脸，眼睛等）

dll目录： D:\\opencv\\opencv\\build\\java\\x64\\opencv_java2413.dll （dll库）

xml目录：D:\\opencv\\opencv\\sources\\data\\haarcascades\\haarcascade_frontalface_alt.xml（目录中有各类识别文件）

下面给出图片的各种操作的源码和运行结果，源码中有各个操作的解释，看代码就可以理解，java代码如下

package com.zmx.opencvtest;


import org.opencv.core.*;
import org.opencv.core.Point;
import org.opencv.imgcodecs.Imgcodecs;
import org.opencv.imgproc.Imgproc;
import org.opencv.objdetect.CascadeClassifier;

import javax.imageio.ImageIO;
import javax.swing.*;
import java.awt.*;
import java.awt.image.BufferedImage;
import java.io.File;
import java.io.IOException;



/**
 * Created by Administrator on 2017/8/17.
 */
public class DetectFaceTest 

    static
        // 载入opencv的库
        String opencvpath = System.getProperty("user.dir") + "\\\\opencv\\\\x64\\\\";
        String opencvDllName = opencvpath + Core.NATIVE_LIBRARY_NAME + ".dll";
        System.load(opencvDllName);
    


    /**
     * opencv实现人脸识别
     * @param imagePath
     * @param outFile
     * @throws Exception
     */
    public static void detectFace(String imagePath,  String outFile) throws Exception
    

        System.out.println("Running DetectFace ... ");
        // 从配置文件lbpcascade_frontalface.xml中创建一个人脸识别器，该文件位于opencv安装目录中
        CascadeClassifier faceDetector = new CascadeClassifier(
                "D:\\\\opencv\\\\opencv\\\\sources\\\\data\\\\haarcascades\\\\haarcascade_frontalface_alt.xml");

        Mat image = Imgcodecs.imread(imagePath);

        // 在图片中检测人脸
        MatOfRect faceDetections = new MatOfRect();

        faceDetector.detectMultiScale(image, faceDetections);

        System.out.println(String.format("Detected %s faces", faceDetections.toArray().length));

        Rect[] rects = faceDetections.toArray();
        if(rects != null && rects.length > 1)
            throw new RuntimeException("超过一个脸");
        
        // 在每一个识别出来的人脸周围画出一个方框
        Rect rect = rects[0];

        Imgproc.rectangle(image, new Point(rect.x-2, rect.y-2),
                          new Point(rect.x + rect.width, rect.y + rect.height),
                          new Scalar(0, 255, 0));

        Imgcodecs.imwrite(outFile, image);
        System.out.println(String.format("人脸识别成功，人脸图片文件为： %s", outFile));


    


    /**
     * opencv实现人眼识别
     * @param imagePath
     * @param outFile
     * @throws Exception
     */
    public static void detectEye(String imagePath,  String outFile) throws Exception 


        CascadeClassifier eyeDetector = new CascadeClassifier(
                "D:\\\\opencv\\\\opencv\\\\sources\\\\data\\\\haarcascades\\\\haarcascade_eye.xml");

        Mat image = Imgcodecs.imread(imagePath);  //读取图片

        // 在图片中检测人脸
        MatOfRect faceDetections = new MatOfRect();

        eyeDetector.detectMultiScale(image, faceDetections, 2.0,1,1,new Size(20,20),new Size(20,20));

        System.out.println(String.format("Detected %s eyes", faceDetections.toArray().length));
        Rect[] rects = faceDetections.toArray();
        if(rects != null && rects.length <2)
            throw new RuntimeException("不是一双眼睛");
        
        Rect eyea = rects[0];
        Rect eyeb = rects[1];


        System.out.println("a-中心坐标 " + eyea.x + " and " + eyea.y);
        System.out.println("b-中心坐标 " + eyeb.x + " and " + eyeb.y);

        //获取两个人眼的角度
        double dy=(eyeb.y-eyea.y);
        double dx=(eyeb.x-eyea.x);
        double len=Math.sqrt(dx*dx+dy*dy);
        System.out.println("dx is "+dx);
        System.out.println("dy is "+dy);
        System.out.println("len is "+len);

        double angle=Math.atan2(Math.abs(dy),Math.abs(dx))*180.0/Math.PI;
        System.out.println("angle is "+angle);

        for(Rect rect:faceDetections.toArray()) 
            Imgproc.rectangle(image, new Point(rect.x, rect.y), new Point(rect.x
                    + rect.width, rect.y + rect.height), new Scalar(0, 255, 0));
        
        Imgcodecs.imwrite(outFile, image);

        System.out.println(String.format("人眼识别成功，人眼图片文件为： %s", outFile));

    


    /**
     * 裁剪图片并重新装换大小
     * @param imagePath
     * @param posX
     * @param posY
     * @param width
     * @param height
     * @param outFile
     */
    public static void imageCut(String imagePath,String outFile, int posX,int posY,int width,int height )

        //原始图像
        Mat image = Imgcodecs.imread(imagePath);

        //截取的区域：参数,坐标X,坐标Y,截图宽度,截图长度
        Rect rect = new Rect(posX,posY,width,height);

        //两句效果一样
        Mat sub = image.submat(rect);   //Mat sub = new Mat(image,rect);

        Mat mat = new Mat();
        Size size = new Size(300, 300);
        Imgproc.resize(sub, mat, size);//将人脸进行截图并保存

        Imgcodecs.imwrite(outFile, mat);
        System.out.println(String.format("图片裁切成功，裁切后图片文件为： %s", outFile));

    


    /**
     *
     * @param imagePath
     * @param outFile
     */
    public static void setAlpha(String imagePath,  String outFile) 
        /**
         * 增加测试项
         * 读取图片，绘制成半透明
         */
        try 

            ImageIcon imageIcon = new ImageIcon(imagePath);

            BufferedImage bufferedImage = new BufferedImage(imageIcon.getIconWidth(),
                              imageIcon.getIconHeight(), BufferedImage.TYPE_4BYTE_ABGR);

            Graphics2D g2D = (Graphics2D) bufferedImage.getGraphics();

            g2D.drawImage(imageIcon.getImage(), 0, 0, imageIcon.getImageObserver());

            //循环每一个像素点，改变像素点的Alpha值
            int alpha = 100;
            for (int j1 = bufferedImage.getMinY(); j1 < bufferedImage.getHeight(); j1++) 
                for (int j2 = bufferedImage.getMinX(); j2 < bufferedImage.getWidth(); j2++) 
                    int rgb = bufferedImage.getRGB(j2, j1);
                    rgb = ( (alpha + 1) << 24) | (rgb & 0x00ffffff);
                    bufferedImage.setRGB(j2, j1, rgb);
                
            
            g2D.drawImage(bufferedImage, 0, 0, imageIcon.getImageObserver());

            //生成图片为PNG
            ImageIO.write(bufferedImage, "png",  new File(outFile));

            System.out.println(String.format("绘制图片半透明成功，图片文件为： %s", outFile));

        
        catch (Exception e) 
            e.printStackTrace();
        

    

    /**
     * 为图像添加水印
     * @param buffImgFile 底图
     * @param waterImgFile 水印
     * @param outFile 输出图片
     * @param alpha   透明度
     * @throws IOException
     */
    private static void watermark(String buffImgFile,String waterImgFile,String outFile, float alpha) throws IOException 
        // 获取底图
        BufferedImage buffImg = ImageIO.read(new File(buffImgFile));

        // 获取层图
        BufferedImage waterImg = ImageIO.read(new File(waterImgFile));

        // 创建Graphics2D对象，用在底图对象上绘图
        Graphics2D g2d = buffImg.createGraphics();

        int waterImgWidth = waterImg.getWidth();// 获取水印层图的宽度

        int waterImgHeight = waterImg.getHeight();// 获取水印层图的高度

        // 在图形和图像中实现混合和透明效果
        g2d.setComposite(AlphaComposite.getInstance(AlphaComposite.SRC_ATOP, alpha));

        // 绘制
        g2d.drawImage(waterImg, 0, 0, waterImgWidth, waterImgHeight, null);

        g2d.dispose();// 释放图形上下文使用的系统资源

        //生成图片为PNG
        ImageIO.write(buffImg, "png",  new File(outFile));

        System.out.println(String.format("图片添加水印成功，图片文件为： %s", outFile));
    


    /**
     * 图片合成
     * @param image1
     * @param image2
     * @param posw
     * @param posh
     * @param outFile
     * @return
     */
    public static void simpleMerge(String image1, String image2, int posw, int posh, String outFile) throws IOException

        // 获取底图
        BufferedImage buffImg1 = ImageIO.read(new File(image1));

        // 获取层图
        BufferedImage buffImg2 = ImageIO.read(new File(image2));

        //合并两个图像
        int w1 = buffImg1.getWidth();
        int h1 = buffImg1.getHeight();

        int w2 = buffImg2.getWidth();
        int h2 = buffImg2.getHeight();

        BufferedImage imageSaved = new BufferedImage(w1, h1, BufferedImage.TYPE_INT_ARGB); //创建一个新的内存图像

        Graphics2D g2d = imageSaved.createGraphics();

        g2d.drawImage(buffImg1, null, 0, 0);  //绘制背景图像

        for (int i = 0; i < w2; i++) 
            for (int j = 0; j < h2; j++) 
                int rgb1 = buffImg1.getRGB(i + posw, j + posh);
                int rgb2 = buffImg2.getRGB(i, j);

                /*if (rgb1 != rgb2) 
                    rgb2 = rgb1 & rgb2;
                */
                imageSaved.setRGB(i + posw, j + posh, rgb2); //修改像素值
            
        

        ImageIO.write(imageSaved, "png", new File(outFile));

        System.out.println(String.format("图片合成成功，合成图片文件为： %s", outFile));

    

    public static void main(String[] args) throws Exception 

        //人脸识别
        detectFace("E:\\\\person.jpg", "E:\\\\personFaceDetect.png");

        //人眼识别
        detectEye("E:\\\\person.jpg",  "E:\\\\personEyeDetect.png");

        //图片裁切
        imageCut("E:\\\\person.jpg","E:\\\\personCut.png", 50, 50,100,100);

        //设置图片为半透明
        setAlpha("E:\\\\person.jpg", "E:\\\\personAlpha.png");


        //为图片添加水印
        watermark("E:\\\\person.jpg","E:\\\\ling.jpg","E:\\\\personWaterMark.png", 0.2f);


        //图片合成
        simpleMerge("E:\\\\person.jpg", "E:\\\\ling.jpg", 45, 50, "E:\\\\personMerge.png");

    

     上述运行的结果如下：

（1）原图： E:\\person.jpg     水印的图片：E:\\ling.jpg

      

Running DetectFace ... 
Detected 1 faces
（2）人脸识别成功，人脸图片文件为： E:\\personFaceDetect.png

（3）Detected 4 eyes
a-中心坐标 93 and 102
b-中心坐标 59 and 107
dx is -34.0
dy is 5.0
len is 34.36568055487916
angle is 8.36588612403259
人眼识别成功，人眼图片文件为： E:\\personEyeDetect.png

图片裁切成功，裁切后图片文件为： E:\\personCut.png

绘制图片半透明成功，图片文件为： E:\\personAlpha.png

图片添加水印成功，图片文件为： E:\\personWaterMark.png

图片合成成功，合成图片文件为： E:\\personMerge.png

总结一下，opencv检测眼睛的还是有一些瑕疵，但是人脸检测效果还可以，我的测试就是如此，大家可以参考实现自己的代码。