Java借助OpenCV实现人脸识别登录完整示例
OpenCV
效果预览
概述
OpenCV(开源计算机视觉库)是在BSD(开源协议)许可下发布的。它是一个高度优化的库,专注于实时应用程序。它具有C ++,Python和Java接口,支持Windows,Linux,Mac OS,iOS和Android。
下载与安装
下载地址:
https://opencv.org/releases/
下载到本地后,双击进行安装即可
目录说明
安装目录如下
build :基于window构建
sources:开源,提供源码
build目录说明
这里是Java开发关注java目录即可
x64与x86代表给不同的系统使用
opencv-460.jar给java操作openvc的程序包
由于是64位系统,所以关注x64目录
DLL(Dynamic Link Library)文件为动态链接库文件,又称“应用程序拓展”,是软件文件类型。DLL文件,放置于系统中。当执行某一个程序时,相应的DLL文件就会被调用
OpenCV的基本使用
官网文档地址:
https://docs.opencv.org/4.6.0/df/d65/tutorial_table_of_content_introduction.html
中文文档:
http://wiki.opencv.org.cn/index.php
教程参考:
https://www.w3cschool.cn/opencv/
教程参考:
https://www.yiibai.com/opencv/opencv_adding_text.html
项目集成
这里使用IDEA进行开发,导入opencv-460.jar库
使用快捷键 Ctrl+Shift+Alt+S打开
选择库项,导入Java库。
除了上述方式,还可以将
opencv-460.jar
安装到本地仓库或私有仓库,然后在pom.xml中引入依赖。
图片人脸检测
public static void main(String[] args){imageFaceDetection();}/**
* 图片人脸检测
*/
public static void imageFaceDetection(){// 加载OpenCV本地库
System.loadLibrary(Core.NATIVE_LIBRARY_NAME);// 从配置文件lbpcascade_frontalface.xml中创建一个人脸识别器,文件位于opencv安装目录中
CascadeClassifier faceDetector =newCascadeClassifier("D:\\Development\\opencv\\sources\\data\\haarcascades\\haarcascade_frontalface_alt.xml");// 读取测试图片
String imgPath ="D:\\user\\test.png";
Mat image = Imgcodecs.imread(imgPath);if(image.empty()){thrownewRuntimeException("图片内存为空");}// 检测脸部
MatOfRect face =newMatOfRect();// 检测图像中的人脸
faceDetector.detectMultiScale(image, face);// 匹配Rect矩阵
Rect[] rects = face.toArray();
System.out.println("识别人脸个数: "+ rects.length);// 识别图片中的所以人脸并分别保存
int i =1;for(Rect rect : face.toArray()){
Imgproc.rectangle(image,newPoint(rect.x, rect.y),newPoint(rect.x + rect.width, rect.y + rect.height),newScalar(0,255,0),3);// 进行图片裁剪imageCut(imgPath,"D:\\user\\"+ i +".jpg", rect.x, rect.y, rect.width, rect.height);
i++;}// 图片中人脸画框保存到本地
Imgcodecs.imwrite("D:\\user\\test1.png", image);// 展示图片
HighGui.imshow("人脸识别", image);
HighGui.waitKey(0);}/**
* 裁剪人脸
*
* @param readPath 读取文件路径
* @param outPath 写出文件路径
* @param x 坐标X
* @param y 坐标Y
* @param width 截图宽度
* @param height 截图长度
*/
public static void imageCut(String readPath, String outPath, int x, int y, int width, int height){// 原始图像
Mat image = Imgcodecs.imread(readPath);// 截取的区域
Rect rect =newRect(x, y, width, height);// Mat sub = new Mat(image,rect);
Mat sub = image.submat(rect);
Mat mat =newMat();
Size size =newSize(width, height);// 人脸进行截图并保存
Imgproc.resize(sub, mat, size);
Imgcodecs.imwrite(outPath, mat);}
人脸对比相似度
对比1.jpg与1-1.jpg
对比1.jpg与3.jpg
// 初始化人脸探测器staticCascadeClassifier faceDetector;static{// 加载OpenCV本地库System.loadLibrary(Core.NATIVE_LIBRARY_NAME);// 从配置文件lbpcascade_frontalface.xml中创建一个人脸识别器,文件位于opencv安装目录中
faceDetector =newCascadeClassifier("D:\\Development\\opencv\\sources\\data\\haarcascades\\haarcascade_frontalface_alt.xml");}publicstaticvoidmain(String[] args){double comparison =faceRecognitionComparison("D:\\user\\1.jpg","D:\\user\\1-1.jpg");System.out.println("对比结果:"+ comparison);if(comparison >0.85){System.out.println("人脸匹配成功");}else{System.out.println("人脸不匹配识别");}double comparison2 =faceRecognitionComparison("D:\\user\\1.jpg","D:\\user\\3.jpg");System.out.println("对比结果:"+ comparison2);if(comparison2 >0.85){System.out.println("人脸匹配成功");}else{System.out.println("人脸不匹配识别");}// 终止当前运行的 Java 虚拟机。System.exit(0);}/**
* 人脸识别比对
*/publicstaticdoublefaceRecognitionComparison(String image1,String image2){Mat mat1 =conv_Mat(image1);Mat mat2 =conv_Mat(image2);Mat mat3 =newMat();Mat mat4 =newMat();// 颜色范围MatOfFloat ranges =newMatOfFloat(0f,256f);// 直方图大小, 越大匹配越精确 (越慢)MatOfInt histSize =newMatOfInt(1000);Imgproc.calcHist(Arrays.asList(mat1),newMatOfInt(0),newMat(), mat3, histSize, ranges);Imgproc.calcHist(Arrays.asList(mat2),newMatOfInt(0),newMat(), mat4, histSize, ranges);// 比较两个密集或两个稀疏直方图returnImgproc.compareHist(mat3, mat4,Imgproc.CV_COMP_CORREL);}/**
* 灰度化人脸
*/publicstaticMatconv_Mat(String img){// 读取图像Mat mat1 =Imgcodecs.imread(img);Mat mat2 =newMat();// 灰度化:将图像从一种颜色空间转换为另一种颜色空间Imgproc.cvtColor(mat1, mat2,Imgproc.COLOR_BGR2GRAY);// 探测人脸:检测到的对象作为矩形列表返回MatOfRect faceDetections =newMatOfRect();
faceDetector.detectMultiScale(mat1, faceDetections);// rect中人脸图片的范围for(Rect rect : faceDetections.toArray()){Mat face =newMat(mat1, rect);return face;}returnnull;}
对比结果如下
对比结果:1.0
人脸匹配成功
对比结果:0.2501351968792374
人脸不匹配识别
识别视频中的人脸
// 初始化人脸探测器staticCascadeClassifier faceDetector;static{// 加载OpenCV本地库System.loadLibrary(Core.NATIVE_LIBRARY_NAME);// 从配置文件lbpcascade_frontalface.xml中创建一个人脸识别器,文件位于opencv安装目录中
faceDetector =newCascadeClassifier("D:\\Development\\opencv\\sources\\data\\haarcascades\\haarcascade_frontalface_alt.xml");}publicstaticvoidmain(String[] args){videoFaceRecognition();// 终止当前运行的 Java 虚拟机。System.exit(0);}/**
* 从视频中识别人脸
*/publicstaticvoidvideoFaceRecognition(){// 读取视频文件VideoCapture capture =newVideoCapture();
capture.open("D:\\user\\test.mp4");if(!capture.isOpened()){thrownewRuntimeException("读取视频文件失败");}Mat video =newMat();int index =0;while(capture.isOpened()){// 抓取、解码并返回下一个视频帧写入Mat对象中
capture.read(video);// 显示从视频中识别的人脸图像HighGui.imshow("视频识别人脸",getFace(video));// 获取键盘输入
index =HighGui.waitKey(100);// 如果是 Esc 则退出if(index ==27){
capture.release();return;}}}/**
* 从视频帧中识别人脸
*
* @param image 待处理Mat图片,即视频中的某一帧
* @return 处理后的图片
*/publicstaticMatgetFace(Mat image){MatOfRect face =newMatOfRect();// 检测输入图像中不同大小的对象。检测到的对象作为矩形列表返回。
faceDetector.detectMultiScale(image, face);Rect[] rects = face.toArray();System.out.println("识别人脸个数: "+ rects.length);if(rects.length >0&&Math.random()*10>8){Imgcodecs.imwrite("D:\\user\\"+ UUID.randomUUID()+".png", image);}if(rects !=null&& rects.length >=1){// 为每张识别到的人脸画一个圈for(int i =0; i < rects.length; i++){/**
* 绘制一个简单的、粗的或填充的直角矩形
*
* img 图像
* pt1 - 矩形的顶点
* pt2 - 与 pt1 相对的矩形的顶点
* color – 矩形颜色或亮度(灰度图像)意味着该函数必须绘制一个填充的矩形。
*/Imgproc.rectangle(image,newPoint(rects[i].x, rects[i].y),newPoint(rects[i].x + rects[i].width, rects[i].y + rects[i].height),newScalar(0,255,0));/**
* 绘制一个文本字符串,放在识别人脸框上
*
* img -- 图像
* text -- 要绘制的文本字符串
* org – 图像中文本字符串的左下角
* fontFace – 字体类型,请参阅#HersheyFonts
* fontScale – 字体比例因子乘以特定字体的基本大小
* color - 文本颜色
* thickness ——用于绘制文本的线条粗细
* lineType – 线型
* bottomLeftOrigin – 当为 true 时,图像数据原点位于左下角。否则,它位于左上角
*/Imgproc.putText(image,"test",newPoint(rects[i].x, rects[i].y),Imgproc.FONT_HERSHEY_SCRIPT_SIMPLEX,1.0,newScalar(0,255,0),1,Imgproc.LINE_AA,false);}}return image;}
摄像头识别人脸
// 初始化人脸探测器staticCascadeClassifier faceDetector;static{// 加载OpenCV本地库System.loadLibrary(Core.NATIVE_LIBRARY_NAME);// 从配置文件lbpcascade_frontalface.xml中创建一个人脸识别器,文件位于opencv安装目录中
faceDetector =newCascadeClassifier("D:\\Development\\opencv\\sources\\data\\haarcascades\\haarcascade_frontalface_alt.xml");}publicstaticvoidmain(String[] args)throwsException{cameraFaceRecognition();// 终止当前运行的 Java 虚拟机。System.exit(0);}/**
* 摄像头实时人脸识别
*
* @throws Exception
*/publicstaticvoidcameraFaceRecognition()throwsException{// 打开摄像头获取视频流,0 打开默认摄像头VideoCapture videoCapture =newVideoCapture(0);// 检查是否支持摄像头 true:代表摄像头可以打开 false:不可以打开System.out.println(videoCapture.isOpened());// 获取摄像头高度int height =(int) videoCapture.get(Videoio.CAP_PROP_FRAME_HEIGHT);// 获取摄像头宽度int width =(int) videoCapture.get(Videoio.CAP_PROP_FRAME_WIDTH);if(height ==0|| width ==0){thrownewException("摄像头不存在");}Mat video =newMat();int index =0;if(videoCapture.isOpened()){while(true){
videoCapture.read(video);HighGui.imshow("实时人脸识别",getFace(video));// 键盘输入
index =HighGui.waitKey(50);// 是Esc则退出,比强制退出好if(index ==27){// 写入人脸Imgcodecs.imwrite("D:\\user\\"+"face.png", video);
videoCapture.release();return;}}}}/**
* 从视频帧中识别人脸
*
* @param image 待处理Mat图片,即视频中的某一帧
* @return 处理后的图片
*/publicstaticMatgetFace(Mat image){MatOfRect face =newMatOfRect();// 检测输入图像中不同大小的对象。检测到的对象作为矩形列表返回。
faceDetector.detectMultiScale(image, face);Rect[] rects = face.toArray();System.out.println("识别人脸个数: "+ rects.length);if(rects !=null&& rects.length >=1){// 为每张识别到的人脸画一个圈for(int i =0; i < rects.length; i++){// 绘制一个简单的、粗的或填充的直角矩形Imgproc.rectangle(image,newPoint(rects[i].x, rects[i].y),newPoint(rects[i].x + rects[i].width, rects[i].y + rects[i].height),newScalar(0,255,0));// 绘制一个文本字符串,放在识别人脸框上Imgproc.putText(image,"test",newPoint(rects[i].x, rects[i].y),Imgproc.FONT_HERSHEY_SCRIPT_SIMPLEX,1.0,newScalar(0,255,0),1,Imgproc.LINE_AA,false);}}return image;}
自定义窗口
OpenCV带的HighGUI图形用户界面感觉可配置参数太少,因此可自定义窗口用于代替。
importorg.opencv.core.Point;importorg.opencv.core.*;importorg.opencv.imgcodecs.Imgcodecs;importorg.opencv.imgproc.Imgproc;importorg.opencv.objdetect.CascadeClassifier;importorg.opencv.videoio.VideoCapture;importorg.opencv.videoio.Videoio;importjavax.swing.*;importjava.awt.*;importjava.awt.image.BufferedImage;publicclassMyJPanelextendsJPanel{privateBufferedImage mImg;// 初始化人脸探测器staticCascadeClassifier faceDetector;staticVideoCapture videoCapture;staticJFrame frame;static{// 加载OpenCV本地库System.loadLibrary(Core.NATIVE_LIBRARY_NAME);// 从配置文件lbpcascade_frontalface.xml中创建一个人脸识别器,文件位于opencv安装目录中
faceDetector =newCascadeClassifier("D:\\Development\\opencv\\sources\\data\\haarcascades\\haarcascade_frontalface_alt.xml");}publicvoidpaintComponent(Graphics g){if(mImg !=null){
g.drawImage(mImg,0,0, mImg.getWidth(), mImg.getHeight(),this);}}/**
* 摄像头识别人脸
*/publicstaticvoidcameraFaceRecognition()throwsException{try{// 打开摄像头获取视频流,0 打开默认摄像头
videoCapture =newVideoCapture(0);// 检查是否支持摄像头 true:代表摄像头可以打开 false:不可以打开System.out.println(videoCapture.isOpened());// 获取摄像头高度int height =(int) videoCapture.get(Videoio.CAP_PROP_FRAME_HEIGHT);// 获取摄像头宽度int width =(int) videoCapture.get(Videoio.CAP_PROP_FRAME_WIDTH);if(height ==0|| width ==0){thrownewException("摄像头不存在");}//使用Swing生成GUI
frame =newJFrame("人脸识别");
frame.setDefaultCloseOperation(WindowConstants.DISPOSE_ON_CLOSE);MyJPanel panel =newMyJPanel();//设置中心显示
frame.setContentPane(panel);
frame.setVisible(true);
frame.setSize(width + frame.getInsets().left + frame.getInsets().right, height + frame.getInsets().top + frame.getInsets().bottom);
frame.setLocationRelativeTo(null);// 创建矩阵Mat capImg =newMat();// 创建一个临时矩阵Mat temp =newMat();while(frame.isShowing()){//从摄像头读取一帧数据,保存到capImg矩阵中。
videoCapture.read(capImg);//转换为彩色图Imgproc.cvtColor(capImg, temp,Imgproc.COLOR_RGBA2BGRA);// 人脸识别
capImg =getFace(capImg);// 本地图片保存Imgcodecs.imwrite("D:\\user\\1.jpg", capImg);//转为图像显示
panel.mImg = panel.matToImage(capImg);// 重绘此组件
panel.repaint();}}finally{// 关闭摄像头
videoCapture.release();
frame.dispose();}}/**
* 从视频帧中识别人脸
*
* @param image 待处理Mat图片,即视频中的某一帧
* @return 处理后的图片
*/publicstaticMatgetFace(Mat image){MatOfRect face =newMatOfRect();// 检测输入图像中不同大小的对象。检测到的对象作为矩形列表返回。
faceDetector.detectMultiScale(image, face);Rect[] rects = face.toArray();System.out.println("识别人脸个数: "+ rects.length);if(rects !=null&& rects.length >=1){// 为每张识别到的人脸画一个圈for(int i =0; i < rects.length; i++){// 绘制一个简单的、粗的或填充的直角矩形Imgproc.rectangle(image,neworg.opencv.core.Point(rects[i].x, rects[i].y),neworg.opencv.core.Point(rects[i].x + rects[i].width, rects[i].y + rects[i].height),newScalar(0,255,0));// 绘制一个文本字符串,放在识别人脸框上Imgproc.putText(image,"test",newPoint(rects[i].x, rects[i].y),Imgproc.FONT_HERSHEY_SCRIPT_SIMPLEX,1.0,newScalar(0,255,0),1,Imgproc.LINE_AA,false);}}return image;}/**
* 转换图像
*/privateBufferedImagematToImage(Mat mat){int dataSize = mat.cols()* mat.rows()*(int) mat.elemSize();byte[] data =newbyte[dataSize];
mat.get(0,0, data);int type = mat.channels()==1?BufferedImage.TYPE_BYTE_GRAY :BufferedImage.TYPE_3BYTE_BGR;if(type ==BufferedImage.TYPE_3BYTE_BGR){for(int i =0; i < dataSize; i +=3){byte blue = data[i +0];
data[i +0]= data[i +2];
data[i +2]= blue;}}BufferedImage image =newBufferedImage(mat.cols(), mat.rows(), type);
image.getRaster().setDataElements(0,0, mat.cols(), mat.rows(), data);return image;}}
摄像头拍摄视频写入本地
publicstaticvoidmain(String[] args)throwsException{MyJPanel.cameraFaceRecognition();// 终止当前运行的 Java 虚拟机。System.exit(0);}
// 初始化人脸探测器staticCascadeClassifier faceDetector;staticBufferedImage mImg;static{// 加载OpenCV本地库System.loadLibrary(Core.NATIVE_LIBRARY_NAME);// 从配置文件lbpcascade_frontalface.xml中创建一个人脸识别器,文件位于opencv安装目录中
faceDetector =newCascadeClassifier("D:\\Development\\opencv\\sources\\data\\haarcascades\\haarcascade_frontalface_alt.xml");}publicstaticvoidmain(String[] args)throwsException{writeVideo();// 终止当前运行的 Java 虚拟机。System.exit(0);}/**
* 摄像头拍摄视频写入本地
*/publicstaticvoidwriteVideo()throwsException{// 打开摄像头获取视频流,0 打开默认摄像头VideoCapture videoCapture =newVideoCapture(0);// 检查是否支持摄像头 true:代表摄像头可以打开 false:不可以打开System.out.println(videoCapture.isOpened());// 获取摄像头高度int height =(int) videoCapture.get(Videoio.CAP_PROP_FRAME_HEIGHT);// 获取摄像头宽度int width =(int) videoCapture.get(Videoio.CAP_PROP_FRAME_WIDTH);if(height ==0|| width ==0){thrownewException("摄像头不存在");}Mat video =newMat();int index =0;Size size =newSize(videoCapture.get(Videoio.CAP_PROP_FRAME_WIDTH), videoCapture.get(Videoio.CAP_PROP_FRAME_HEIGHT));VideoWriter writer =newVideoWriter("D:\\user\\1.mp4",VideoWriter.fourcc('D','I','V','X'),30.0, size,true);while(videoCapture.isOpened()){//从摄像头读取一帧数据,保存到capImg矩阵中。
videoCapture.read(video);
writer.write(video);HighGui.imshow("视频人脸识别", video);// 获取键盘输入
index =HighGui.waitKey(100);// 是Esc则退出,若强制退出将导致录制视频无法播放if(index ==27){
videoCapture.release();
writer.release();return;}}}
Spring Boot集成OpenCV
添加依赖
<dependencies><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-web</artifactId></dependency><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-devtools</artifactId><scope>runtime</scope><optional>true</optional></dependency><dependency><groupId>org.projectlombok</groupId><artifactId>lombok</artifactId><optional>true</optional></dependency><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-test</artifactId><scope>test</scope></dependency><dependency><groupId>com.alibaba</groupId><artifactId>fastjson</artifactId><version>1.2.76</version></dependency><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-thymeleaf</artifactId></dependency></dependencies>
项目集成OpenCV
项目集成OpenCV参考上述
OpenCV的基本使用
中的
项目集成
请求接口
@Controller@RequestMapping("/user")publicclassUserFaceLogin{@AutowiredprivateMyJPanel myJPanel;@RequestMapping("/login")publicStringlogin()throwsException{// 调用摄像头显示boolean result = myJPanel.cameraFaceRecognition();if(result){return"/success.html";}else{return"/error.html";}}}
配置application.yml
开发环境与生产环境需区分
opencv:
lib:
linuxxmlpath:/usr/local//opencv/haarcascades/haarcascade_frontalface_alt.xml
windowxmlpath:D:\Development\opencv\sources\data\haarcascades\haarcascade_frontalface_alt.xml
指定虚拟机参数
-Djava.library.path=D:\Development\opencv\build\java\x64
或
-Djava.library.path=D:\Development\opencv\build\java\x64;D:\Development\opencv\build\x64\vc15\bin
OpenCvUtil
完成初始化工作以及添加人脸匹配功能,更多功能扩展此工具类即可。
@ComponentpublicclassOpenCvUtilimplementsCommandLineRunner{// 初始化人脸探测器staticCascadeClassifier faceDetector;@Value("${opencv.lib.linuxxmlpath}")privateString linuxXmlPath;@Value("${opencv.lib.windowxmlpath}")privateString windowXmlPath;/**
* 判断是否是Windows系统
*/privatestaticfinalboolean IS_WINDOWS =System.getProperty("os.name").toLowerCase().contains("win");@Overridepublicvoidrun(String... args){System.loadLibrary(Core.NATIVE_LIBRARY_NAME);String path ="";if(IS_WINDOWS){
path = windowXmlPath;}else{
path = linuxXmlPath;}/**
* 初始化人脸探测器
*/
faceDetector =newCascadeClassifier(path);}publicstaticintmatch(String loginImagePath,String comparedImagePath){Mat mat1 =conv_Mat(loginImagePath);if(mat1 ==null){return0;}Mat mat2 =conv_Mat(comparedImagePath);Mat mat3 =newMat();Mat mat4 =newMat();// 颜色范围MatOfFloat ranges =newMatOfFloat(0f,256f);// 直方图大小, 越大匹配越精确 (越慢)MatOfInt histSize =newMatOfInt(1000);Imgproc.calcHist(Arrays.asList(mat1),newMatOfInt(0),newMat(), mat3, histSize, ranges);Imgproc.calcHist(Arrays.asList(mat2),newMatOfInt(0),newMat(), mat4, histSize, ranges);// 比较两个密集或两个稀疏直方图Double score =Imgproc.compareHist(mat3, mat4,Imgproc.CV_COMP_CORREL);System.out.println("score "+ score);if(score >=0.8){return1;}return0;}publicstaticMatconv_Mat(String img){// 读取图像Mat mat1 =Imgcodecs.imread(img);Mat mat2 =newMat();// 灰度化:将图像从一种颜色空间转换为另一种颜色空间Imgproc.cvtColor(mat1, mat2,Imgproc.COLOR_BGR2GRAY);// 探测人脸:检测到的对象作为矩形列表返回MatOfRect faceDetections =newMatOfRect();
faceDetector.detectMultiScale(mat1, faceDetections);// rect中人脸图片的范围for(Rect rect : faceDetections.toArray()){Mat face =newMat(mat1, rect);return face;}returnnull;}}
自定义窗口
自定义窗口用于实时获取摄像头拍摄画面
@ComponentpublicclassMyJPanelextendsJPanel{@AutowiredprivateOpenCvUtil openCvUtil;privateBufferedImage mImg;privateVideoCapture videoCapture;privateJFrame frame;publicvoidpaintComponent(Graphics g){if(mImg !=null){
g.drawImage(mImg,0,0, mImg.getWidth(), mImg.getHeight(),this);}}/**
* 摄像头识别人脸
*/publicBooleancameraFaceRecognition()throwsException{try{// 打开摄像头获取视频流,0 打开默认摄像头
videoCapture =newVideoCapture(0);// 检查是否支持摄像头 true:代表摄像头可以打开 false:不可以打开System.out.println(videoCapture.isOpened());// 获取摄像头高度int height =(int) videoCapture.get(Videoio.CAP_PROP_FRAME_HEIGHT);// 获取摄像头宽度int width =(int) videoCapture.get(Videoio.CAP_PROP_FRAME_WIDTH);if(height ==0|| width ==0){thrownewException("摄像头不存在");}// 使用Swing生成GUI
frame =newJFrame("人脸识别");
frame.setDefaultCloseOperation(WindowConstants.DISPOSE_ON_CLOSE);MyJPanel panel =newMyJPanel();//设置中心显示
frame.setContentPane(panel);
frame.setVisible(true);
frame.setSize(width + frame.getInsets().left + frame.getInsets().right, height + frame.getInsets().top + frame.getInsets().bottom);
frame.setLocationRelativeTo(null);// 创建矩阵Mat capImg =newMat();// 创建一个临时矩阵Mat temp =newMat();// 对比图片String comparedImagePath ="D:\\user\\"+"compared.jpg";// 摄像头拍摄图片String loginImagePath ="D:\\user\\"+"login.jpg";int tag =0;while(frame.isShowing()&& tag <5){
tag++;//从摄像头读取一帧数据,保存到capImg矩阵中。
videoCapture.read(capImg);//转换为彩色图Imgproc.cvtColor(capImg, temp,Imgproc.COLOR_RGBA2BGRA);// 人脸识别
capImg =this.getFace(capImg);// 本地图片保存Imgcodecs.imwrite(loginImagePath, capImg);//转为图像显示
panel.mImg = panel.matToImage(capImg);// 重绘组件
panel.repaint();int result =OpenCvUtil.match(loginImagePath, comparedImagePath);if(result ==1){returntrue;}}}catch(Exception e){
e.printStackTrace();}finally{// 关闭窗口if(frame !=null){
frame.dispose();}// 关闭摄像头if(videoCapture !=null){
videoCapture.release();}}returnfalse;}/**
* 从视频帧中识别人脸
*
* @param image 待处理Mat图片,即视频中的某一帧
* @return 处理后的图片
*/publicMatgetFace(Mat image){MatOfRect face =newMatOfRect();// 检测输入图像中不同大小的对象。检测到的对象作为矩形列表返回。
openCvUtil.faceDetector.detectMultiScale(image, face);Rect[] rects = face.toArray();System.out.println("识别人脸个数: "+ rects.length);if(rects !=null&& rects.length >=1){// 为每张识别到的人脸画一个圈for(int i =0; i < rects.length; i++){// 绘制一个简单的、粗的或填充的直角矩形Imgproc.rectangle(image,neworg.opencv.core.Point(rects[i].x, rects[i].y),neworg.opencv.core.Point(rects[i].x + rects[i].width, rects[i].y + rects[i].height),newScalar(0,255,0));// 绘制一个文本字符串,放在识别人脸框上Imgproc.putText(image,"test",newPoint(rects[i].x, rects[i].y),Imgproc.FONT_HERSHEY_SCRIPT_SIMPLEX,1.0,newScalar(0,255,0),1,Imgproc.LINE_AA,false);}}return image;}/**
* 转换图像
*/privateBufferedImagematToImage(Mat mat){int dataSize = mat.cols()* mat.rows()*(int) mat.elemSize();byte[] data =newbyte[dataSize];
mat.get(0,0, data);int type = mat.channels()==1?BufferedImage.TYPE_BYTE_GRAY :BufferedImage.TYPE_3BYTE_BGR;if(type ==BufferedImage.TYPE_3BYTE_BGR){for(int i =0; i < dataSize; i +=3){byte blue = data[i +0];
data[i +0]= data[i +2];
data[i +2]= blue;}}BufferedImage image =newBufferedImage(mat.cols(), mat.rows(), type);
image.getRaster().setDataElements(0,0, mat.cols(), mat.rows(), data);return image;}}
创建页面
创建模拟人脸登录的页面Index.html以及人脸登录成功跳转页面success.html和人脸登录失败跳转页面error.html
index.html
<!DOCTYPE html><html lang="en"><head><meta charset="UTF-8"><title>Title</title></head><body><div id="index"class="tab-pane"><a href="/user/login">人脸登录</a></div></body></html>
success.html
<!DOCTYPE html><html lang="en"><head><meta charset="UTF-8"><title>Title</title></head><body><div><h3>人脸识别登录成功</h3></div></body></html>
error.html
<!DOCTYPE html><html lang="en"><head><meta charset="UTF-8"><title>Title</title></head><body><div><h3>人脸识别登录失败</h3></div></body></html>
启动类配置
在开发过程中遇到一个异常,即使用自定义窗口时,需要修改启动类,设置
.setHeadless(false)
,或添加JVM参数
-Djava.awt.headless=false
来解决。
@SpringBootApplicationpublicclassFaceOpenCvApplication{publicstaticvoidmain(String[] args){SpringApplicationBuilder builder =newSpringApplicationBuilder(FaceOpenCvApplication.class);
builder.headless(false).run(args);}}
常见异常记录
异常1
Exception in thread "main" java.lang.UnsatisfiedLinkError: no opencv_java460 in java.library.path
at java.lang.ClassLoader.loadLibrary(ClassLoader.java:1860)
at java.lang.Runtime.loadLibrary0(Runtime.java:871)
at java.lang.System.loadLibrary(System.java:1122)
将
D:\Development\opencv\build\java\x64\opencv_java460.dll
文件拷贝至下面2个目录,任选其一即可。
异常2
java.lang.Exception: unknown exception
org.opencv.videoio.VideoCapture.VideoCapture_3(NativeMethod)org.opencv.videoio.VideoCapture.<init>(VideoCapture.java:62)com.boxuegu.servlet.UserFaceLogin.doGet(UserFaceLogin.java:25)javax.servlet.http.HttpServlet.service(HttpServlet.java:635)javax.servlet.http.HttpServlet.service(HttpServlet.java:742)org.apache.tomcat.websocket.server.WsFilter.doFilter(WsFilter.java:52)
配置类库路径
进入
D:\Development\opencv\build\x64\vc15\bin
,获取该路径
添加JVM运行参数配置
-Djava.library.path=D:\Development\opencv\build\java\x64
或者
-Djava.library.path=D:\Development\opencv\build\java\x64;D:\Development\opencv\build\x64\vc15\bin
异常3
没重启Tomcat,而是让Tomcat自动重启war包导致
java.lang.UnsatisfiedLinkError:NativeLibraryD:\Development\opencv\build\java\x64\opencv_java460.dll already loaded in another classloader
java.lang.ClassLoader.loadLibrary0(ClassLoader.java:1900)java.lang.ClassLoader.loadLibrary(ClassLoader.java:1850)java.lang.Runtime.loadLibrary0(Runtime.java:871)java.lang.System.loadLibrary(System.java:1122)com.boxuegu.servlet.UserFaceLogin.doGet(UserFaceLogin.java:24)javax.servlet.http.HttpServlet.service(HttpServlet.java:635)javax.servlet.http.HttpServlet.service(HttpServlet.java:742)org.apache.tomcat.websocket.server.WsFilter.doFilter(WsFilter.java:52)
异常4
Exception in thread "main"java.lang.UnsatisfiedLinkError:org.opencv.videoio.VideoCapture.VideoCapture_5(I)J
at org.opencv.videoio.VideoCapture.VideoCapture_5(NativeMethod)
at org.opencv.videoio.VideoCapture.<init>(VideoCapture.java:181)
别忘了加载OpenCV本地库
static{// 加载OpenCV本地库System.loadLibrary(Core.NATIVE_LIBRARY_NAME);}
异常5
java.lang.UnsatisfiedLinkError:org.opencv.objdetect.CascadeClassifier.CascadeClassifier_1(Ljava/lang/String;)J
at org.opencv.objdetect.CascadeClassifier.CascadeClassifier_1(NativeMethod)~[opencv-460.jar:4.6.0]
at org.opencv.objdetect.CascadeClassifier.<init>(CascadeClassifier.java:48)~[opencv-460.jar:4.6.0]
受
spring-boot-devtools
依赖影响,最初排除此依赖,clean项目后正常。后来又加上此依赖,结果又不影响,注意当修改配置后没反应等异常情况还是多clean项目。
异常6
java.awt.HeadlessException
at java.awt.GraphicsEnvironment.checkHeadless(GraphicsEnvironment.java:204)
at java.awt.Window.<init>(Window.java:536)
at java.awt.Frame.<init>(Frame.java:420)
at javax.swing.JFrame.<init>(JFrame.java:233)
修改启动类,设置
.setHeadless(false);
@SpringBootApplicationpublicclassFaceOpenCvApplication{publicstaticvoidmain(String[] args){SpringApplicationBuilder builder =newSpringApplicationBuilder(FaceOpenCvApplication.class);
builder.headless(false).run(args);}}
或者设置JVM虚拟机参数
-Djava.awt.headless=false
版权归原作者 CodeDevMaster 所有, 如有侵权,请联系我们删除。