Java基于人脸识别技术和热成像相结合的智能体温检测系统源码+项目使用说明.zip

package com.cqmckj.twjc; import com.cqmckj.twjc.util.Config; import org.opencv.core.*; import org.opencv.imgproc.Imgproc; import org.opencv.objdetect.CascadeClassifier; import org.opencv.videoio.VideoCapture; import javax.swing.*; import java.awt.*; import java.awt.event.WindowAdapter; import java.awt.event.WindowEvent; import java.awt.image.BufferedImage; /** * Created by yl * 2020/2/7. */ public class Twjc{ private boolean isLoop = true; private JFrame frame; private JLabel img,gImg; public Twjc(){ frame = new JFrame("体温检测"); frame.setLayout(null); frame.setSize(1280,530); Font font = new Font("宋体",0,14); JLabel labelRgb = new JLabel("彩色相机"); labelRgb.setFont(font); labelRgb.setBounds(10, 10, 200, 20); JLabel labelRcx = new JLabel("热成像相机"); labelRcx.setFont(font); labelRcx.setBounds(650, 10, 200, 20); img = new JLabel(); img.setBounds(0, 30, 640, 480); gImg = new JLabel(); gImg.setBounds(640, 30, 640, 480); frame.add(labelRgb); frame.add(labelRcx); frame.add(img); frame.add(gImg); frame.setVisible(true); int screen_width = Toolkit.getDefaultToolkit().getScreenSize().width; int screen_height = Toolkit.getDefaultToolkit().getScreenSize().height; frame.setLocation((screen_width - frame.getWidth()) / 2, (screen_height - frame.getHeight()) / 2); frame.addWindowListener(new WindowAdapter() { public void windowClosing(WindowEvent arg0) { frame.setVisible(false); isLoop = false; System.exit(0); } }); frame.setVisible(true); init(); } static { //加载opencv System.loadLibrary(Core.NATIVE_LIBRARY_NAME); } private void init(){ Scalar markColor = null; Scalar markColorDef = new Scalar(Config.markColorB,Config.markColorG,Config.markColorR); Scalar markColorErr = new Scalar(0,0,255); org.opencv.core.Point circlePoint = new org.opencv.core.Point(); org.opencv.core.Point putTextPoint = new org.opencv.core.Point(); org.opencv.core.Point rectanglePoint1 = new org.opencv.core.Point(); org.opencv.core.Point rectanglePoint2 = new org.opencv.core.Point(); VideoCapture videoCaptureRgb = new VideoCapture(); VideoCapture videoCaptureRcx = new VideoCapture(); CascadeClassifier faceDetector = new CascadeClassifier(Config.detectPath); if (!videoCaptureRgb.open(Config.cameraRgb)) { System.out.println("彩色相机打开失败"); System.exit(0); return; } if (!videoCaptureRcx.open(Config.cameraRcx)) { System.out.println("热成像相机打开失败"); videoCaptureRgb.release(); System.exit(0); return; } Mat rgbImage = new Mat(); Mat rcxImage = new Mat(); MatOfRect faceDetections = new MatOfRect(); while (isLoop) { //限制检测频率 try {Thread.sleep(Config.detectSleep);} catch (InterruptedException e) {} //获得彩色图像 if (!videoCaptureRgb.read(rgbImage)) { continue; } //获得热成像图像 if (!videoCaptureRcx.read(rcxImage)) { continue; } //基于彩色图像进行人脸识别 faceDetector.detectMultiScale(rgbImage, faceDetections); //遍历识别结果 for(Rect rect:faceDetections.toArray()){ //人脸大小过滤 if(rect.width>Config.detectFaceMax || rect.width<Config.detectFaceMin){ continue; } //截取人脸进行温度提取 Mat faceImg = new Mat(rcxImage, rect); BufferedImage buffImg = conver2Image(faceImg); int[] nubRe = findMaxTemp(buffImg); //获得温度 float temp = formatTemp(nubRe[0]); if(temp>Config.twLimit){ markColor=markColorErr; tempErr(faceImg,temp); }else{ markColor=markColorDef; } //绘图 circlePoint.x=rect.x+nubRe[1]; circlePoint.y=rect.y+nubRe[2]; Imgproc.circle(rgbImage,circlePoint,Config.markRadius,markColor,Config.markThickness); putTextPoint.x=rect.x+2; putTextPoint.y=rect.y+20; Imgproc.putText(rgbImage,temp+"",putTextPoint,0, .7, markColor,Config.markThickness,Imgproc.LINE_AA,false); rectanglePoint1.x=rect.x; rectanglePoint1.y=rect.y; rectanglePoint2.x=rect.x + rect.width; rectanglePoint2.y=rect.y + rect.height; Imgproc.rectangle(rgbImage, rectanglePoint1, rectanglePoint2,markColor,Config.markThickness); } //UI显示图像 img.setIcon(new ImageIcon(conver2Image(rgbImage))); gImg.setIcon(new ImageIcon(conver2Image(rcxImage))); } //释放资源 videoCaptureRgb.release(); videoCaptureRcx.release(); faceDetections.release(); rgbImage.release(); rcxImage.release(); } /** * 温度异常回调函数 * @param faceImg 人脸图像 * @param temp 温度 */ private static void tempErr(Mat faceImg,float temp){ //TODO 自行实现 } /** * 热成像转换为温度 * @param temp 色彩平均值 * @return 热成像上下限中的比值，一位小数 */ private static float formatTemp(int temp){ return ((int)((temp/255f)*(Config.rcxMax-Config.rcxMin)*10))/10f; } /** * 从图像中查找最高的温度范围 * @param image 图像 * @return 0：温度最大值 1：x位置 2：y位置 */ private static int[] findMaxTemp(BufferedImage image){ int[] re = {0,0,0}; int width = image.getWidth(); int height = image.getHeight(); int avg,max=-1,maxX=-1,maxY=-1; for(int i = 0;i<width-Config.rcxRange;i++){ for(int j = 0;j<height-Config.rcxRange;j++){ avg = avgPix(image,i,j,Config.rcxRange); if(avg>max){ maxX = i; maxY = j; max = avg; } } } re[0] = max; re[1] = maxX; re[2] = maxY; return re; } /** * 统计一定范围像素点的GRB平均值 * @param image 图像 * @param x 像素点x位置 * @param y 像素点y位置 * @param len 范围，以xy为起点的正方形边长 * @return 平均值 */ private static int avgPix(BufferedImage image,int x,int y,int len){ int sum = 0; for(int i=0;i<len;i++){ for(int j=0;j<len;j++){ sum += grayPix(image.getRGB(x+i,y+j)); } } return sum/(len*len); } /** * rgb三通道的平均值 * @param rgb 值 * @return 平均值 */ private static int grayPix(int rgb){ return ( ((rgb>>16)&0xff) + ((rgb>>8)&0xff) +(rgb&0xff) ) / 3; } /** * opencv图像对象转BufferedImage * @param mat opencv图像对象 * @return BufferedImage */ private static BufferedImage conver2Image(Mat mat) { int width = mat.cols(); int height = mat.rows(); int dims = mat.channels(); int[] pixels = new int[width*height]; byte[] rgbdata = new byte[width*height*dims]; mat.get(0, 0, rgbdata); BufferedImage image = new BufferedImage(width, height, BufferedImage.TYPE_INT_ARGB);