FengXiao_monitor/src/main/java/com/example/demo/Util/Train.java

package com.example.demo.Util;

import org.opencv.core.Mat;
import org.opencv.core.MatOfInt;
import org.opencv.face.FaceRecognizer;
import org.opencv.face.LBPHFaceRecognizer;
import org.opencv.imgcodecs.Imgcodecs;
import org.opencv.imgproc.Imgproc;
import org.opencv.objdetect.CascadeClassifier;

import java.io.File;
import java.io.IOException;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;
import java.util.Map;


public class Train {
    /*public static void main(String[] args) throws IOException {
            trainPath();
    }*/


    public static void trainPath() throws IOException {
        train("D:\\demos\\src\\main\\resources\\static\\imagedb",
                "D:\\demos\\src\\main\\resources\\static\\model");
    }

    /**
     * 训练模型的方法，传入人脸图片所在的文件夹路径，和模型输出的路径
     * 训练结束后模型文件会在模型输出路径里边
     **/
    public static void train(String imageFolder, String saveFolder)
            throws IOException {

        System.loadLibrary("opencv_java410");
        FaceRecognizer faceRecognizer = LBPHFaceRecognizer.create();
        CascadeClassifier faceCascade = new CascadeClassifier();
// opencv的模型
        faceCascade.load("D:/opencv/build/etc/haarcascades/haarcascade_frontalface_alt.xml");
// 读取文件于数组中
        File[] files = new File(imageFolder).listFiles();
        Map<String, Integer> nameMapId = new HashMap<String, Integer>(10);
// 图片集合
        List<Mat> images = new ArrayList<Mat>(files.length);
// 名称集合
        List<String> names = new ArrayList<String>(files.length);
        List<Integer> ids = new ArrayList<Integer>(files.length);
        for (int index = 0; index < files.length; index++ ) {
// 解析文件名 获取名称
            File file = files[index];
            String name = file.getName().split("\\.")[1];
            Integer id = nameMapId.get(name);
            if (id == null) {
                id = names.size();
                names.add(name);
                nameMapId.put(name, id);
                faceRecognizer.setLabelInfo(id, name);
            }

            Mat mat = Imgcodecs.imread(file.getCanonicalPath());
            Mat gray = new Mat();
// 图片预处理
            Imgproc.cvtColor(mat, gray, Imgproc.COLOR_BGR2GRAY);
            images.add(gray);
            System.out.println("add total " + images.size());
            ids.add(id);
        }
        int[] idsInt = new int[ids.size()];
        for (int i = 0; i < idsInt.length; i++) {
            idsInt[i] = ids.get(i).intValue();
        }
// 显示标签
        MatOfInt labels = new MatOfInt(idsInt);
// 调用训练方法
        faceRecognizer.train(images, labels);
// 输出持久化模型文件 训练一次后就可以一直调用
        faceRecognizer.save(saveFolder + "/face_model.yml");
    }


}