一个简单的人脸检测和替换脸部的功能

步骤 1：导入OpenCV库

在项目中导入OpenCV库，你可以在Maven或Gradle中添加相应的依赖。这里以Maven为例：

<dependency>
    <groupId>org.openpnp</groupId>
    <artifactId>opencv</artifactId>
    <version>4.5.2-1</version>
</dependency>

步骤 2：实现人脸检测和替换

import org.opencv.core.Core;
import org.opencv.core.CvType;
import org.opencv.core.Mat;
import org.opencv.core.MatOfRect;
import org.opencv.core.Rect;
import org.opencv.core.Scalar;
import org.opencv.core.Size;
import org.opencv.core.MatOfByte;
import org.opencv.core.MatOfInt4;
import org.opencv.imgcodecs.Imgcodecs;
import org.opencv.objdetect.CascadeClassifier;
import org.opencv.imgproc.Imgproc;
import org.opencv.core.CvType;
import org.opencv.core.CvType.CV_8UC3;

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

public class FaceSwap {

    public static void main(String[] args) {
        System.loadLibrary(Core.NATIVE_LIBRARY_NAME);

        String imagePath = "path/to/your/image.jpg";
        String faceImagePath = "path/to/your/face_image.jpg";

        Mat image = Imgcodecs.imread(imagePath);
        Mat faceImage = Imgcodecs.imread(faceImagePath);

        Mat result = swapFaces(image, faceImage);

        displayImage(matToBufferedImage(result), "Face Swap Result");
    }

    public static Mat swapFaces(Mat image, Mat faceImage) {
        CascadeClassifier faceDetector = new CascadeClassifier("haarcascades/haarcascade_frontalface_default.xml");

        MatOfRect faceDetections = new MatOfRect();
        faceDetector.detectMultiScale(image, faceDetections);

        for (Rect rect : faceDetections.toArray()) {
            Mat faceROI = new Mat(faceImage.size(), CvType.CV_8UC3);
            Imgproc.resize(faceImage, faceROI, new Size(rect.width, rect.height));

            int x = rect.x;
            int y = rect.y;
            int width = rect.width;
            int height = rect.height;

            for (int i = 0; i < height; i++) {
                for (int j = 0; j < width; j++) {
                    double[] pixel = faceROI.get(i, j);
                    image.put(y + i, x + j, pixel);
                }
            }
        }

        return image;
    }

    public static void displayImage(Image img, String title) {
        ImageIcon icon = new ImageIcon(img);
        JFrame frame = new JFrame();
        frame.setLayout(new FlowLayout());
        frame.setSize(img.getWidth(null) + 50, img.getHeight(null) + 50);
        JLabel lbl = new JLabel();
        lbl.setIcon(icon);
        frame.add(lbl);
        frame.setTitle(title);
        frame.setDefaultCloseOperation(JFrame.EXIT_ON_CLOSE);
        frame.setVisible(true);
    }

    public static BufferedImage matToBufferedImage(Mat matrix) {
        int cols = matrix.cols();
        int rows = matrix.rows();
        int elemSize = (int) matrix.elemSize();
        byte[] data = new byte[cols * rows * elemSize];
        int type;
        matrix.get(0, 0, data);
        switch (matrix.channels()) {
            case 1:
                type = BufferedImage.TYPE_BYTE_GRAY;
                break;
            case 3:
                type = BufferedImage.TYPE_3BYTE_BGR;
                byte b;
                for (int i = 0; i < data.length; i = i + 3) {
                    b = data[i];
                    data[i] = data[i + 2];
                    data[i + 2] = b;
                }
                break;
            default:
                return null;
        }
        BufferedImage image = new BufferedImage(cols, rows, type);
        image.getRaster().setDataElements(0, 0, cols, rows, data);
        return image;
    }
}

请替换imagePath和faceImagePath为你实际的图像路径。上述代码使用了OpenCV的Haar级联分类器进行人脸检测，并简单地替换掉检测到的人脸部分。