Goal
Test OpenCV Scharr Filter.OpenCV_ScharrFilter.java
import org.opencv.core.Core;
import org.opencv.core.CvType;
import org.opencv.core.Mat;
import org.opencv.imgcodecs.Imgcodecs;
import org.opencv.imgproc.Imgproc;
import javax.swing.*;
import java.awt.*;
import java.awt.event.ActionEvent;
import java.awt.event.ActionListener;
import java.awt.image.BufferedImage;
public class OpenCV_ScharrFilter {
static {System.loadLibrary(Core.NATIVE_LIBRARY_NAME);}
private JFrame frmjavaSwing;
/**
* Launch the application.
*/
public static void main(String[] args){
EventQueue.invokeLater(new Runnable() {
@Override
public void run() {
try{
OpenCV_ScharrFilter window = new OpenCV_ScharrFilter();
window.frmjavaSwing.setVisible(true);
}catch (Exception e){
e.printStackTrace();
}
}
});
}
/**
* Create the application.
*/
public OpenCV_ScharrFilter(){
init();
}
/**
* Init the contents of the frame.
*/
private void init(){
final Mat src = Imgcodecs.imread(
"D:\\projects\\Java\\OpenCV_Samples\\resource\\imgs\\lena.jpg");
BufferedImage image = matToBufferedImage(src);
frmjavaSwing = new JFrame();
frmjavaSwing.setTitle("opencv Sharr filter練習");
frmjavaSwing.setBounds(100, 100, 560, 620);
frmjavaSwing.setDefaultCloseOperation(JFrame.EXIT_ON_CLOSE);
frmjavaSwing.getContentPane().setLayout(null);
final JLabel lblNewLabel = new JLabel("");
lblNewLabel.setBounds(5, 60, image.getHeight()+10, image.getWidth()+10);
lblNewLabel.setIcon(new ImageIcon(image));
frmjavaSwing.getContentPane().add(lblNewLabel);
JButton btnX = new JButton("水平處理");
btnX.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent arg0) {
BufferedImage newImage=matToBufferedImage(Convolution(src,1));
lblNewLabel.setIcon(new ImageIcon(newImage));
}
});
btnX.setBounds(42, 10, 114, 23);
frmjavaSwing.getContentPane().add(btnX);
JButton btnY = new JButton("垂直處理");
btnY.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent arg0) {
BufferedImage newImage=matToBufferedImage(Convolution(src,2));
lblNewLabel.setIcon(new ImageIcon(newImage));
}
});
btnY.setBounds(211, 10, 102, 23);
frmjavaSwing.getContentPane().add(btnY);
}
public Mat Convolution(Mat source,int type){
Mat kernel =new Mat(3,3, CvType.CV_32F);
if (type==1){
kernel.put(0, 0, new float[]{-3,-10,-3});
kernel.put(1, 0, new float[]{0,0,0});
kernel.put(2, 0, new float[]{3,10,3});
}else if (type==2){
kernel.put(0, 0, new float[]{-3,0,3});
kernel.put(1, 0, new float[]{-10,0,10});
kernel.put(2, 0, new float[]{-3,0,3});
}
Mat destination=new Mat(source.rows(),source.cols(),source.type());
Imgproc.filter2D(source, destination, -1, kernel);
return destination;
}
public 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;
// bgr to rgb
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 image2 = new BufferedImage(cols, rows, type);
image2.getRaster().setDataElements(0, 0, cols, rows, data);
return image2;
}
}