图像处理04

折腾了一大圈，最后还是移植到java上了…………

插段代码留个底。

import net.sourceforge.tess4j.ITesseract;
import net.sourceforge.tess4j.Tesseract;
import org.opencv.core.*;
import org.opencv.imgproc.Imgproc;

import javax.imageio.ImageIO;
import java.awt.image.BufferedImage;
import java.io.ByteArrayInputStream;
import java.io.InputStream;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;
import java.util.Map;

import static org.opencv.imgcodecs.Imgcodecs.imencode;
import static org.opencv.imgcodecs.Imgcodecs.imread;
import static org.opencv.imgcodecs.Imgcodecs.imwrite;
import static org.opencv.imgproc.Imgproc.*;

public class textDetection {
    static {
        System.loadLibrary(Core.NATIVE_LIBRARY_NAME);
    }

    public static ITesseract instance = new Tesseract();

    public static void main(String[] args) {
        System.out.println("hello");

        //ocr初始化
        instance.setDatapath("C:\Program Files (x86)\Tesseract-OCR\tessdata");
        //此处的-psm 7为参数，指定了识别目标为单行从左到右的字母，增加该参数可以提高识别率。
        //但是不知道java怎么加这个参数。
//        List<String> config=new ArrayList<String>();
//        config.add("-psm 7");
//        instance.setConfigs(config);

        //文字提取和识别
        Mat img = imread("test2.jpg");
        List<Mat> pre = preprocess(img);
        imwrite("test_1.jpg", pre.get(0));
        Map<Rect, String> region = dectection(pre.get(0), pre.get(1));
 57     }

//    图片预处理
    //返回值中，0为二值化结果，1为腐蚀膨胀后的结果
    public static List<Mat> preprocess(Mat img) {
        List<Mat> re=new ArrayList<Mat>();
        Mat gray = new Mat();
        cvtColor(img, gray, COLOR_BGR2GRAY);
        Mat th = new Mat();
        Imgproc.threshold(gray, th, 48, 255, THRESH_BINARY_INV);
        re.add(th);

        Mat element1=getStructuringElement(MORPH_RECT, new Size(10,10));
        Mat element2=getStructuringElement(MORPH_RECT, new Size(30,30));

        //腐蚀和膨胀
        Mat erosion=new Mat();
        erode(th, erosion, element1);
        Mat dilation=new Mat();
        dilate(erosion, dilation, element2);
        Mat dilation2=new Mat();
        dilate(dilation, dilation2, element2);
        Mat dilation3=new Mat();
        dilate(dilation2, dilation3, element2);
        re.add(dilation3);

        //中间图输出
//        imwrite("binary.png", th);
//        imwrite("dilation.png", dilation);
//        imwrite("erosion.png", erosion);
//        imwrite("dilation2.png", dilation2);


        return re;
    }

//    文字提取和识别
    //参数img为二值化结果，用于文字识别。position为腐蚀膨胀结果，用于文字提取。
    //返回值为矩形框和其对应的文字识别结果的map
    public static Map<Rect, String> dectection(Mat img, Mat position) {
        List<MatOfPoint> pointlist=new ArrayList<MatOfPoint>();
        Mat mat=new Mat();
        findContours(position, pointlist, mat, RETR_EXTERNAL, CHAIN_APPROX_SIMPLE);
        int i=1;
        Map<Rect, String> re=new HashMap<Rect, String>();
        for (MatOfPoint point:pointlist) {
            Rect rect = boundingRect(point);
            System.out.println(i+" "+rect.x+" "+rect.y+" "+rect.width+" "+rect.height);
            i++;
            rectangle(img, new Point(rect.x, rect.y), new Point(rect.x+rect.width, rect.y+rect.height), new Scalar(255, 0, 0));
            Mat sub_mat=img.submat(rect.y, rect.y+rect.height, rect.x, rect.x+rect.width);
            BufferedImage bimage=getImage(sub_mat);
            try {
                String answer = instance.doOCR(bimage);
                //System.out.println(answer);
                re.put(rect, answer);
            } catch (Exception e) {
                e.printStackTrace();
            }
        }
        //imwrite("test_2.jpg", img);
        return re;
    }

//    mat转bufferedimage
    public static BufferedImage getImage(Mat img) {
        MatOfByte mob= new MatOfByte();
        imencode(".jpg", img, mob);
        byte[] byteArray = mob.toArray();
        BufferedImage bufImage = null;
        try {
            InputStream in = new ByteArrayInputStream(byteArray);
            bufImage = ImageIO.read(in);
        } catch (Exception e) {
            e.printStackTrace();
        }
        return bufImage;
    }
166 
    //两点距离
    public static Double dis(Double[] a, Double[] b) {
        return Math.sqrt(dis_s(a, b));
    }

    //两点距离的平方
    public static Double dis_s(Double[] a, Double[] b) {
        return ((a[0]-b[0])*(a[0]-b[0])+(a[1]-b[1])*(a[1]-b[1]));
    }

}