Opencv实现身份证OCR识别的示例详解
作者：落日流年
这篇文章主要为大家详细介绍了如何使用Opencv实现身份证OCR识别功能,文中的示例代码讲解详细,具有一定的学习价值,感兴趣的小伙伴可以跟随小编一起了解一下
Opencv 配置IDEA可参考：Java调用opencv IDEA环境配置的教程详解
opencv位置：
OpencvUtil类：
package com.x.common.utils;
import org.opencv.core.*;
import org.opencv.core.Point;
import org.opencv.imgcodecs.Imgcodecs;
import org.opencv.imgproc.Imgproc;
import org.opencv.objdetect.CascadeClassifier;
import javax.imageio.ImageIO;
import java.awt.*;
import java.awt.image.BufferedImage;
import java.awt.image.DataBufferByte;
import java.io.ByteArrayInputStream;
import java.io.InputStream;
import java.util.ArrayList;
import java.util.Collections;
import java.util.Comparator;
import java.util.List;
public class OpencvUtil {
private static final int BLACK = 0;
private static final int WHITE = 255;
/**
* 灰化处理
* @return
*/
public static Mat gray (Mat mat){
Mat gray = new Mat();
Imgproc.cvtColor(mat, gray, Imgproc.COLOR_BGR2GRAY,1);
return gray;
}
/**
* 二值化处理
* @return
*/
public static Mat binary (Mat mat){
Mat binary = new Mat();
Imgproc.adaptiveThreshold(mat, binary, 255, Imgproc.ADAPTIVE_THRESH_MEAN_C, Imgproc.THRESH_BINARY_INV, 25, 10);
return binary;
}
/**
* 模糊处理
* @param mat
* @return
*/
public static Mat blur (Mat mat) {
Mat blur = new Mat();
Imgproc.blur(mat,blur,new Size(5,5));
return blur;
}
/**
*膨胀
* @param mat
* @return
*/
public static Mat dilate (Mat mat,int size){
Mat dilate=new Mat();
Mat element = Imgproc.getStructuringElement(Imgproc.MORPH_RECT, new Size(size,size));
//膨胀
Imgproc.dilate(mat, dilate, element, new Point(-1, -1), 1);
return dilate;
}
/**
* 腐蚀
* @param mat
* @return
*/
public static Mat erode (Mat mat,int size){
Mat erode=new Mat();
Mat element = Imgproc.getStructuringElement(Imgproc.MORPH_RECT, new Size(size,size));
//腐蚀
Imgproc.erode(mat, erode, element, new Point(-1, -1), 1);
return erode;
}
/**
* 边缘检测
* @param mat
* @return
*/
public static Mat carry(Mat mat){
Mat dst=new Mat();
//高斯平滑滤波器卷积降噪
Imgproc.GaussianBlur(mat, dst, new Size(3,3), 0);
//边缘检测
Imgproc.Canny(mat, dst, 50, 150);
return dst;
}
/**
* 轮廓检测
* @param mat
* @return
*/
public static List<MatOfPoint> findContours(Mat mat){
List<MatOfPoint> contours=new ArrayList<>();
Mat hierarchy = new Mat();
Imgproc.findContours(mat, contours, hierarchy, Imgproc.RETR_LIST, Imgproc.CHAIN_APPROX_SIMPLE);
return contours;
}
/**
* 清除小面积轮廓
* @param mat
* @param size
* @return
*/
public static Mat drawContours(Mat mat,int size){
List<MatOfPoint> cardContours=OpencvUtil.findContours(mat);
for (int i = 0; i < cardContours.size(); i++)
{
double area=OpencvUtil.area(cardContours.get(i));
if(area<size){
Imgproc.drawContours(mat, cardContours, i, new Scalar( 0, 0, 0),-1 );
}
}
return mat;
}
/**
* 人脸识别
* @param mat
* @return
*/
public static Mat face(Mat mat){
CascadeClassifier faceDetector = new CascadeClassifier(
System.getProperty("user.dir")+"\\opencv\\haarcascades\\haarcascade_frontalface_alt.xml");
// 在图片中检测人脸
MatOfRect faceDetections = new MatOfRect();
//指定人脸识别的最大和最小像素范围
Size minSize = new Size(100, 100);
Size maxSize = new Size(500, 500);
//参数设置为scaleFactor=1.1f, minNeighbors=4, flags=0 以此来增加识别人脸的正确率
faceDetector.detectMultiScale(mat, faceDetections, 1.1f, 4, 0, minSize, maxSize);
Rect[] rects = faceDetections.toArray();
if(rects != null && rects.length == 1){
// 在每一个识别出来的人脸周围画出一个方框
Rect rect = rects[0];
return mat;
}else{
return null;
}
}
/**
* 循环进行人脸识别
* */
public static Mat faceLoop(Mat src){
Mat face=new Mat();
//默认人脸识别失败时图像旋转90度
int k=90;
while (k>0){
for(int i=0;i<360/k;i++){
//人脸识别
face= OpencvUtil.face(src);
if(face==null){
src = rotate3(src,k);
}else{
break;
}
}
if(face!=null){
break;
}else{
k=k-30;
}
}
return src;
}
/**
* 剪切身份证区域
* @param mat
*/
public static Mat houghLinesP(Mat begin,Mat mat){
//灰度
mat=OpencvUtil.gray(mat);
//二值化
mat=OpencvUtil.binary(mat);
//腐蚀
mat=OpencvUtil.erode(mat,5);
//边缘检测
mat=OpencvUtil.carry(mat);
//降噪
mat=OpencvUtil.navieRemoveNoise(mat,1);
//膨胀
mat=OpencvUtil.dilate(mat,3);
//轮廓检测,清除小的轮廓部分
List<MatOfPoint> contours=OpencvUtil.findContours(mat);
for(int i=0;i<contours.size();i++){
double area=OpencvUtil.area(contours.get(i));
if(area<5000){
Imgproc.drawContours(mat, contours, i, new Scalar( 0, 0, 0), -1);
}
}
Mat storage = new Mat();
Imgproc.HoughLinesP(mat, storage, 1, Math.PI / 180, 10, 0, 10);
double[] maxLine = new double[]{0,0,0,0};
//获取最长的直线
for (int x = 0; x < storage.rows(); x++)
{
double[] vec = storage.get(x, 0);
double x1 = vec[0], y1 = vec[1], x2 = vec[2], y2 = vec[3];
double newLength = Math.sqrt(Math.abs((x1 - x2)* (x1 - x2)+(y1 - y2)* (y1 - y2)));
double oldLength = Math.sqrt(Math.abs((maxLine[0] - maxLine[2])* (maxLine[0] - maxLine[2])+(maxLine[1] - maxLine[3])* (maxLine[1] - maxLine[3])));
if(newLength>oldLength){
maxLine = vec;
}
}
//计算最长线的角度
double angle = getAngle(maxLine[0],maxLine[1],maxLine[2],maxLine[3]);
//旋转角度
mat = rotate3( mat,angle);
begin = rotate3( begin,angle);
Imgproc.HoughLinesP(mat, storage, 1, Math.PI / 180, 10, 10, 10);
List<double[]> lines=new ArrayList<>();
//在mat上划线
for (int x = 0; x < storage.rows(); x++)
{
double[] vec = storage.get(x, 0);
double x1 = vec[0], y1 = vec[1], x2 = vec[2], y2 = vec[3];
Point start = new Point(x1, y1);
Point end = new Point(x2, y2);
//获取与图像x边缘近似平行的直线
if(Math.abs(start.y-end.y)<5){
if(Math.abs(x2-x1)>20){
lines.add(vec);
}
}
//获取与图像y边缘近似平行的直线
if(Math.abs(start.x-end.x)<5){
if(Math.abs(y2-y1)>20){
lines.add(vec);
}
}
}
//获取最大的和最小的X,Y坐标
double maxX=0.0,minX=10000,minY=10000,maxY=0.0;
for(int i=0;i<lines.size();i++){
double[] vec = lines.get(i);
double x1 = vec[0], y1 = vec[1], x2 = vec[2], y2 = vec[3];
maxX=maxX>x1?maxX:x1;
maxX=maxX>x2?maxX:x2;
minX=minX>x1?x1:minX;
minX=minX>x2?x2:minX;
maxY=maxY>y1?maxY:y1;
maxY=maxY>y2?maxY:y2;
minY=minY>y1?y1:minY;
minY=minY>y2?y2:minY;
}
if(maxX<mat.cols()&&minX>0&&maxY<mat.rows()&&minY>0){
List<Point> list=new ArrayList<>();
Point point1=new Point(minX+10,minY+10);
Point point2=new Point(minX+10,maxY-10);
Point point3=new Point(maxX-10,minY+10);
Point point4=new Point(maxX-10,maxY-10);
list.add(point1);
list.add(point2);
list.add(point3);
list.add(point4);
mat=shear(begin,list);
}else{
mat=begin;
}
return mat;
}
/**
* 计算角度
* @param px1
* @param py1
* @param px2
* @param py2
* @return
*/
public static double getAngle(double px1, double py1, double px2, double py2) {
//两点的x、y值
double x = px2-px1;
double y = py2-py1;
double hypotenuse = Math.sqrt(Math.pow(x, 2)+Math.pow(y, 2));
//斜边长度
double cos = x/hypotenuse;
double radian = Math.acos(cos);
//求出弧度
double angle = 180/(Math.PI/radian);
//用弧度算出角度
if (y<0) {
angle = -angle;
} else if ((y == 0) && (x<0)) {
angle = 180;
}
while (angle<0){
angle = angle +90;
}
return angle;
}
/**
* 累计概率hough变换直线检测
* @param mat
*/
public static Mat houghLines(Mat mat){
Mat storage = new Mat();
Imgproc.HoughLines(mat, storage, 1, Math.PI / 180, 50, 0, 0, 0, 1);
for (int x = 0; x < storage.rows(); x++) {
double[] vec = storage.get(x, 0);
double rho = vec[0];
double theta = vec[1];
Point pt1 = new Point();
Point pt2 = new Point();
double a = Math.cos(theta);
double b = Math.sin(theta);
double x0 = a * rho;
double y0 = b * rho;
pt1.x = Math.round(x0 + 1000 * (-b));
pt1.y = Math.round(y0 + 1000 * (a));
pt2.x = Math.round(x0 - 1000 * (-b));
pt2.y = Math.round(y0 - 1000 * (a));
if (theta >= 0)
{
Imgproc.line(mat, pt1, pt2, new Scalar(255), 3);
}
}
return mat;
}
/**
* 根据四点坐标截取模板图片
* @param mat
* @param pointList
* @return
*/
public static Mat shear (Mat mat,List<Point> pointList){
int x=minX(pointList);
int y=minY(pointList);
int xl=xLength(pointList)>mat.cols()-x?mat.cols()-x:xLength(pointList);
int yl=yLength(pointList)>mat.rows()-y?mat.rows()-y:yLength(pointList);
Rect re=new Rect(x,y,xl,yl);
return new Mat(mat,re);
}
/**
* 图片旋转
* @param splitImage
* @param angle
* @return
*/
public static Mat rotate3(Mat splitImage, double angle){
double thera = angle * Math.PI / 180;
double a = Math.sin(thera);
double b = Math.cos(thera);
int wsrc = splitImage.width();
int hsrc = splitImage.height();
int wdst = (int) (hsrc * Math.abs(a) + wsrc * Math.abs(b));
int hdst = (int) (wsrc * Math.abs(a) + hsrc * Math.abs(b));
Mat imgDst = new Mat(hdst, wdst, splitImage.type());
Point pt = new Point(splitImage.cols() / 2, splitImage.rows() / 2);
// 获取仿射变换矩阵
Mat affineTrans = Imgproc.getRotationMatrix2D(pt, angle, 1.0);
//System.out.println(affineTrans.dump());
// 改变变换矩阵第三列的值
affineTrans.put(0, 2, affineTrans.get(0, 2)[0] + (wdst - wsrc) / 2);
affineTrans.put(1, 2, affineTrans.get(1, 2)[0] + (hdst - hsrc) / 2);
Imgproc.warpAffine(splitImage, imgDst, affineTrans, imgDst.size(),
Imgproc.INTER_CUBIC | Imgproc.WARP_FILL_OUTLIERS);
return imgDst;
}
/**
* 图像直方图处理
* @param mat
* @return
*/
public static Mat equalizeHist(Mat mat){
Mat dst = new Mat();
List<Mat> mv = new ArrayList<>();
Core.split(mat, mv);
for (int i = 0; i < mat.channels(); i++)
{
Imgproc.equalizeHist(mv.get(i), mv.get(i));
}
Core.merge(mv, dst);
return dst;
}
/**
* 8邻域降噪,又有点像9宫格降噪;即如果9宫格中心被异色包围，则同化
* @param pNum 默认值为1
*/
public static Mat navieRemoveNoise(Mat mat,int pNum) {
int i, j, m, n, nValue, nCount;
int nWidth = mat.cols();
int nHeight = mat.rows();
// 如果一个点的周围都是白色的，而它确是黑色的，删除它
for (j = 1; j < nHeight - 1; ++j) {
for (i = 1; i < nWidth - 1; ++i) {
nValue = (int)mat.get(j, i)[0];
if (nValue == 0) {
nCount = 0;
// 比较以(j ,i)为中心的9宫格，如果周围都是白色的，同化
for (m = j - 1; m <= j + 1; ++m) {
for (n = i - 1; n <= i + 1; ++n) {
if ((int)mat.get(m, n)[0] == 0) {
nCount++;
}
}
}
if (nCount <= pNum) {
// 周围黑色点的个数小于阀值pNum,把该点设置白色
mat.put(j, i, WHITE);
}
} else {
nCount = 0;
// 比较以(j ,i)为中心的9宫格，如果周围都是黑色的，同化
for (m = j - 1; m <= j + 1; ++m) {
for (n = i - 1; n <= i + 1; ++n) {
if ((int)mat.get(m, n)[0] == 0) {
nCount++;
}
}
}
if (nCount >= 7) {
// 周围黑色点的个数大于等于7,把该点设置黑色;即周围都是黑色
mat.put(j, i, BLACK);
}
}
}
}
return mat;
}
/**
* 连通域降噪
* @param pArea 默认值为1
*/
public static Mat contoursRemoveNoise(Mat mat,double pArea) {
//mat=floodFill(mat,mat.new Point(mat.cols()/2,mat.rows()/2),new Color(225,0,0));
int i, j, color = 1;
int nWidth = mat.cols(), nHeight = mat.rows();
for (i = 0; i < nWidth; ++i) {
for (j = 0; j < nHeight; ++j) {
if ((int) mat.get(j, i)[0] == BLACK) {
//用不同颜色填充连接区域中的每个黑色点
//floodFill就是把一个点x的所有相邻的点都涂上x点的颜色，一直填充下去，直到这个区域内所有的点都被填充完为止
Imgproc.floodFill(mat, new Mat(), new Point(i, j), new Scalar(color));
color++;
}
}
}
//统计不同颜色点的个数
int[] ColorCount = new int[255];
for (i = 0; i < nWidth; ++i) {
for (j = 0; j < nHeight; ++j) {
if ((int) mat.get(j, i)[0] != 255) {
ColorCount[(int) mat.get(j, i)[0] - 1]++;
}
}
}
//去除噪点
for (i = 0; i < nWidth; ++i) {
for (j = 0; j < nHeight; ++j) {
if (ColorCount[(int) mat.get(j, i)[0] - 1] <= pArea) {
mat.put(j, i, WHITE);
}
}
}
for (i = 0; i < nWidth; ++i) {
for (j = 0; j < nHeight; ++j) {
if ((int) mat.get(j, i)[0] < WHITE) {
mat.put(j, i, BLACK);
}
}
}
return mat;
}
/**
* Mat转换成BufferedImage
*
* @param matrix
* 要转换的Mat
* @param fileExtension
* 格式为 ".jpg", ".png", etc
* @return
*/
public static BufferedImage Mat2BufImg (Mat matrix, String fileExtension) {
MatOfByte mob = new MatOfByte();
Imgcodecs.imencode(fileExtension, matrix, 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;
}
/**
* BufferedImage转换成Mat
*
* @param original
* 要转换的BufferedImage
* @param imgType
* bufferedImage的类型 如 BufferedImage.TYPE_3BYTE_BGR
* @param matType
* 转换成mat的type 如 CvType.CV_8UC3
*/
public static Mat BufImg2Mat (BufferedImage original, int imgType, int matType) {
if (original == null) {
throw new IllegalArgumentException("original == null");
}
if (original.getType() != imgType) {
BufferedImage image = new BufferedImage(original.getWidth(), original.getHeight(), imgType);
Graphics2D g = image.createGraphics();
try {
g.setComposite(AlphaComposite.Src);
g.drawImage(original, 0, 0, null);
} finally {
g.dispose();
}
}
DataBufferByte dbi =(DataBufferByte)original.getRaster().getDataBuffer();
byte[] pixels = dbi.getData();
Mat mat = Mat.eye(original.getHeight(), original.getWidth(), matType);
mat.put(0, 0, pixels);
return mat;
}
/**
* 人眼识别
* @param mat
* @return
*/
public static List<Point> eye(Mat mat){
List<Point> eyeList=new ArrayList<>();
CascadeClassifier eyeDetector = new CascadeClassifier(
System.getProperty("user.dir")+"\\opencv\\haarcascades\\haarcascade_eye.xml");
// 在图片中检测人眼
MatOfRect eyeDetections = new MatOfRect();
//指定人脸识别的最大和最小像素范围
Size minSize = new Size(20, 20);
Size maxSize = new Size(30, 30);
eyeDetector.detectMultiScale(mat, eyeDetections, 1.1f, 3, 0, minSize, maxSize);
Rect[] rects = eyeDetections.toArray();
if(rects != null && rects.length == 2){
Point point1=new Point(rects[0].x,rects[0].y);
eyeList.add(point1);
Point point2=new Point(rects[1].x,rects[1].y);
eyeList.add(point2);
}else{
return null;
}
return eyeList;
}
/**
* 获取轮廓的顶点坐标
* @param contour
* @return
*/
public static List<Point> getPointList(MatOfPoint contour){
MatOfPoint2f mat2f=new MatOfPoint2f();
contour.convertTo(mat2f,CvType.CV_32FC1);
RotatedRect rect=Imgproc.minAreaRect(mat2f);
Mat points=new Mat();
Imgproc.boxPoints(rect,points);
return getPoints(points.dump());
}
/**
* 获取轮廓的面积
* @param contour
* @return
*/
public static double area (MatOfPoint contour){
MatOfPoint2f mat2f=new MatOfPoint2f();
contour.convertTo(mat2f,CvType.CV_32FC1);
RotatedRect rect=Imgproc.minAreaRect(mat2f);
return rect.boundingRect().area();
}
/**
* 获取点坐标集合
* @param str
* @return
*/
public static List<Point> getPoints(String str){
List<Point> points=new ArrayList<>();
str=str.replace("[","").replace("]","");
String[] pointStr=str.split(";");
for(int i=0;i<pointStr.length;i++){
double x=Double.parseDouble(pointStr[i].split(",")[0]);
double y=Double.parseDouble(pointStr[i].split(",")[1]);
Point po=new Point(x,y);
points.add(po);
}
return points;
}
/**
* 获取最小的X坐标
* @param points
* @return
*/
public static int minX(List<Point> points){
Collections.sort(points, new XComparator(false));
return (int)(points.get(0).x>0?points.get(0).x:-points.get(0).x);
}
/**
* 获取最小的Y坐标
* @param points
* @return
*/
public static int minY(List<Point> points){
Collections.sort(points, new YComparator(false));
return (int)(points.get(0).y>0?points.get(0).y:-points.get(0).y);
}
/**
* 获取最长的X坐标距离
* @param points
* @return
*/
public static int xLength(List<Point> points){
Collections.sort(points, new XComparator(false));
return (int)(points.get(3).x-points.get(0).x);
}
/**
* 获取最长的Y坐标距离
* @param points
* @return
*/
public static int yLength(List<Point> points){
Collections.sort(points, new YComparator(false));
return (int)(points.get(3).y-points.get(0).y);
}
//集合排序规则（根据X坐标排序）
public static class XComparator implements Comparator<Point> {
private boolean reverseOrder; // 是否倒序
public XComparator(boolean reverseOrder) {
this.reverseOrder = reverseOrder;
}
public int compare(Point arg0, Point arg1) {
if(reverseOrder)
return (int)arg1.x - (int)arg0.x;
else
return (int)arg0.x - (int)arg1.x;
}
}
//集合排序规则（根据Y坐标排序）
public static class YComparator implements Comparator<Point> {
private boolean reverseOrder; // 是否倒序
public YComparator(boolean reverseOrder) {
this.reverseOrder = reverseOrder;
}
public int compare(Point arg0, Point arg1) {
if(reverseOrder)
return (int)arg1.y - (int)arg0.y;
else
return (int)arg0.y - (int)arg1.y;
}
}
}OCRUtil类：
package com.xinjian.x.common.ocr;
import net.sourceforge.tess4j.ITesseract;
import net.sourceforge.tess4j.Tesseract;
import net.sourceforge.tess4j.util.LoadLibs;
import java.awt.image.BufferedImage;
import java.io.File;
public class OCRUtil {
/**
* 识别图片信息
* @param img
* @return
*/
public static String getImageMessage(BufferedImage img,String language){
String result="end";
try{
ITesseract instance = new Tesseract();
File tessDataFolder = LoadLibs.extractTessResources("tessdata");
instance.setLanguage(language);
instance.setDatapath(tessDataFolder.getAbsolutePath());
result = instance.doOCR(img);
//System.out.println(result);
}catch(Exception e){
System.out.println(e.getMessage());
}
return result;
}
}language为语言包名称eng或者chi_sim,chi_sim语言包可能与jar包不匹配需要注意
<!--OCR Tesseract-->
<dependency>
<groupId>net.java.dev.jna</groupId>
<artifactId>jna</artifactId>
<version>4.1.0</version>
</dependency>
<dependency>
<groupId>net.sourceforge.tess4j</groupId>
<artifactId>tess4j</artifactId>
<version>2.0.1</version>
<exclusions>
<exclusion>
<groupId>com.sun.jna</groupId>
<artifactId>jna</artifactId>
</exclusion>
</exclusions>
</dependency>Main 方法：
package com.xinjian.x.modules.orc;
import com.xinjian.x.common.ocr.OCRUtil;
import com.xinjian.x.common.utils.OpencvUtil;
import org.opencv.core.*;
import org.opencv.core.Point;
import org.opencv.imgcodecs.Imgcodecs;
import org.opencv.imgproc.Imgproc;
import java.awt.image.BufferedImage;
import java.util.ArrayList;
import java.util.List;
import static com.xinjian.x.common.utils.OpencvUtil.rotate3;
import static com.xinjian.x.common.utils.OpencvUtil.shear;
public class OrcTest {
static {
System.loadLibrary(Core.NATIVE_LIBRARY_NAME);
//注意程序运行的时候需要在VM option添加该行 指明opencv的dll文件所在路径
//-Djava.library.path=$PROJECT_DIR$\opencv\x64
}
public static void main(String[] args){
String path="D:/Users/xinjian09/Desktop/c.jpg";
Mat mat= Imgcodecs.imread(path);
cardUp(mat);
}
/**
* 身份证反面识别
*/
public static void cardDown(Mat mat){
//灰度
mat=OpencvUtil.gray(mat);
//二值化
mat=OpencvUtil.binary(mat);
//腐蚀
mat=OpencvUtil.erode(mat,3);
//膨胀
mat=OpencvUtil.dilate(mat,3);
//检测是否有居民身份证字体，若有为正向，若没有则旋转图片
for(int i=0;i<4;i++){
String temp=temp(mat);
if(!temp.contains("居")&&!temp.contains("民")){
mat= rotate3(mat,90);
}else{
break;
}
}
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/result.jpg", mat);
String organization=organization (mat);
System.out.print("签发机关是："+organization);
String time=time (mat);
System.out.print("有效期限是："+time);
}
public static String temp (Mat mat){
Point point1=new Point(mat.cols()*0.30,mat.rows()*0.25);
Point point2=new Point(mat.cols()*0.30,mat.rows()*0.25);
Point point3=new Point(mat.cols()*0.90,mat.rows()*0.45);
Point point4=new Point(mat.cols()*0.90,mat.rows()*0.45);
List<Point> list=new ArrayList<>();
list.add(point1);
list.add(point2);
list.add(point3);
list.add(point4);
Mat temp= shear(mat,list);
List<MatOfPoint> nameContours=OpencvUtil.findContours(temp);
for (int i = 0; i < nameContours.size(); i++)
{
double area=OpencvUtil.area(nameContours.get(i));
if(area<100){
Imgproc.drawContours(temp, nameContours, i, new Scalar( 0, 0, 0), -1);
}
}
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/temp.jpg", temp);
BufferedImage nameBuffer=OpencvUtil.Mat2BufImg(temp,".jpg");
String nameStr=OCRUtil.getImageMessage(nameBuffer,"chi_sim");
nameStr=nameStr.replace("\n","");
return nameStr;
}
public static String organization (Mat mat){
Point point1=new Point(mat.cols()*0.36,mat.rows()*0.68);
Point point2=new Point(mat.cols()*0.36,mat.rows()*0.68);
Point point3=new Point(mat.cols()*0.80,mat.rows()*0.80);
Point point4=new Point(mat.cols()*0.80,mat.rows()*0.80);
List<Point> list=new ArrayList<>();
list.add(point1);
list.add(point2);
list.add(point3);
list.add(point4);
Mat name= shear(mat,list);
List<MatOfPoint> nameContours=OpencvUtil.findContours(name);
for (int i = 0; i < nameContours.size(); i++)
{
double area=OpencvUtil.area(nameContours.get(i));
if(area<100){
Imgproc.drawContours(name, nameContours, i, new Scalar( 0, 0, 0), -1);
}
}
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/organization.jpg", name);
BufferedImage nameBuffer=OpencvUtil.Mat2BufImg(name,".jpg");
String nameStr=OCRUtil.getImageMessage(nameBuffer,"chi_sim");
nameStr=nameStr.replace("\n","");
return nameStr+"\n";
}
public static String time (Mat mat){
Point point1=new Point(mat.cols()*0.38,mat.rows()*0.82);
Point point2=new Point(mat.cols()*0.38,mat.rows()*0.82);
Point point3=new Point(mat.cols()*0.85,mat.rows()*0.92);
Point point4=new Point(mat.cols()*0.85,mat.rows()*0.92);
List<Point> list=new ArrayList<>();
list.add(point1);
list.add(point2);
list.add(point3);
list.add(point4);
Mat time= shear(mat,list);
List<MatOfPoint> timeContours=OpencvUtil.findContours(time);
for (int i = 0; i < timeContours.size(); i++)
{
double area=OpencvUtil.area(timeContours.get(i));
if(area<100){
Imgproc.drawContours(time, timeContours, i, new Scalar( 0, 0, 0), -1);
}
}
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/time.jpg", time);
//起始日期
Point startPoint1=new Point(0,0);
Point startPoint2=new Point(0,time.rows());
Point startPoint3=new Point(time.cols()*0.47,0);
Point startPoint4=new Point(time.cols()*0.47,time.rows());
List<Point> startList=new ArrayList<>();
startList.add(startPoint1);
startList.add(startPoint2);
startList.add(startPoint3);
startList.add(startPoint4);
Mat start= shear(time,startList);
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/start.jpg", start);
BufferedImage yearBuffer=OpencvUtil.Mat2BufImg(start,".jpg");
String startStr=OCRUtil.getImageMessage(yearBuffer,"eng");
startStr=startStr.replace("-","");
startStr=startStr.replace(" ","");
startStr=startStr.replace("\n","");
//截止日期
Point endPoint1=new Point(time.cols()*0.47,0);
Point endPoint2=new Point(time.cols()*0.47,time.rows());
Point endPoint3=new Point(time.cols(),0);
Point endPoint4=new Point(time.cols(),time.rows());
List<Point> endList=new ArrayList<>();
endList.add(endPoint1);
endList.add(endPoint2);
endList.add(endPoint3);
endList.add(endPoint4);
Mat end= shear(time,endList);
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/end.jpg", end);
BufferedImage endBuffer=OpencvUtil.Mat2BufImg(end,".jpg");
String endStr=OCRUtil.getImageMessage(endBuffer,"chi_sim");
if(!endStr.contains("长")&&!endStr.contains("期")){
endStr=OCRUtil.getImageMessage(endBuffer,"eng");
endStr=endStr.replace("-","");
endStr=endStr.replace(" ","");
}
return startStr+"-"+endStr;
}
/**
* 身份证正面识别
*/
public static void cardUp (Mat mat){
Mat begin=mat.clone();
//截取身份证区域，并校正旋转角度
mat = OpencvUtil.houghLinesP(begin,mat);
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/houghLinesP.jpg", mat);
//循环进行人脸识别,校正图片方向
mat=OpencvUtil.faceLoop(mat);
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/face.jpg", mat);
//灰度
mat=OpencvUtil.gray(mat);
//二值化
mat=OpencvUtil.binary(mat);
//腐蚀
mat=OpencvUtil.erode(mat,1);
//膨胀
mat=OpencvUtil.dilate(mat,1);
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/bbb.jpg", mat);
//获取名称
String name=name(mat);
System.out.print("姓名是："+name);
//获取性别
String sex=sex(mat);
System.out.print("性别是："+sex);
//获取民族
String nation=nation(mat);
System.out.print("民族是："+nation);
//获取出生日期
String birthday=birthday(mat);
System.out.print("出生日期是："+birthday);
//获取住址
String address=address(mat);
System.out.print("住址是："+address);
//获取身份证
String card=card(mat);
System.out.print("身份证号是："+card);
}
public static String name(Mat mat){
Point point1=new Point(mat.cols()*0.18,mat.rows()*0.11);
Point point2=new Point(mat.cols()*0.18,mat.rows()*0.24);
Point point3=new Point(mat.cols()*0.4,mat.rows()*0.11);
Point point4=new Point(mat.cols()*0.4,mat.rows()*0.24);
List<Point> list=new ArrayList<>();
list.add(point1);
list.add(point2);
list.add(point3);
list.add(point4);
Mat name= shear(mat,list);
name=OpencvUtil.drawContours(name,50);
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/name.jpg", name);
BufferedImage nameBuffer=OpencvUtil.Mat2BufImg(name,".jpg");
String nameStr=OCRUtil.getImageMessage(nameBuffer,"chi_sim");
nameStr=nameStr.replace("\n","");
return nameStr+"\n";
}
public static String sex(Mat mat){
Point point1=new Point(mat.cols()*0.18,mat.rows()*0.25);
Point point2=new Point(mat.cols()*0.18,mat.rows()*0.35);
Point point3=new Point(mat.cols()*0.25,mat.rows()*0.25);
Point point4=new Point(mat.cols()*0.25,mat.rows()*0.35);
List<Point> list=new ArrayList<>();
list.add(point1);
list.add(point2);
list.add(point3);
list.add(point4);
Mat sex= shear(mat,list);
sex=OpencvUtil.drawContours(sex,50);
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/sex.jpg", sex);
BufferedImage sexBuffer=OpencvUtil.Mat2BufImg(sex,".jpg");
String sexStr=OCRUtil.getImageMessage(sexBuffer,"chi_sim");
sexStr=sexStr.replace("\n","");
return sexStr+"\n";
}
public static String nation(Mat mat){
Point point1=new Point(mat.cols()*0.39,mat.rows()*0.25);
Point point2=new Point(mat.cols()*0.39,mat.rows()*0.36);
Point point3=new Point(mat.cols()*0.55,mat.rows()*0.25);
Point point4=new Point(mat.cols()*0.55,mat.rows()*0.36);
List<Point> list=new ArrayList<>();
list.add(point1);
list.add(point2);
list.add(point3);
list.add(point4);
Mat nation= shear(mat,list);
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/nation.jpg", nation);
BufferedImage nationBuffer=OpencvUtil.Mat2BufImg(nation,".jpg");
String nationStr=OCRUtil.getImageMessage(nationBuffer,"chi_sim");
nationStr=nationStr.replace("\n","");
return nationStr+"\n";
}
public static String birthday(Mat mat){
Point point1=new Point(mat.cols()*0.18,mat.rows()*0.35);
Point point2=new Point(mat.cols()*0.18,mat.rows()*0.35);
Point point3=new Point(mat.cols()*0.55,mat.rows()*0.48);
Point point4=new Point(mat.cols()*0.55,mat.rows()*0.48);
List<Point> list=new ArrayList<>();
list.add(point1);
list.add(point2);
list.add(point3);
list.add(point4);
Mat birthday= shear(mat,list);
birthday=OpencvUtil.drawContours(birthday,50);
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/birthday.jpg", birthday);
//年份
Point yearPoint1=new Point(0,0);
Point yearPoint2=new Point(0,birthday.rows());
Point yearPoint3=new Point(birthday.cols()*0.29,0);
Point yearPoint4=new Point(birthday.cols()*0.29,birthday.rows());
List<Point> yearList=new ArrayList<>();
yearList.add(yearPoint1);
yearList.add(yearPoint2);
yearList.add(yearPoint3);
yearList.add(yearPoint4);
Mat year= shear(birthday,yearList);
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/year.jpg", year);
BufferedImage yearBuffer=OpencvUtil.Mat2BufImg(year,".jpg");
String yearStr=OCRUtil.getImageMessage(yearBuffer,"eng");
//月份
Point monthPoint1=new Point(birthday.cols()*0.44,0);
Point monthPoint2=new Point(birthday.cols()*0.44,birthday.rows());
Point monthPoint3=new Point(birthday.cols()*0.55,0);
Point monthPoint4=new Point(birthday.cols()*0.55,birthday.rows());
List<Point> monthList=new ArrayList<>();
monthList.add(monthPoint1);
monthList.add(monthPoint2);
monthList.add(monthPoint3);
monthList.add(monthPoint4);
Mat month= shear(birthday,monthList);
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/month.jpg", month);
BufferedImage monthBuffer=OpencvUtil.Mat2BufImg(month,".jpg");
String monthStr=OCRUtil.getImageMessage(monthBuffer,"eng");
//日期
Point dayPoint1=new Point(birthday.cols()*0.69,0);
Point dayPoint2=new Point(birthday.cols()*0.69,birthday.rows());
Point dayPoint3=new Point(birthday.cols()*0.80,0);
Point dayPoint4=new Point(birthday.cols()*0.80,birthday.rows());
List<Point> dayList=new ArrayList<>();
dayList.add(dayPoint1);
dayList.add(dayPoint2);
dayList.add(dayPoint3);
dayList.add(dayPoint4);
Mat day= shear(birthday,dayList);
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/day.jpg", day);
BufferedImage dayBuffer=OpencvUtil.Mat2BufImg(day,".jpg");
String dayStr=OCRUtil.getImageMessage(dayBuffer,"eng");
String birthdayStr=yearStr+"年"+monthStr+"月"+dayStr+"日";
birthdayStr=birthdayStr.replace("\n","");
return birthdayStr+"\n";
}
public static String address(Mat mat){
Point point1=new Point(mat.cols()*0.17,mat.rows()*0.47);
Point point2=new Point(mat.cols()*0.17,mat.rows()*0.47);
Point point3=new Point(mat.cols()*0.61,mat.rows()*0.76);
Point point4=new Point(mat.cols()*0.61,mat.rows()*0.76);
List<Point> list=new ArrayList<>();
list.add(point1);
list.add(point2);
list.add(point3);
list.add(point4);
Mat address= shear(mat,list);
address=OpencvUtil.drawContours(address,50);
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/address.jpg", address);
BufferedImage addressBuffer=OpencvUtil.Mat2BufImg(address,".jpg");
return OCRUtil.getImageMessage(addressBuffer,"chi_sim")+"\n";
}
public static String card(Mat mat){
Point point1=new Point(mat.cols()*0.34,mat.rows()*0.75);
Point point2=new Point(mat.cols()*0.34,mat.rows()*0.75);
Point point3=new Point(mat.cols()*0.89,mat.rows()*0.91);
Point point4=new Point(mat.cols()*0.89,mat.rows()*0.91);
List<Point> list=new ArrayList<>();
list.add(point1);
list.add(point2);
list.add(point3);
list.add(point4);
Mat card= shear(mat,list);
card=OpencvUtil.drawContours(card,50);
Imgcodecs.imwrite("D:/Users/xinjian09/Desktop/card.jpg", card);
BufferedImage cardBuffer=OpencvUtil.Mat2BufImg(card,".jpg");
return OCRUtil.getImageMessage(cardBuffer,"eng")+"\n";
}
}以上就是Opencv实现身份证OCR识别的示例详解的详细内容，更多关于Opencv身份证OCR识别的资料请关注脚本之家其它相关文章！