java实现图⽚⽂字识别ocr
最近在开发的时候需要识别图⽚中的⼀些⽂字，⽹上找了相关资料之后，发现google有⼀个离线的⼯具，以下为java使⽤的demo
在此之前，使⽤这个⼯具需要在本地安装OCR⼯具：
下⾯⼀个是⼀定要安装的离线包，建议默认安装
上⾯⼀个是中⽂的语⾔包，如果⽹络可以FQ的童鞋可以在安装的时候就选择语⾔包在线安装，有多种语⾔可供选择，默认只有英⽂的
exe安装好之后，把上⾯⼀个⽂件拷到安装⽬录下tessdata⽂件夹下
如C:\Program Files (x86)\Tesract-OCR\tessdata下
然后下⾯两个是可选包，如果图⽚不做临时⽂件处理的话，可以不需要带的
金缕衣⾸先是⼀个临时⽂件⽣成⽤的类以防源⽂件损坏，参考某位博友的例⼦@Gunner
package org.;
import java.awt.image.BufferedImage;
import java.io.File;
import java.io.IOException;
import java.util.Iterator;
import java.util.Locale;
import javax.imageio.IIOImage;
import javax.imageio.ImageIO;
import javax.imageio.ImageReader;
import javax.imageio.ImageWriteParam;
import javax.imageio.ImageWriter;
import adata.IIOMetadata;
import javax.imageio.stream.ImageInputStream;
import javax.imageio.stream.ImageOutputStream;
import dia.imageio.plugins.tiff.TIFFImageWriteParam;
public class ImageIOHelper {
private Locale locale=Locale.CHINESE;
/**
* ur t locale Construct
* @param locale
*/
public ImageIOHelper(Locale locale){
this.locale=locale;
}
/**
* default construct using default locale Locale.CHINESE
*/
public ImageIOHelper(){
}
/**
* create tempFile of Image in order to prevent damaging original file
* @param imageFile
* @param imageFormat like png,jps .etc
* @return TempFile of Image
耳机模式在哪里关闭* @throws IOException
*/
public File createImage(File imageFile, String imageFormat) throws IOException {
乌镇在哪个省哪个市
Iterator<ImageReader> readers = ImageReadersByFormatName(imageFormat);
ImageReader reader = ();
ImageInputStream iis = ateImageInputStream(imageFile);
reader.tInput(iis);
梦见切肉
IIOMetadata streamMetadata = StreamMetadata();
TIFFImageWriteParam tiffWriteParam = new TIFFImageWriteParam(Locale.CHINESE); tiffWriteParam.tCompressionMode(ImageWriteParam.MODE_DISABLED);
Iterator<ImageWriter> writers = ImageWritersByFormatName("tiff");
ImageWriter writer = ();
BufferedImage bi = ad(0);
IIOImage image = new IIOImage(bi,ImageMetadata(0));
File tempFile = tempImageFile(imageFile);
ImageOutputStream ios = ateImageOutputStream(tempFile);
writer.tOutput(ios);
writer.write(streamMetadata, image, tiffWriteParam);
ios.clo();
iis.clo();
writer.dispo();
reader.dispo();
return tempFile;
}
/**
* add suffix to tempfile
* @param imageFile
* @return
* @throws IOException
*/
private File tempImageFile(File imageFile) throws IOException {
String path = Path();
StringBuffer strB = new StringBuffer(path);
strB.inrt(path.lastIndexOf('.'),"_text_recognize_temp");
String String().replaceFirst("(?<=//.)(//w+)$", "tif");化归思想
return new String());
}
}
下⾯是真正识别的内容：
package org.;
import java.io.BufferedReader;
import java.io.File;
import java.io.FileInputStream;
import java.io.IOException;
import java.io.InputStreamReader;
import java.util.ArrayList;
import java.util.List;
import java.util.Locale;
import org.jdesktop.swingx.util.OS;
haier洗衣机/**
* TEXT Recognize Utils
* @author ink.Flower
*
*/
public class OCRUtil {
private final String LANG_OPTION = "-l"; //英⽂字母⼩写l，并⾮数字1
private final String EOL = Property("line.parator");
private String tessPath = "C://Program Files (x86)//Tesract-OCR";//ocr默认安装路径
private String transname="chi_sim";//默认中⽂语⾔包，识别中⽂
/**
* Construct method of OCR ,t Tesract-OCR install path
* @param tessPath Tesract-OCR install path
* @param transFileName traningFile name aineddata
*/
public OCRUtil(String tessPath,String transFileName){
}
/**
* Construct method of OCR,default path is "C://Program Files (x86)//Tesract-OCR"
*/
public OCRUtil(){ }
public String getTessPath() {
return tessPath;
}
public void tTessPath(String tessPath) {
}
public String getTransname() {
return transname;
}
public void tTransname(String transname) {
}
public String getLANG_OPTION() {
return LANG_OPTION;
}
云存储云计算public String getEOL() {
return EOL;
}
/**
* recognize text in image
* @param imageFile
* @param imageFormat
* @return text recognized in image
* @throws Exception
*/
public String recognizeText(File imageFile,String imageFormat)throws Exception{
File tempImage = new ImageIOHelper().createImage(imageFile,imageFormat);
return ocrImages(tempImage, imageFile);
}
/**
* recognize text in image
* @param imageFile
* @param imageFormat
* @param locale
* @return text recognized in image
* @throws Exception
*/
public String recognizeText(File imageFile,String imageFormat,Locale locale)throws Exception{
File tempImage = new ImageIOHelper(locale).createImage(imageFile,imageFormat);
return ocrImages(tempImage, imageFile);
}
/**
*
* @param tempImage
* @param imageFile
* @return
* @throws IOException
* @throws InterruptedException
*/
private String ocrImages(File tempImage,File imageFile) throws IOException, InterruptedException{ File outputFile = new ParentFile(),"output");
List<String> cmd = new ArrayList<String>();
if(OS.isWindowsXP()){
cmd.add(tessPath+"//tesract");
}el if(OS.isLinux()){
cmd.add("tesract");
}el{
cmd.add(tessPath+"//tesract");
}
cmd.add("");
cmd.Name());
cmd.add(LANG_OPTION);
cmd.add(transname);
ProcessBuilder pb = new ProcessBuilder();
pb.ParentFile());
cmd.t(1, Name());
Process process = pb.start();
int w = process.waitFor();
tempImage.delete();//删除临时正在⼯作⽂件
if(w==0){
BufferedReader in = new BufferedReader(new InputStreamReader(new AbsolutePath()+".txt"),"UTF-8"));今生共相伴歌词
String str;
while((str = in.readLine())!=null){
strB.append(str).append(EOL);
}
in.clo();
}el{
String msg;
switch(w){
ca 1:
msg = "Errors accessing files.There may be spaces in your image's filename.";
break;
ca 29:
msg = "Cannot recongnize the image or its lected region.";
break;
ca 31:
msg = "Unsupported image format.";
break;
default:
msg = "Errors occurred.";
}
tempImage.delete();
throw new RuntimeException(msg);
}
new AbsolutePath()+".txt").delete();
String();
}
}
在实验中发现，如果对有多个⽂字的⼤图进⾏直接识别的话，效果可能⽐较差，所以可以参考另⼀篇切图的博⽂，将图⽚取⼀块之后再识别
这样成功率会提⾼很多。
以上为离线识别版本，效率因图⽽已，具体使⽤的时候可以总结分析，希望对⼤家的学习有所帮助，也希望⼤家多多⽀持。
