java 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，如果网络可以翻墙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)\Tesseract-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

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.ink.image.textrz;

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 javax.imageio.metadata.IIOMetadata;
import javax.imageio.stream.ImageInputStream;
import javax.imageio.stream.ImageOutputStream;  

import com.sun.media.imageio.plugins.tiff.TIFFImageWriteParam;  

public class ImageIOHelper {
    private Locale locale=Locale.CHINESE;
    /**
     * user set 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
     * @throws IOException
     */
    public File createImage(File imageFile, String imageFormat) throws IOException {
        Iterator<ImageReader> readers = ImageIO.getImageReadersByFormatName(imageFormat);
        ImageReader reader = readers.next();
        ImageInputStream iis = ImageIO.createImageInputStream(imageFile);
        reader.setInput(iis);
        IIOMetadata streamMetadata = reader.getStreamMetadata();
        TIFFImageWriteParam tiffWriteParam = new TIFFImageWriteParam(Locale.CHINESE);
        tiffWriteParam.setCompressionMode(ImageWriteParam.MODE_DISABLED);
        Iterator<ImageWriter> writers = ImageIO.getImageWritersByFormatName("tiff");
        ImageWriter writer = writers.next();
        BufferedImage bi = reader.read(0);
        IIOImage image = new IIOImage(bi,null,reader.getImageMetadata(0));
        File tempFile = tempImageFile(imageFile);
        ImageOutputStream ios = ImageIO.createImageOutputStream(tempFile);
        writer.setOutput(ios);
        writer.write(streamMetadata, image, tiffWriteParam);
        ios.close();
        iis.close();
        writer.dispose();
        reader.dispose();
        return tempFile;
    }
    /**
     * add suffix to tempfile
     * @param imageFile
     * @return
     * @throws IOException
     */
    private File tempImageFile(File imageFile) throws IOException {
        String path = imageFile.getPath();
        StringBuffer strB = new StringBuffer(path);
        strB.insert(path.lastIndexOf('.'),"_text_recognize_temp");
        String s=strB.toString().replaceFirst("(?<=//.)(//w+)$", "tif");
        Runtime.getRuntime().exec("attrib "+"\""+s+"\""+" +H"); //设置文件隐藏
        return new File(strB.toString());
    }    

}  

下面是真正识别的内容：

package org.ink.image.textrz;

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;    

/**
 * 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 = System.getProperty("line.separator");
    private String tessPath = "C://Program Files (x86)//Tesseract-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 ,set Tesseract-OCR install path
     * @param tessPath Tesseract-OCR install path
     * @param transFileName traningFile name like eng.traineddata
     */
    public OCRUtil(String tessPath,String transFileName){
        this.tessPath=tessPath;
        this.transname=transFileName;
    }
    /**
     * Construct method of OCR,default path is "C://Program Files (x86)//Tesseract-OCR"
     */
    public OCRUtil(){     }

    public String getTessPath() {
        return tessPath;
    }
    public void setTessPath(String tessPath) {
        this.tessPath = tessPath;
    }
    public String getTransname() {
        return transname;
    }
    public void setTransname(String transname) {
        this.transname = transname;
    }
    public String getLANG_OPTION() {
        return LANG_OPTION;
    }
    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 File(imageFile.getParentFile(),"output");
        Runtime.getRuntime().exec("attrib "+"\""+outputFile.getAbsolutePath()+"\""+" +H"); //设置文件隐藏
        StringBuffer strB = new StringBuffer();
        List<String> cmd = new ArrayList<String>();
        if(OS.isWindowsXP()){
            cmd.add(tessPath+"//tesseract");
        }else if(OS.isLinux()){
            cmd.add("tesseract");
        }else{
            cmd.add(tessPath+"//tesseract");
        }
        cmd.add("");
        cmd.add(outputFile.getName());
        cmd.add(LANG_OPTION);
        cmd.add(transname);
        ProcessBuilder pb = new ProcessBuilder();
        pb.directory(imageFile.getParentFile());
        cmd.set(1, tempImage.getName());
        pb.command(cmd);
        pb.redirectErrorStream(true);
        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 FileInputStream(outputFile.getAbsolutePath()+".txt"),"UTF-8"));
            String str;
            while((str = in.readLine())!=null){
                strB.append(str).append(EOL);
            }
            in.close();
        }else{
            String msg;
            switch(w){
            case 1:
                msg = "Errors accessing files.There may be spaces in your image's filename.";
                break;
            case 29:
                msg = "Cannot recongnize the image or its selected region.";
                break;
            case 31:
                msg = "Unsupported image format.";
                break;
            default:
                msg = "Errors occurred.";
            }
            tempImage.delete();
            throw new RuntimeException(msg);
        }
        new File(outputFile.getAbsolutePath()+".txt").delete();
        return strB.toString();
    }
}    

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

http://www.cnblogs.com/inkflower/p/6642089.html    ←我是链接

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

另外，博主也在网上看到百度也有图片文字识别的工具

百度OCR企业版

以上这个貌似是付费的，如果要免费的可以网上找找，另外API也可以找找