java视觉识别：java文字识别技术 亲测 识别率很高

java视觉识别：java文字识别技术 亲测 识别率很高测试类TestOCR ：OCR 类：3.到网上搜索下载java图形处理所需的2个包：jai_imageio-1.1-alpha.jar，swingx-1.6.1.jar4.java程序清单：ImageioHelper 类：

java文字识别程序的关键是寻找一个可以调用的OCR引擎。tesseract-ocr就是一个这样的OCR引擎，在1985年到1995年由HP实验室开发，现在在Google。tesseract-OCR 3.0发布，支持中文。不过tesseract-ocr 3.0不是图形化界面的客户端，别人写的FreeOCR图形化客户端还不支持导入新的 3.0 traineddata。但这标志着，现在有自由的中文OCR软件了。

java中使用tesseract-ocr3.01的步骤如下：

1.下载安装tesseract-ocr-setup-3.01-1.exe（3.0以上版本才增加了中文识别）

2.在安装向导中可以选择需要下载的语言包。

3.到网上搜索下载java图形处理所需的2个包：jai_imageio-1.1-alpha.jar，swingx-1.6.1.jar

4.java程序清单：

ImageioHelper 类：

1. 
   
3. import java.awt.image.BufferedImage;
4. import java.io.File;
5. import java.io.IOException;
6. import java.util.Iterator;
7. import java.util.Locale;
9. import javax.imageio.IIOImage;
10. import javax.imageio.ImageIO;
11. import javax.imageio.Imagereader;
12. import javax.imageio.ImageWriteParam;
13. import javax.imageio.ImageWriter;
14. import javax.imageio.metadata.IIOMetadata;
15. import javax.imageio.stream.ImageInputStream;
16. import javax.imageio.stream.ImageOutputStream;
18. import com.sun.media.imageio.plugins.tiff.TIFFImageWriteParam;
20. public class ImageIOHelper {
22. public static File createImage(File imageFile String imageFormat) {
23. File tempFile = null;
24. try {
25. Iterator readers = ImageIO.getImageReadersByFormatName(imageFormat);
26. ImageReader reader = readers.next();
28. ImageInputStream iis = ImageIO.createImageInputStream(imageFile);
29. reader.setInput(iis);
30. //Read the stream metadata
31. IIOMetadata streamMetadata = reader.getStreamMetadata();
33. //Set up the writeParam
34. TIFFImageWriteParam tiffWriteParam = new TIFFImageWriteParam(Locale.CHINESE);
35. tiffWriteParam.setCompressionMode(ImageWriteParam.MODE_DISABLED);
37. //Get tif writer and set output to file
38. Iterator writers = ImageIO.getImageWritersByFormatName("tiff");
39. ImageWriter writer = writers.next();
41. BufferedImage bi = reader.read(0);
42. IIOImage image = new IIOImage(bi null reader.getImageMetadata(0));
43. tempFile = tempImageFile(imageFile);
44. ImageOutputStream ios = ImageIO.createImageOutputStream(tempFile);
45. writer.setOutput(ios);
46. writer.write(streamMetadata image tiffWriteParam);
47. ios.close();
49. writer.dispose();
50. reader.dispose();
52. } catch (IOException e) {
53. e.printStackTrace();
54. }
55. return tempFile;
56. }
58. private static File tempImageFile(File imageFile) {
59. String path = imageFile.getPath();
60. StringBuffer strB = new StringBuffer(path);
61. strB.insert(path.lastIndexOf('.') 0);
62. return new File(strB.toString().replaceFirst("(?<=//.)(//w )$" "tif"));
63. }
65. }

OCR 类：

1. 
   
2. package com.hhp.util;
4. import java.io.BufferedReader;
5. import java.io.File;
6. import java.io.FileInputStream;
7. import java.io.InputStreamReader;
8. import java.util.ArrayList;
9. import java.util.List;
10. import org.jdesktop.swingx.util.OS;
12. public class OCR {
13. private final String LANG_OPTION = "-l"; //英文字母小写l，并非数字1
14. private final String EOL = System.getProperty("line.separator");
15. private String tessPath = "C://Program Files (x86)//Tesseract-OCR";
16. //private String tessPath = new File("tesseract").getAbsolutePath();
18. public String recognizeText(File imageFile String imageFormat)throws Exception{
19. File tempImage = ImageIOHelper.createImage(imageFile imageFormat);
20. File outputFile = new File(imageFile.getParentFile() "output");
21. StringBuffer strB = new StringBuffer();
22. List cmd = new ArrayList();
23. if(OS.isWindowsXP()){
24. cmd.add(tessPath "//tesseract");
25. }else if(OS.isLinux()){
26. cmd.add("tesseract");
27. }else{
28. cmd.add(tessPath "//tesseract");
29. }
30. cmd.add("");
31. cmd.add(outputFile.getName());
32. cmd.add(LANG_OPTION);
33. cmd.add("chi_sim");
34. //cmd.add("eng");
36. ProcessBuilder pb = new ProcessBuilder();
37. pb.directory(imageFile.getParentFile());
39. cmd.set(1 tempImage.getName());
40. pb.command(cmd);
41. pb.redirectErrorStream(true);
43. Process process = pb.start();
44. //tesseract.exe 1.jpg 1 -l chi_sim
45. int w = process.waitFor();
47. //删除临时正在工作文件
48. tempImage.delete();
50. if(w==0){
51. BufferedReader in = new BufferedReader(new InputStreamReader(new FileInputStream(outputFile.getAbsolutePath() ".txt") "UTF-8"));
53. String str;
54. while((str = in.readLine())!=null){
55. strB.append(str).append(EOL);
56. }
57. in.close();
58. }else{
59. String msg;
60. switch(w){
61. case 1:
62. msg = "Errors accessing files.There may be spaces in your image's filename.";
63. break;
64. case 29:
65. msg = "Cannot recongnize the image or its selected region.";
66. break;
67. case 31:
68. msg = "Unsupported image format.";
69. break;
70. default:
71. msg = "Errors occurred.";
72. }
73. tempImage.delete();
74. throw new RuntimeException(msg);
75. }
76. new File(outputFile.getAbsolutePath() ".txt").delete();
77. return strB.toString();
78. }
79. }

测试类TestOCR ：

1. 
   
2. import java.io.File;
3. import java.io.IOException;
5. import com.hhp.util.OCR;
7. public class OcrTest {
9. public static void main(String[] args) {
10. String path = "C://temp//OCRcode//4.jpg";
11. System.out.println("ORC Test Begin......");
12. try {
13. String valCode = new OCR().recognizeText(new File(path) .jpg");
14. System.out.println(valCode);
15. } catch (IOException e) {
16. e.printStackTrace();
17. } catch (Exception e) {
18. e.printStackTrace();
19. }
20. System.out.println("ORC Test End......");
21. }
23. }

经过测试，tesseract-ocr 3.01的文字识别率很高，对于网站中常见的验证码识别率也很高。