AIM ：将PDF转换为base64，其中PDF可以是普通PDF或扫描PDF。
我正在使用Tesseract OCR将扫描的PDF转换为文本文件。由于我在Java工作，我正在使用terr4j库。
我认为程序的流程如下：
Get PDF file ---> Convert each page to image using Ghost4j ---> Pass each image to tess4f for OCR ---> convert whole text to base64。
我可以使用以下代码将PDF文件转换为图像：
package helpers;
import java.io.File;
import java.io.FileInputStream;
import java.io.FileNotFoundException;
import java.io.IOException;
import java.awt.Image;
import java.awt.image.RenderedImage;
import java.util.List;
import javax.imageio.ImageIO;
import org.ghost4j.document.DocumentException;
import org.ghost4j.document.PDFDocument;
import org.ghost4j.analyzer.FontAnalyzer;
import org.ghost4j.renderer.RendererException;
import org.ghost4j.renderer.SimpleRenderer;
import net.sourceforge.tess4j.*;
class encoder {
public static byte[] createByteArray(File pCurrentFolder, String pNameOfBinaryFile) {
String pathToBinaryData = pCurrentFolder.getAbsolutePath()+"/"+pNameOfBinaryFile;
File file = new File(pathToBinaryData);
if (!file.exists()) {
System.out.println(pNameOfBinaryFile+" could not be found in folder "+pCurrentFolder.getName());
return null;
}
FileInputStream fin = null;
try {
fin = new FileInputStream(file);
} catch (FileNotFoundException e) {
e.printStackTrace();
}
byte fileContent[] = new byte[(int) file.length()];
try {
if (fin != null)
fin.read(fileContent);
} catch (IOException e) {
e.printStackTrace();
}
return fileContent;
}
public void covertToImage(File pdfDoc) {
PDFDocument document = new PDFDocument();
try {
document.load(pdfDoc);
} catch (IOException e) {
e.printStackTrace();
}
SimpleRenderer renderer = new SimpleRenderer();
renderer.setResolution(300);
List<Image> images = null;
try {
images = renderer.render(document);
} catch (IOException e) {
e.printStackTrace();
} catch (RendererException e) {
e.printStackTrace();
} catch (DocumentException e) {
e.printStackTrace();
}
try {
if (images != null) {
// for testing only 1 page
ImageIO.write((RenderedImage) images.get(10), "png", new File("/home/cloudera/Downloads/1.png"));
}
} catch (IOException e) {
e.printStackTrace();
}
}
}
public class encodeFile {
public static void main(String[] args) {
/* This part is for pure PDF files i.e. not scanned */
//byte[] arr = encoder.createByteArray(new File("/home/cloudera/Downloads/"), "test.pdf");
//String result = javax.xml.bind.DatatypeConverter.printBase64Binary(arr);
//System.out.println(result);
/* This part create the image for a page of scanned PDF file */
new encoder().covertToImage(new File("/home/cloudera/Downloads/isl99201.pdf")); // results in 1.png
/* This part is for OCR */
Tesseract instance = new Tesseract();
String res = instance.doOCR(new File("/home/cloudera/Downloads/1.png"));
System.out.println(res);
}
}
运行它会产生以下错误：
当我尝试从PDF创建图像时会发生这种情况。我已经看到，如果我从build.sbt中删除tess4j，则创建的图像没有任何错误，但我必须使用它。
Connected to the target VM, address: '127.0.0.1:46698', transport: 'socket'
Exception in thread "main" java.lang.AbstractMethodError: com.sun.jna.Structure.getFieldOrder()Ljava/util/List;
at com.sun.jna.Structure.fieldOrder(Structure.java:884)
at com.sun.jna.Structure.getFields(Structure.java:910)
at com.sun.jna.Structure.deriveLayout(Structure.java:1058)
at com.sun.jna.Structure.calculateSize(Structure.java:982)
at com.sun.jna.Structure.calculateSize(Structure.java:949)
at com.sun.jna.Structure.allocateMemory(Structure.java:375)
at com.sun.jna.Structure.<init>(Structure.java:184)
at com.sun.jna.Structure.<init>(Structure.java:172)
at com.sun.jna.Structure.<init>(Structure.java:159)
at com.sun.jna.Structure.<init>(Structure.java:151)
at org.ghost4j.GhostscriptLibrary$display_callback_s.<init>(GhostscriptLibrary.java:63)
at org.ghost4j.Ghostscript.buildNativeDisplayCallback(Ghostscript.java:381)
at org.ghost4j.Ghostscript.initialize(Ghostscript.java:336)
at org.ghost4j.renderer.SimpleRenderer.run(SimpleRenderer.java:105)
at org.ghost4j.renderer.AbstractRemoteRenderer.render(AbstractRemoteRenderer.java:86)
at org.ghost4j.renderer.AbstractRemoteRenderer.render(AbstractRemoteRenderer.java:70)
at helpers.encoder.covertToImage(encodeFile.java:62)
at helpers.encodeFile.main(encodeFile.java:86)
Disconnected from the target VM, address: '127.0.0.1:46698', transport: 'socket'
Process finished with exit code 1
将任何图像传递给tess4j时发生此错误：
Connected to the target VM, address: '127.0.0.1:46133', transport: 'socket'
Exception in thread "main" java.lang.UnsatisfiedLinkError: Unable to load library 'tesseract': Native library (linux-x86-64/libtesseract.so) not found in resource path (....)
at com.sun.jna.NativeLibrary.loadLibrary(NativeLibrary.java:271)
at com.sun.jna.NativeLibrary.getInstance(NativeLibrary.java:398)
at com.sun.jna.Library$Handler.<init>(Library.java:147)
at com.sun.jna.Native.loadLibrary(Native.java:412)
at com.sun.jna.Native.loadLibrary(Native.java:391)
at net.sourceforge.tess4j.util.LoadLibs.getTessAPIInstance(LoadLibs.java:78)
at net.sourceforge.tess4j.TessAPI.<clinit>(TessAPI.java:40)
at net.sourceforge.tess4j.Tesseract.init(Tesseract.java:360)
at net.sourceforge.tess4j.Tesseract.doOCR(Tesseract.java:273)
at net.sourceforge.tess4j.Tesseract.doOCR(Tesseract.java:205)
at net.sourceforge.tess4j.Tesseract.doOCR(Tesseract.java:189)
at helpers.encodeFile.main(encodeFile.java:89)
Disconnected from the target VM, address: '127.0.0.1:46133', transport: 'socket'
Process finished with exit code 1
我正在Intellij使用64位CentOS 6.6上的SBT。通过一些互联网搜索，我能够理解上述问题，但我面临两个限制：
正在使用的JNA库默认为最新版本，即4.1.0。我在互联网上读到了JNA与其他库之间的不兼容性。所以我试着指定旧版本的3.4.0。但是build.sbt一直拒绝这样做。
我在64位系统上，tessearct可以使用32位系统。我应该如何将其整合到项目中？
以下是build.sbt中处理所有必需库的部分：
"org.ghost4j" % "ghost4j" % "0.5.1",
"org.bouncycastle" % "bctsp-jdk14" % "1.46",
"net.sourceforge.tess4j" % "tess4j" % "2.0.0",
"com.github.jai-imageio" % "jai-imageio-core" % "1.3.0"
"net.java.dev.jna" % "jna" % "3.4.0", // does not make any difference as only 4.1.0 is installed.
请帮我解决这个问题。
更新：我将"net.java.dev.jna" % "jna" % "3.4.0" force()添加到build.sbt，这解决了我的第一个问题。
答案 0 :(得分：0)
这个问题的解决方案在于我在github上找到的Tesseract-API。我将其分成my Github帐户并为扫描图像添加了test并进行了一些代码重构。这种方式使库开始正常运行。我用于测试的扫描文档是here。
我在Travis上成功构建它，现在它在32位和64位系统上运行良好。