java springboot 使用tess4j的ocr功能实现解析纯图pdf

依赖

        <!-- tess4j 解析图片 -->
        <dependency>
            <groupId>net.sourceforge.tess4j</groupId>
            <artifactId>tess4j</artifactId>
            <version>4.5.4</version>
        </dependency>
        <!-- 引入 PDFBox 相关的依赖 解析pdf -->
        <dependency>
            <groupId>org.apache.pdfbox</groupId>
            <artifactId>pdfbox</artifactId>
            <version>2.0.29</version>
        </dependency>
        <dependency>
            <groupId>org.apache.pdfbox</groupId>
            <artifactId>fontbox</artifactId>
            <version>2.0.29</version>
        </dependency>
        <dependency>
            <groupId>org.apache.pdfbox</groupId>
            <artifactId>xmpbox</artifactId>
            <version>2.0.29</version>
        </dependency>
        <dependency>
            <groupId>org.apache.pdfbox</groupId>
            <artifactId>preflight</artifactId>
            <version>2.0.29</version>
        </dependency>
        <dependency>
            <groupId>org.apache.pdfbox</groupId>
            <artifactId>pdfbox-tools</artifactId>
            <version>2.0.29</version>
        </dependency>
        <dependency>
            <groupId>org.apache.pdfbox</groupId>
            <artifactId>jempbox</artifactId>
            <version>1.8.17</version>
        </dependency>

下载训练数据资源

地址 : https://github.com/tesseract-ocr/tessdata

点此下载，解压

图片.png

将加载目录设置到这一级

图片.png

springboot 配置读取训练资源

application.yml

# 训练数据文件夹的路径
tess4j:
  datapath: C:/Users/98473/Downloads/tessdata-main/tessdata-main

加载配置

import net.sourceforge.tess4j.Tesseract;
import org.springframework.beans.factory.annotation.Value;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

/**
 * @author Jenson
 * @version 1.0
 */
@Configuration
public class TesseractOcrConfig {
    @Value("${tess4j.datapath}")
    private String dataPath;

    @Bean
    public Tesseract tesseract() {
        Tesseract tesseract = new Tesseract();
        // 设置训练数据文件夹路径
        tesseract.setDatapath(dataPath);
        // 设置为 中文简体 和 英文, +代表使用多语言识别
        tesseract.setLanguage("eng+chi_sim");
        return tesseract;
    }
}

解析纯图pdf为文本

 //解析pdf
    String analysisPdf(String filePath) throws IOException, TesseractException {

        PDDocument doc = PDDocument.load(new File(filePath));
        // 获取PDF文档中的文本内容
        String text = new PDFTextStripper().getText(doc);
        if (!StringUtils.hasText(text)) {
            // 可能是纯图pdf
            log.info("----> 作为纯图pdf解析");
            StringBuilder stringBuilder = new StringBuilder();
            int pageNum = doc.getNumberOfPages();
            for (int i = 0; i < pageNum; i++) {
                PDPage page = doc.getPage(i);
                PDResources resources = page.getResources();
                Iterable<COSName> cosNameIterable = resources.getXObjectNames();
                for (COSName cosName : cosNameIterable) {
                    if (resources.isImageXObject(cosName)) {
                        PDImageXObject pdxObject = (PDImageXObject) resources.getXObject(cosName);
                        ByteArrayOutputStream byteArrayOutputStream = new ByteArrayOutputStream();
                        ImageIO.write(pdxObject.getImage(), "PNG", byteArrayOutputStream);
                        // ocr识别文本
                        ByteArrayInputStream byteArrayInputStream = new ByteArrayInputStream(byteArrayOutputStream.toByteArray());
                        BufferedImage bufferedImage = ImageIO.read(byteArrayInputStream);
                        // 对图片进行文字识别
                        String imageText = tesseract.doOCR(bufferedImage);
                        stringBuilder.append(imageText);

                        byteArrayOutputStream.close();
                        byteArrayInputStream.close();
                    }
                }
            }
            text = stringBuilder.toString();
        }

        // 输出文本内容
        return text;

    }

参考

https://blog.csdn.net/benben_521ben/article/details/127332675
https://blog.csdn.net/mxt51220/article/details/133809953

最后编辑于：2024.03.26 10:37:03