Tesseract centos环境安装，基于springboot图片提取文字

零的学习笔记

已于 2024-04-17 17:17:41 修改

阅读量639

点赞数

分类专栏： java类文章标签： centos java linux

于 2023-02-28 17:56:17 首次发布

本文链接：https://blog.csdn.net/qq_27421665/article/details/129266725

版权

java类专栏收录该内容

1 篇文章 0 订阅

订阅专栏

下载tesseract-orc

https://github.com/tesseract-ocr/tesseract/tags

下载leptonica

wget http://www.leptonica.org/source/leptonica-1.78.0.tar.gz

解压leptonica
tar -xvf leptonica-1.78.0.tar.gz
配置编译安装leptonica
进文件夹

./configure
make
make install

安装automake和libtool

yum install automake
yum install libtool

配置leptonica环境变量

vi /etc/profile

export LD_LIBRARY_PATH=/usr/local/lib
export LIBLEPT_HEADERSDIR=/usr/local/include
export PKG_CONFIG_PATH=/usr/local/lib/pkgconfig

source /etc/profile

安装tesseract-ocr
进文件夹

./autogen.sh
./configure
make
make install

编译的时候遇到问题
Your-compiler-does-not-have-the-necessary-C17-support
执行

yum install -y centos-release-scl
yum install devtoolset-8-gcc*
scl enable devtoolset-8 bash

重新编译

测试是否安装成功,执行

ldconfig
tesseract --version

将下载的语言包复制到/usr/local/share/tessdata
最新包

<dependency>
   <groupId>net.sourceforge.tess4j</groupId>
   <artifactId>tess4j</artifactId>
   <version>5.6.0</version>
</dependency>
springboot 解决 jna 找不到对应的文件

<dependency>
   <groupId>net.java.dev.jna</groupId>
   <artifactId>jna</artifactId>
   <version>5.13.0</version>
</dependency>

demo

@PostMapping("/upload")
public String testFile(MultipartFile file){
    try {
        Tesseract instance = new Tesseract();
        instance.setLanguage("eng+chi_sim");
        instance.setHocr(false);
        instance.setPageSegMode(ITessAPI.TessPageSegMode.PSM_AUTO_OSD);
        instance.setOcrEngineMode(TessAPI.TessOcrEngineMode.OEM_LSTM_ONLY);
        PDDocument document = PDDocument.load(file.getInputStream());
        PDFRenderer pdfRenderer = new PDFRenderer(document);
        BufferedImage bim = pdfRenderer.renderImageWithDPI(0, 200, ImageType.RGB);
        String result = instance.doOCR(bim);
        return result;
    } catch (IOException | TesseractException e) {
        e.printStackTrace();
    }
    return "";
}