谷歌开源OCR,tesseract-ocr使用笔记

官方教程地址:https://github.com/tesseract-ocr/tesseract/wiki/Compiling

测试版本为

root@9a2a063f9534:/tesseract/testing# tesseract -v
tesseract 4.00.00dev-697-gcdc3533
 leptonica-1.74.4
  libjpeg 8d (libjpeg-turbo 1.4.2) : libpng 1.2.54 : libtiff 4.0.6 : zlib 1.2.8

 Found AVX2
 Found AVX
 Found SSE

一、Docker + Ubuntu

git clone git@github.com:tesseract-ocr/tesseract.git
cd tesseract
docker pull ubuntu:latest
docker build -t google-ocr:latest .
docker run -itd --name ocr google-ocr:latest /bin/bash
docker exec -it ocr /bin/bash

进入环境后,需要训练功能要执行下面第二条

apt-get install -y  g++ autoconf automake libtool autoconf-archive pkg-config libpng-dev libjpeg8-dev libtiff5-dev zlib1g-dev git
#training
apt-get install -y libicu-dev libpango1.0-dev libcairo2-dev 

Leptonica

Tesseract    Leptonica    Ubuntu
4.00       1.74.2      Must build from source

官网给出必须源码安装,所以去找源码安

cd /tmp
git clone https://github.com/DanBloomberg/leptonica.git
cd leptonica
autoreconf -vi
./autobuild 
./configure
make
make install

安装主体ocr

cd /tesseract
./autogen.sh
LIBLEPT_HEADERSDIR=/usr/include 
./configure --with-extra-libraries=/usr/local/lib
make install

测试安装是否成功

tesseract
tesseract -v

下载字库模型,选自己需要的就行了

字库地址:https://github.com/tesseract-ocr/tessdata
手册地址:https://github.com/tesseract-ocr/tesseract/wiki/Data-Files

将字库放入指定路径

export TESSDATA_PREFIX=/tesseract/tessdata
cp xxx.traindata /tesseract/

执行测试

cd /tesseract/testing
#english
tesseract phototest.tif result -l eng
#chinese
tesseract chi.jpg result1 -l chi_sim

检查输出

cat result.txt
cat result1.txt

可通过训练提高精度,训练方法见官方文档,这个我没试过。

附录:

python 调用接口:https://www.pyimagesearch.com/2017/07/10/using-tesseract-ocr-python/

python 官方调用依赖:https://github.com/madmaze/pytesseract

原文地址:https://www.cnblogs.com/mangoVic/p/8074624.html