java：pdfbox 3.0 去除扫描版PDF中文本水印

官网下载 https://pdfbox.apache.org/download.html
下载 pdfbox-app-3.0.3.jar

cd D:\pdfbox
运行 java -jar pdfbox-app-3.0.3.jar

java -jar pdfbox-app-3.0.3.jar
Usage: pdfbox [COMMAND] [OPTIONS]
Commands:debug          Analyzes and inspects the internal structure of a PDF documentdecrypt        Decrypts a PDF documentencrypt        Encrypts a PDF documentdecode         Writes a PDF document with all streams decodedexport:images  Extracts the images from a PDF documentexport:xmp     Extracts the xmp stream from a PDF documentexport:text    Extracts the text from a PDF documentexport:fdf     Exports AcroForm form data to FDFexport:xfdf    Exports AcroForm form data to XFDFimport:fdf     Imports AcroForm form data from FDFimport:xfdf    Imports AcroForm form data from XFDFoverlay        Adds an overlay to a PDF documentprint          Prints a PDF documentrender         Converts a PDF document to image(s)merge          Merges multiple PDF d*.ocuments into onesplit          Splits a PDF document into number of new documentsfromimage      Creates a PDF document from imagesfromtext       Creates a PDF document from textversion        Gets the version of PDFBoxhelp           Display help information about the specified command.
See 'pdfbox help <command>' to read about a specific subcommand

运行 java -jar pdfbox-app-3.0.3.jar debug

# 导出扫描版PDF文件中每页的图片文件

java -jar pdfbox-app-3.0.3.jar export:images -prefix=test -i your_book.pdf
导出
Writing image: test-1.jpg
Writing image: test-2.jpg
Writing image: test-3.png
……

# from 多个 image 合并生成 pdf

java -jar pdfbox-app-3.0.3.jar fromimage -o=book1.pdf -i=test-1.jpg -i=test-2.jpg -i=test-3.png -i=test-4.jpg
生成 book1.pdf  视觉效果太差，而且命令行长度限制了图片文件数（一般扫描书都有几百页）。

还是要自己编程搞定。