Ocrmypdf on 夜云泊 https://lifeislife.cn/tags/ocrmypdf/ feedId:57980998056508425+userId:73222296380546048 Recent content in Ocrmypdf on 夜云泊 Hugo -- 0.166.0 zh Tue, 19 Sep 2023 19:51:18 +0000 ocrmypdf 让 PDF 可搜索 https://lifeislife.cn/posts/ocrmypdf-%E8%AE%A9pdf%E5%8F%AF%E6%90%9C%E7%B4%A2/ Tue, 19 Sep 2023 19:51:18 +0000 https://lifeislife.cn/posts/ocrmypdf-%E8%AE%A9pdf%E5%8F%AF%E6%90%9C%E7%B4%A2/ <p>买的一些课程配套资料都是 PDF 格式的，为了防止盗版都事先用的图片转成的 PDF，这样 PDF 里的内容既没法复制也没法搜索，在查找资料里的关键词的时候就很不方便，所以就想着把这些 PDF 转成可搜索的 PDF。找到了一款工具叫做 ocrmypdf，可以把 PDF 转成可搜索的 PDF，而且还支持中文，这里记录一下使用方法。详细使用文档可以参考官方文档<a href="https://ocrmypdf.readthedocs.io/en/latest/installation.html"> OCRmyPDF documentation</a>。</p> <h1 id="安装">安装</h1> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl">sudo apt install ocrmypdf </span></span></code></pre></div><h1 id="使用">使用</h1> <h2 id="指定-ocr-的语言">指定 OCR 的语言</h2> <p>安装语言包</p> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl">sudo apt install tesseract-ocr-chi-sim </span></span></code></pre></div><p>查看是否安装成功</p> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl">$ tesseract --list-langs </span></span><span class="line"><span class="cl">List of available languages <span class="o">(</span>3<span class="o">)</span>: </span></span><span class="line"><span class="cl">chi_sim </span></span><span class="line"><span class="cl">eng </span></span><span class="line"><span class="cl">osd </span></span></code></pre></div><p>注意参数 <code>-l</code> 后面的语言包名称是下划线，而不是短横线。</p> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl">ocrmypdf -l chi_sim input.pdf output.pdf </span></span></code></pre></div><div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl">$ ocrmypdf -l chi_sim --redo-ocr input.pdf output.pdf </span></span><span class="line"><span class="cl">Scanning contents: 100%<span class="p">|</span>██████████████████████████████████████████████████████████████████████████████████████████████████████<span class="p">|</span> 752/752 <span class="o">[</span>00:14&lt;00:00, 51.36page/s<span class="o">]</span> </span></span><span class="line"><span class="cl">Start processing <span class="m">24</span> pages concurrently </span></span><span class="line"><span class="cl"> <span class="m">33</span> redoing OCR </span></span><span class="line"><span class="cl"> <span class="m">26</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">54</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">88</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">119</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">203</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">256</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">265</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">347</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">376</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">383</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">386</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">402</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">404</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">403</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">412</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">415</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">410</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">439</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">519</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">526</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">587</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">591</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">595</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">607</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">644</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">661</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">682</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">720</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl"> <span class="m">742</span> <span class="o">[</span>tesseract<span class="o">]</span> lots of diacritics - possibly poor OCR </span></span><span class="line"><span class="cl">OCR: 100%<span class="p">|</span>████████████████████████████████████████████████████████████████████████████████████████████████████████████████<span class="p">|</span> 752.0/752.0 <span class="o">[</span>03:41&lt;00:00, 3.40page/s<span class="o">]</span> </span></span><span class="line"><span class="cl">Postprocessing... </span></span><span class="line"><span class="cl">Some input metadata could not be copied because it is not permitted in PDF/A. You may wish to examine the output PDF<span class="err">&#39;</span>s XMP metadata. </span></span><span class="line"><span class="cl">PDF/A conversion: 100%<span class="p">|</span>███████████████████████████████████████████████████████████████████████████████████████████████████████<span class="p">|</span> 752/752 <span class="o">[</span>01:09&lt;00:00, 10.80page/s<span class="o">]</span> </span></span><span class="line"><span class="cl">Recompressing JPEGs: 0image <span class="o">[</span>00:00, ?image/s<span class="o">]</span> </span></span><span class="line"><span class="cl">Deflating JPEGs: 100%<span class="p">|</span>██████████████████████████████████████████████████████████████████████████████████████████████████████<span class="p">|</span> 756/756 <span class="o">[</span>00:00&lt;00:00, 920.21image/s<span class="o">]</span> </span></span><span class="line"><span class="cl">JBIG2: 0item <span class="o">[</span>00:00, ?item/s<span class="o">]</span> </span></span><span class="line"><span class="cl">Optimize ratio: 1.20 savings: 17.0% </span></span><span class="line"><span class="cl">Output file is okay but is not PDF/A <span class="o">(</span>seems to be No PDF/A metadata in XMP<span class="o">)</span> </span></span></code></pre></div><p>转换的结果还不错，页面排版不会改变，保持原样，但是搜索文字时可能需要用空格分开。</p> 买的一些课程配套资料都是 PDF 格式的，为了防止盗版都事先用的图片转成的 PDF，这样 PDF 里的内容既没法复制也没法搜索，在查找资料里的关键词的时候就很不方便，所以就想着把这些 PDF 转成可搜索的 PDF。找到了一款工具叫做 ocrmypdf，可以把 PDF 转成可搜索的 PDF，而且还支持中文，这里记录一下使用方法。详细使用文档可以参考官方文档 OCRmyPDF documentation。

安装

sudo apt install ocrmypdf

使用

指定 OCR 的语言

安装语言包

sudo apt install tesseract-ocr-chi-sim

查看是否安装成功

$ tesseract --list-langs
List of available languages (3):
chi_sim
eng
osd

注意参数 -l 后面的语言包名称是下划线，而不是短横线。

ocrmypdf -l chi_sim input.pdf output.pdf

$  ocrmypdf -l chi_sim  --redo-ocr  input.pdf output.pdf
Scanning contents: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████| 752/752 [00:14<00:00, 51.36page/s]
Start processing 24 pages concurrently
   33 redoing OCR                                                                                                                                                 
   26 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
   54 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
   88 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  119 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  203 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  256 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  265 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  347 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  376 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  383 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  386 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  402 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  404 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  403 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  412 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  415 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  410 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  439 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  519 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  526 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  587 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  591 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  595 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  607 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  644 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  661 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  682 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  720 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
  742 [tesseract] lots of diacritics - possibly poor OCR                                                                                                          
OCR: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 752.0/752.0 [03:41<00:00,  3.40page/s]
Postprocessing...
Some input metadata could not be copied because it is not permitted in PDF/A. You may wish to examine the output PDF's XMP metadata.                              
PDF/A conversion: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████| 752/752 [01:09<00:00, 10.80page/s]
Recompressing JPEGs: 0image [00:00, ?image/s]
Deflating JPEGs: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████| 756/756 [00:00<00:00, 920.21image/s]
JBIG2: 0item [00:00, ?item/s]
Optimize ratio: 1.20 savings: 17.0%
Output file is okay but is not PDF/A (seems to be No PDF/A metadata in XMP)

转换的结果还不错，页面排版不会改变，保持原样，但是搜索文字时可能需要用空格分开。

]]>