> For the complete documentation index, see [llms.txt](https://docs.vntranslator.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.vntranslator.com/user-guide/ocr/ocr-engines/tesseract-ocr.md).

# Tesseract OCR

{% embed url="<https://www.youtube.com/watch?v=5Uzl211TxjY>" %}

### **Download & Install Tesseract**

* Visit the [Tesseract at UB Mannheim](https://github.com/UB-Mannheim/tesseract/wiki)
* Select the **tesseract-ocr-w64-setup-v5.3.x.exe (64 bit**) file to download the Tesseract executable installer
* Once downloaded, open the executable file and follow the installation prompts

{% hint style="info" %}
Make sure you have installed the tesseract-64bit in C:\Program Files\Tesseract-OCR.

If you are using **Windows OCR** or **Tesseract OCR**, make sure to select the correct OCR language for your game.

**For Japanese games:**

* Windows OCR: `Japanese`
* Tesseract OCR: `Japanese (jpn.traineddata)`

**For English games:**

* Windows OCR: `English (United States)` or `English (United Kingdom)`
* Tesseract OCR: `English (eng.traineddata)`
  {% endhint %}

### Trained Data Files (Languages)

You can download the `.traineddata` file for the language you need and place it in Tesseract OCR installation directory `C:\Program Files\Tesseract-OCR\tessdata`\\`[here]` \
(this should be the same as where the tessdata directory is installed)

> **tessdata** <https://github.com/tesseract-ocr/tessdata> \
> Speed : Faster than tessdata-best \
> Accuracy : Slightly less accurate than tessdata-best

> **tessdata-best** `(Recommended for video games)` <https://github.com/tesseract-ocr/tessdata_best> \
> Speed : Slowest \
> Accuracy : Most accurate

> **tessdata-fast** <https://github.com/tesseract-ocr/tessdata_fast> \
> Speed : Fastest \
> Accuracy : Least accurate

### Page Segmentation Modes

The PSM allows you to select a segmentation method dependent on your particular image and the environment in which it was captured

<table><thead><tr><th width="69" align="center"> </th><th>Page segmentation modes</th></tr></thead><tbody><tr><td align="center">1</td><td>Orientation and script detection (OSD) only.</td></tr><tr><td align="center">2</td><td>Automatic page segmentation with OSD.</td></tr><tr><td align="center">3</td><td>Automatic page segmentation, but no OSD, or OCR. (not implemented)</td></tr><tr><td align="center">4</td><td>Fully automatic page segmentation, but no OSD. (Default)</td></tr><tr><td align="center">5</td><td>Assume a single column of text of variable sizes.</td></tr><tr><td align="center">6</td><td>Assume a single uniform block of vertically aligned text.</td></tr><tr><td align="center">7</td><td>Assume a single uniform block of text.</td></tr><tr><td align="center">8</td><td>Treat the image as a single text line.</td></tr><tr><td align="center">9</td><td>Treat the image as a single word.</td></tr><tr><td align="center">10</td><td>Treat the image as a single word in a circle.</td></tr><tr><td align="center">11</td><td>Treat the image as a single character.</td></tr><tr><td align="center">12</td><td>Sparse text. Find as much text as possible in no particular order.</td></tr><tr><td align="center">13</td><td>Sparse text with OSD.</td></tr><tr><td align="center">14</td><td>Raw line. Treat the image as a single text line, bypassing hacks that are Tesseract-specific.</td></tr></tbody></table>