> For the complete documentation index, see [llms.txt](https://docs.wehost.co.in/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.wehost.co.in/blog/depixelating-information-in-the-document-finding-what-was-not-supposed-to-be-found.md).

# Depixelating information in the document (finding what was not supposed to be found)

### In short

* While playing the GreenHorn box on Hack The Box, I encountered a situation where a PDF contained a pixelated password
  * <https://www.hackthebox.com/achievement/machine/409699/617>

<figure><img src="/files/lk2fiGksLpInls3rvaEI" alt=""><figcaption></figcaption></figure>

* Upon deeper investigation, I found ways to depixelate the image
  * with
    * <https://github.com/spipm/Depix>
* the result&#x20;

<figure><img src="/files/prKjcWfy2OrGbz7xRh1O" alt=""><figcaption></figcaption></figure>

* potential password

```
sidefromsidetheothersidesidefromsidetheotherside
```

### Diving deeper

* I will dive deeper into how it works another day
* Today, I'll focus on how to depixelate the image
* I extracted the image from the PDF. *(Note: Do not take a screenshot! Always extract the image directly from the PDF to preserve quality.)*
  * I initially made the mistake of thinking it *would be fine* to use a screenshot.
  * As a result, Depix was unable to depixelate the image
* After I passed it to Depix

```
python3 depix.py -p pix.png -s ./images/searchimages/debruinseq_notepad_Windows10_closeAndSpaced.png -o output.png
```

* the result from

<figure><img src="/files/jz8OYq5z7I4fY4xgCcxC" alt=""><figcaption></figcaption></figure>

* to

<figure><img src="/files/toS2J4TfAPboD76vEiu8" alt=""><figcaption></figcaption></figure>

## References

* <https://github.com/spipm/Depix>