Linux OCR and Text Hooking Tools

2 comments

1. you can just run the windows game and textextractor in wine. Most visual novels don’t have native linux builds anyway.

2. Sup brother.

   I was using gazou also, but it’s currently having problems with leptonica for me. I also use agent.

   Now to not just tell what you already know, there is kanjitomo, which will only work on x11. (I use wayland, so I don’t use this)

   ​

   Then there is [https://github.com/kha-white/mokuro](https://github.com/kha-white/mokuro) for manga. You have to process the manga, which requires you to download it, but if you do it works wonders. Makes the entire thing selectable in your browser, so you can just hover over words and use 10ten or Yomichan or whatever it is you use.

   Next there is tesseract, which you can use as a replacement for whatever you were using gazou for.

   Last, but not least, the easiest answer is the exact same text hookers you use on Windows. If for example you were using texthookers for VN’s on windows, if you go to VN reddit and get those they work on Linux through wine. The very important thing is though is that wine processes are isolated. For example, if you open a visual novel on steam using proton, you can’t just double click the text hooker and expect to find your game. You have to use protontricks to run the texthooker in the same prefix as the game. Same principle if you use system wine, just make sure to open the texthooker in the same prefix and it should work flawlessly.

   If you have any questions about a specific use case let me know (For example, you wanna play vn’s and need a text hooker, or whatever.) On Windows and Linux both there really is no “general” solution besides something like kanji tomo (which I actually don’t like very much anyway) so there usually is a “best for doing xxx” option.