I’m learning japanese through visual novels and want to do the same with mangas. What’s the best app, akin to Textractor for vns, that can extract the sentences from the pages so I can look up the kanji and all?
I’m using an open source MacOS app called bob (Github link shared down below) to do text extraction and translation, and it works great. However couple of things need to be aware:
– Even though the installation instructions are in English, but the app UI is all in Chinese kanji, took me a while to figure out how to get it to work.
– User will need to get Google API key in order to use the app (the ones come out-of-box does not work with Japanese)
In short, you’ll need to be pretty tech-savvy and have a lot of patient to make it to work. 😂 kanjitomo is an alternative, it’s easier to setup, but it does not work as well as bob.
Both use [manga-ocr](https://github.com/kha-white/manga-ocr) as the underlying OCR, which you can also use if you’re jujst looking for an OCR program without anything else.
3 comments
I’m using an open source MacOS app called bob (Github link shared down below) to do text extraction and translation, and it works great. However couple of things need to be aware:
– Even though the installation instructions are in English, but the app UI is all in Chinese kanji, took me a while to figure out how to get it to work.
– User will need to get Google API key in order to use the app (the ones come out-of-box does not work with Japanese)
In short, you’ll need to be pretty tech-savvy and have a lot of patient to make it to work. 😂 kanjitomo is an alternative, it’s easier to setup, but it does not work as well as bob.
https://www.kanjitomo.net/
https://github.com/ripperhe/Bob/blob/master/README.en.md
I haven’t tried a ton out but poricom works great
[mokuro](https://github.com/kha-white/mokuro) if you want to preprocess your manga and then have all text be scannable, [Poricom](https://github.com/bluaxees/Poricom) if you want a reader program that uses OCR.
Both use [manga-ocr](https://github.com/kha-white/manga-ocr) as the underlying OCR, which you can also use if you’re jujst looking for an OCR program without anything else.