>>42383239If you're patient and enjoy tedious work you can separate the pdf into individual pages and run them through an OCR one sentence at a time and fix the errors, then throw them into a machine translator. If you just toss the pdf in all at once you get weird errors like this.