Converting a very large (long) image to multiple pages

Question

I have a png file which is 1080 pixels wide and 100705 pixels long (not a typo) or if you prefer, it has a print size of 91.44 mm wide x 8526.36 mm long, which has some very sharply rendered text on it. I also have it in pdf format and in there it is one incredibly long page.

Every OCR program I have tried falls in a heap - generally the error has been some form of "Image too big".

What I ultimately want to do is to scrape all the text off the image. I'm guessing that I will have to chop the image or the pdf into pages first.

Does anyone have any suggestions for anything Linux based that will do the job? Happy to hear any alternative suggestions.

Thanks and Regards Nigel.