If you install tesseract OCR then OpenKM becomes with image OCR capabilities.
But I recommend always a better solution, make external documents scanning and OCR and the upload to OpenKM. I suggest for example some software like Omnipage pro ( really cheap and good OCR engine ). Always is good idea some human make some certification of scanning result. Some of our clients are happy with this cheap solution, sanning documents to pdf text that can be directly readed by openkm without any OCR need into.
Of course you can uploading images and OpenKM with teserract will indexing it, but I think you'll get better results with the first option I suggest. We're talking about 100-400 euros depending the ommnipage pro version. You can connect OpenKM as a webdav shared network resource and directly uploading files into. I think it's the most easiest scenario you can imagine for new documents.
If you got older scanned documents as images then you need first option. But I'm not pleased to give you bad news about it, some lower scanning resolution are not acceptable for any OCR, you need minimal image resolution to OCR engine gets some acceptable results. I've been looking some years scanned documents with really lower resolution that are not acceptable for any OCR engine, bad news.
If you're on a clean scenario or for uploading new documents, I suggest take in mind first option I suggested to you.
Tell us what you've done, we're interesting in results ( good or bad, your expertice on it, it'll be useful to other users and us ).
