CuneiDjVu is a graphical frontend to a set of the Windows console utilities providing the DjVu OCR capability based on the CuneiForm-Linux OCR Engine
- This software creates the OCR layer in DjVu files.
Fed it a 512 page 7.18MB djvu file, all b&w except for colour covers, created in DjVuToy using lossless cormpression. Result screen after a few moments: "Processing failed to finish. No pages in DjVu were processed." DjVu file opens and reads images fine in various DjVu readers including multipurpose ones like Sumatra. iirc from some searches a while ago, CuneiForm has a propensity to crash although I've seen it stated that its recognition is better than tesseract. Having experienced the ouput of tesseract with tiffdjvuocr I think CuneiForm could do no worse since in my case tesseract ouput 75% garbage when a ten year old version of Acrobat produced near 100% correct text layer ocr from the same tifs. Anyway, that said, consider an update to use the latest cuneiform with maybe a user option to use tesseract?