I am looking for OCR software to digitalize scans. Some are prints, some are hand-written (like this). I want to have an easier time finding documents. So I want to generate one text file for each image (of a few hundred images).
Online services are OK, but I prefer offline software. Python libraries are also OK.
Requirements
What I've tried
- http://www.onlineocr.net/: Does not work at all with handwriting, Does not recognize tables (tested with this file)
- http://www.free-ocr.com/: Same problem as http://onlineocr.net
- ocrfeeder: Crappy results (tested with the same image as before). As it uses tesseract internally, I don't think I have to try tesseract.
- GOCR: Same as ocrfeeder
- Tesseract:
tesseract morc-data/morc-3-1.jpg out.txt -l deu
- same as ocrfeeder