The Common OCR Service Interface. COSI is an API that allows developpers to easily bring OCR (Optical Character Recognition) capabilities to imageprocessing applications. COSI supports existing OCR tools such as Tesseract, GOCR or GNU Ocrad.
Tifftool is a high-performance tool to clean scanned documents in preparation for onscreen display or for OCR. Features include skew correction, orientation correction, despeckle, page alignment, split pages and batch processing.