Improving OCR using internal document redundancy

Open in new window