A scan is an image before it is information
To a search engine, a scanned page may initially be pixels rather than words. OCR analyses those pixels and produces text that can support search, indexing and classification.
OCR is not infallible
Low resolution, skew, handwriting, unusual layouts, damaged originals and poor contrast can reduce accuracy. Important extracted values should be checked when errors would carry operational consequences.
OCR becomes more useful inside a workflow
Extracted text can feed search, metadata suggestions and classification, while the original document remains the source record. That combination is more useful than treating OCR output as a replacement for the original.
Explore OCR in Lumen.
LUMEN DOCUMENT MANAGEMENT
Put structure, search and governance around business records.
Lumen combines document capture, OCR, metadata, permissions, approvals, audit history and assistive AI capabilities.
Explore Lumen