Automatic document classification- barcode and layout

Docs4ECM supports automatic document classification with barcode classification and automatic image layout classification for maximum process automation.

  • Classification of content based on barcode type, value and specified regular expression
  • Support for Barcode separation to separate batches of paper documents while scanning
  • Barcodes automatically populate indexing values eliminating the need for users to manually enter data, ultimately saving time
  • Layout classification based on image

Image processing and enhancements

After receiving images, Docs4ECM performs a range of image processing functions to improve the quality of document images for further recognition or archiving. Image processing technologies include: deskewing, rotation, distortion correction, text line straightening, splitting facing pages, adaptive binarization, and more.

  • Image scaling, cropping, line straightening, background removal etc..
  • Autodetection of page orientation (90, 180, and 270 degrees)
  • Automated image de-skewing (up to +/- 20 degrees)
  • Image despeckling (or image clean-up)

OCR of ID cards - MRZ and complete OCR (signature, user image)

Docs4ECM brings automation to scanning, recognition and extraction of the required data from ID cards, passports and driving licenses in Croatian and other languages.

  • Automatic document type recognition: ID, passport, license
  • Full MRZ data reading
  • Full ID reading: image, signature and other fields outside of MRZ

OCR of Payment orders - barcode and full OCR

Docs4ECM OCR system uses an advanced digital image processing filter, and optical character recognition post-processing methods in order to deliver best quality OCR results for further processing.

  • Dual stream capture to enhance OCR from B&W images while retaining color images for archiving.
  • Full barcode data reading and population
  • Pre-configured algorithms, type of data (name, surname, amount etc.) to validate extracted information
  • Support for dictionaries and lists of allowed terms
  • Post-processing to improve the standard of accuracy
  • Custom logic for text replacement in post-processing
  • Inline control before delivery OCR data (e.g. compare the readings to a database)