Solutions
OCR
Verification Infrastructure for KR Data.
Korean-data-native OCR for identity verification, paper voucher classification, and kiosk workflows.
ComFac's OCR Engine
A self-developed OCR engine captures, classifies, extracts, and protects Korean document data for connected services.
Recognition pipeline
- 01
Auto focus and capture
Guide the camera, find the document boundary, and capture a usable image.
- 02
Image classification
Identify the document type and isolate the fields required by the service.
- 03
Detection and machine learning
Extract text and return structured identity or voucher data.
- 04
Masking and protected output
Mask sensitive fields and deliver only the data the workflow needs.
High recognition performance
Models interpret document layouts and field patterns.
Broad extensibility
The engine adapts to document types and operating environments.
Automation support
Structured results move directly into the next service action.
Designed for real operating environments
One OCR layer supports mobile identity checks, automated voucher handling, and unattended kiosks.
Identity verification
Capture Korean ID cards on mobile and return protected verification data.
Paper voucher sorting
Classify paper vouchers through a dedicated recognition engine.
Kiosk ID recognition
Connect Windows kiosks and webcams to identity recognition workflows.
From guided capture to protected response
Client, application server, and OCR engine each handle a defined part of the recognition and verification flow.
Client application
- 01Select document type
- 02Guide camera and capture
- 03Check image quality
- 04Send image and consent
Application server
- 01Receive image and metadata
- 02Parse and store request
- 03Send recognition request
- 04Validate and store result
OCR engine
- 01Receive protected image
- 02Extract text with ML
- 03Mask sensitive fields
- 04Return structured result
Use one OCR engine across identity checks, voucher processing, and unattended terminals.