Capture cheques from desktop scanners, mobile devices and self-service machines. Extract and validate every field, control exceptions, then post straight into core banking, clearing and ERP workflows.
One controlled pipeline
Channel independentDesktop cheque scanners
Mobile cheque capture
Self-service scanners
Cheque processing workflow
Capture · extract · validate · decide · audit
More than cheque OCR
OCR is one step. A production cheque processing system also has to reject bad images, compare fields, prevent duplicate deposits, manage corrections, survive downstream outages and prove what happened to every item.
This page covers the end-to-end orchestration around capture, validation, exceptions, posting, and reconciliation. For the field-reading and MICR boundary, see the cheque OCR and validation workflow.
Run the same cheque workflow across teller counters, branches, remote business customers, back-office teams and self-service deposits. Keep channel rules configurable while sharing extraction, risk checks, exception handling and core banking integration.
Capture incoming cheques close to where they are received, then send validated data into treasury, accounts receivable and ERP processes. Match deposits to customers and invoices without retyping the cheque at every handoff.
Native capture support
Each channel has different hardware and user needs. The item that reaches your operation should still follow one set of data, risk and audit rules.
Teller, branch, lockbox and back-office capture with scanner lifecycle and image-quality controls.
Front-and-back image capture for business banking, field teams and remote deposit workflows.
Cheque deposits through kiosks and cash deposit machines using the same processing pipeline.
End-to-end workflow
The workflow is designed around safe straight-through processing: automate clean items, isolate uncertainty and never lose the chain of evidence.
The capture layer collects front and back images, associates them with the device, user and deposit, and checks crop, orientation, contrast and image completeness before the item enters processing.
MICR, printed OCR and handwritten-field recognition run as separate reads. Payee, date, legal amount, courtesy amount, account data and cheque number return as structured fields with confidence scores.
The workflow compares the amount in words with the numeric amount, applies stale and post-dated rules, checks the cheque number, detects duplicates and routes uncertain results for review.
Clean items move to clearing, core banking, ERP or treasury systems. Item IDs, acknowledgements and retry controls keep downstream failures from creating duplicate postings.
Operations can follow each cheque from capture through correction, approval, posting, rejection, return and archive, with the original values and every manual decision preserved.
Use APIs for account validation, posting and status; controlled files for bank and clearing exchange; and events for audit, analytics and operational monitoring. The integration model follows the destination instead of forcing every system through one brittle interface.
These are integration patterns, not a fixed endpoint or message schema. The item identity, acknowledgements, retry policy, and field mapping are agreed with the bank, clearing, ERP, or treasury system during implementation.
Integration patterns
Each interface carries a stable item identity so acknowledgements, retries and exceptions remain traceable across systems.
Illustrative pattern: send the extracted account and amount context to the bank service, keep its response with the cheque item, and route a rejection to repair. The exact fields and interface follow the target system.
Illustrative pattern: match a validated item to a customer, invoice, or deposit record, then preserve the source item identity and downstream response. An unmatched item waits in an accounts-receivable queue.
Illustrative pattern: send approved items through the bank’s accepted API, file, or event interface, capture item-level acknowledgements, and repair rejected items without losing the original decision history.
The safest automation is selective. High-confidence items can move quickly while uncertain, high-value or rejected items wait in the right queue with the evidence an operator needs to decide.
See the MICR, OCR, handwriting, validation and exception workflow.
Learn moreConnect cheque deposits with the wider branch and self-service channel.
Learn moreRead the detailed capture-to-posting pipeline for banks and corporates.
Learn more