Skip to main content
BankingGovernment

Printed and handwritten cheque OCR with MICR validation and controlled review

Extract cheque fields, validate the MICR line and route uncertain items for review at self-service kiosks and desktop scanners.

OCR is one stage of cheque processing. The Azimut SDK can read printed and handwritten fields, validate the MICR line and keep uncertain or inconsistent items in a review path before the application hands them to clearing or core banking. Recognition results depend on the field, image quality, cheque stock, capture channel, and configured thresholds, so accuracy should be measured on a representative deployment corpus rather than treated as a universal percentage.

How it works

How the cheque ocr & fraud detection flow works

The SDK handles supported device operations. Your application coordinates the workflow with the services responsible for identity, payment and account decisions.

1

Image capture — kiosk or desktop scanner

The SDK accepts cheque images from CDM-embedded scanners at self-service kiosks and from desktop document scanners at teller or back-office workstations. Front and rear images, image-quality checks, and capture metadata depend on the connected hardware and channel.

2

MICR line reading and validation

The SDK reads MICR data from the supported scanner. Confirm the E-13B or CMC-7 format, the fields available for the cheque layout and the validation rules configured for the deployment. Cross-field checks can compare the captured values where the required data is available.

3

OCR field extraction — printed and handwritten

The processing path can extract payee, date, written amount, and numeric amount from printed and handwritten cheque fields. Where the configured OCR integration supplies field-level confidence, agree how those scores inform review thresholds. Missing required fields or failed checks also need an explicit review outcome. Results should be benchmarked by field on representative cheque images.

4

Cross-field validation

The workflow compares related values such as the written and numeric amounts, cheque number and date, then applies the rules agreed for the deployment. A mismatch can be held for review instead of being passed downstream as a clean item.

5

Image, duplicate, and risk checks

Image-quality failures, duplicate indicators, amount discrepancies, and other configured risk rules can produce discrete results for the application or an exception queue. The exact checks depend on the capture hardware, connected services, and operating policy.

6

Clearing and core banking posting

Approved results can be exchanged with clearing, core banking, ERP, or reconciliation systems through the interface agreed for the deployment. A rejected or unavailable downstream system should leave the item in a traceable repair state so it can be resolved without an ambiguous re-submission.

What the SDK extracts from each cheque

Cheque data extraction returns the configured fields as structured data. Confirm which confidence scores, if any, the chosen processing integration supplies. The same processing boundary can serve bank check images from a kiosk-embedded scanner or a desktop document scanner, subject to the connected capture integration.

Payee name

The pay-to name, extracted from printed or handwritten cheques for downstream validation.

Date

Cheque date read and normalised, with stale-dated and post-dated cheques flagged for the application to act on.

Legal amount (written)

The amount written in words, validated against the numeric amount where that rule is configured.

Courtesy amount (numeric)

The numeric amount, cross-checked against the legal amount so written/numeric mismatches surface before clearing.

MICR line

The configured MICR fields, such as account, routing or sort data, and cheque serial, read and validated for the deployment’s cheque format.

Image and processing status

Capture quality, validation outcomes and exception state inform the application’s decision for each item. Available results depend on the configured integration.

Cheque OCR integration

Read the fields. Act on the validation result.

The scanner integration returns captured cheque data and can include a separate OCR result. In that result, content holds the extracted fields and validations holds the configured checks. Your application must inspect both before deciding whether an item can continue. OCR may be disabled or return no result. A successful scan alone does not approve a deposit.

Example OCR response
{
  "chequeOcrResponse": {
    "content": {
      "payee": "Example Supplies",
      "amountInWords": "One thousand two hundred fifty",
      "amountInNumerals": "1520.00"
    },
    "validations": {
      "amount": {
        "error": true,
        "errorMessage": "Written and numeric amounts differ"
      }
    }
  }
}

Follow the item into review

In this example, the written amount is 1,250 and the numeric amount is 1,520. The amount check reports an error. The application should hold the item for review, retain both extracted values and let an authorised operator compare them with the source image. Choosing one amount automatically would hide the discrepancy.

Keep the capture reference, original values, review decision and downstream acknowledgement associated with the same cheque item. Before connecting the workflow, confirm which fields and checks your scanner and OCR configuration return, how absent results are represented, and which outcomes require review. Treat a missing check as incomplete evidence: an error flag of false alone does not establish that a check ran. This response fragment does not define an HTTP endpoint or a posting instruction.

Confidence scores are a separate capability to confirm. A validation error flag or a scanner quality score is not a calibrated probability that an OCR field is correct.

Evaluate the cheques your operation receives

Agree a labelled sample across the cheque formats, handwriting, scanners and image conditions you expect. Keep a separate set for acceptance testing after configuration. Compare each required field with its verified value, and report incorrect automatic accepts, items sent for review and capture failures alongside correct reads. Separate printed and handwritten results, and retain the sample count for each group.

Include amount mismatches, missing fields, unreadable MICR and incomplete images. Agree the expected outcome before the run so that a rejected image or an item correctly held for review is distinguishable from a wrong value sent downstream. Record processing time separately from customer positioning and operator review time.

Plan the capture-to-posting integration evaluation

Request a cheque integration walkthroughBring your scanner model, operating system, required fields and target banking or ERP interface. Agree a secure sample-transfer process with the team before sharing cheque images or account data.

What the SDK handles

Hardware

  • Kiosk-embedded cheque scanner (front/back where supported)
  • Desktop document scanner (teller and back-office)
  • MICR reader (E-13B and CMC-7)
  • Cash acceptor / CDM
  • Receipt printer

Integrations

  • Core banking APIs
  • Cheque clearing networks (MICR E-13B / CMC-7)
  • Image-quality and validation services
  • Exception and review workflows
  • Fraud and duplicate-risk services where connected
  • 1LINK and RAAST (Pakistan)

Customers

Deployed with

Production deployments running cheque ocr & fraud detection flows in real environments.

Bank Alfalah
BankingPakistan

Bank Alfalah

Cash and cheque deposits at Digital Branch kiosks. OCR extraction and clearing integrated through the SDK.

Bank Al Habib
BankingPakistan

Bank Al Habib

Live production cheque deposits via CDMs, with active operational support for clearing and exception handling.

Diamond Trust Bank
BankingKenya

Diamond Trust Bank

Cheque deposit flows at self-service kiosks. SDK handles scanning, data extraction, and core banking posting.

Banque Atlantique
BankingWest Africa

Banque Atlantique

Cash and cheque deposits via CDMs across the West African network.

Related solutions

From the blog

Frequently asked questions

There is no single accuracy figure that applies to every cheque field or deployment. Handwriting, cheque stock, image quality, capture channel, and thresholds all affect results. The safe evaluation is a representative field-level test that records correct reads, incorrect automatic accepts, review rate, capture failures and downstream outcomes, plus confidence where available; uncertain reads should be routed to review.

The configured extraction can include payee name, date, written (legal) amount, numeric (courtesy) amount, and MICR fields, each with confidence where supported. The deployment can also return capture quality and validation outcomes for review and downstream decisions.

The same processing boundary can serve CDM-embedded scanners at self-service kiosks and desktop document scanners at teller and back-office workstations, subject to the supported capture integration and the image data it provides.

The integration can read MICR fields and apply the checks configured for the deployment. Confirm scanner support for the required E-13B or CMC-7 format, field mapping for the cheque layout and which cross-field checks are available. Missing data and failed checks need an agreed review outcome.

The workflow can expose configurable checks such as written-versus-numeric amount discrepancies, image-quality failures, duplicate indicators, and downstream risk rules. Each result can be sent to the application or an exception queue; the exact controls depend on the deployment and connected services.

Yes — "cheque" and "check" are regional spellings for the same instrument. The capture format, MICR configuration, validation rules, and clearing interface still need to match the jurisdiction and target bank or network.

Ready to deploy cheque ocr & fraud detection on your kiosks?

Talk to our team about your hardware, integration requirements, and deployment timeline.

Get in Touch