Capture with your existing setup
Use images from a branch scanner, kiosk adapter or mobile app. Confirm the exact device, driver and image format before assuming compatibility.
Check scanner integrationBANK CHECK OCR API
Reduce re-keying between cheque capture and operational review. Extract handwritten amounts, payee names and dates from scanner images or phone photos, then connect the fields and MICR details to your approval and reconciliation workflow.
Compare the extracted values and apply your workflow rules.
Illustrative scenarios with fictional data. These controls do not call the OCR API or approve a deposit.
CAPTURE TO DECISION
A readable image is only the start. Teams still need to check amounts, resolve uncertain fields and connect the result to an operational record. Define each hand-off before changing the capture process.
Use images from a branch scanner, kiosk adapter or mobile app. Confirm the exact device, driver and image format before assuming compatibility.
Check scanner integrationRead printed and handwritten cheque details. Keep the source image available alongside the extracted values and any device MICR read.
Try the online OCR toolRoute missing or conflicting values to a reviewer. Your application or the ChequeDB management workflow handles corrections, approvals and follow-up.
Explore review and approvalsThe API supplies extraction for the agreed integration. Scanner control, review permissions, bank acceptance and downstream posting each need an explicit owner.
WORKED EXAMPLE
Imagine a cheque whose numeric box says 1,750.00 while its written amount says one thousand two hundred fifty. Both fields may be read correctly; the cheque still contains a disagreement.
| Stage | Evidence to retain | Next step |
|---|---|---|
| Capture | The source image and a stable item reference, such as SAMPLE-0042. | Associate the image with the processing attempt. |
| Extract | Numeric amount: 1,750.00. Written amount: one thousand two hundred fifty. | Compare the amounts before passing a value downstream. |
| Review | The reviewer sees both reads beside the source image. | If the image confirms the conflict, keep it unresolved and follow the bank’s exception process. |
| Record the decision | Reviewer, time, reason and supporting evidence. | Keep the item on hold until the authorised decision is recorded. |
| Reconcile later | The separately confirmed posting or bank reference and status. | Do not treat OCR success or internal approval as proof of deposit or settlement. |
Illustrative workflow using fictional data, not a live API response or a measured result. If OCR misreads a clear image, record the original value, correction and reason. A genuine disagreement on the cheque needs a different decision.
EXTRACTED FIELDS
Get cheque data as named fields in a JSON response. Validate field coverage on your bank layouts, languages and capture conditions.
| Cheque field | JSON field | What it captures |
|---|---|---|
| Amount in numerals | amount_in_numerals | The handwritten or printed number in the amount box. |
| Amount in words | amount_in_words | The amount written across the cheque, including handwritten text. |
| Payee | payee | The person or business the cheque is payable to. |
| Date | date | The date written on the cheque. |
| Bank | bank_name | The issuing bank’s name. |
| Branch | branch | The bank branch shown on the cheque. |
| Cheque number | cheque_number | The cheque’s identifying number. |
| Drawer | drawer_name | The name of the person or business issuing the cheque. |
MICR DATA CAPTURE
Read MICR characters from the cheque image or use a magnetic read from a compatible scanner. Confirm the clearing format and enabled MICR output for your deployment.
Identify bank routing information, such as a US routing number or a UK sort code. Preserve the format required by the applicable clearing scheme.
Capture the account identifier as text and preserve its leading zeros when mapping it to your records.
Capture the cheque or serial number and associate it with the source image and workflow record.
MICR fields and their order vary by country. Agree their response mapping separately from the eight OCR fields shown above.
YOUR CHEQUES. YOUR APPLICATION.
Use phone photos or scanned images as the input. Confirm the bank layouts, languages and fields needed for your deployment.
Connect images from your banking app, upload form or scanner adapter. Test glare, faint ink and skew as well as clean captures.
Evaluate the cheque templates, currencies, languages and clearing formats used by your customers. Confirm the fields your integration needs.
Discuss the processing environment that fits your organisation. Agree network access, data retention and operating responsibilities with our team.
PLAN YOUR PILOT
Start with one capture source and one downstream workflow. Define the test set and acceptance criteria together, then review the results with both engineering and operations.
Bring representative, de-identified images, bank layouts, languages and capture details. Agree required fields, hosting, access and retention before sharing documents.
Ask for field-level results, wrong values accepted without review, unresolved items and response times. Time the reviewer’s work as well as extraction.
Use the benchmark scorecardReview the field mapping, exception decisions and sample downstream record. Record the remaining integration work and the criteria for expanding the pilot.
Plan a cheque OCR pilotPilot scope and deliverables are agreed with your team. API documentation, authentication, limits and enabled components depend on the provisioned integration.
QUESTIONS & ANSWERS
Yes. The OCR response includes the amount in words and the amount in numerals, as well as payee, date, bank, branch, cheque number and drawer. Recognition depends on handwriting, layout, language and image quality, so evaluate the fields on your own cheque samples.
Upload an image to the online cheque OCR and date tool, or contact our team to evaluate the API with your documents. The online tool uses server processing and applies your selected date rule. API access and limits are provisioned for your deployment.
MICR reading concerns the control line, which can contain bank routing, account, cheque number and other scheme-specific information, including encoded amounts in some cases. The visible payee, written amounts and date require separate extraction.
Run an evaluation using your bank layouts, languages and capture devices. Measure correct reads for each field, items needing review and response times at your expected volume. Our team can help define the evaluation criteria for your deployment.
Images from an existing capture setup can be evaluated with the OCR API. Confirm the exact scanner, driver or vendor SDK, operating system and image output. Device control and compatibility are separate from extracting fields from the captured image.
Agree a representative sample set, field-level scorecard, exception walkthrough and downstream field mapping. Include incorrect reads, unresolved items, reviewer effort and response times. Record any remaining integration work before deciding whether to expand the pilot.
OCR supplies the extracted fields and configured checks. Your bank’s systems determine eligibility, authenticity checks, acceptance, clearing and settlement.
BUILD WITH CHEQUEDB
Bring your capture setup, required fields and review rules. Agree a pilot that measures correct reads, unresolved items and the evidence your team needs to make a decision.