On-Premises Cheque OCR

Read printed and handwritten cheques inside your own infrastructure

Run OCR, handwriting recognition, validation, search, and review on infrastructure you control. Turn captured cheque images into structured, searchable records without making public-cloud OCR part of the processing path.

Book Demo See What Chequedb Reads

Printed fields use OCR. Handwritten fields use cheque-specific handwriting recognition. MICR handles the codeline. Confidence and review connect the results.

Clear system boundary

Scanner software captures the image. Chequedb creates the record.

Keep the capture layer that fits your scanner estate. Chequedb reads the image, cross-checks the fields, routes uncertainty, and stores the result with audit history.

Scanner middleware captures cheque images, Chequedb recognises and validates the fields, and business systems use searchable audit-ready records.
1

Capture layer

OEM APIs, TWAIN, scanner middleware, mobile capture, or image import provide the front and rear cheque images.

2

Cheque intelligence layer

Printed OCR, handwriting recognition, MICR, amount reconciliation, confidence scoring, and image-quality checks create structured fields.

3

Operations layer

Search, review, approvals, corrections, ERP hand-off, and audit evidence turn extracted fields into controlled work.

Printed and handwritten fields

Handwriting is a first-class input, not an edge case

Printed text is the straightforward part. Cheques also carry cursive payee names, handwritten dates, numeric amounts, and legal amounts in words. Chequedb applies field-specific recognition and returns confidence for each result.

Cheque fields mapped to printed OCR, handwriting recognition, and MICR reading.

Printed OCR

Read machine-printed bank names, account-holder details, reference numbers, and dates.

Handwriting recognition

Read payee names, dates, memos, amounts in figures, and amounts written in words.

Amount reconciliation

Compare the courtesy amount with the legal amount and flag disagreement for review.

Confidence-led review

Show the image crop, extracted value, confidence, and reason whenever a field needs review.
Learn how handwritten cheque OCR works
Searchable cheque records

Search the content of every cheque

Image folders hold files. Chequedb indexes what is on them. Search by payee, amount, date, cheque number, account detail, status, exception, linked invoice, or reviewer action.

The record keeps the source image, extracted fields, corrections, and workflow history together.

Explore Searchable Archive

Example record searches

Search extracted fields and workflow state

Handwritten payee contains "Summit Supply"
Amount above 500,000 and still not deposited
Written and numeric amounts do not match
Low-confidence payee corrected after approval
Customer-controlled deployment

Keep recognition, records, and review under your control

Deploy cheque-processing services and data stores inside infrastructure your organisation controls. Define the network boundary, storage location, user access, retention, integration paths, and software update policy for the deployment.

Customer-controlled processing and storage
Local or private-network API access
Identity, ERP, archive, and monitoring integration
Controlled software and model updates

Exact topology, connectivity, and operational responsibilities are agreed for each deployment.

Your infrastructure boundary

Processing and records stay together

OCR and handwriting
Structured records
Search and review
Customer storage
Private API and approved internal integrations
Processing flow

From captured image to audit-ready record

Each stage keeps the original evidence connected to the extracted value and the final decision.

Customer-controlled cheque OCR pipeline from image enhancement through recognition, validation, review, and searchable storage.
1. Capture

Accept front and rear images from a scanner, camera, or image-import workflow.

2. Read

Apply MICR, printed OCR, and handwriting recognition to the appropriate fields.

3. Validate

Check image quality, confidence, amounts, dates, and field agreement.

4. Review

Route uncertain fields with the relevant image crop and reason code.

5. Search and integrate

Store the resolved record and send approved data to internal systems.

Frequently Asked Questions

Can Chequedb read printed and handwritten cheques?

Yes. Chequedb applies OCR to printed text, cheque-specific handwriting recognition to handwritten fields, and MICR methods to the codeline. Each field carries a confidence value, and low-confidence values can be routed for review.

What remains on premises?

Cheque recognition, structured records, search, and review can run inside customer-controlled infrastructure. Network, storage, integration, support, and update boundaries are documented for each deployment.

Does Chequedb replace scanner middleware?

Not necessarily. Scanner middleware and OEM APIs manage capture hardware. Chequedb sits above the capture layer, reads the image, validates fields, and creates the searchable cheque record. Review cheque scanning software for the wider capture workflow.

What can users search?

Users can search indexed cheque fields and workflow context, including payee, amount, date, cheque number, account details, status, exceptions, corrections, approvals, and linked accounting records. See the searchable cheque archive for examples.

What happens when handwriting is unclear?

Low-confidence fields go to review with the original image crop, extracted value, confidence, and reason. Corrections are recorded in the audit history instead of replacing the original result without explanation.

Keep cheque intelligence inside your infrastructure

See Chequedb read printed and handwritten cheques, create searchable records, and route uncertain fields without depending on public-cloud OCR.

Book an On-Premises Demo Explore the Bank Check OCR API