What Is MICR Technology? Magnetic Ink Character Recognition for Cheque Processing

MICR stands for Magnetic Ink Character Recognition. It reads the encoded E-13B or CMC-7 control line on a cheque, typically supplying routing, account and cheque-number identifiers for automated processing. MICR does not read the payee, date, signature or written amount; OCR and ICR handle visible printed and handwritten fields.

Reviewed by the Chequedb product team on 3 September 2026 against the standards and first-party sources linked below.

Chequedb uses MICR as one control input in a wider cheque workflow. The platform normalises the MICR read, adds confidence and mismatch signals, combines it with OCR and ICR results, and routes uncertain items for review. For the production workflow, see Chequedb MICR reader software for cheque processing.

This page explains how MICR technology works, the main font standards, what the MICR line commonly carries, and where MICR fits in a modern cheque-processing pipeline.

What Does MICR Stand For?

MICR stands for Magnetic Ink Character Recognition. The phrase describes the original physical process:

• Magnetic: the printed material contains particles that produce a magnetic signal when magnetised
• Ink: the characters are printed using magnetic ink or toner
• Character: each permitted character has a specified shape
• Recognition: a reader identifies the character from its signal or, in an optical workflow, from its image

The ASC X9 description of X9.100-20 specifies E-13B character shapes, dimensions, magnetic signal levels and tolerances. The ASC X9 technical glossary describes MICR as a machine-language specification for the paper-based payment system using magnetic-ink E-13B characters.

How MICR Technology Works

In magnetic reading, E-13B or CMC-7 characters pass through a magnetising field and then over a read head. Each valid character produces a signal shaped by its geometry and magnetic-ink distribution. Reader software compares that signal with the permitted character patterns and returns the decoded control-line data.

Signal quality can be affected by printing, alignment, document condition, contamination, transport and the reader itself. A production workflow should therefore retain confidence or reject information and route uncertain reads for review instead of assuming that every decoded character is correct.

Optical MICR reads the printed MICR characters from an image. It can be useful when only an image is available, but it observes visual shape rather than magnetic properties. A magnetic-versus-optical disagreement can be a useful review signal; it is not, by itself, proof of alteration or fraud.

This is also why MICR and OCR are not interchangeable. MICR targets a tightly specified control line. The Library of Congress definition of OCR describes optical character recognition as converting visual images of characters into computer-readable characters.

MICR Font Standards: E-13B and CMC-7

Two major MICR character families appear in cheque-processing systems. The applicable font and field layout depend on the jurisdiction, clearing scheme and issuing institution.

E-13B

E-13B includes ten digits and four special symbols. ASC X9.100-20 covers the U.S. print and test requirements for the character shapes and magnetic signal. ISO 1004-1:2013 is the ISO print specification for E13B.

The special symbols delimit scheme-specific parts of the control line, such as transit, on-us and amount fields. Teams should parse the line according to the relevant clearing rules rather than assume one universal field order.

CMC-7

CMC-7 characters use vertical strokes with defined shapes, dimensions and tolerances. ISO 1004-2:2013 specifies ten digits, five symbols and 26 letters for magnetic-ink recognition. Its use and field layout also depend on the relevant national or clearing-system rules.

Font detection

MICR software can classify an observed line as E-13B or CMC-7 and parse the permitted symbols for that scheme. Detection should still expose uncertainty: a low-quality image, weak magnetic signal, damaged character or unexpected layout can require manual review.

What the MICR Line Contains

A cheque MICR line commonly carries three core identifiers:

1. Routing, transit or institution identifier: identifies the paying institution or clearing route according to the scheme.
2. Account identifier: identifies the account associated with the item.
3. Cheque or serial number: identifies the particular item within the account's cheque sequence.

Some schemes or processing stages also encode other control information, including transaction codes or an amount field. Field order, permitted length and delimiter use are not universal. In the United States, the Federal Reserve Banks' Operating Circular 3 addresses MICR printing, routing numbers and records associated with cash items.

What the MICR line does not contain

The control line does not normally contain:

• The payee name
• The cheque date
• The written or numeric amount as shown on the face, unless a scheme or processing stage separately encodes an amount field
• The signature
• The memo or remittance reference
• Front or back image evidence

Reading the MICR line tells a processing system how to identify and route the item. It does not establish that the visible amount is correct, the payee is authorised, the date is acceptable, the signature is genuine or the item should post automatically.

MICR, OCR and ICR: Different Responsibilities

MICR is purpose-built control-line recognition. OCR and ICR are image-recognition methods used for the visible document:

• MICR reads E-13B or CMC-7 control data from a magnetic signal or an image of the MICR line.
• OCR converts visible machine-printed characters into machine-readable text.
• ICR recognises hand-printed or cursive characters, with results that depend on handwriting, field constraints and image quality.

IBM's Datacap text-recognition guidance distinguishes OCR for machine-printed text from ICR for hand-printed or cursive text and recommends validation rather than assuming recognition is perfect.

The useful architecture is therefore not “MICR versus OCR”. It is MICR plus the appropriate image-recognition layers, followed by field validation, discrepancy checks and controlled review.

Magnetic and Optical MICR Reads

MICR control-line data can reach software through two related paths:

1. Magnetic reading: a dedicated read head detects the magnetic signal from the printed characters.
2. Optical reading: software recognises the MICR character shapes from the cheque image.

Where both are available, comparing the results can surface a damaged line, encoding error, image problem or possible alteration. The correct operational response is to produce a reason code and review the evidence. A mismatch is a signal to investigate, not a conclusive fraud finding.

The ASC X9 advisory on magnetic ink explains that ANSI X9.100-20 remains the relevant magnetic-ink standard for U.S. checks. The Federal Reserve circular likewise references the adopted ANSI standard for MICR placement and location.

Why MICR Alone Is Not Enough

A syntactically valid MICR line does not prove that:

• The visible payee has not changed
• The written and numeric amounts agree
• The cheque date satisfies the applicable policy
• The signature is genuine
• The item has not already been presented through another channel
• The cheque image and control record belong to the same physical item

Those questions require other controls: OCR and ICR for visible fields, amount comparison, image and document review, duplicate detection, positive-pay matching, date-policy rules and, where appropriate, human approval.

MICR should be treated as a control input, not the entire cheque decision. Chequedb's role is to combine that input with the rest of the document and preserve how exceptions were reviewed.

Where MICR Software Fits in the Pipeline

A MICR reader performs the magnetic or optical read. MICR software turns the observed characters into structured data and controlled workflow steps:

1. Capture: receive scanner images, a magnetic MICR read, or both.
2. Classify and parse: identify E-13B or CMC-7, locate delimiters and extract scheme-specific fields.
3. Normalise: map routing, account and cheque identifiers into consistent output fields without discarding the source read.
4. Validate: apply format checks, reference-data checks and magnetic-versus-optical comparisons where those inputs exist.
5. Score and route: attach confidence and reason codes, then send uncertain or conflicting items to a review queue.
6. Record the decision: retain the source values, validation results, reviewer action and downstream outcome for reconciliation and audit review.

Chequedb's MICR reader software and exception workflow owns this product intent. This article remains the definition and standards reference.

MICR SDK Integration

For teams integrating supported cheque scanners, an SDK may handle:

• Scanner initialisation and device communication through interfaces such as TWAIN or ISIS
• Front and back image capture
• Transfer of a magnetic MICR read when the device provides one
• MICR-line localisation and optical recognition from an image
• Font classification and delimiter parsing
• Image-quality and confidence signals

Hardware capabilities differ. A scanner that only returns images cannot supply a magnetic signal, and software should not label an optical-only result as a magnetic read. Integration testing should cover real documents, degraded samples, unexpected layouts and the manual-review path.

Standards and Technical Sources

• ASC X9 — X9.100-20 and X9.100-160-1 descriptions for E-13B character, signal, placement and printing requirements
• ASC X9 TR 100 glossary for financial-services MICR terminology
• ISO 1004-1:2013 for E13B print specifications
• ISO 1004-2:2013 for CMC7 print specifications
• Federal Reserve Banks Operating Circular 3 for U.S. check collection and MICR context
• Library of Congress OCR Data for a concise OCR definition
• IBM Datacap text-recognition guidance for the OCR/ICR distinction and recognition caveats

These sources describe standards and technical concepts. The operational rules for a particular cheque depend on the applicable institution, jurisdiction and clearing scheme.

Summary

• MICR means Magnetic Ink Character Recognition.
• It reads a specified E-13B or CMC-7 control line used for routing, account and cheque identifiers.
• OCR reads visible machine-printed characters; ICR handles handwriting.
• No recognition layer is a complete cheque-validation decision on its own.
• Magnetic and optical disagreements are review signals, not proof of fraud.
• Chequedb combines MICR, OCR and ICR outputs with validation, exception routing and an audit-ready decision history.

For the production implementation, see Chequedb MICR reader software for cheque processing. For the broader extraction workflow, see cheque data extraction with OCR, MICR and audit-ready validation.