Iris Recognition Glossary

A comprehensive A-Z reference of 46+ terms used in iris recognition, biometric security, and identity verification. Ideal for engineers, procurement teams, and system integrators evaluating iris biometric solutions.

A

Algorithm: A set of mathematical rules that processes iris image data to extract features, encode templates, and perform matching. HOMSH's PhaseIris algorithm uses phase-based feature extraction to achieve FAR below 10^-7.
Authentication: The process of verifying a person's claimed identity. In iris biometrics, authentication is a 1:1 comparison between a live iris scan and a stored template linked to a credential such as an ID card or PIN.

B

Binocular Iris Camera: A dual-camera module that captures both left and right irises simultaneously, improving recognition accuracy and speed. HOMSH's D30 and D60 terminals use binocular iris cameras for dual-eye enrollment and matching.
Biometric Template: A compact digital representation of a biometric feature (iris, fingerprint, face) extracted by an algorithm. Iris templates (IrisCodes) are typically 256-512 bytes and cannot be reverse-engineered into the original image.

C

Capture Distance: The range of distances at which an iris camera can acquire a usable image. Standard modules operate at 15-30cm. HOMSH's advanced terminals support capture distances up to 100cm with auto-focus optics.
CE Certification: European conformity marking indicating that a product meets EU health, safety, and environmental protection standards. All HOMSH iris recognition products carry CE certification for sale in the European Economic Area.
Collarette: A zigzag boundary line in the iris that separates the pupillary zone (inner) from the ciliary zone (outer). The collarette is one of many unique structural features used by iris recognition algorithms for identification.
Crypts: Diamond-shaped openings or depressions in the iris stroma visible as dark spots in NIR imaging. Crypts are among the most distinctive and stable features used for iris pattern encoding.

D

Daugman, John: The Cambridge University professor who invented the IrisCode algorithm in the 1990s, establishing the mathematical foundation for modern iris recognition. His work on Gabor wavelet-based phase encoding remains the basis of most commercial iris recognition systems, including HOMSH's PhaseIris algorithm.
Dual-Eye Enrollment: The practice of enrolling both left and right irises during registration. Dual-eye enrollment improves system reliability because if one eye cannot be captured (due to injury, disease, or obstruction), the other eye serves as a backup.

E

Encoding: The process of converting raw iris texture patterns into a binary template (IrisCode) using mathematical transforms such as Gabor filters. Encoding reduces a high-resolution iris image to a compact, matchable representation.
Enrollment: The initial registration process where a person's iris image is captured, processed, and stored as a template in the system database. HOMSH devices complete enrollment in 2-5 seconds, with the MC21 module achieving 0.7-second enrollment.

F

FAR (False Accept Rate): The probability that the system incorrectly accepts an unauthorized person as a match. Expressed as a percentage or ratio (e.g., 10^-7 means 1 false accept per 10 million attempts). Lower FAR indicates higher security. HOMSH systems achieve FAR below 10^-7.
FCC Certification: Federal Communications Commission certification required for electronic devices sold in the United States, ensuring electromagnetic compatibility and limited radio frequency interference. All HOMSH products carry FCC certification.
FPGA (Field-Programmable Gate Array): A reconfigurable hardware chip that executes algorithms at the silicon level, far faster than software on a CPU. HOMSH's Qianxin FPGA chip performs 1:N iris matching at over 1 million comparisons per second with lower power consumption than GPU or cloud-based solutions.
FRR (False Reject Rate): The probability that the system incorrectly rejects an authorized person. FAR and FRR are inversely related — increasing security (lower FAR) typically increases FRR. HOMSH's PhaseIris algorithm achieves FRR below 0.5% while maintaining FAR below 10^-7.
Furrows: Radial and concentric grooves or folds in the iris stroma that create unique patterns visible under NIR illumination. Furrows are among the stable features used by iris recognition algorithms for template generation.

G

Gabor Filter: A mathematical filter used to extract local phase information from iris texture at multiple frequencies and orientations. Gabor filtering is the core feature extraction method in most iris recognition algorithms, including John Daugman's original IrisCode and HOMSH's PhaseIris.
GB/T20979-2019: China's national standard for iris recognition systems, specifying performance requirements for FAR, FRR, capture time, and environmental conditions. HOMSH's iris recognition products are certified to this standard.

H

Hamming Distance: A measure of the difference between two binary IrisCodes, calculated as the proportion of bits that differ. A Hamming distance near 0 indicates a match (same person); a distance near 0.5 indicates non-match (different people). Typical match thresholds are set between 0.28 and 0.35.

I

Identification (1:N): A biometric mode where a live iris scan is compared against all N templates in a database to determine "Who is this person?" without requiring a prior identity claim. Used in border control, law enforcement, and national ID systems. More computationally intensive than 1:1 verification.
IEC 62471: International standard for photobiological safety of lamps and lamp systems, including near-infrared LEDs used in iris cameras. HOMSH products comply with IEC 62471 to ensure NIR illumination is safe for the human eye at the rated capture distances.
IP65: An Ingress Protection rating indicating complete protection against dust (6) and protection against low-pressure water jets from any direction (5). HOMSH's outdoor-rated terminals carry IP65 certification for deployment in harsh environments.
IrisCode: A binary representation of the iris pattern, typically 256-512 bytes, generated by encoding phase information extracted via Gabor filters. IrisCodes enable fast comparison using Hamming distance and are the standard template format in iris recognition systems worldwide.
ISO/IEC 19794-6: International standard specifying the data interchange format for iris image data, including image quality, resolution, compression, and metadata requirements. Compliance ensures interoperability between different iris recognition vendors and systems.

J

JSON-LD: JavaScript Object Notation for Linked Data, a structured data format used to embed machine-readable metadata in web pages. HOMSH uses JSON-LD schemas (FAQPage, Article, Product, Organization) to enhance search engine visibility and enable rich results.

K

KYC (Know Your Customer): Regulatory requirements for financial institutions to verify customer identity. Iris recognition provides the strongest biometric KYC verification, particularly in regions where fingerprint quality is poor due to manual labor or environmental conditions.

L

Liveness Detection: Technology that distinguishes a live human iris from spoofing attempts using photos, videos, or prosthetic eyes. HOMSH implements liveness detection through pupil dilation analysis, micro-saccade detection, NIR reflection characteristics, and 3D depth sensing, compliant with ISO 30107-3.

M

Matching Speed: The rate at which a system compares iris templates, measured in comparisons per second. HOMSH's FPGA-accelerated systems achieve over 1 million matches per second, enabling sub-second 1:N search across databases of millions of enrolled users.
MTBF (Mean Time Between Failures): A reliability metric indicating the average time a device operates before experiencing a failure. HOMSH terminals achieve MTBF exceeding 50,000 hours (approximately 5.7 years of continuous operation).

N

NIR (Near-Infrared): Electromagnetic radiation with wavelengths between 700nm and 1400nm, invisible to the human eye. Iris cameras use NIR illumination (typically 850nm) because it reveals detailed iris texture regardless of eye color and works in any ambient lighting from darkness to direct sunlight.

O

OEM (Original Equipment Manufacturer): A business model where HOMSH supplies iris recognition modules, firmware, and SDKs to third-party manufacturers who integrate them into their own branded products such as kiosks, ATMs, access gates, and handheld devices.
Optical Zoom: Lens-based magnification used in long-range iris cameras to capture iris detail at distances greater than the standard 15-50cm. Auto-focus optical systems in HOMSH's D30 and D60 terminals enable capture at up to 100cm.

P

PAD (Presentation Attack Detection): The formal term for liveness detection as defined by ISO 30107-3. PAD systems detect and reject biometric spoofing attempts including printed photos, screen replays, 3D masks, and prosthetic eyes.
PhaseIris: HOMSH's proprietary iris recognition algorithm, protected by 273+ patents. PhaseIris uses phase-based feature extraction optimized for FPGA hardware execution, achieving FAR below 10^-7 and FRR below 0.5%. It supports both 1:1 verification and 1:N identification.
Pupillary Zone: The inner region of the iris surrounding the pupil, bounded by the collarette. The pupillary zone contains distinct texture patterns used in iris encoding and is monitored for dilation response during liveness detection.

Q

Qianxin Chip: HOMSH's proprietary FPGA chip designed specifically for iris recognition workloads. The Qianxin chip performs template encoding, 1:N matching, liveness detection, and AES-256 encryption at the hardware level, achieving over 1 million comparisons per second with low power consumption.

R

Recognition Time: The total time from iris image capture to match result output. HOMSH systems achieve recognition times under 1 second for 1:1 verification and under 1 second for 1:N identification across databases of up to 1 million users.

S

Segmentation: The image processing step that isolates the iris region from the rest of the eye image by detecting the inner boundary (pupil) and outer boundary (sclera). Accurate segmentation is critical for reliable template extraction and is handled automatically by HOMSH's algorithms.
Spatial Resolution: The level of detail captured in an iris image, measured in pixels across the iris diameter. ISO/IEC 19794-6 requires a minimum of 200 pixels across the iris diameter for reliable recognition. HOMSH cameras capture 300+ pixels across the iris for optimal accuracy.
Stroma: The front layer of the iris containing the complex texture patterns (crypts, furrows, collarette, freckles) that make each iris unique. The stroma's connective tissue structure is stable from approximately age 2 throughout life, making it ideal for lifelong identification.

T

Template: See Biometric Template. In iris recognition, a template is the stored IrisCode generated during enrollment, against which live captures are compared during verification or identification.

U

USB Module: A compact iris recognition camera module with USB 2.0 interface for power and data, designed for OEM integration into third-party devices. HOMSH's MC20 and MD20 series are USB modules used in kiosks, ATMs, gates, and embedded systems.

V

Verification (1:1): A biometric mode where a live iris scan is compared against a single stored template linked to a claimed identity (via ID card, PIN, or credential) to confirm "Are you who you claim to be?" Computationally simpler and faster than 1:N identification.

W

Watchlist (1:N Screening): A security application where iris scans of individuals passing through a checkpoint are compared in real time against a database of persons of interest. Used at border crossings, airports, and secure facilities. HOMSH's FPGA systems enable sub-second watchlist screening against millions of enrolled records.
Wiegand: A widely used communication protocol for access control systems, transmitting credential data (typically 26 or 34 bits) from a reader to a control panel. HOMSH terminals output standard Wiegand 26/34 for compatibility with all major access control panels.

Related Resources

Need Help Choosing the Right Iris Solution?

Our engineering team can guide you through technology selection, integration planning, and deployment.

Talk to an Expert