Iris Recognition FAQ
Common questions about iris recognition accuracy, technology, and real-world deployment.
How accurate is iris recognition compared to fingerprints?
Iris recognition achieves a False Accept Rate (FAR) below 0.0001% — that's 1 error in 1 billion attempts. Fingerprint recognition has a FAR of up to 0.1%, making iris recognition 1,000× more accurate. For high-stakes applications like border control, financial transactions, or national ID systems, this difference is critical.
Does iris recognition work in dusty or wet environments?
Yes. HOMSH iris recognition hardware is specifically engineered for harsh conditions — construction sites, mining operations, border crossings, and outdoor deployments. The iris is a protected internal organ and is not affected by dust, moisture, cuts, calluses, or temperature extremes. Fingerprints, by contrast, degrade significantly in these environments.
Can iris recognition be spoofed with a photo or fake eye?
Modern iris recognition systems including HOMSH's use active liveness detection to reject photos, printed eyes, and synthetic replicas. Our PhaseIris™ algorithm detects pupil dilation, micro-tremors, and near-infrared reflection patterns that are impossible to replicate with static images.
What is FPGA-powered iris recognition?
FPGA (Field Programmable Gate Array) is a hardware chip that executes iris recognition algorithms at the silicon level, far faster than software running on a CPU or GPU. HOMSH's proprietary Qianxin FPGA chip completes 1:N matching across millions of enrolled identities in under one second, with lower power consumption and higher reliability than cloud-based solutions.
How does iris recognition work at long distance?
Standard iris recognition requires the subject to be 15–30cm from the camera (contact range). HOMSH's advanced terminals support recognition at up to 1 meter with standard hardware. Long-range iris recognition (3–10 meters) is technically feasible with specialized optics and is an active research area.
Which industries use iris recognition?
Key industries include: Border control and immigration (national ID verification, e-gates), Financial services (vault access, high-value transaction authentication), Construction and mining (site access control where fingerprints fail), Healthcare (patient identity without contact), Government and defense (classified access control), and Prisons and corrections (inmate identification).
Can iris recognition be used for large-scale 1:N matching?
Yes. HOMSH's FPGA-accelerated system supports 1:N matching across databases of millions of enrolled identities in under one second. This is critical for border control and national ID systems where the system must verify a person against an entire country's population.
What are the environmental conditions for HOMSH iris hardware?
HOMSH terminals operate in temperatures from -20°C to 60°C, humidity from 10% to 95% (non-condensing), and are rated IP65 for dust and water resistance. They function in direct sunlight and low-light conditions using active near-infrared (NIR) illumination.
Does HOMSH support OEM integration?
Yes. HOMSH supplies compact iris recognition modules (PCB-level) specifically designed for OEM integration into third-party devices, kiosks, gates, ATMs, and embedded systems. We provide SDK, API documentation, and technical support for integration.
What is the PhaseIris™ algorithm?
PhaseIris™ is HOMSH's proprietary iris recognition algorithm, protected by 273+ patents. It uses phase-based feature extraction from the iris texture, encoded into a compact IrisCode that enables fast, accurate matching. The algorithm achieves FAR < 0.0001% and is optimized for FPGA execution.
How far away can iris recognition work?
Standard iris recognition systems operate at 15-50cm (contact range). HOMSH's advanced terminals such as the D30 and D60 series support recognition at distances up to 100cm using auto-focus iris cameras. Long-range iris recognition at 3-10 meters is technically feasible with specialized telephoto optics and higher-resolution sensors, though it remains an active area of development for walkthrough-style deployments.
Does iris recognition work with glasses or contact lenses?
Yes. Modern iris recognition systems, including all HOMSH products, are designed to work reliably with prescription glasses, reading glasses, and most contact lenses. The near-infrared (NIR) illumination used by iris cameras penetrates clear and lightly tinted lenses. However, heavily tinted sunglasses, mirrored coatings, or deeply colored cosmetic contact lenses may obstruct the iris pattern and require removal for accurate capture.
What is FAR and FRR in iris recognition?
FAR (False Accept Rate) measures how often the system incorrectly accepts an unauthorized person. FRR (False Reject Rate) measures how often it incorrectly rejects an authorized person. These two metrics are inversely related: lowering FAR increases security but may raise FRR. HOMSH's PhaseIris™ algorithm achieves FAR below 0.0001% (1 in 1 million) while maintaining FRR under 0.5%, making it one of the most accurate biometric systems available. By comparison, fingerprint systems typically have FAR of 0.1% and FRR of 2-5%.
Can iris recognition work in bright sunlight?
Yes. HOMSH iris recognition hardware uses active near-infrared (NIR) illumination at 850nm wavelength, which is invisible to the human eye. Because NIR imaging relies on its own light source rather than ambient visible light, the system functions reliably in direct sunlight (up to 6,000+ Lux), complete darkness, and everything in between. The MD20 module, for example, is rated for 0-6,000 Lux working illumination.
How many users can an iris scanner store?
Storage capacity varies by product. HOMSH's compact OEM modules range from 500 users (MC20-500) to 10,000 users (MC20-10000 and MD30). Access control terminals like the D30 and D60 support 10,000+ enrolled users locally. For enterprise and national-scale deployments, HOMSH's FPGA-accelerated server systems support databases of millions of enrolled identities with 1:N matching completed in under one second.
What is the difference between 1:1 and 1:N matching?
In 1:1 (verification) mode, the system compares a live iris scan against a single stored template — typically linked to an ID card, PIN, or credential. This confirms 'Are you who you claim to be?' In 1:N (identification) mode, the system compares a live scan against an entire database of N enrolled templates to determine 'Who are you?' without requiring any prior claim of identity. 1:N is computationally more demanding and is used in border control, law enforcement, and national ID systems. HOMSH's FPGA hardware accelerates 1:N matching to under 1 second even across millions of records.
Is iris recognition GDPR compliant?
Iris recognition can be deployed in full compliance with GDPR and other data protection regulations. Key requirements include: obtaining explicit consent before enrollment, storing encrypted biometric templates rather than raw iris images, implementing data minimization and purpose limitation, providing users the right to deletion, and conducting a Data Protection Impact Assessment (DPIA). HOMSH systems store only compact mathematical templates (IrisCodes), not photographs of the eye, and support on-device processing to minimize data transmission.
How long does iris enrollment take?
Iris enrollment with HOMSH devices typically takes 2-5 seconds for a first-time user. The process involves the user looking at the device while the camera captures a high-resolution near-infrared image of one or both irises. The PhaseIris™ algorithm then extracts the unique iris pattern and encodes it into a compact template (IrisCode). The MC21 module achieves enrollment in as fast as 0.7 seconds. Most systems capture both eyes for redundancy, which adds minimal time but significantly improves matching reliability.
What happens if someone closes their eyes during iris scanning?
If a person closes their eyes, the iris scanner cannot capture the iris pattern and will prompt the user to open their eyes and look at the camera. HOMSH devices include real-time eye detection that provides audio and visual feedback to guide the user into the correct position. The system requires a clear view of the iris — including the pupil and surrounding texture — to perform recognition. Partial eye closure or excessive blinking will trigger a re-capture prompt, not a false match.
Can iris recognition be fooled by photos or fake eyes?
Modern iris recognition systems with liveness detection, including all HOMSH products, are highly resistant to spoofing attacks. The PhaseIris™ algorithm detects presentation attacks by analyzing pupil dilation response, micro-tremors, near-infrared reflection characteristics, and 3D depth — features that cannot be replicated by printed photos, screen displays, or prosthetic eyes. HOMSH devices comply with ISO 30107-3 (Presentation Attack Detection) standards, ensuring that only a live human iris is accepted.