US 12361755 B1Patent grantUnited States

Encrypted Facial Identification

Official patent title

Face identification system and method for identifying face image

Arabic title: نظام للتعرف على الوجه وطريقة للتعرف على صورة الوجه

Invention

Problem

Face-identification systems must protect sensitive biometric images while preserving recognition accuracy, processing speed, and resistance to interception or impersonation.

Why it matters

Biometric templates cannot be easily replaced after compromise, so privacy-preserving processing and rigorously validated recognition are important for trustworthy access decisions.

Approach

The system converts facial-image pixels to binary form, transforms them with two-dimensional cellular automata, maps the result through gray code, and encrypts it with a generated key. PCA-derived encrypted features are compared with an encrypted database using Euclidean distance before an identification or access notification is sent.

Who may benefit

Potential beneficiaries include access-control vendors, identity-verification providers, security integrators, campus and enterprise facilities, biometric research teams, and privacy engineering groups.

Potential value

The architecture keeps facial processing in an encrypted-image pipeline that combines cellular-automata transformation, key-based encryption, PCA feature reduction, and database matching.

Background

Face recognition is used for device authentication, physical access, surveillance, and identity verification, but stored or transmitted facial data creates lasting privacy and security risk. Conventional systems can also lose accuracy under changes in pose, lighting, expression, or camera quality. The publication integrates encryption with PCA-based recognition so features can be handled in an encrypted workflow. The supplied record describes architecture and evaluation figures, but it does not establish independent cryptographic review, spoof resistance, demographic fairness, operational false-match rates, or production-scale reliability.

Technology overview

A camera captures a face, after which a pixel-bit converter produces binary values. A 3×3 two-dimensional cellular automata function applies update rules, a gray-code generator maps states to grayscale values, and an encryption algorithm uses a key that may derive from a chaotic logistic sequence. Feature circuitry applies PCA to generate eigenface representations. Euclidean distance compares the encrypted projection with known encrypted features, and a notification subsystem reports identity or grants or denies access.

Potential applications

Encrypted biometric access-control systems.
Secure identity verification across networked devices.
Privacy-focused facial authentication research.
Campus or enterprise entry-notification workflows.

Evidence-supported advantages

Facial images are transformed and encrypted before feature matching.
The pipeline combines cellular automata, gray code, and key-based encryption.
PCA reduces feature dimensionality for database comparison.
The architecture separates capture, encryption, extraction, recognition, and notification.
Access-grant or denial notification is integrated into the workflow.

Development stage

A system architecture, processing flow, and evaluation plots are described; independent security, liveness, fairness, and deployment-scale accuracy validation are not established.

Commercial opportunity

The design may interest access-control and identity-platform vendors. Deployment requires independent cryptanalysis, key-management review, attack and liveness testing, representative accuracy and bias evaluation, latency and scalability benchmarks, secure template lifecycle controls, privacy impact assessment, legal compliance, human fallback, and operational monitoring. Security and recognition claims are not independently verified here.

Patent classifications

WIPO IPC

G06V40/16Image or video recognition or understanding
G06V40/50Image or video recognition or understanding

CPC

G06V40/168Image or video recognition or understanding
G06V40/169Image or video recognition or understanding
G06V40/172Image or video recognition or understanding
G06V40/50Image or video recognition or understanding
H04L9/0866Transmission of digital information, e.g. telegraphic communication
H04L9/0869Transmission of digital information, e.g. telegraphic communication
H04L9/3231Transmission of digital information, e.g. telegraphic communication

Inventors

First inventorHaifa Abdullah Salem Alqahtani
InventorMohamed Alkhatib

Keywords

face recognition
biometric encryption
cellular automata
gray code
principal component analysis
eigenfaces
Euclidean distance
access control

Patent document and drawings

The patent publication is mapped to this record. Patent drawings remain within that publication; no separately cleared public media package has been supplied.