A post-quantum secure abnormal road condition efficient detection method

By constructing a quantum-safe K-means hierarchical clustering tree and a pre-trained CNN model, encrypted processing of vehicle location and road condition image features is achieved, solving the problems of lack of privacy protection and real-time data analysis in existing technologies, improving the real-time performance and accuracy of anomaly detection, and adapting to the real-time handling needs of traffic management.

CN122223960APending Publication Date: 2026-06-16NORTHWESTERN POLYTECHNICAL UNIV +1
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
NORTHWESTERN POLYTECHNICAL UNIV
Filing Date
2026-03-11
Publication Date
2026-06-16

AI Technical Summary

Technical Problem

The lack of privacy protection in current road condition detection and the inability to analyze encrypted data in real time lead to data processing bottlenecks and end-to-end latency, affecting the performance and reliability of anomaly detection.

Method used

A post-quantum-secure approach is adopted, utilizing K-means hierarchical clustering trees and pre-trained CNN models to detect abnormal road conditions through encrypted feature vectors, thereby achieving privacy protection for vehicle location and road condition image features, and performing real-time analysis in the cloud.

Benefits of technology

It achieves post-quantum level privacy protection, improves data security and retrieval efficiency, ensures the real-time and accuracy of anomaly detection, adapts to the real-time handling needs of traffic management, and enhances the long-term availability and social acceptance of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122223960A_ABST
    Figure CN122223960A_ABST
Patent Text Reader

Abstract

The application specifically relates to a post-quantum secure abnormal road condition efficient detection method, and belongs to the field of information security.The method comprises five steps of system initialization, index tree construction and encryption, encrypted query generation, similar abnormal vector matching, abnormal early warning and processing: a traffic management department generates key parameters, trains a convolutional neural network model for feature extraction, constructs an encrypted K-means hierarchical clustering tree and uploads the cloud server; a vehicle collects abnormal road condition images, encrypts feature vectors and positions and then uploads queries; the cloud server completes similar vector matching on the encrypted K-means hierarchical clustering tree and returns the result; the traffic management department restores the original image locally according to the result, issues an early warning according to the abnormal level and executes corresponding disposal measures. The application realizes efficient detection of abnormal road conditions through post-quantum secure ciphertext vector similarity retrieval, reduces the retrieval calculation complexity to sub-linear, the retrieval accuracy is more than 95% and the average retrieval delay is less than 0.27 ms, and privacy protection and detection efficiency are considered.
Need to check novelty before this filing date? Find Prior Art