Biological multi-modal feature recognition method based on dynamic fuzzy reasoning

By adopting dynamic fuzzy inference and multimodal feature recognition technology in biometric systems, the problems of unstable identification accuracy and susceptibility to counterfeiting in the prior art are solved, and higher recognition accuracy and robustness are achieved.

CN119625850AActive Publication Date: 2025-03-14HEFEI UNIV OF TECH +2
View PDF 10 Cites 1 Cited by

Patent Information

Application Number
CN202411537223.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-31
Publication Date
2025-03-14
Estimated Expiration
2044-10-31

AI Technical Summary

Technical Problem

Existing biometric technologies are susceptible to environmental factors in high-security environments, have unstable recognition accuracy, and are vulnerable to counterfeiting attacks, and have a high misidentification rate.

Method used

The biological multimodal feature recognition method based on dynamic fuzzy inference is adopted to improve the accuracy and robustness of the recognition by preprocessing and feature extraction of face, fingerprint and iris images by preprocessing and feature extraction of three-dimensional Euclidean distances, and fuzzy inference and dynamic threshold adjustment techniques.

Benefits of technology

It enhances the accuracy of the identification system, effectively reduces the false recognition rate, improves the robustness and anti-forgery ability of the system, and adapts to efficient and stable under various operating conditions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119625850A_ABST
    Figure CN119625850A_ABST
Patent Text Reader

Abstract

The invention provides a biological multi-modal feature recognition method and device based on dynamic fuzzy reasoning. The method comprises the steps that an obtained face image, a fingerprint image and an iris image of a target object are preprocessed; performing feature extraction on the preprocessed face image, the fingerprint image and the iris image, and determining a three-dimensional Euclidean distance of the face, the fingerprint and the iris between the target object and an object to be verified in a database according to a feature extraction result; performing fuzzy reasoning based on the three-dimensional Euclidean distance of the face, the fingerprint and the iris to obtain a recognition score; comparing the identification score with a preset decision threshold value, when the identification score is greater than or equal to the decision threshold value, determining that the target object is matched with the to-be-verified object, otherwise, determining that the target object is not matched with the to-be-verified object; wherein the decision threshold value is obtained by training through a reward / punishment mechanism based on preset target recognition precision. The method can effectively cope with various environment changes, improves the recognition accuracy, and is suitable for application scenes needing extremely high safety.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Dual-modal bio-feature signing-in system based on human face and handwritten signature

    CN107507286A

  • Biological identification method and related product

    CN107545166A

  • User identity identification method, device and terminal equipment

    CN108776795A

  • Multi-biological feature fusion recognition method based on image quality evaluation

    CN108921191A

  • Multiple fuzzy reasoning weighted KNN positioning method

    CN111050282A

Cited By

  • Dimming glass display control method based on biological characteristic dynamic encryption

    CN120891661A