Identity authentication method integrating multiple type features for palm print image recognition through hierarchical topic model

Abstract

The invention discloses an identity authentication method integrating multiple type features for palm print image recognition through a hierarchical topic model. According to the method, palm print images are converted into documents through analogy, palm print feature components are converted into words through analogy, and palms are converted into topics through analogy; first, topic model parameters and the number of the topics are calculated according to the palm print images for training; second, the hierarchical topic model is utilized to perform layering and grouping on the palm print images for training, and a topological structure of the hierarchical topic model is optimized according to recognition of palm print images for testing; and third, the trained topic model is used to classify to-be-detected palm print images into corresponding groups, features of the to-be-detected palm print images are matched with features of the palm print images in the corresponding groups one by one, and therefore identity authentication is realized. Through the method, acquired recognition accuracy is high, recognition speed is high, and the instantaneity requirement of a large-scale application place for identity authentication through palm print recognition can be met.

Description

Technical field [0001] The invention belongs to the field of palmprint image processing, and relates to a method for identity authentication through palmprint image recognition, and in particular to a classification by identifying multilayer LDA (Latent Dirichlet Allocation (LDA)) The palm print image is used for identity authentication. Background technique [0002] With the rapid growth of information interaction in modern society, personal identification has become an urgent problem to be solved. It has been highly valued and widely used in electronic transactions, public security, commercial finance, medical health, national defense and military departments. Traditional identification methods include identification items (such as keys, ID cards, seals, bank cards, passports, etc.) and identification knowledge (such as user names, passwords, passwords, etc.). Once the identification items and identification knowledge are Misappropriating, forging or forgetting, its identity c...

AI Technical Summary

Problems solved by technology

The above method is more suitable for occasions where the palmprint to be detected is matched one by one with the registered palmprint to confirm the identity, but for a large-scale recognition system, one-by-one matching cannot meet the real-time requirements of identity recognition.

Images