Identity Recognition Method Based on Multimodal Biometric Fusion Based on Evolutionary Strategy

Abstract

The invention discloses an identity recognition method based on multi-modal biological feature fusion based on an evolutionary strategy. The specific steps of the method are as follows: firstly, image data of three biological features of human face, static hand shape and palmprint are collected, and the extracted Feature extraction for each biological feature and feature authentication using the nearest neighbor classifier; then use the evolutionary strategy to adaptively adjust the weights of each classifier; finally use the adaptively adjusted weights to classify the face and static hand shape The recognition results output by the three classifiers corresponding to the three biometric features of the palmprint are weighted and fused to obtain the final recognition result. The present invention uses the evolutionary strategy to self-adaptively adjust the weight, and fuses the classifier with multiple biological characteristics through the weighting method on the decision-making layer to reduce the influence of the surrounding environment on the recognition result, so that the safety and adaptability of the recognition method are improved. Compared with other fusion methods, it has been greatly improved, and the fault tolerance of the recognition method has also been improved.

Description

technical field [0001] The invention relates to the fields of computer data processing and identity feature identification, in particular to an identity identification method based on multi-modal biological feature fusion based on an evolutionary strategy. Background technique [0002] With the continuous development of information technology, personal information security has become an important issue in the research of information technology. In recent years, due to the uniqueness and stability of biometrics, the use of biometrics for identity authentication has become more and more popular. But most of the ones that are widely used in our real life are based on a single biometric feature, such as fingerprints, faces, gestures, and irises. However, single-mode biometric identification has several limitations: (1) Noise influence in the data acquisition process: the biometric data collected by the acquisition system generally contain noise, and these biometric data contain...

AI Technical Summary

Problems solved by technology

However, single-mode biometric identification has several limitations: (1) Noise influence in the data acquisition process: the biometric data collected by the acquisition system generally contain noise, and these biometric data containing noise will affect the system. Accuracy of identification results; (2) Non-universal: some people may not be born with certain biological characteristics or cannot be collected and identified due to external causes of certain biological characteristics being destroyed; (3) Security: Pretenders may Misappropriating and copying a certain feature to impersonate a legitimate user

Although this method can solve some of the shortcomings of single biometric identification, it still has the main problem of unstable recognition accuracy and recognition speed in different environments.

Images