An intelligent vein authentication method and system with autonomous learning ability

Abstract

The invention discloses an intelligent vein authentication method and system with self-learning ability. In the method, the full convolutional neural network deep learning model is combined with repeated, regular global training and irregular local training. Training, before the full convolutional neural network deep learning model goes online, use the accumulated samples to conduct a global training, and save the optimized feature extraction, feature recognition and feature classification parameters; after the model goes online, use the newly registered samples to The model conducts irregular local training to complete the optimization of feature classification parameters; under certain conditions, the newly accumulated samples are used to conduct regular global training on the model, and the feature extraction, feature recognition and feature classification parameters are optimized again, ensuring that This model becomes an intelligent authentication system with self-learning ability and enhances the accuracy of vein authentication at the same time.

Description

Technical field [0001] The present invention relates to the field of image recognition and processing, and more particularly to a smart vein authentication method and system with autonomous learning capabilities. Background technique [0002] In recent years, with the rapid development of biological identification technology, the fingerprint certification method has been widely used. At the same time, there is a more confidentiality of iris, sound, human face, vein, refers to the intravenous type of intravenous, and more and more. Among them, the venous authentication method is more widely used due to the advantages of high confidentiality, unproduction, convenient authentication, and small certification equipment. In the prior art, the finger vein authentication generally uses various image segmentation algorithms from the acquired finger image, forming a venous image, and then extracts the individual vein characteristics from the context of the context of the context. Finally, ...

AI Technical Summary

Problems solved by technology

However, in practical applications, the samples to be identified and classified are dynamically increased. Only the parameters generated during the supervised learning process before the system goes online cannot accurately identify all the newly added samples to be identified, which affects the recognition accuracy of new samples.

(2) High requirements on the amount of learning data and long learning time

The deep learning model is relatively complex, and parameter optimization requires a large amount of training data for support, and multiple iterations are required to complete the training

However, the current finger vein sample data has neither a standard training data set nor a sufficient amount of training data.

(3) The accuracy of certification has not yet reached the level of precision

Images