Fingerprint feature extraction device, storage medium storing instructions to perform fingerprint feature extraction method, and method for training neural network to extract feature information of fingerprint

A non-transitory computer-readable storage medium storing computer executable instructions, wherein the instructions, when executed by a processor, cause the processor to perform a fingerprint feature extraction method, the method comprising:

• preparing a pre-trained neural network;
• detecting a fingerprint input of a target input on a fingerprint contact surface included in a fingerprint feature extraction device;
• modifying feature information of a registered fingerprint in the fingerprint feature extraction device based on the feature information of the fingerprint input,
• wherein the modifying includes:
• comparing the feature information of the fingerprint input with the feature information of the registered fingerprint and calculating a similarity between the feature information of the fingerprint input with the feature information of the registered fingerprint;
• determining whether the similarity exceeds a first threshold;
• storing the fingerprint input if the similarity exceeds the first threshold;
• generating modification information to match corresponding minutiae between the feature information of the fingerprint input and the feature information of the registered fingerprint;
• and
• modifying the feature information of the registered fingerprint based on the modification information;
• pairing the modified feature information of the registered fingerprint with the feature information of the fingerprint input;
• and
• inputting results of pairing to the pre-trained neural network to output a fingerprint feature extraction result.

A method of training a neural network for fingerprint feature extraction using a fingerprint feature extraction device, the method comprising:

• modifying feature information of a registered fingerprint in the fingerprint feature extraction device based on feature information of a fingerprint input of a target by the fingerprint feature extraction device,
• wherein the modifying comprises:
• comparing the feature information of the fingerprint input with the feature information of the registered fingerprint, and calculating a similarity between the feature information of the fingerprint input with the feature information of the registered fingerprint;
• determining whether the similarity exceeds a first threshold;
• storing the fingerprint input if the similarity exceeds the first threshold;
• generating modification information to match corresponding minutiae between the feature information of the fingerprint input and the feature information of the registered fingerprint;
• and
• modifying the feature information of the registered fingerprint based on the modification information;
• pairing modified feature information of the registered fingerprint with the feature information of the fingerprint input;
• and
• training the neural network to output a fingerprint feature extraction result with respect to the fingerprint input by using results of pairing as training input data and the modified feature information as training label data.