An electronic contract handwritten signature authentication method based on twin neural networks

Abstract

The invention discloses an electronic contract handwritten signature authentication method based on twin neural networks, which relates to the technical field of electronic contract handwritten signature authentication. The invention comprises the following steps: S1, a plurality of user handwritten signatures are left behind in a real name authentication link, and the handwritten signatures are saved as image files; 2, selecting two of a plurality of handwritten signature image files and one of that non-user handwritten signature image file, inputting the two handwritten signature image filesinto a twin neural network model for vector calculation, obtaining a loss function, and obtaining a training set cost function through the loss function; 3, train other handwritten signature image files of that user through an optimization algorithm, and outputting an authentication model; S4: When signing on line, the handwritten signature on the touch screen input device is saved as an image immediately, the identification model is input, the identification result is output, and the preservation and the identification of the handwritten signature are completed. The invention can realize theidentification of the handwritten signature in the electronic contract signing, and the electronic signature is replaced or perfected by the handwritten signature.

Description

technical field [0001] The invention relates to the technical field of electronic contract handwritten signature identification, and more specifically relates to a method for electronic contract handwritten signature identification based on a twin neural network. Background technique [0002] The civil legal act of electronic contract is the legal act of two or more civil subjects. The purpose of establishing, changing, and terminating property civil rights and obligations between the parties is electronic. A new form of contract. The entire signing process is in electronic form, and electronic contracts are negotiated, signed and performed through e-mail, communication tools, etc. This contract method greatly saves transaction costs and improves economic benefits. [0003] Different from the conclusion of traditional contracts, the conclusion of electronic contracts takes place in a virtual space. The two parties to the transaction generally do not meet each other, and the...

AI Technical Summary

Problems solved by technology

[0022] The purpose of the present invention is: in order to solve the existing third-party electronic contract platform, the parties to the contract need reliable third-party authentication when signing electronic signatures online, and only the certified electronic signatures have the same legal effect as handwritten signatures. The present invention provides a A method for identifying handwritten signatures of electronic contracts based on twin neural networks, which can realize the identification of handwritten signatures in electronic contract signing, and replace or improve electronic signatures with handwritten signatures

Images