End-to-end text image watermark model establishment method based on deep learning

Abstract

An end-to-end text image watermark model establishment method based on deep learning comprises the following steps: S1, preparing a text image set suitable for deep learning, dividing the text image set into a training set and a test set which are different from each other, and carrying out a batch data training process; S2, putting the training set into an embedding and extracting algorithm framework of a text image watermark, carrying out training on a model, and training a text image watermark model by utilizing deep learning; and S3, putting the prepared test set into the text image watermark model for testing. The end-to-end text image watermark model establishment method based on deep learning disclosed by the invention has the characteristics of high invisibility, strong robustness and excellent practicability. In the actual use process, the cost is lower, the detection precision is higher, and the consumed time is shorter.

Description

technical field [0001] The invention relates to the field of digital media copyright protection, in particular to an end-to-end method for establishing a text image watermark model based on deep learning. Background technique [0002] With the rapid development of computer technology, e-office and e-government with the characteristics of easy dissemination and low cost have been widely used. When people use digital media efficiently and conveniently, they also face risks such as leakage and piracy of digital media content. Therefore, digital media content needs to be properly processed. Digital watermarking technology is an effective copyright protection technology, which can protect digital media content, and even track leaks. The data content with text images as the carrier is the main part of digital text dissemination, and its formats include electronic documents, electronic signatures, electronic invoices, electronic contracts, electronic certificates and other carrier...

AI Technical Summary

Problems solved by technology

[0005] The present invention aims at the deficiencies in the prior art, that is, to solve the problems of invisibility, robustness, non-blind extraction and poor practicability of the traditional spatial domain watermarking algorithm and transform domain watermarking algorithm, and provides an end-to-end depth-based Learning method for building text image watermarking model

Images