Image coding method for image security confidentiality and anti-counterfeiting identification
The invention discloses an image coding method for image security and confidentiality and anti-counterfeiting identification, which uses an artificial intelligence image coding technology based on deep learning. A display encoded image is generated using, as an input to a decoding generator, a fusio...
Saved in:
Main Authors | , , |
---|---|
Format | Patent |
Language | Chinese English |
Published |
23.05.2023
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The invention discloses an image coding method for image security and confidentiality and anti-counterfeiting identification, which uses an artificial intelligence image coding technology based on deep learning. A display encoded image is generated using, as an input to a decoding generator, a fusion feature extracted from a local implicit correlation feature and a global high-dimensional correlation feature of an encrypted image after key encryption. Therefore, the anti-counterfeiting capability can be greatly enhanced while the content of the original image is not changed.
公开了一种图像安全保密和防伪识别的图像编码方法,其使用基于深度学习的人工智能图像编码技术,利用对于密钥加密后的加密图像的局部隐含关联特征和全局性高维关联特征的特征提取的融合特征作为解码生成器的输入来生成显示编码图像。这样,能够在不改变原始图像内容的同时使得防伪能力得到大幅增强。 |
---|---|
Bibliography: | Application Number: CN202211631187 |