Eye fundus image nondestructive digital copyright management method based on fingerprint and zero-watermark mixed framework

Abstract

The invention discloses an eye fundus image nondestructive digital copyright management method based on a fingerprint and zero-watermark mixed framework. The method comprises the following steps: firstly, performing 2D-DCT transformation on a normalized eye fundus image, setting different weights for coefficients obtained by transformation, extracting eye fundus image features with high identifiability and high robustness as fingerprints for content-based searching; and secondly, building a mapping relation between the fingerprints and watermarks according to a zero-watermark copyright protection method based on a VSS visual cryptography algorithm for copyright authentication. Compared with the conventional method, the method disclosed by the invention has the advantages that the whole process of the method is nondestructive and meets the requirement for accurate diagnosis; the method has the effect of protecting the copyright of the eye fundus image for a long time, and the problem of one-off copyright authentication based on digital watermarks; and the eye fundus image nondestructive digital copyright management method is applicable to digital copyright management of the eye fundus image, and has a certain clinical significance for auxiliary diagnosis of eye fundus-related diseases.

Description

technical field [0001] The invention belongs to the field of information security, and relates to a non-destructive digital rights management method for fundus images based on a mixed architecture of fingerprints and zero watermarks. Background technique [0002] Fundus images are images taken from different angles of the inner wall of the eyeball with a fundus camera, and are a standard objective diagnostic image. Various fundus diseases such as diabetic retinopathy, glaucoma, hypertension, age-related macular degeneration, etc. can be reflected in fundus images. Therefore, the processing and analysis of fundus images is of great significance to the early screening, diagnosis, treatment and auxiliary diagnosis and treatment of these diseases. As a part of medical digital content, fundus images, as well as their inherent clinical medical value and research value, indicate that fundus images should be effectively protected. However, according to the survey, there is current...

AI Technical Summary

Problems solved by technology

However, the extraction of digital fingerprints is the inherent characteristics of digital images, and there is no clear copyright information, such as publisher information, use right approval judgment information, etc., which may cause unnecessary copyright disputes

[0008] Since fundus images are mostly used in the medical industry, the requirements for digital rights management are relatively high. None of the existing digital rights management methods can fully meet the many requirements for digital rights management of fundus images.

Images