학술논문

Robust Zero-Watermarking Algorithm for Medical Images Based on Direction Gradient Histogram and DCT
Document Type
Conference
Source
2023 IEEE 6th International Conference on Pattern Recognition and Artificial Intelligence (PRAI) Pattern Recognition and Artificial Intelligence (PRAI), 2023 IEEE 6th International Conference on. :575-583 Aug, 2023
Subject
Computing and Processing
Histograms
Watermarking
Medical services
Feature extraction
Robustness
Internet
Encryption
HOG
medical images
zero-watermarking
DCT
robustness
perceptual hashing
Language
Abstract
With the rapid development of Internet technology, using the Internet to transmit medical images has gradually become an efficient and convenient transmission method. It improves the efficiency of medical services, enabling medical institutions and doctors to share medical images and diagnostic information quickly. However, there are also some potential security risks in transmitting medical images over the Internet, such as data leakage and tampering. This article proposes a medical image zero-watermarking algorithm based on the direction gradient histogram (HOG) and discrete cosine transform (DCT) to solve the above problems. The proposed algorithm first extracts the feature vector of the medical image and then utilizes the perceptual hash algorithm to obtain its corresponding feature sequence. The watermark image is then encrypted using chaotic scrambling through logical mapping. The zero-watermarking technique is employed for the embedding and extraction of the watermark. Experimental results demonstrate the effectiveness of the proposed algorithm in successfully extracting and restoring the watermark even under challenging conditions.