数字水印
计算机科学
水印
稳健性(进化)
加密
离散余弦变换
多媒体
计算机视觉
计算机安全
嵌入
人工智能
图像(数学)
生物化学
化学
基因
作者
Samrah Mehraj,Subreena Mushtaq,Shabir A. Parah,Kaiser J. Giri,Javaid A. Sheikh,Amir H. Gandomi,Mohammad Hijji,Brij B. Gupta,Khan Muhammad
标识
DOI:10.1109/tce.2022.3217974
摘要
Heritage multimedia, which comprises images, audio, and videos, are precious artifacts of a region. In addition to enabling a better understanding of earlier generations, heritage media provides information about their creative approach, style of living, and diversity of historical and archaeological ideologies. Heritage is an important resource that boosts the local economy, generates sustainable communities, and improves tourism and business sectors. Once heritage images are dissipated to consumers, they may be transmitted through wire/wireless systems, where the data may be accessed by both authorized and unauthorized consumers. With the rapid advancement of technology and 5G networks, establishing an approach that protects cultural heritage media from unauthorized consumers is necessary. This study presents a robust and blind watermarking-framework for cultural images (RBWCI) utilizing the discrete cosine transform domain for ownership verification and copyright protection. The embedding scenario of our technique relies on the difference between coefficients obtained from the two channels of the YCbCr color model. The watermark bit is embedded by taking the difference between two preselected mid-frequency coefficients (Cb-Cr). Moreover, chaotic and deoxyribonucleic acid encryption are employed to ensure double-layer watermark security. The proposed algorithm demonstrates a peak signal-to-noise ratio between 42-43 dB and structural similarity index metric value of approximately 1 for different test images when not under attack. The robustness of the RBWCI was revealed by comparing it with various state-of-the-art schemes, making it suitable for different consumer applications.
科研通智能强力驱动
Strongly Powered by AbleSci AI