算术下溢
像素
嵌入
信息隐藏
图像(数学)
图像质量
隐写术
数学
人工智能
对偶(语法数字)
隐写分析技术
计算机科学
算法
计算机视觉
模式识别(心理学)
艺术
文学类
程序设计语言
作者
Luyen Cao Thi,Nguyễn Kim Sao,At PhamVan
标识
DOI:10.1016/j.jisa.2023.103470
摘要
Compared to conventional reversible data hiding (RDH) methods, dual-image RDH methods have greater embedding capacity, better stego image quality and higher security. Among existing dual-image RDH methods, three methods of Lu et al., Yao et al. and Kim et al. are remarkable. These three methods have much higher embedding capacities than other methods, while still maintaining stego image quality. However, these three methods all have a limitation that in order to avoid underflow/overflow, they have to ignore pixels around two pixels with values 0 and 255. In many cases, the number of such pixels is not small and this limits the embedding capacity of all three methods above. In this paper, the maximum number of bits that can be embedded in each pixel is evaluated correctly. Embedding a number of bits less than or equal to the maximum number at each pixel will not cause the underflow/overflow. Therefore the proposed method can embed data bits in all pixels of an original image. As a result, the proposed method has not only a larger embedding capacity, but also better stego image quality than mentioned above methods. With an embedding ratio of 3.45 bits per a pixel, the proposed method has a greater embedding capacity than existing dual image reversible data hiding methods while maintaining good image quality.
科研通智能强力驱动
Strongly Powered by AbleSci AI