Invisibility and Readability of Temporally and Spatially Intensity Modulated Metaimage for Information Hiding on Digital Signage Display System

Invisibility and Readability of Temporally and Spatially Intensity Modulated Metaimage for Information Hiding on Digital Signage Display System
复制标题

数字标牌显示系统信息隐藏的时空强度调制元图像的不可见性和可读性

DOI:
10.1109/tia.2019.2930940
复制
发表时间:
2019
影响因子:
4.4
通讯作者:
Kazutake Uehira
Kazutake Uehira
中科院分区:
工程技术2区
文献类型:
--
作者:
Hiroshi. Unno;Kazutake Uehira

文献摘要

相似文献

本文描述了一种显示技术,它使我们能够通过使用时间、空间、颜色强度调制的元图像来将元图像不可见地嵌入到所显示的封面图像的颜色分量中,并提取它。我们首先研究了在显示的封面图像中,时间和空间强度调制的元图像在颜色分量方面的不可见性。实验结果表明,元图像嵌入到蓝色分量中的不可见性远远高于红色分量和绿色分量。我们的实验还发现,红色和绿色分量不能用于嵌入了元图像的颜色分量。我们还提出并评估了一种新的技术来提高用摄像机捕获的显示封面图像中不可见嵌入的元图像的可读性。所提出的技术使用两种方法。提出了一种计算估计灰度元图像的方法,通过使用嵌入或未嵌入元图像的颜色分量之间的相关性,从其直方图中获得二值化阈值的值。提出了一种计算二值图像估计值的方法,利用得到的二值化阈值和嵌入图像的尺寸特征参数,大大降低了由封面图像引起的噪声。通过使用这些方法,大大提高了元图像的可读性。
This paper describes a display technique that enables us to invisibly embed a metaimage into a color component of a displayed cover image by using a temporally, spatially, color-intensity-modulated metaimage and to extract it. We first examined invisibility in terms of color components for a temporally and spatially intensity-modulated metaimage in a displayed cover image. The results obtained from these experiments revealed that the invisibility with which the metaimage was embedded in the blue color component was much higher than that of the red and green components. Our experiments also revealed that the red and green components could not be used for a color component in which the metaimage was embedded. We also proposed and evaluated a new technique of improving the readability of an invisibly embedded metaimage in a displayed cover image captured with a video camera. The proposed technique uses two methods. A proposed method of calculating an estimated grayscale metaimage enabled us to obtain a value for the binarizing threshold from its histogram by using the correlation between color components in which the metaimage was either embedded or not embedded. A proposed method of calculating an estimated binary metaimage could drastically reduce noise caused by the cover image by using the obtained value of the binary threshold and size feature parameters of the embedded metaimage. The readability of the metaimage was drastically improved by using these methods.