Information hiding in motion data of virtual characters. (30th November 2020)

• Record Type: Journal Article
• Authors: Wu, Xu; Guo, Shihui; Xing, Gao; Liao, Minghong; Chang, Chin-Chen; Yau, Wei-Chuen
• Is Part Of: Expert systems with applications. Volume 159(2020)
• Publication Date: 2020-11-30
• Subjects: Information hiding -- Magic matrix -- Motion data; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2020.113516
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: We proposed to use 3D motion data as a novel medium for information hiding. We evaluate each skeleton joint and optimally select a group of skeleton joints. Table-of-Significance is proposed to reflect the significance of skeleton joint. Abstract: As an import branch of information security, image/video-based information hiding system has been investigated extensively in the past few decades. Meanwhile, 3D applications are increasingly popular in various areas such as animation, virtual, and augmented reality. This work explores the feasibility of data hiding in 3D applications and proposes a novel method to encrypt a message in the motion data of 3D virtual character. For the human visual system, it is challenging to notice the subtle details in dynamic scenarios at a fast frame rate. Therefore, the message encryption in 3D character movements offers an intuitive solution to information hiding. Our work can be divided into two stages. The first stage optimally selects a group of joints which introduces minimal modification on the global presentation of the character motion to maintain the maximum visual consistency between the embedding motion data and the original motion data. The second stage is to encrypt the secret data based on the Sudoku magic matrix and hide the data in the optimal selection of joints. The experimental results show that our system has a high hiding capacity and excellent visual quality.