Article Preview
Top1. Introduction
Embedding secret randomly or deterministically in one or more of the three channels viz. RGB in comparison to direct embedding can significantly improve the secret message against attacks during transmission and it is reported to be highly robust against plain image attack (Murillo-Escobar et al., 2015). Information of the three channels viz.RGB based on pixel value differencing (PVD) has been reported (Prasad & Pal, 2017). Implantation in one bit of each byte from all of the R, G and B planes of a color cover image are incorporated. More secure technique is reported with random substitution in one of the 4 LSB positions. (Saikia & Tuithung 2019).
An RGB image steganography in three color planes (Red, Green, Blue) have been reported by (Ghosh, Debnath & Banik, 2019). Color value of an image pixel has been hidden within the same color plane of the cover image using Discrete Cosine Transformation (DCT) to enhance security. Introducing chaos dynamics in the process of encryption adds to its protection. Chaotic functions are nonlinear dynamic functions used to generate random sequences with enhanced confidence and are suitable in information security due to some value-added characteristics e.g. ergodicity, noise like structure and sensitivity to initial conditions (Kwok & Tang, 2007; Kocarev & Lian, 2011).
Recently, various techniques are available using a number of chaotic functions. Combination of LSB substitution and Hamming code in steganography has been used to generate steganographic schemes (Saidi et al., 2017). Skew tent map has been used for generating random sequences and used as a key step in embedding procedure. Results obtained indicate that the technique is immune against attacks. One dimensional logistic map in two different techniques (Alam et al., 2017; Alam et al., 2018) has been used with neural network and meta heuristic Ant Colony Optimization (ACO) algorithms for edge detection in image. Secret data is then embedded in edge positions to improve imperceptibility. Both the techniques are applied to secure confidential information through steganography.
Image steganography for hiding Turkish text work has been reported in (Kasapbaşi, 2019) using logistic map and Huffman compression technique. The use of logistic map in LSB image steganography has been reported in (Rajendran & Doraipandian 2017). Major advantage of the chaos-based techniques is improved security with highly random chaotic sequences for generation of keys using chaotic functions which are nonlinear yet deterministic. However, randomness analysis of the sequences generated with chaotic map was not reported in most of these literatures.
More secured algorithms can be designed with 3D chaotic maps. 3D piecewise logistic map which is expanded form of 1D logistic map, has shown better performance in terms of chaotic and ergodic properties (Sahari & Boukemara 2018). A 3D map based on Cosine-Arcsine system, for generating sequences with strong chaotic characteristics, having maximum Lyapunov exponent (λ)= 20.58 (Sharif et al., 2017). For a nonlinear dynamical system, it is mathematically known that if λ > 0 the orbit is unstable and chaotic. Hence, nearby points, no matter how close, will diverge to any arbitrary separation. This technique is then applied to hide confidential information in the form of bits, in cover image using spatial steganography (Sharif et al., 2017.
A transform domain image steganographic method based on 3D sine chaotic map, resulted in good quality stego-image in terms of performance in embedding and extraction processes. The results are reported to be quite robust against the common attacks (Valandar et al., 2019). Recently, an approach for data-hiding in videos using 3D Chaotic Maps has been reported (Narayanan et al., 2019). The authors represented a video as a 3-dimensional image, with the third axis constituting the frames of the video.