We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Polynomial fitting: enhancing the stego quality of DCT-based Steganography schemes.
- Authors
Baziyad, Mohammed; Rabie, Tamer; Kamel, Ibrahim; Benkhelifa, Mahdi
- Abstract
The strong energy compaction property of the Discrete Cosine Transform (DCT) has been widely proposed as a solution to the capacity-quality trade-off problem in steganography. However, several recent steganography schemes did not fully exploit the strong compaction property of the DCT and hide in the whole large region that contains the less significant coefficients. Instead, hiding regions were usually approximated by regular regions, such as squares or rectangles. The reason was justified to reduce the communication overhead required to be sent to the receiver to be able to locate and extract the secret data. For example, for a squared hiding region, the receiver will only need the information of the dimension of the squared hiding area to be able to allocate and extract the secret data. This paper proposes a novel scheme to maximize the region used to hide the secret message, yet keep the communication cost within the required limit. The proposed scheme provides a mathematical model for a control mechanism for trading the communication cost with the stego quality. To reduce the communication cost, the proposed scheme suggests establishing a polynomial curve between the significant and less significant DCT coefficients, and then hide under the curve in the full area of the less significant DCT coefficients. In this case, only the polynomial coefficients are required to be sent to assist the receiver to locate the secret data. In addition to the reasonable communication cost, hiding in the whole area of the less significant coefficients will ensure a smooth transition between the DCT coefficients and the hidden data mimicking the effect of introducing a Gaussian function in the frequency domain which will smooth noisy areas in the stego image. Moreover, to maximize the stego quality, the secret data is embedded by superimposing a scaled-down version of the secret signal on the less significant coefficients instead of the conventional method of replacing the less significant coefficients with the secret data. This hiding procedure ensures maintaining the structure of the less significant DCT coefficients instead of completely removing them. Therefore, the experimental results have proven that the proposed hiding scheme is able to effectively handle the capacity-quality trade-off and achieve superior performance over competitive hiding schemes in terms of the hiding capacity and the stego quality.
- Subjects
CRYPTOGRAPHY; DISCRETE cosine transforms; POLYNOMIALS; GAUSSIAN function; MATHEMATICAL models
- Publication
Multimedia Tools & Applications, 2022, Vol 81, Issue 30, p43999
- ISSN
1380-7501
- Publication type
Article
- DOI
10.1007/s11042-022-13004-w