Cryptography Reference
In-Depth Information
image quality. In 2004, Tzeng, Yang and Tsai proposed a method [13] based on
a color-ordering relationship and a color-mapping function. The first is used
to rearrange the order of the color according to the luminance of the color, and
the second is used to embed data by the ordering of the color. The proposed
method also considers the necessity of reducing local image content changes to
achieve a good compromise between data-hiding capacity and image quality.
Du and Hsu proposed an adaptive data hiding scheme [14] that embeds the
secret data in the compressed codes of Vector Quantization (VQ) [15, 16].
This method uses a combination number to embed secret data. It offers a
high embedding capacity, but it does not fit palette images. The modified
color is easily detectable by human eyes. In this chapter, we propose a new
scheme that offers a high embedding capacity and good image quality. The
clustering technique is adopted in our proposed scheme for high embedding
capacity. In addition, we use two embedding schemes, the Cluster Ordering-
and-Mapping Technique and combination technique. The remainder of this
chapter is organized as follows. In Section 5.2, we give a brief review of Tzeng
et al.s method. Our proposed method follows in Section 5.3. The experimental
results are shown in Section 5.4. The conclusions are in Section 5.5.
5.2 Related Works
We briefly review some data-hiding schemes. In Subsection 5.2.1, we shall
introduce codeword grouping scheme by Chang et al. [12], and review the
color ordering and mapping developed by Tzeng et al. [13] in Subsection 5.2.2.
5.2.1 The Codeword Grouping Scheme
Chang et al. proposed an Adaptive Steganography for index-based images
using a codeword grouping scheme that supports Adaptive Hiding Capacity.
The palette-based image is one kind of Index-Based Images. The Codeword
Grouping Scheme will group the colors of the palette into many sub-clusters
according to the distance between colors that is less than the grouping thresh-
old. Since the number of sub-cluster members determines the hiding capacity,
the sub-clusters with larger members will be grouped first. Here, the number
of sub-cluster members is restricted to the power of 2. From that, the colors
of the palette are grouped into many sub-clusters with difference members
according to difference grouping thresholds. For example, a palette with 256
colors will be grouped into 4-member, 2-member and 1-member sub-clusters
according to grouping thresholds T
M 4
, T
M 2
. If the colors belong to the 4-
member sub-cluster, the distance in between of them is less than T
M 4
.Ifthe
colors belong to the same 2-member sub-cluster, the distance between them
is less than T
M 2
. The remaining colors are grouped individually as 1-member
sub-clusters. In the embedding procedure, the number of sub-cluster members
indicates how many secret bits can be embedded. If the number of sub-cluster
