Color images steganalysis using rgb channel geometric transformation measures

Abstract

In recent years, information security has received a great deal of attention. To give an example, steganography techniques are used to communicate in a secret and invisible way. Digital color images have become a good medium for digital steganography because of their easy manipulation as carriers via Internet, e‐mails, or used on websites. The main goal of steganalysis is to detect the presence of hidden messages in a digital media. The proposed method is a further extension of the authors' previous work: steganalysis based on color feature correlation and machine learning classification. Fusing features with those obtained from color‐rich models allows increasing the detectability of hidden messages in the color images. Our new proposition uses two types of features, computed between color image channels. The first type of feature reflects local Euclidean transformations, and the second one reflects mirror transformations. These geometric measures are obtained by the sine and cosine of gradient angles between all the color channels. Features are extracted from co‐occurrence correlation matrices of measures. We demonstrate the efficiency of the proposed framework on three steganography algorithms designed to hide messages in images represented in the spatial domain: S‐UNIWARD, WOW, and Synch‐HILL. For each algorithm, we applied a range of different payload sizes. The efficiency of the proposed method is demonstrated by the comparison with the previous authors work and the spatial color‐rich model and color filter array‐aware features for steganalysis. Copyright © 2016 John Wiley & Sons, Ltd.

Color Image Steganalysis Based On Steerable Gaussian Filters Bank

“…Abdulrahman et al [2] proposed the RGB Geometric Color Rich Model (GCRM). The authors show that if one channel has been affected by a steganography method, the inter channel correlation will measure the local modifications.…”

“…Those geometrical measurements allow to generate 6000 features, based on local Euclidean and mirror transformations, when using co-occurrence matrices with a fixed truncation T =1 and different values for the quantization q ∈ {0.1, 0.3, 0.5, 0.7, 0.9, 1}. Concatenate these features with those from CRM [10] increases the detectability of hidden messages in color images [2], [23].…”

“…Three algorithms have been tested: S-UNIWARD, WOW and Synch-HILL with different relative payloads sizes: {0.01, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5}. Furthermore, the proposed method is tested against three other approaches: CRM [10], CFARM [9] and GCRM [2]. Table 2 demonstrates that the proposed method registered the highest performance.…”

“…For these three methods, the machine learning algorithm is the Ensemble Classifier [16], and the rich features are: the Spatial Color Rich Model abbreviated to CRM [10], the CFA-aware Rich Model abbreviated to CFARM [9], and the RGB Geometric Color Rich Model abbreviated to GCRM (RGB for the Red, Green and Blue channels) [2]. Note that among these three features, the GCRM which is an extension of the CCRM (Correlation Color Rich Model [1]) seems to be the equivalent or better than the CRM and CFARM when recent adaptive grey-level embedding algorithms are used for embedding independently in each color channel.…”

“…Section 2 describes the latest recent methods in color steganalysis by recalling the CRM [10], the CFARM [9] and the GCRM [2]. Then, we present a detailed description of our proposed method in Section 3; the steerable Gaussian filters computation is demonstrated and the feature set is presented.…”