A Novel Secure Color Image Steganography Based on Denoising Methods in DWT and IWT Techniques

Steganographic techniques have been used for centuries. Using stegnography, secret message can be embedded inside a piece of unsuspicious information and sending it without anyone’s knowledge about the existence of the secret message. This paper introduces A novel secure color image steganography utilizing image denoising algorithm by wavelet thresholding. Wavelet transform is employed to represent represent spatial domain image into time frequency domain. At first The preprocessed color cover image is transmitted from RGB to YIQ color model in order to extract its I-component. At the same time, Arnold Transformation is performed to scramble the secret message then both cover-I and secret message are decomposed using Discrete Wavelet Transform (DWT) and Integer Wavlet Transform(IWT). In general secret data is hidden in noisy components of cover medium, this implies that calculating a threshold based on wavelet coefficients of cover image to determine the noisy components. Afterwards the normalized secret message Alpha blending with the cover-I. This proposed method improves the capacity, Peak Signal to Noise Ratio (PSNR), provides high security and certain robustness. Keywords-DWT, IWT, Denoising Methods, Arnold Transformation, Image quality Metrics. I.INTRODUCTION With the recent advances in multimedia communications and its influence in our electronic world, the importance of information security has been dramatically increased. Existing technologies in the field of information security systems offer concealing the occurrence of communication for anyone except the intended recipient. In this way, steganography provides a reliable solution for embedding a secrete data into a cover media imperceptibly. Basically, the ultimate objectives of steganography are undetectability, robustness, and high capacity of the hidden data that separate it from related techniques such as watermarking and cryptography. Also, the hidden message can be recovered using appropriate keys without any knowledge of the original cover media. In general, steganography algorithms usually struggle with achieving a high embedding rate, large capacity, and good imperceptibility. This work aims to present an efficient steganography technique in image files. The most common steganographic techniques in digital images focus on spatial domain methods-which generally use a direct Least Significant Bit (LSB) replacement techniqueand frequency domain methods such as Discrete Cosine Transform (DCT), Fourier Transform (FT), Discrete Wavelet Transform (DWT) and so on. II. RELATED WORK It is presented source of the related work in our proposed work. H.S. Manjunatha Reddy & K.B. Raja [1] have proposed a Wavelet based Non LSB Steganography (WNLS) in which the cover image is segmented into 4 x 4 cells and DWT/IWT is applied on each cell. The 2 x 2 cell of HH band of DWT/IWT are considered and manipulated with payload bit pairs using identity matrix to generate stego image. The key is used to extract payload bit pairs at the destination. PSNR Rate is high for this proposed system. Weiqi Luo et al., [2] have proposed an edge adaptive scheme which can select the embedding regions according to the size of secret message and the difference between two consecutive pixels in the cover image. The proposed method enhances the security as well as preserving higher visual quality of stego images. Namita Tiwari1 & Madhu Shandilya [3] have proposed a technique to investigate two methods of RGB image steganography one is pixel indicator technique and other is triple-A algorithem. They uses the same principle of LSB, where the secret is hidden in the least significant bits of the pixels, with more randomization in selection of the number of bits used and the color. This randomization is expected to increase the security of the system and also increase the capacity. These techniques can be applied to RGB images where each pixel is represented by three bytes to indicate the intensity of red, green, and blue in that pixel. This work showed attractive results especially in the capacity of the data-bits to be hidden with relation to the RGB image pixels. Ashish Chawla & Pranjal Shukla [4] have proposed a modified secure and high capacity based steganography scheme of hiding a large-size secret image into a small-size cover image. Matrix Rotation is performed to scrambles the International Conference on Engineering Trends and Science & Humanities (ICETSH-2015) ISSN: 2348 – 8387 www.internationaljournalssrg.org Page 51 secret image. DWT is performed in both images and followed by Alpha blending operation. Then the Inverse Discrete Wavelet Transformation (IDWT) is applied to get the stego image. Proposed algorithm for modified steganography is highly secured with certain strength in addition to good perceptual invisibility. N Sathisha et al., [5] have proposed Non Embedding Steganography using Average Technique in Transform domain (NESATT). The Lifting Wavelet Transform (LWT) is applied on both cover image and payload image. The Diagonal band (CD) of cover image and Approximation band (PA) of payload are segmented into N x N blocks. The average value of each blocks are scale downed by key and embedded into corresponding block of Horizontal band (CH) of cover image. The inverse LWT is applied on stego object. The capacity and PSNR values are high in the case of proposed algorithm M. Ravi Shankar Reddy & Sri. J. Swami Naik [6] have proposed a technique for the hiding of text messages into a digital image in spatial domain. In our approach in each pixel two bits of message part is embedded in such a way that the fourth bit place, second bit plane and also the least significant bits are allowed. Proposed message hiding approach is robust and very useful in real world applications. Yambem Jina Chanu et al., [7] have proposed an Adaptive Stegnographic technique, which embeds secret message bits in the edges of the images is proposed. Manisha Boora & Monika Gambhir [8] have proposed a scheme for hiding a larger size secret-image into smaller size coverimage. Arnold Transformation is performed to obtain scrambled secret image. DWT is performed on both cover image and secret image and this is followed by alpha blending operation. This proposed algorithm is highly secured with good perceptual invisibility. III. METHODOLOGY A. Pre-processing Histograms are functions describing information extracted from the image. The histogram function is defined over all possible intensity levels. For each intensity level, its value is equal to the number of the pixels with that intensity. Adaptive histogram equalization uses the histogram equalization mapping function supported over a certain size of a local window to determine each enhanced density value. Therefore regions occupying different gray scale ranges can be enhanced simultaneously. B. Discrete Wavelet Transform (DWT) This transform is based on filtering the image coefficients on a raw-by-raw and column-by-column basis repetitively. Two-dimensional wavelets decomposition is same as onedimensional one which leads to a decomposition of approximation coefficients at level j in four components: the approximation (LL) at level j+1, and the details in three details in three orientations; horizontal (HL), vertical (LH), and diagonal (HH) as shown in Fig. 1. The low-pass and high pass filters of the wavelet transform naturally break a signal into similar and discontinuous or rapidly-changing sub signals, respectively. The slow changing aspects of a signal are preserved in approximation which includes low-pass band. On the other hand, a detail band consists of high frequency coefficients, which contains edge details of spatial domain image, noise components, and the quickly changing parts. Therefore we can embed data in the regions that human vision is less sensitive to, such as the high resolution detail bands (LH , HL ,and HH) where generally contains noisy components. As a matter of fact, the basic idea of denoising is to reduce high frequencies, because noise is always presented by high frequencies. As a result, most denoising techniques tend to suppress high frequencies. In steganographic point of view, these noisy wavelet coefficients have the potency to hide secret data. C Arnold transform Arnold Transform is commonly known as cat face transform and is only suitable for N×N images digital images. It is defined as