High Capacity Secure Image Steganography Based on Contourlet Transform Kolsoom Shahryari 1 , Mehrdad Gholami 2 1 Department of Computer and IT, University of Kurdistan Sanandaj, Kurdistan, Iran Shahryari.kolsoom@gmail.com 2 Department of Electrical Engineering, AmirKabir University of Technology Tehran, Tehran, Iran Mehrdad_gholami@aut.ac.ir Abstract In this paper we propose an image steganography technique which embeds secret data without making explicit modifications to the image. The proposed method simultaneously provides both imperceptibility and undetectability. We decompose image by contourlet transform and determine nonsmooth regions. Embedding data in these regions cause less degradation in image quality. Contourlet sub-bands are divided into 3×3 blocks. Central coefficient of each block is considered for embedding if they belong to edgy regions. Experiments show that this method can achieve high embedding capacity while remains undetectable by Farid's universal steganalysis technique. Keywords: Contourlet Transform, Steganalysis, Steganography. 1. Introduction Steganography is the art and science of secret communication, which aims to hide a secret data securely in transmission. So, steganography algorithms should hide a message in the media such as audio, image and video in such a way that the very presence of the hidden message remains secret. Contrary to steganography methods, there are steganalysis techniques that try to detect secret communication. The most important requirements of a steganography system is security that include undetectability and imperceptibility. Literature reviews show that there are many steganography methods that use Discrete Cosine Transform (DCT) and Discrete Wavelet Transform (DWT) for hiding secret message in images. F5 [1],Outguess [2],Model-based [3], Perturbed Quantization[4] and YASS [5] hide secret message by modification of DCT coefficient. Another steganography methods based on wavelet transform have been proposed. In [6] a wavelet domain steganography algorithm that is compatible with JPEG 2000 was proposed. In this method the secret data embedded in quantized wavelet coefficient. [7], [8], [9] were proposed another steganography methods based on wavelet. Recently some research about steganography in contourlet domain has been done. Sajedi in [10] proposed an adaptive steganography method based on contourlet transform that provides large embedding capacity. The contourlet coefficients with larger magnitude that correspond to the edges are selected for embedding. This method is superior to a similar wavelet based approach. In this paper, we propose a new steganography method that uses contourlet transform for hiding secret data in image. In this method, we first apply contourlet transform on image and then hide secret data in proper contourlet coefficients. The human visual system has been tuned so as to capture the essential information of a natural scence using a least number of visual active cells. So, an efficient image representation should be based on a local, directional and multiresolution expansion. In this method we embed data in non-smooth region of carrier image. So, the visual degradation is minimal. We examine our algorithm by a well-known steganalysis method and we found that it couldn't discriminate between cover and stego images with rate better than random guess. In section 2, we describe our algorithm. The experiments results are given in section 3. At last, in section 4, we get conclusions. ACSIJ Advances in Computer Science: an International Journal, Vol. 2, Issue 4, No.5 , September 2013 ISSN : 2322-5157 www.ACSIJ.org 62 Copyright (c) 2013 Advances in Computer Science: an International Journal. All Rights Reserved.