Data Embedding Technique for Image Steganography in Cloud Computing

Abstract

The cloud computing security is the set to protect the data, applications stored on the cloud. The security on cloud is the concerned issues by the cloud providers as well as to the cloud users as the cloud network is prone to number of attacks such as network-based, virtual machine based, storage based, and application based attacks etc. Data embedding is the technique used for hiding the information by using the steganography and cryptography technique in a hybrid way. Different techniques have been used for providing security to the digital data by using cryptography and steganography techniques. The cryptography algorithm encrypts the sensitive data but leave the mark that data is processed. On the other side, in steganography, the sensitive data is hide behind the cover media and leave no mark and imperceptible to un-authorized parties that data is processed. In steganography, security is dependent on how much visual quality is preserved after data embedding. Hence, Least Significant Bit (LSB) is the most used technique for data hiding in which cover media LSB bits are replaced with data bits and required 8 pixels. To improve the capacity, 2/3/4 bits of LSB has been replaced with data bits that influenced the visual quality of cover media. Hence, a layer of cryptography is added on the sensitive data before data hiding in the cover media that provides confidentiality and imperceptibility. The main goal of this dissertation is to provide high level of security for sensitive data that is stored on the cloud database. In the proposed technique, Firstly, the sensitive data is encrypted using Advanced Encryption Standard(AES) algo- rithm. Next, in steganography, cover media is processed to form nonogram puzzle inside the symmetric shapes. The selection of nonogram puzzle and symmetric shape is done on the fly based on output of random number generator. Lastly, the encrypted information is hidden inside the nonogram puzzle by using 2-bit LSB technique. For experimental analysis on standard dataset images, the pro- posed technique is simulated and analyzed in terms of visual quality, PSNR and Embedding capacity.