Abstract: method of embedding secret message-extracting message into

Abstract: Cloud Computing is a powerful, flexible, cost efficient platform for providing consumer /it services over the Internet. However Cloud Computing has various level of risk because most important information is maintained and managed by third party vendors, which means harder to maintain security for user’s data .Steganography is one of the ways to provide security for secret data by inserting in a image or video. In this most of the algorithms are based on the Least Significant Bit (LSB), but the hackers easily detects it embeds directly. An Efficient and secure method of embedding secret message-extracting message into or from color image using Artificial Neural Network will be proposed. The proposed method will be tested, implemented and analyzed for various color images of different sizes and different sizes of secret messages. The performance of the algorithm will be analyzed by calculating various parameters like PSNR, MSE and the results are good compared to existing algorithms.

Keywords: Artificial Neural Network, Steganography, PSNR,MSEIntroduction:In cloud computing environment, the security of data is the important parameter. Various approaches like cryptographic techniques, watermarking and hiding strategies were created with a specific end goal to secure the data. It was insufficient to secure the substance of the hidden message from outside phishers and programmers because data is maintained by a third party from different places at different locations. So it is necessary to have a method which can hold the presence of hidden message secrecy.

Don't waste your time
on finding examples

We can write the essay sample you need

The system used to actualize this is called as Steganography. Steganography can be characterized as a workmanship or art of invisible writing, which enabled hiding of secret data into another media. The word Steganography is comes from the Greek words “Stegos” meaning “cover” and “Grafia” meaning “written work” signifying it as “invisible writing”. Consequently steganography is a novel procedure which is utilized to shroud the hidden message and keep it from identification.

The proposed technique enable embedding of secret message into single color cover image using ANN in spatial domain along with encryption to enhance security level. The Organization of the paper is as follows. In section 2 the related work is discusses. In section 3 our proposed method is described. Finally the results are presented in section 4.II.

RELATED WORKSIn Suneetha D et.al’s 1 has effectively proposed a new algorithm using LSB based image stegnaography in which secret data is hidden in the combinations of two bits . The secrete message is converted into binary is in the form of 0 and 1. For hiding 0 bit use some combination of two bits and for I use another combination of bits. Hence it’s difficult for attackers to retrieve data. Results reveal that high security is provided with acceptable PSNR valuesIn Kiran Kumar R.

et.al 2 demonstrated a performance analysis on LSB technique by embedding the secret image bits in Fibonacci edge based pixels of cover image and corresponding PSNR values are notedDown. Results emphasize that PSNR value gradually decreases with increase in the embedding bits.

In Kiran .R et.al’s 3 proposed steganography approach which aims to increase theembedding capacity along with stego image quality by using an optimal LSBs method, on the basis partion based edge pixels decided which are best suitable for embedding secret data. Methodologyproposed is based on dividing the image into 9 equal parts, one for embedding the secret message and applies change to the value of some bits that have the secret bits obtained by the simple form of LSBSubstitution technique. The advantages of the presented method is increasing the amount of secret message in each pixel of the cover image and improving the quality of the stego image.

In Suneetha .D 4 demonstrated a spatial domain technique of hiding secret image bits in different parts of a cover image .This technique helps to embed the secret data with minimum distortion to the cover file, By using this algorithm it is used for construction of blind steganalysis and accurate targeted method for various forms of images.

Experiment analysis of new method shows PSNR is greater than other LSBs replacement.In Siddharth Singh et.al’s 5 proposed a robust steganography approach using DCT, chaotic sequence generator and Arnold transform, here the random sequence generator is used for hiding data in middle band DCT coefficient of cover image is generated using chaotic system.

Securityfactor is improved by using Arnold transform to scramble hidden data before hiding. Experimental analysis demonstrate algorithm achieve more secure, robust to JPEG compression, LPF filtering and various crop attacks then various other approaches using DCT domain.In Sadeq AlHamouz et.al’s 6 proposed approach using back propagation neural network. In this papers two images are used one is secret images and the cover images, both are color images.

The algorithm uses two different phases one is data embedding process and other one is data extracting process. The hiding bit positions are calculated using Fibonacci linear feedback shift register. The experimental results are compared with several exciting algorithms that high PSNR value is achieved with more processing time.III. MATERIALS AND METHODS3.

1. Artificial neural network approachIn this proposed Steganography algorithm cascaded feed forward neural network is used along with it Levenberg Marquardt training algorithm. The cascade feed forward neural networks are similar to feed forward networks. It consists of several layers. The first layer has a connection from the network input.

Each sussequeent layers has the connection from the previous layer. The function newcf is used to create cascade forward networks. For example a four layer network has connections from layer 1 to layer 2, layer 2 to layer 3, layer 3 to layer 4 and layer 1 to layer 4. The four layer network has connections from the input layer to all four layers. The additional layers might improve the speed.


I'm Owen!

Would you like to get a custom essay? How about receiving a customized one?

Check it out