An Improved Algorithm for Gaussian Noise Removal in Stego Image using Shift Invariant Wavelet Transform

IJSTE - International Journal of Science Technology & Engineering Volume 3 Issue 01 July 2016 ISSN (online): 2349-784X An Improved Algorithm for Gaussian Noise Removal in Stego Image using Shift Invariant Wavelet Transform Pooja Pandey Praveen Chouksey M. Tech Scholar Assistant Professor Department of Computer Science & Engineering Department of Computer Science & Engineering Dr. C.V. Raman University, Bilaspur, Chattisgarh-India Dr. C.V. Raman University, Bilaspur, Chattisgarh-India Abstract Images are often corrupted with noise during processing of retrieval from storage media. Image also corrupted during transmission and reception. In digital communication image transmitted in which information security is very important task so hidden information is embedded in to cover image called stego image. The information in the stego image is in the form of bit which even corrupted and error occurs in it. For purpose of photography when digital camera is used under poor lighting conditions, many black spot or dots can be observed in captured image. Stego image also degraded. Degradation comes from blurring as well as noise due to electronic and photo-metric sources. Reduction in Bandwidth of the image caused by the imperfect image formation process is Blurring. Keywords: Stego Image, LSB method, Image Denoising, DWT, Filtering I. INTRODUCTION In recent times, the need for digital communication has increased dramatically and as a result, the Internet has become essentially means more effective and faster communication to digital communication. At the same time, data on the Internet has become susceptible to copyright infringement, espionage, piracy, etc., which therefore requires secret communication. As a result, a new domain dedicated to information security has evolved and is known as data hiding. Steganography is a relatively novel addition to the area of data hiding but traces its origin to long ago in history. Steganography employs medium such as image, audio, video, or text file to conceal any information in it, so that does not draw any interest and looks like an innocuous medium. Cover medium such as digital image, video and photo became the obvious choice. Stego media are the media, which contain the secret information while cover media are the plain file. Recently, the images have been a popular choice as a means to cover mainly because of its redundancy in the representation and the ability to penetrate applications in daily life. Over the years, many algorithms have been proposed to hide data in images and developing new algorithms are a topic of current research. In this thesis, some of the most popular and effective among image steganography algorithms are analyzed for their mechanisms, advantages and disadvantages, which could be a valuable guide for future research scope openings. Gaussian Noise: II. NOISES & THEIR TYPES Most denoising algorithms assume zero mean additive white Gaussian noise (AWGN) because it is symmetric, continuous, and has a smooth density distribution It is an evenly distributed all over the signal component. That means a pixel in any image having Gaussian noise is the sum of random Gaussian distributed noise and true pixel value. Such kind of noise distributed Normally to the original image. The image is independent of the noise it is applied to. Speckle Noise: Speckle noise is a multiplicative noise. The value of noisy pixel is the multiplication of noise value and the original pixel value. Coherent imaging systems i.e laser, acoustics and SAR (Synthetic Aperture Radar) imagery affected by such type of noise. The source of this noise is attributed to random interference between the coherent returns Salt & Pepper Noise: Fat- tail distributed or impulsive noise is sometime called salt- and pepper noise or spike noise. The image having this type of noise contains bright Pixel in dark regions and dark pixel in bright regions. This noise is caused by analog to digital converter error and also by bit transmission error. Salt and pepper noise is an impulse type of noise. It has only two possible values. The All rights reserved by www.ijste.org 5

probability of each is typically less than 0.1. This noise giving the image a salt and pepper like appearance means corrupted pixels are set alternatively to the minimum or to the maximum value. Shot Noise: This type of noise comes due to variation in the number of photon sensed at a given exposure level in sensor. Shot noise has a root mean square value proportional to the square root of intensity of image, and noise at the different pixel is independent on each other. Shot noise follows the passion distribution, which is not very different from Gaussian distribution. This is also known as photon shot noise. There is also an additional DARK CURRENT SHOT NOISE which comes due to large leakage current in the sensor. It is also called the Poisson noise. Quantization Noise: It is also called uniform noise. This noise is caused by quantizing the pixels of a sensed image to a number of discrete levels. It has approximately uniform distribution. Brownian Noise: It is also known as brown noise or red noise. This type of noise is produced by the Brownian motion of the signal. Hence its alternative name is RANDOM WALK NOISE. Brownian noise comes under the category of fractal or 1/f noises. The mathematical model for 1/f noise is fractional Brownian motion. Fractal Brownian motion is a non-stationary stochastic process that follows a normal distribution. It is obtained by integrating white noise. III. IMAGE DENOISING Image denoising involves the modification of the image data to produce a visually high quality image. Removing of unwanted noise in order to restore the original image. The primary goal of noise reduction is to remove noise without losing much detail in an image. Image denoising is different from image enhancement. Image enhancement is an objective process and image denoising is a subjective process. In image denoising an attempt are made to recover an image that has been degraded by using prior knowledge of the degradation process. Image enhancement, on other hand manipulation of image characteristics to make it more appealing to human eye. There is some overlap between these two processes. IV. PROBLEM IDENTIFICATION In previous work we are taking a simple digital image without any secret data embedded in it. The algorithm only deals with simple images.the image denoising technique is used for Gaussian noise removal from the simple image. Limitation of Exisiting Work: The performance of communication system of only simple image not the stego image transported over a noisy wireless link The internet and multimedia communication in the present wireless age is need of everyone. Therefore network security is very important. The previous work deals only with simple images and denoising technique for it. Security is the major threat for network. If two persons need to exchange message secretly through data network, it is difficult in modern days as there are many techniques which captures and alter the message sent. V. METHODOLOGY In this work, a new technique of LSB steganography has been proposed which is an improvised version of one bit LSB technique. One of the reasons that intruders can be successful is that most of the information they acquire from a system is in a form that they can read and comprehend. Intruders may reveal the information to others, modify it to misrepresent an individual or organization, or use it to launch an attack. One solution to this problem is to use steganography. Stenography is a technique of hiding information in digital media. LSB Method: The LSB is the lowest significant bit in the byte value of the image pixel. The LSB based image steganography embeds the secret in the least significant bits of pixel values of the cover image. The concept of LSB Embedding is simple. It exploits the fact that the level of precision in many image formats is far greater that that perceivable by average human vision. Therefore, an altered image with slight variation in its colours will be indistinguishable from the original by a human being, Just by looking at it. All rights reserved by www.ijste.org 6

Fig. 3.1: Improved Algorithm Flowchart VI. RESULT & DISCUSSION In this research work an investigation has been made on suitability wavelet thresholding and translation invariant methods of image denoising to remove noise using orthogonal wavelet basis. Then shift invariant denoising method also evaluate in terms of PSNR and MSE. This chapter contains the results, obtained after following the discrete wavelet and shift invariant denoising algorithm as discussed in the previous chapter. The results have been demonstrated in the form of comparison tables. At the All rights reserved by www.ijste.org 7

chapter end, a graphical representation has also been done for a quick analysis of results. All the techniques have been tested for standard cameraman test images. The noisy image is obtained then this image has denoised by the following wavelet Family for both image denoising Technique Haar Db1 Symmlet Coiflet The PSNR value of the noisy image shows that higher the PSNR value is closer to the original image. As per analysis of table1, 2, 3 and 4 the PSNR value and MSE of noisy image for different wavelet analyser. All experiment has been done for camera man image with noisy PSNR (db): 20.3754, MSE: 596.4104 mean: 0.0, variance: 0.01. Snap Shots: Shows a snap shot of MATLAB gui which shows a true readings of MSE and PSNR value taken at the same time while the table has drawn. Shows all the snap shots as below: Fig. 4.1: Baby 1- LSB TECHNIQUE All rights reserved by www.ijste.org 8

Fig. 4.2: Baby 1 Image denoising Fig. 4.3: Baby 2 LSB TECHNIQUE All rights reserved by www.ijste.org 9

Fig. 4.4: Baby 2 Image denoising VII. CONCLUSION & FUTURE SCOPE The simulation analysis, the wavelet transform in stego image denoising in particular stationary images, can effectively remove Gaussian noise and improve SNR. With regard to complexity of stego image structure, shift invariant wavelet transform denoising can play the advantages compared to traditional denoising, invariant wavelet can better demonstrate its advantages. From the simulation results, we also obtain that use the principle of sure shrink threshold can effectively reduce noise, and can retain a useful component of image. Translation invariant capability of Attenuating Gibbs oscillation and adaptation to discontinuities gave an advantage to provide better result. Future Work: However selection of the actual denoising procedure plays an important role, it is essential develop to experiment and compare the methods. Although our proposed algorithm give the encouraging denoising result In future research, it is also possible to get the better denoising performance. For example it is also possible to develop thresholding function which is more coherent and better related to neighboring coefficient and also represent better hierarchical dependency between different wavelet decomposition level.finally it is also possible to combine our method with other to get high quality of result. Along with these points we can also consider some points as a future work. REFERENCES [1] R. C. Gonzalez and R.E Wood. Digital image processing Printice Hall, Upper saddle river, N.J 2 nd edition 2002. [2] S.Kother Mohideen, Dr. S. Arumuga Perumal and Dr. M.Mohamed Sathik, Image De-noising using Discrete Wavelet transform, IJCSNS International Journal of Computer Science and Network Security, VOL.8 No.1, January 2008. [3] D. Giaouris and J.W. Finch Denoising using wavelets on electric drive applications,school of Electrical, Electronic & Computer Engineering, Newcastle University, Newcastle upon Tyne NE1 7RU, UK 2007. [4] Vidhyalavanya. R, and Madheswaran.M, An Improved Denoising Algorithm using Parametric Multiwavelets for Image Enhancement, International Journal of Advanced Science and Technology Vol. 16, March, 2010. [5] Sachin D Ruikar and Dharmpal D Doye, Wavelet Based Image Denoising Technique, International Journal of Advanced Computer Science and Applications,Vol. 2, No.3, March 2011. [6] S.Sudha, G.R.Suresh and R. Sukanesh, Speckle Noise Reduction in Ultrasound Images by Wavelet Thresholding based on Weighted Variance, International Journal of Computer Theory and Engineering, Vol. 1, No. 1, pp.1793-8201, April 2009. All rights reserved by www.ijste.org 10

[7] Abhishek Raj, Alankrita, Akansha Srivastava, and Vikrant Bhateja, Computer Aided Detection of Brain Tumor in Magnetic Resonance Images, International Journal of Engineering and Technology, Vol. 3, No. 5, October 2011. [8] Eric J. Balster, Member, Yuan F. Zheng, Fellow, and Robert L. Ewing, Senior Member, Feature-Based Wavelet Shrinkage Algorithm for Image Denoising, IEEE Transactions On Image Processing, VOL. 14, NO. 12, DECEMBER 2005. [9] Jean-Luc Starck, Jalal Fadili, and Fionn Murtagh, The Undecimated Wavelet Decomposition and its Reconstruction, IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 16, NO. 2, FEBRUARY 2007. [10] Manish Goyal and Glenetan singh sekhon, Hybrid Threshold Technique for Speckle Noise Reduction using wavelets for Grey scale images, IJCST Vol. 2, Issue 2, June 2011. [11] Sachin Ruikar and D D Doye, Image Denoising Using Wavelet Transform, 2010 International Conference on Mechanical and Electrical Technology. [12] S.S.Patil, A.B.Patil, S.C.Deshmukh and M.N.Chavan, Wavelet Shrinkage Techniques for Images, International Journal of Computer Applications (0975 8887) Volume 7 No.1, September 2010. [13] V.V.K.D.V.Prasad, P.Siddaiah, and B.Prabhakara Rao, A New Wavelet Based Method for Denoising of Biological Signals, IJCSNS International Journal of Computer Science and Network Security, VOL.8 No.1, January 2008. [14] Erik B. Sudderth, Michael I. Jordan, Jyri J. Kivinen, Image denoising with nonparametric hidden markov trees, 2007 IEEE international conference on image processing. [15] S. M. Mahbubur Rahman, M. Omair Ahmad, M. N. S. Swamy, Locally adaptive wavelet based image denoising using the gram-charlier prior function, IEEE International Conference on Image processing 2007. [16] Nilamani Bhoi and Dr. Sukadev Meher, Total Variation based Wavelet Domain Filter for Image Denoising, ICETE First International Conference on Emerging Trends in Engineering and Technology 2008. All rights reserved by www.ijste.org 11