Comparison of LSB and Proposed ModifiedDWT Algorithm for Image Steganography

Abstract

Information is wealth of any organization and in present era where information is transferred through digital media and internet, it becomes a top priority for any organization to protect this wealth. Steganography is one such technique in which presence of secret message cannot be detected and we can use it as a tool for security purpose fortransmitting the confidential information in a secure way.This work aims to do comparison betweenthe Least Significant Bit (LSB) Insertion algorithm and proposed Modified Discrete Wavelet Transform (DWT)algorithm. Least Significant Bits (LSB) Insertion method is a spatial domain technique in which the cover-image pixelsare manipulated to embed the secret information. This techniqueis simple and easy to implement.The Proposed Modified Discrete Wavelet Transform (DWT) algorithm in this work is a frequency domain technique. This proposed modified method hides secret bits in multiple matrixes obtained by fragmenting cover and secret image into several parts making sure that the embedding impact on the cover image is minimum and secrecy is high with pinnacle peak signal to noise ratio (PSNR). This proposed modified DWT algorithm divides the image into 4 sub bands which laterdivides in ‘S’, ‘U’ and ‘V’ matrix spaces with the help of Singular Value decomposition (SVD) each for R,G,B color spaces in order to embed the secret informationand generating the decryption keysmaking information embedding more firm.The performance of LSB insertion and proposed modified DWT algorithm are analysed on the basis of the parameters mean square error (MSE) and PSNRwhich shows that thePSNR values of encrypted and decrypted images when compared with standard results [11] hashigher PSNR value for proposed modified DWT algorithm than LSBinsertion algorithmindicating the good quality of encrypted and decrypted image with the increment in the security.