Edge Adaptive Image Steganography Based on LSB Matching Revisited. Article ( PDF Available) in IEEE Transactions on Information Forensics. In this paper, we expand the LSB matching revisited image steganography and propose an edge adaptive scheme which can select the. Journal of Computer Applications (JCA) ISSN: , Volume IV, Issue 1, Edge Adaptive Image Steganography Based On LSB Matching Revisited 1 .
|Published (Last):||21 December 2010|
|PDF File Size:||4.30 Mb|
|ePub File Size:||1.57 Mb|
|Price:||Free* [*Free Regsitration Required]|
Skip to main content. Log In Sign Up. Edge adaptive image steganography based on LSB matching revisited. Usually, the larger payload embedded in a cover, the more of embedding positions within a cover image mainly depends on a pseudorandom number generator without considering the detectable artifacts would be introduced into the stego. In many relationship between the image content itself and the size of the applications, the most important requirement for steganography secret message.
In this paper, we consider dig- low security based on our analysis and extensive experiments, especially for those images with many smooth regions.
In this ital images as covers and investigate an adaptive and secure data paper, we expand the LSB matching revisited iamge steganog- hiding scheme in the spatial least-significant-bit LSB domain.
In embedding regions according to the size of secret message and this embedding scheme, only matchinf LSB plane bbased the cover image the difference between two consecutive pixels in the cover image. When the dorandom number generator PRNG. As a result, some struc- embedding rate increases, more edge regions can be released tural asymmetry never decreasing even pixels and increasing adaptively for data hiding by adjusting just a few parameters. Statistically, the probability of increasing or decreasing for each modified pixel value is the same and so I.
S It aims to embed secret data into a digital cover media, such as digital audio, image, video, etc. Up to now, several steganalytic algorithms e. On the other side, steganalysis aims to expose the presence of In , Harmsen and Pearlman showed that LSBM works as a hidden secret messages in those stego media.
If there exists a low-pass filter on the histogram of the image, which means that steganalytic algorithm which can guess whether a given media the histogram of the stego image contains fewer high-frequency is a cover or not with a higher probability than random guessing, components compared with the histogram of its cover.
Based on the steganographic system is considered broken. In practice, this property, the authors introduced a detector using the center of mass COM of the histogram characteristic function HCF.
Manuscript received October 16, ; accepted December 13, The associate down-sampled image and the adjacency histogram instead of editor coordinating bbased review of this manuscript and approving it for publica- the traditional histogram, which are effective for grayscale im- tion was Dr. In a recent work , Li et al.
Edge Adaptive Image Steganography Based on LSB Matching Revisited – Semantic Scholar
The experimental results showed that the Color versions of one or more of the figures in this paper are available online at http: New Jersey Institute of Technology. Downloaded on May 27, at By doing this, these methods can spread the feature based on the alteration rate of the number of neighbor- secret data over the whole stego image randomly even at low hood pixel values.
The experimental results demonstrated that embedding rate. However, based on our analysis and extensive the method was more effective on uncompressed grayscale im- experiments, we find that such embedding schemes do not ages.
Besides those specific detectors, some universal stegana- perform well in terms of the security or visual quality of the lytic algorithms such as , , and  can also be used for stego images. Similar to the problem of pixel values independently, LSB matching revisited LSBMR cover image selection , we should preferentially use those  uses a pair of pixels as an embedding unit, in which the LSB subimages with good hiding characteristics while leaving the of the first pixel carries one bit of secret message, and the re- others unchanged.
Therefore, deciding how to select the regions lationship odd—even combination of the two pixel values car- is the key issue of our proposed scheme. Generally, the regions ries another bit of secret message. In such a way, the modifi- located at the sharper edges present more complicated statis- cation rate of pixels can decrease from 0.
It is also shown that such a new In this paper, we propose an edge adaptive scheme and apply scheme can avoid the LSB replacement style asymmetry, and it to the LSBMR-based method. The experimental results eval- thus it should make the detection slightly more difficult than the uated on thousands of natural images using different kinds of LSBM approach based on our experiments.
Section III shows the details of its neighbors. Until now, several edge adaptive schemes such as data embedding and data extraction in our scheme. Section IV — have been investigated. In , Hempstalk proposed presents experimental results and discussions. Finally, con- a hiding scheme by replacing the LSB of a cover according cluding remarks and future work are given in Section V.
Although this method can embed most secret data along sharper edges and can achieve more visually impercep- II. In , Singh et al. As expose the limitations of these existing schemes. Finally we mentioned in , the maximum embedding capacity of such propose some strategies to overcome these limitations.
Furthermore, the In the LSB replacement and LSBM approaches, the embed- threshold is predetermined and thus it cannot change adap- ding process is very similar. Given a secret bit stream to be tively according to the image contents and the message to be embedded, a traveling order in the cover image is first gener- embedded. For LSB replacement, the secret bit which the number of embedded bits is determined by the dif- simply overwrites the LSB of the pixel, i.
The larger the dif- while the higher bit planes are preserved. For the LSBM ference, the larger the number of secret bits that can be em- scheme, if the secret bit is not equal to the LSB of the given bedded. Usually, PVD-based approaches can provide a larger pixel, then 1 is added randomly to the pixel while keeping the embedding capacity on average, larger than 1 bpp.
Based on altered pixel in the range of. In such a way, the LSB of our extensive experiments, however, we find that the existing pixels along the traveling order will match the secret bit stream PVD-based approaches cannot make full use of edge informa- after data hiding both for LSB replacement and LSBM.
There- tion for data hiding, and they are also poor at resisting some fore, the extracting process is exactly the same for the two ap- statistical analyses. Authorized licensed use limited to: The black pixels denote that those pixel values in the corresponding positions have been modified after data hiding.
What is more, it does not introduce the LSB re- placement style asymmetry. Similarly, in data extraction, it first generates a traveling order by a PRNG with a shared key. And then for each embedding unit along the order, two bits can be extracted. The first secret bit is the LSB of the first pixel value, and the second bit can be obtained by calculating the relation- ship between the two pixels as shown above.
Our human vision is sensitive to slight changes in the smooth regions, while it can tolerate more severe changes in the edge regions. Several PVD-based methods such as — have been proposed to enhance the embedding capacity without in- troducing obvious visual artifacts into the stego images. The basic idea of PVD-based approaches is to first divide the cover image into many nonoverlapping units with two consecutive LSBMR applies a pixel pair in the cover image pixels and then deal with the embedding unit along a pseudo- as an embedding unit.
After message embedding, the unit is random order which is also determined by a PRNG. The larger Authorized licensed use limited to: LSB of three cover images. It can be observed that the LSB is not completely random. Some of the LSB planes would even present texture information just like those in the higher bit planes a Example 1.
Furthermore, secret bits that can be embedded into the unit. To a certain extent, the edge information such as the location and the statistical existing PVD-based approaches are edge adaptive since more moments is highly dependent on image content, which may secret data is embedded in those busy regions. However, sim- make detection even more difficult.
This is why our proposed ilar to the LSBM and LSBMR approaches, pixel pair selection scheme will first embed the secret bits into edge regions as far is mainly dependent on a PRNG, which means that the modi- as possible while keeping other smooth regions as they are.
As fied pixels will still be spread around the whole stego image as shown in Fig. It is observed that many smooth re- this property. However, the HBC method just modifies the LSBs gions will be altered inevitably after data hiding even when the while keeping the most significant bits unchanged; thus it can difference between two consecutive pixels is zero meaning the be regarded as an edge adaptive case of LSB replacement, and subimages are located over flat regionswhile many available the LSB replacement style asymmetry will also occur in their sharp edge regions have not been fully exploited.
We will show some experimental evidence to expose the Most existing steganographic approaches usually assume that limitation of the HBC method in Section IV-C1. The reason is that all the nonoverlapping freely using a PRNG.
However, such an assumption is not al- 8 8 blocks within JPEG images are arranged regularly due ways true, especially for images with many smooth regions. If spatial-domain steganographic Fig. It can methods were performed on JPEG decompressed images, it be clearly observed that the LSB can reflect the texture infor- would inevitably lead to JPEG incompatibilities , namely mation of the cover image to some extent.
Based on extensive the additional secret message would destroy the unique fin- experiments, we find that uncompressed natural images usually gerprints introduced by the previous JPEG compression with contain some flat regions it may be as small as 5 5 and it a given quantization table. We can even potentially detect a is hard to noticeand the LSB in those regions have the same hidden message as short as one bit from the JPEG stegos.
Edge Adaptive Image Steganography Based on LSB Matching Revisited
In the data embedding stage, the scheme first initializes LSB plane as illustrated in Fig. Therefore, the selected reviisted. Finally, it does some postprocessing to ob- it is expected that fewer detectable artifacts and visual artifacts tain the stego image.
Matchong the scheme needs to revise the Authorized licensed use limited to: Please note that the parameters may be different for different image content and secret message. We need them as side information to guarantee the validity of data extraction. In prac- tice, such side information 7 bits in our work can be embedded into a predetermined region of the image.
In data extraction, the scheme first extracts the side informa- tion from the stego image. Based on the side information, it then does some preprocessing and identifies the regions that have been used for data hiding.
Finally, it obtains the secret message according to the corresponding extraction algorithm. In this paper, we apply such a region adaptive scheme to the spatial LSB domain.
We use the absolute difference between two adjacent pixels as the criterion for region selection, and use LSBMR as the data hiding algorithm. Ldb details of the data embedding and data extraction algorithms are as follows. Revisuted Embedding where and denote two secret bits to be embedded.
The cover image of size of is first di- The function is defined as. For is a random value in and denotes the each small block, we rotate it by a random degree in the pixel pair after data hiding. The resulting image is rearranged as a row vectoror the new difference may be less by raster scanning.
And then the vector is divided into than the threshold. In such cases,1 baded need to readjust nonoverlapping embedding units with every two consec- them as by utive pixelswhereas- suming is an even number. Two benefits can be obtained by the random rotation.