Add to Watchlist

Optical watermarking system by digital holography


Citation of segment
Embed Code
Purchasing a DVD Cite video

Formal Metadata

Title Optical watermarking system by digital holography
Title of Series 7th International Symposium on Display Holography (ISDH 2006)
Part Number 59
Number of Parts 61
Author Lin, Li-Chien
Chen, Chiung-Liang
License CC Attribution 3.0 Unported:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor.
DOI 10.5446/21296
Publisher River Valley TV
Release Date 2012
Language English

Content Metadata

Subject Area Computer Science
Abstract We propose and analyze an optical watermarking system based on a digital holographic technique. This system constructs a digital holographic watermark by the modified Mach-Zehnder interferometric architecture, and watermark detection is implemented using an optically VanderLugt 4-f correlator with a weighting matched filter mask. The correlation signal at the output plane can be used to recognize whether there is an authorized mark pattern in the watermarked image. In addition, we examine the nonlinear effect of the optical device on the correlation output. The output of the nonlinear device is considered as a mask of matched filter in the Fourier transform domain of an optically VanderLugt 4-f correlator. Through theoretical results, we provide an analytical solution for the correlation output of the nonlinear watermarking system.
of the morning my friend and this is my equation to prove it presented by paper here 2 years and that the topic of this article watermarking assisted by digital holographic the this is
all lined up to the 1st introduction and then I introduce my sister and they're includes 3 part of my system this constructed invading and detection then follow that and I have come to the simulation and the optical history is mental results and then we have some nonlinear problem of this system I tried to solve and the last part of this to stop this company that that Congress so 1st of many happened
to duty of to what market system of thinks the topic here but merit difference from this prize for right so I have to create clearly his friends the problem with this and then you can follow to to speak so the problem is that since the Internet and wireless communication come to our lives and the people can access a lot of images in the in every day so the the cause of problem that the security power such such as in the picture here but if I created this picture is mine I created said all come I I want to make sure that this is mine I just put them not just my men here all men them here to to to crime and this is my speech created otherwise the give up put this picture on the internet that people can easy to copy and there's so we call that security the image and in this security problem of image this is a very an important technique called the watermarking so you had to to put mark in in this image there was this following the perception the shit of image so so we just 1 we in the water market problem with just the height is something in this picture all and you can see for example this feature I already high something in this picture but you can it is almost a century the sense of these 2 features but it's actually something wishing here and you can see it so if you want to call the then I just How can I resolve the mark in this feature and the that the if I have key to identify there is a lot in this feature then I can track and this is my teacher so that is the watermarking problem at and I've been doing this topic for cover years but I didn't come to some new ideas because this is a very important problem in this in digital signal processing a lot of a lot of a lot of people already have it's in this area so I can could not a new idea and that is so that had to my colleagues that's the professor time here feels will show up here today and she told me that maybe you can try something else all this so he drew maybe you can try Coptic to increment this kind of problem so I can't figure out what kind of systems that implement this is probably not so we called it optical watermark consistent by using the to the paragraph come that's why I am here
today so in our system which had to use up to to increments so the 1st part is construction and the baby so I mean that mark the image and then if I had to then I can to the model is that in in this image and if you don't have the key you are not able to assist of good so but in this is and I used to of politics process the 1st 1 is the the the 2 orographic I use that this but to construct a a would come up here and then they the image and then of occasion I just use of 2 2 months and coordinated as so I
just use this picture is to describe the whole process here I have a lot and I use the whole land that to generate a watermark there the watermark can be conveyed to the image so the point is a requirement but when we use the whole to do this kind of what this because the watermark the retirement of the world if you if must be sure the random why because the show the red that give us the capacity energy capacity to invade into the watermarking it's also is is more perceptually invisible if I want to in base to this image so the 1st half of the world all all of process used this use the whole that's his public from that process to this market image 2 of pseudo-random watermark this all construction of the watermark then is that in the I just in base this is not what and innovative this watermark to the cover image and then we have what might image results so you can see if we just invade a small amount of the watermark then you cannot see the difference between the cover image and the watermarked image but height In this image that here is solid detector I use them and the police are solid detector so if the if you have a a key and you have the correct amount then you can get correlation here in the collision of the correlation press that's we call we then divide that market research in in this image at so this is the whole
process the and then there
is the best thing to into a vast part of my system the 1st part is the holographic to construct the feature that they're they're not image here the mapping into here is much better than a random place then for a transform environments In here to adjust the interference with the red reference being then we got the interference pattern here and duties on z the older about process we got the beach focus this acid in a cement watermark and so in this process we have to image the 1st ones among image and this random thing can be can be viewed as the sheet so if you want to go into your home things you just use key that's the key Our marking image here so we need this kid to open them up so people have the key then the country the picture is here so I use this process to generate features for about past my what the
then i'll invaded interestingly uh much crap articles that intended to 88 to the top image and result is my watermarking image this is my
identification process a user must not stand correlated and you find the correct correct marked image and the correct key then I can find a correlation case in my topic this just as structure of the marks and correlated the so this is
all a simulation of the result by using a computer and use it as you can see all all opportunity here so we can see that correlation which you we use this to edit and then 5 the no there's
some parametric was very important in our system such as the weighting that you made the weighting of the what amounted lady to the what is that we have here so if the waiting is smaller then correlation it is kind of always but if the weighting is higher than the correlation in is training and then you can easily to at the it
ended up not the not too good another criterion that we can evaluate the performance the difference between the watermarked image and the all regional host the image so if it is invest in perceptually invisible that you have you achieve a good quality of the watermarked image the so this
simulation results then you
can see that this feature that if we have a correct mock then we can have shut correlation if we the incorrect much then there is no correlation here this all
optical experimental set I want to censor professors tend to give about the 2 goods the structure of the optical set up so the tomorrow market image and that the chief this here that this is for a transform so this construction after and in the she should be 1 this is recorded our for about us watermark and put it into the personal computer here so this is the construction of all watermark and then it's up and implications then we use the same market image he and then we just put that the watermarked image here to to get the correlation between the watermarked image and that the watermarked so here we can get to the correlation pick here and to analyze thank this personal computer and this is all
biological sector the and
that that's the experiment the result so the you can see that this is spherical structure that we can see we're where we find it the market and the is alkylation present here is of course in case of and
this all the optical results and that you as you can see we also a sentimental for the weighting of the watermark and that if the weighting is small then the peak is if those are we increase the weighting than the pig is higher and higher the this is the result
I believe that weighting is a lot the perceptual quality would so this is a there's a conference call between you know between these 2 perfect so the the you you
have correct mocking image then you cannot choose to the correlation and the best
optical used random and the result and when we try to analyze this system we find more of some problem because we've we've asked I you can see the system doesn't include the forward transform and interference process all this past that you knowledge enormous magic is viable find some linear effect between them energy of the watermark embedding into the image and the correlation we find some nonlinear effect so we go back to our system and check
out quite there is some there is the nonlinear effect that then finding which we find out that the this is called we put the watermarked image in the bispectrum I'm on the rate In architecture and that's because there but about image he included information about image and that what so it is wise to rent a and so there's some only factor in this speech alignment later so if we have the tools to to find out the relationship between this speech and I'm looking at the input and output that we can analyze the whole system we going nowhere so we tried to do this job and there is come we
will use some kind of Armenia the creation to to model the input and output of the of the spatial my moderator and is come out some
mathematical derivation and their the and and that the result the soul and
this feature just show you that are mathematical duration is the exactly the same as the simulation result so this kind of we we have tools to analyze to
analyze the system by and but this is very important because there's some crucial from Parramatta in this system and then we also need some you can imagine to look past identified our what so as we have to truly to analyze the system this system this is is our
final result and we also
walking along these walk and still so and we use the model to analyze the performance of the system and to determine some important parameter of the system so and that's and all my
problems and you for for your attention Thank you think of the remaining users and questions forward interest in some questions especially if much better binary out of it mention of Internet images that they can neutron stars that Digital Forum or the inside using the internet image of colleges of has only if there is that the simple case that is easy to times poignant and right so how such as ECG yeah it's just that they are transferred to the beach then it is easy to storage and this has this of yeah it was very important in size is changed what we really yeah that's good if there's some uncertainties holding image so what does it have any important because if you destroy that would not be made cannot be c can that possess the quality of the image tolerate this very important issue for the with a yeah yeah yeah I mean if we tried to destroy the watermark but the major will be totally destroyed so even if you get the same used use less because the majority or a destroy we use the back to the energy in the form of a which you would basically you believe in comparison with the use of the book you come down to that distance OK you additional random that introduces the area and the the correlation between your and target from for the jointly the world with the world of would come in the user use more is more what the workers and in book so actually this problem
already solved by digital computers and then that the problem is already made very mature so the higher than the reuse objects then you can see that united you for the occasion the judge if I just put the put the was here then we can extend the fact in very very short of time because Spartacus easy paraconsistent high but it's still your but you know spatial people displacement but since with was the i in the laboratory as basis so we didn't conceded by giving you cis is common concrete then I think this is is small problem we it's a little this form all so of these yet again the solution this all the time but I might be a useful tool the way go in by the vision of the the make but against limitation of the mall it was all the is so I went in yeah that's true yeah that's true the RO-TD is still should be taken as a consideration yeah about and think you
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation
Technical drawing
Computer animation
Computer animation
Computer animation
Computer animation
Computer animation


  308 ms - page object


AV-Portal 3.8.0 (dec2fe8b0ce2e718d55d6f23ab68f0b2424a1f3f)