|
|
|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: RE: [WM]: Purpose of using lena image!!!hanghui_75@yahoo.com
Dear all,
Yes, I totally agree with Dr Johnson that there should be experiments that are meaningful and mimic the real world scenarios. For
this purpose, researchers should experiment with a larger data set; large enough to establish a generalized, applicable result.
I have also noticed, in lena image, the half-tone screen in the lower order bits like Prof. Delp has pointed out earlier in this
thread.Since, everyone has access to the limited {F16…Lena} set, these images are seen in almost every image processing paper,
although far better data sets are available and could be used. There are some other datasets available but most of them require
special permissions and are limited to a specific community or research lab.
I would just like to add a suggestion in this thread. Since we all know the problem, why don't we move towards a change (solution).
All i want to say i= s that we can add a large database (about 2000 or more) of images on this community's "resources" link; or we
can have a dedicated website for this purpose. Images may be both RGB and gray scale and, if possible, animated a= s well. We may
also categorize them according to the applications; like medical imagery, aerial imagery, general purpose imagery etc. There could
b= e a proper reference for each elemet in the dataset along with information about their properties like their histogram images,
ideal PSNR etc. This database should be free and should serve as a standard source of reference for research. Most of the gurus of
image processing (especially
watermarking) are members of this watermarking community. Names like Barni, Piva, Gonzalez, Hernandez, Li, Kim, P.Moulin, Delp,
Johnson needs no introduction. I would like to request them to come forward and support us with their technical help and experience.
I also volunteer myself for any services or experimentation in this regard.
Our needs change with time. People created those images (lena, baboon, F16... ) three decades back. Now, I personally feel, we have
a wider horizo= n to ponder over with advancement in image capturing devices and application range of image processing.
with kindest regards,
Imran Usman
Graduate Student
Faculty of Computer Science and Engineering Ghulam Ishaq Khan Institute of Engineering Sciences and Technology, Topi, Pakistan
On 4/10/06, JJTC. COM <jjtc.com@gmail.com> wrote:
>
> Researchers must consider repeatable results by the community of their
> research efforts. Researchers should also use experiments that are
> meaningful and show more than just a few examples.
>
> Prof. Delp correctly points out the limitations of the over used F16,
> Cameraman, Baboon, Lena, etc. (I will note this set as {F16…Lena} in
> the remainder of this e-mail). One reason for collections of
> test media existing (or being developed) across multiple disciplines
> is so researchers can have common sets from which to select. These
> common sets can help to establish repeatable results. Since nearly
> everyone has access to the limited {F16…Lena} set being discussed in
> this thread, these images have been used over and over again. With
> the improvements of image processing and the diversity of digital
> media available to researchers, many more updated/current formats are
> available.
>
> Researchers should perform many experiments and discuss the results.
> One frustration expressed by many reviewers is that researchers have a
> tendency to express claims based on very little experimentation using
> limited test sets.
>
> In the past, the {F16…Lena} set has played a role in providing
> experimentation on sources that are widely available. These images
> have a "long" history in research but it is time to move on. Do not
> limit
> yourselves to the {F16…Lena} set, make your own sets, and use
> community-shared sets. Be creative!
>
> --
>
> Neil
>
> Neil F. Johnson, Ph.D.
> http://www.jjtc.com
> nfj@jjtc.com
>
> >
> >Subject: Re: [SPAM] RE: [WM]: Purpose of using lena image!!!
> > From: "Edward J. Delp" <ace@ecn.purdue.edu>
> > Date: Mon, 10 Apr 2006 01:53:20 -0400
> > To: <watermarking@watermarkingworld.org>
> > Cc: watermarking@watermarkingworld.org
> >
> >> How about the F-16, Camera man, Baboon, Pepper and Boat? Are they
> >> as bad as Lena?
> >
> >Yes some of these also started "life" as half-tone images.
> >
> >I do not understand why people want to use these images for data
> hiding when you can see the half-tone screen in the lower order
> >bits.
> >
> >Also do you know who owns these images?
> >
> >I find it amazing that people say that "watermarking can be used for
> content protection" and then use images that have unknown
> >ownership.
> >
> >They are much better data sets available.
> >--
> >
> >=======================
> >3
> >D=======================3
> >D=============
> >Professor Edward J. Delp
> >Purdue University
> >School of Electrical and Computer Engineering
> >465 Northwestern Avenue
> >West Lafayette, IN 47907-2035
> >USA
> >
> >Email: ace@ecn.purdue.edu
> >
> >Telephone: +1 765 494 1740
> >Fax: +1 765 494 0880
> >
> >URL: http://www.ece.purdue.edu/~ace
> ><http://www.ece.purdue.edu/%7Eace>
> >
> >Map Location:
> >Latitude: N 40 25.767
> >Longitude: W 86 54.744
> >Datum: WGS72
> >
> >"Badges? We ain't got no badges. We don't need no badges.
> >I don't have to show you any stinking badges!"
> >
> >"Go tell the Spartans,
> >Passerby,
> >That here, obedient to their laws,
> >We lie."
>
______________________________________________________________________________
Watermarking Mailing List - http://www.watermarkingworld.org/ml.html
To unsubscribe send email to "majordomo@watermarkingworld.org" with
"unsubscribe watermarking YOURMAIL" in the body.
______________________________________________________________________________
|
