1996-02-26 - Re: How to digitally watermark

Header Data

From: “E. ALLEN SMITH” <EALLENSMITH@ocelot.Rutgers.EDU>
To: Dave.Birch@eworld.com
Message Hash: db9f20f5760f67d863d887dbec51aff53d844ddce76081c84eefd52f412bde7d
Message ID: <01I1O9T9OH3CAKTN7G@mbcl.rutgers.edu>
Reply To: N/A
UTC Datetime: 1996-02-26 21:27:40 UTC
Raw Date: Tue, 27 Feb 1996 05:27:40 +0800

Raw message

From: "E. ALLEN SMITH" <EALLENSMITH@ocelot.Rutgers.EDU>
Date: Tue, 27 Feb 1996 05:27:40 +0800
To: Dave.Birch@eworld.com
Subject: Re: How to digitally watermark
Message-ID: <01I1O9T9OH3CAKTN7G@mbcl.rutgers.edu>
MIME-Version: 1.0
Content-Type: text/plain


>Adam Shostack <adam@homeport.org> wrote..
>>Creating watermarks that can't be removed without degrading
>>image quality is not especially difficult.  The two tricky bits are
>>durability and collusion protection.

	The difficulty I can see is with the "without degrading image quality"
part. Given ever-improving image enhancement and processing techniques, someone
probably could - or at least will be able to in the near future - take any
image in which the watermark didn't degrade the image by itself and remove
the watermark without _perceptible_ alterations in image quality. 

>I hope someone, somewhere finds this interesting...

	Yes, actually; I was just at a seminar on computer image interpretation
for the biomedical field.
	-Allen





Thread