Beruflich Dokumente
Kultur Dokumente
Columbia University
Image and Advanced Television Laboratory
Room 801 Schapiro Research Building
530 West 120th Street
New York, NY 10027-6699
USA
E-mail: {mars, sfchang}@ctr.columbia.edu
Abstract Digital signatures can be used for more than just image authentica-
tion. In particular when combined with secure timestamp, a digital
A methodology for designing content based digital signatures signature can be used as a proof of first authorship.
which can be used to authenticate images is presented. A continu-
ous measure of authenticity is presented which forms the basis of A watermark, on the other hand, is a code secretly embedded into
this methodology. Using this methodology signature systems can be the image. The watermark allows for verification of the origin of an
designed which allow certain types of image modification (e.g. image. However, a watermark alone is not enough to prove first
lossy compression) but which prevent other types of manipulation. authorship, since an image could be marked with multiple water-
Some experience with content based signatures is also presented. marks. It has also been pointed out [6] that digital watermarks are
not well suited to protecting the authenticity of an image.
The idea of signature based authentication is extended to video, and
a system to generate signatures for video sequences is presented. 3.0 Content Based Signatures
This signature also allows smaller segments of the secured video to
be verified as unmanipulated. The key to developing a robust digital signature for images is to
examine what the digital signature should protect. Ideally the signa-
1.0 Motivation ture should protect the message conveyed by the content of the
image, and not the particular representation of that content. Thus
Powerful, and easy to use image manipulation software has made it the robust signature can be used to verify the authenticity of an
possible to alter digital images. It has been suggested that the image which has been modified by processing that does not affect
authenticity of digital images can be preserved by having a camera the content of the image. Examples of this type of processing are
“sign” the image using a digital signature. [1] However, applying a removal of noise or lossy compression. However, manipulation of
signature scheme directly to the image has some drawbacks. For the image which changes the content, such as removal of a person
many applications, image compression is desired to reduce trans- from a scene, can still be detected by the use of this signature.
mission bandwidth, storage space, etc. Authenticity, the ability to
detect image manipulation, is also desired. These two function are Additionally, the use of a content based signature fits well with
at odds with each other since lossy compression is a form of manip- other content based image processing, such as content based coding
ulation. Our goal is to develop a way to be able to prove some form and queries. By using the same content for both the signature and
of authenticity, while still allowing desired forms of manipulation, the compression algorithm, the signature will be able to authenti-
such as lossy compression. Ideally, a robust signature scheme cate images highly compressed using content based coding. With
should not declare an image modified under these circumstances. content based queries a signature can be the basis of a query.
Previous work on image authentication falls into two groups, digital Often people think of authenticity as a binary quantity, either an
signatures[1] and digital watermarks[3]. A digital signature is based image is authentic or it is not authentic. However, this not always
upon the idea of public key encryption. A private key is used to what people want when they are concerned with detecting image
encrypt a hashed version of the image. This encrypted file then manipulation. We propose a continuous interpretation of authentic.
forms a unique “signature” for the image since only the entity sign- An image which is bit by bit identical to the original image is con-
ing the image has knowledge of the private key used. An associated sidered completely authentic (authenticity measure of 1.0). An
public key can be used to decrypt the signature. The image under image which has nothing in common with the original image would
question can be hashed using the same hashing function as used be considered unauthentic (authenticity measure of 0.0). All other
originally. If these hashes match then the image is authenticated. images would be partially authentic. Partially authentic is a loosely
defined concept and measurement of the authenticity is subjective,
and changes from application domain to application domain. One
way of thinking of this authenticity measure is as an authenticity
vs. modification curve (see Figure 1). For example a curve could Original Content Hash for Encryption Content
Data Based
be drawn relating authenticity to the bit rate of a compressed Image Extraction + Digital
image. Thus for each different type of modification there would Reduction Signature
Io Co = fc(Io) Ho = fh(Co) S
be a corresponding curve. The old concept of authenticity would
be represent as a Dirac delta function or a unit step function for all
of the possible types of modification.
Private Key
Kpr
JPEG Compression
Authentication