proff@iq.org
Tue, 13 Oct 1998 04:07:26 +1000 (EST)
> One idea we had was very much along the same lines as you describe. We
> have a series of pictures that are divided up by a grid; the user
> selects a point on each of the pictures (which is recorded as the
> nearest grid point). The output is the uuencoding of the list of grid
> coordinates.
I worked on something similiar to this to. Unfortunately, it seems that
there is an astoundingly over broad patent on image-based authentication
schemes:
http://www.patents.ibm.com/details?patent_number=5608387
I claim:
1. A personal identification device comprising a store of
personal identity statements and a store of complex images
including key complex images linked with personal identity
statements and false complex images not associated with personal
identity statements, a complex image being an image that is
recognizable when already known but not readily capable of
unique description to a person to whom it is not known so that
when a particular complex image is included in a set of similar
but not identical complex images, the particular complex image
is distinguishable from other complex images in the set, by
the human senses of a person familiar with the particular
complex image, within a predetermined time interval when the
whole set is displayed, but it is not readily possible to
describe the particular complex image in terms which are
sufficiently precise that another human being unfamiliar with
the particular complex image can subsequently identify it among
other complex images of the set using normal unaided human
senses to discern the particular image, and within a time
interval similar to the predetermined time interval, and provided
that reference to and comparison with other complex images of
the displayed set is not permitted as part of a description.
Cheers,
Julian.
The following archive was created by hippie-mail 7.98617-22 on Sat Apr 10 1999 - 01:15:21