Appeal No. 2001-0872
Application No. 08/880,137
through 9) concerning the teachings of Chou and Kopec. According to the examiner, Chou teaches
“computing probability as a layout parameter” (answer, page 4), and “using the values from training
images (templates, col. 4, lines 15-35)” (answer, page 5).
Appellants argue (brief, pages 9 through 12) that the references of record neither teach nor
would have suggested to one of ordinary skill in the art to train a processor-controlled machine to
determine an unknown value of a text image layout parameter used with a two-dimensional image
model. Although appellants acknowledge (brief, page 10) that “Chou does disclose a ‘training
mode’ . . . at column 4, lines 9-33, col. 9, lines 45-51, and col. 10, lines 1-16,” they do indicate,
however, that “Chou’s ‘training mode’ refers to the possibility of re-training probabilities associated
with the objects in Chou’s library of bitmapped templates” (brief, page 11). Appellants’ argue
(brief, page 11) that:
It is important to observe that these trained and retrained probabilities do not
correspond in any way to the layout parameters trained in Applicants’ current
invention. Probabilities, quite simply, are not layout parameters as defined by
Applicants. See Applicants’ specification, which expressly states, “a text image
layout parameter does not include a probability parameter that may be associated
with or included in a 2D image model, since a probability parameter is not a quantity
on which a layout procedure depends to map an image constituent from its symbolic
description of its location in a text image to an actual physical location in the image.”
Page 34, lines 6-10. Contrary to the Examiner’s unsubstantiated assertion, Chou does
not teach parameterizing x-y coordinates, point size, or any other layout parameters
in any manner similar to Applicants’.
In response to appellants’ arguments, the examiner now contends (answer, page 11) that:
Examiner was relying upon Kopec to teach layout parameters are determined from
the individual image constituents such as “f and g” on page 69. And training of such
would be obvious in light of Chou because the layout parameters must be calculated
4
Page: Previous 1 2 3 4 5 6 7 Next
Last modified: November 3, 2007