Appeal No. 1998-0671
Application 08/285,328
Algorithms that convert gray images to binary or other
number of gray levels attempting to preserve gray density
exist, and include among them error diffusion . . . .
See also specification, p. 2, last paragraph, which essentially
repeats the first paragraph above. Thus, Eschbach '653 discloses
that a problem with screening is that the amount of gray within an
original image is not maintained exactly over an area and that
throwing away the error caused by thresholding causes a loss of image
information and that it was known to use error diffusion to preserve
gray density. There is no dispute that error diffusion performs the
steps of "(d) generating an error value as a result of the reduction
process . . ." and "(e) diffusing the error value to multi-level grey
scale pixel values of adjacent pixels."
The Survey article also discloses that error diffusion would be
used to provide gray scale rendition where there is coarse gray scale
quantization (p. 298): "The fundamental strategy is simply that of
direct spatial distribution of the errors created by coarse gray
scale quantization, and it can be applied to two or more gray level
marking processes." The Examiner did not rely on this teaching of
the Survey article.
In our opinion, it would have been obvious to one of ordinary
skill in the art to use error diffusion to distribute the error
- 17 -
Page: Previous 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Next
Last modified: November 3, 2007