Appeal 2007-1733
Application 09/978,275
(Bk) by performing multi-resolution decomposition of the input image signal
through Laplacian pyramid decomposition or wavelet transformation. (Id. 9
and 27.)
3. The smoothing filter (3) processes the band-limited signals and
smoothes them out. (Id. 27.)
4. The characteristic calculation unit (2) obtains a pixel vector or a
density vector at the pixel of interest in one of the plurality of band limited
image signals, and detects an orientation of an edge as a noise characteristic
of each of the plurality of band-limited images. Further, the characteristic
calculation unit (2) obtains an index value indicating the degree of noise
suppression for the signal based on information (e.g. menu of radiography,
age of patient, radiography and normalization conditions) indicating the
exposure dose with which the radiographic image has been produced. (Id. 7,
11, 27, and 28.)
5. Calculated noise suppression index values are utilized to adapt
the smoothing filter (3) characteristics (mask size, center angle) to the values
of the exposure dose with which the radiographic image was produced. (Id.
28 and 53).
The Prior Art Relied Upon
6. As depicted in Figures 3a and 4a, Vuylsteke teaches a noise reduction
apparatus (3) having a circuit decomposition unit (30) that performs a
pyramidal decomposition of a radiographic image by applying a transform to
the image to yield a set of detail images (31) at multiple resolution levels
and a residue (31’). Each detail image includes a set of transform
coefficients expressing the relative contribution to the original image of the
corresponding basis function out of a set of predetermined basis functions,
5
Page: Previous 1 2 3 4 5 6 7 8 9 10 11 12 Next
Last modified: September 9, 2013