Image analysis – Pattern recognition – Feature extraction
Reexamination Certificate
1999-04-20
2002-07-02
Couso, Jose L. (Department: 2621)
Image analysis
Pattern recognition
Feature extraction
C382S162000, C382S266000, C382S274000, C348S625000
Reexamination Certificate
active
06415053
ABSTRACT:
BACKGROUND OF THE INVENTION
This invention relates to an image processing method and an image processing apparatus. More specifically, the invention relates to the image processing method and the image processing apparatus employing that method for performing edge portion extraction so as to perform image processing operations such as sharpness enhancement, grain suppression and the like on a color image signal (image data) obtained by photoelectrically reading a color image or a subject, as well as the image processing method and the image processing apparatus employing that method for subjecting the obtained image data to specified image processing operations such as sharpness enhancement, grain suppression and the like to produce image data for output.
A new technology has recently been introduced and this is a printer that relies upon a digital color image reproduction system, where the color image recorded on a film such as a negative film, a reversal film, a color print or the like is read photoelectrically by a photoelectric conversion device such as a CCD or the like to obtain image signals corresponding to each of three primary colors, namely, red (R), green (G) and blue (B) and subsequently the obtained image signals are converted into digital image data signals which are then subjected to various image processing operations to be reproduced on recording materials such as color paper and the like and on display device such as a CRT (cathode ray tube) and the like. The printer operating on this digital color image reproducing system has been commercialized as a digital photoprinter.
In the digital photoprinter, color images can be reproduced in such a way that they have desired colors and gradations by subjecting obtained original image signals to image processing operations, even if the color images are recorded on such as the negative film, the reversal film or the color print under unfavorable photographing conditions such as under-exposure, over-exposure or the like. The color images recorded on the negative film, the reversal film, color prints and the like can optionally be reproduced as different color images in colors and gradations.
Examples of these image processing operations include a method of enhancing sharpness (sharpness enhancement) of images by subjecting the image data signals representing a given image to image processing operations and a grain suppression method for decreasing grainy appearance of images by suppressing the grains of the film. Prior to performing the sharpness enhancement or grain suppression, the edge portion or a grain portion (or flat portion) must be extracted. Various methods for extracting the edge portion and the like have been known.
An exemplary known method for extracting the edge portion is of a type that adopts template matching. In this method, templates comprising some form, for example, a 3×3 matrix, are prepared with reference to a function distribution of pixel values of surrounding pixels of a pixel of interest to be decided which of the prepared template a given image is nearest to. Now, the pixel values of the pixel of interest and its surrounding pixels are set as x
j
(j=1, . . . , n); k pieces of templates are prepared; a value of the i
th
template is set as w
j
i
(j=1, . . . , n). A sum of products p
i
is calculated when i =1, . . . , k by the following equation:
P
i
=&Sgr;
j
w
j
i
·x
j
(&Sgr; is a sum when j=1
, . . . , n
)
It is decided that the template which gives the maximum of p
i
matches with the given image.
Consider, for example, a case where eight templates a
1
to a
8
as shown in
FIG. 10
are given to the image composed of the pixel of interest and eight pieces of its surrounding pixels as shown in FIG.
11
A. Each of these templates a
1
to a
8
corresponds to each of directions to edges as shown by arrows b
1
to b
8
. Operators shown in such templates are called as ‘Sobel’ operators with which each of the sums of products p
1
to p
8
is calculated according to the above equation. The sum of products p
1
of the template a
1
is calculated as follows:
p
1
=1×1+2×2+1×3+0×9+0×10+0×4+(−1)×8+(−2)×7+(−1)×5=−19
Other cumulated sums p
2
to p
8
corresponding to respective other templates a
2
to a
8
are calculated in the same manner. The results are as follows:
p
2
=−8
p
3
=11
p
4
=20
p
5
=19
p
6
=8
p
7
=−11
p
8
=−20
As a result, p
4
shows the maximum value so that the direction obliquely downward from right to left that the arrow b
4
of the template a
4
indicates shows the direction to the edge (density gradient, that is, direction of gradient) of the pixel of interest
10
, as shown in
FIG. 11B
, and, in this case, the value
20
of the sum of products p
4
shows the intensity of the gradient. In such way, the gradient of the pixel of interest (vector quantity) is calculated. Subsequently, with reference to pixels
1
and
5
that are adjacent to the pixel of interest
10
and positioned in the direction of 90 degrees from its gradient direction, each of images with a 3×3 matrix around each of the pixels is calculated employing the above templates a
1
to a
8
in the same manner as above to obtain the gradient direction of each of the adjacent pixels
1
and
5
. If at least one of the pixels
1
and
5
is in the direction of 45 degrees or less from the direction of the pixel of interest
10
, the pixel of interest is decided as the edge portion, that is, having connectivity.
Though the conventional method for extracting the edge portion described above is of course capable of extracting the edge portion to some extent, it in cases detects the grain portion as the edge portion falsely, thereby causing deterioration of quality of a finished image.
As illustrated in
FIG. 9A
, an example shows that, besides white lines which were original edge portions, the grain portions were falsely decided as the edge portions so that a number of white grains have remained.
As already described, in recent years, the digital color image reproduction technique has been known that photoelectrically reads the image in a frame recorded on a photographic film by a reading sensor such as a CCD sensor or the like, subjects the thus read digital image data to image processing operations including enlargement, reduction and other various corrections and allows the image to be formed on a recording material by a laser beam modulated on the basis of the digital image data that has been processed imagewise.
In such technique that digitally reads the image in the frame by the reading sensor such as the CCD sensor or the like, following steps have been taken in order to realize an accurate reading of the image: the image in the frame is read preliminarily (so-called prescan is performed); reading conditions (for example, light quantity to be irradiated to the image in the frame, charge storage time of CCD or the like) are determined in accordance with density and the like of the image in the frame; and the image in the frame is read again under the thus determined reading conditions (so-called fine scan is performed).
At the time of the prescan, correction values to be adapted to various correction processing on color gradation, hypertone, hypersharpness, granularity and the like are each detected on the basis of the read image so that the image read by the fine scan is subjected to correction processing based on each of the thus detected correction values.
The granularity that is one of the objects to be subjected to the correction processing is referred to a rate of graininess or coarseness of the image on the film. Higher the granularity finer the image. The granularity is referred in one case to a rate when compared with that of a specified reference film and in another case to a relative rate within one image on the same film.
The granularity is one of the important fac
Birch & Stewart Kolasch & Birch, LLP
Couso Jose L.
Desire Gregory
Fuji Photo Film Co. , Ltd.
LandOfFree
Image processing method and apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image processing method and apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing method and apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2906509