Image analysis – Histogram processing
Reexamination Certificate
2000-10-13
2004-08-03
Lee, Thomas D. (Department: 2624)
Image analysis
Histogram processing
C382S169000, C382S274000, C358S003270
Reexamination Certificate
active
06771814
ABSTRACT:
CROSS-REFERENCE TO RELATED APPLICATION
This application is based upon and claims the benefit of priority from the prior Japanese Patent Application No. 11-294474, filed Oct. 15, 1999; the entire contents of which are incorporated herein by reference.
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image processing device and an image processing method capable of performing image processing to obtain output image data without any occurrence of a chromaticity differentiation loss and jump in brightness and with enhancing a contrast of the image when a printing device such as a printer outputs the image data on a printing paper. When a dark picture or a light picture is output from a color image output means such as a printer which is narrow in color reproduction range, the chromaticity differentiation loss occurs in a dark part and the jump in brightness occurs in a light part in image data. Conventionally, such chromaticity differentiation loss and/or jump in brightness are prevented by enhancing gradation of an image represented by the image data obtained by a color image read-out means such as an image scanner by carrying out contrast enhancement processing on the image data.
2. Description of the Related Art
Various types of printers and copier machines have been developed previously, these printers of an ink-jet type and a laser type are capable of inputting images data transferred from a personal computer and the like and then printing the image data on a printing paper and so on, and these copier machines are capable of reading optional image by using an optical reader and then printing the image data onto a printing paper.
The image printing devices such as these printers and copier machines perform a contrast enhancement process based on a histogram uniformity method in order to avoid the missing of detailed edge information about the edge of an original image.
There is a local histogram uniformity method as one of general contrast enhancement methods. Because this method performs the contrast enhancement in accordance with a local information about an image, this method can be efficiently applied to the process for natural images that require local information.
However, because this method calculates a mapping curve (a density-value conversion curve) per pixel that is obtained by accumulating the histogram of density values, this method causes a drawback to require an enormous time for the operation.
In order to eliminate this conventional drawback, for example, there is a high-speed local contrast enhancement method for natural images as a prior technique. In this technique, the mapping curve is obtained per region, not per pixel, in order to decrease the processing time to make the histogram, namely in order to perform the making of the histogram at a high speed.
Next, a description will be given of an outline of the conventional technique as written above:
(1) At first, a plurality of density conversion curves have been designed in advance;
(2) Second, suppose the histograms are concentrated around a mean density, and select the density conversion curve according to the mean density (selects the density conversion curve so that the contrast around the mean density may be enhanced); and
(3) Finally, a linear interpolation for the density values is performed when the density conversion curves selected in adjacent regions are different.
By the way, in the conventional technique described above, although it has also been written that the density conversion curve is made per pixel, it is commonly and widely used to make the density conversion curve per block.
(A) Dividing input image into blocks, each block has a uniform size that has been experimentally determined.
(B) Following processes (B-1) to (B-3) are performed per block:
(B-1) Making a density histogram (in this case, each block is a reference region);
(B-2) Clipping the density histogram with a clip value that has been experimentally determined in order to obtain the density histogram after the olipping; and
(B-3) Making an accumulated histogram obtained by accumulating the density histograms after the clipping.
(C) Performing a density conversion per pixel in each block based on the accumulated histogram as the density conversion curve.
In particular, when the density conversion curve for the block including the target pixel is different from the density conversion curves for adjacent blocks, the following linear interpolation processes (C-1) to (C-3) are performed for the density values.
(C-1) Converting the density value for a target pixel by using the density conversion curve that is made in the block B
1
including the target pixel, and obtaining the density value “g
1
” after the conversion;
(C-2) Converting the density value of the target pixel by using the density conversion curves selected in each of the blocks B
2
, B
3
, and B
4
that are mostly adjacent to this target pixel, and obtaining the density values g
2
, g
3
, and g
4
after the conversion; and
(C-3) Calculating the density value g(x,y) after the linear interpolation based on the following equation (1). (Each of the density values g
1
, g
2
, g
3
, and g
4
after the conversion is weighted according to the distance from the center of each of the blocks B
1
, B
2
, B
3
, and B
4
to the target pixel.)
g
(
x,y
)={(
J−j
)/
J
)}{(
I−i
)
g
1
/
I+ig
2
/
I}+j/J
{(
I−i
)
g
3
/
I+ig
4
/
I}
(1).
For the definition of each variable in the above-equation (1), see the detailed explanation for the same equation (1) that will be described in the “DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT” section.
However, in the conventional technique “A high-speed local oontrast enhancement method for natural images” as written above, the reduction of the entire contrast of a target image occurs. For example, when a region A of a small area where the density value is low and a region B of a large area where the density value is high are mixed, and when the size of a reference area is then optimized based on the region A, the entire contrast of the region B is decreased and the contrast of a local area in the region B is enhanced.
On the contrary, when the size of the reference area is optimized based on the region B, because the size of the region A is very smaller than that of the reference area, the density information of the region A cannot be almost used when the histogram is calculated. This causes to decrease the slope of the density curve at a low density region and to decrease the contrast. After all, because the regions where the mapping curve is obtained are same, there is a drawback that it is difficult to set each region having an optimize size in the entire image to be processed.
Furthermore, it is preferable to determine a size of the reference area as a parameter to be used for determining the degree of enhancement for the detailed information, and a clip value as another parameter to be used for determining the degree of enhancement in the contrast according to the feature of a local region of the image. However, because there is no determination method to obtain these parameters in the prior technique, constant values as the parameters that have been experimentally obtained in advance are used. Therefore, it is desirable to automatically determine these parameters according to the feature of the local area of the target image.
According to the decreasing of the reference area, the wide-view contrast is also decreased because the local contrast is enhanced. On the contrary, according to the increasing of the reference area, the local contrast is also decreased because the wide-view contrast is enhanced. Furthermore, according to the increasing of the clip value, the degree of the enhancement is also increased, and according to the decreasing of the clip value, the degree of the enhancement is decreased.
Moreover, when the target image to be processed is switched, the above calculation for obtaining the optimum parameters must
Lee Thomas D.
Nath & Associates PLLC
Riso Kagaku Corporation
LandOfFree
Image processing device and image processing method 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 device and image processing method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing device and image processing method will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3280833