Image analysis – Color image processing – Compression of color images
Reexamination Certificate
2000-05-05
2004-09-07
Mehta, Bhavesh M. (Department: 2625)
Image analysis
Color image processing
Compression of color images
C382S248000, C382S251000, C382S281000, C345S604000, C348S395100, C348S398100, C348S403100, C348S438100, C358S426010
Reexamination Certificate
active
06788811
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to a coding apparatus, a decoding apparatus, a coding method, a decoding method, and a computer-readable recording medium for executing the methods, which are used in any image formation apparatus such as a digital copying machine and a facsimile device.
BACKGROUND OF THE INVENTION
In recent years, there has been an increase in demand for a color image, and a higher resolution has been in progress to response to requests for high quality of images. Under these situations, the demands for compressing a color image in order to save memory has been increasing.
Conventionally, a method for compressing a color image is executed by using a technique of first performing color conversion on an image to separate image information into a lightness signal and color signals, and then performing quantization on the signals so that the color signals are reduced higher than the lightness signal. This technique is executed based on the idea that highly important information is tried to efficiently be preserved by focusing on the information for the lightness signals through color conversion.
For example,
FIG. 40
shows a technique in “Coding of color image based on uniform color spaces” (Oshima, Monma, and Kubota) described in No. 24 of the papers prepared for “The 14
th
National Conference of the Image Electronics Institute in 1986” that was held on Jun. 3 and 4, 1986. In
FIG. 40
, of the data for a color image subjected to color conversion (RGB→Lab conversion) in an RGB-to-Lab conversion section
101
, only color signals a and b that do not give influence over the image quality even when the signals are degraded are subjected to 4-pixel averaging by 2×2-pixel block in color-difference averaging sections
112
and
122
. The lightness signal and the color signals of image data are then sub-sampled in a brightness sampling section
103
and a color-difference sampling sections
113
,
123
, respectively, and are coded in a coding section
134
.
The technology described in JP63-9282A is a technique based on the same idea as described above, and the technique is shown in FIG.
41
. In
FIG. 41
, in the data for a color image subjected to color conversion in a color conversion section
202
, a lightness signal is orthogonally transformed in a brightness orthogonal transformation section
203
to a DC component and AC components, and both of the DC and AC components are quantized and coded in a quantization/coding section
204
. Whereas, referring to the color signals, each typical color (a block average in the embodiment), that represents the information within a block with one color, is computed in color-difference typical color computation sections
213
,
223
, and only the typical colors are quantized and coded in quantization/coding sections
214
and
224
.
Further, the technology described in JP06-245079A is shown in FIG.
42
. In
FIG. 42
, of the data for a color image subjected to color conversion by predetermined block for each color, both a lightness signal and color signals are transformed to DC components and AC components in a brightness DC/AC transformation section
303
and in color-difference DC/AC transformation sections
313
and
323
, respectively. The AC components of the lightness signal and the color signals are quantized using a ratio between the amplitudes of the two in quantization/coding sections
304
,
314
, and
324
. Further, the DC components of the color signals are quantized in the quantization/coding sections
314
and
324
using nonlinear quantization in which information with a smaller color difference is quantized more finely.
As explained above, the conventional coding of a color image has been realized, as the basic flow of processing, by allowing information to be biased through color conversion and then allowing the information to be further biased by using a position correlation for each lightness signal and color signals as required.
However, this conventional type of coding has not always been the most appropriate system for the coding of a color image, which has been found wasteful from the viewpoint of a circuit scale. More specifically, there is a problem such that {circle around (1)} color conversion circuit, that may have required its small scale or a small number thereof, increases in number because the position correlation that is more direct as compared to a correlation between colors is not used in the upstream of the processing. There is also a problem such that {circle around (2)} different quantization tables are required for a lightness signal and color signals because color conversion is executed and then quantization is performed on the signals.
SUMMARY OF THE INVENTION
It is an object of this invention to provide a coding apparatus, a decoding apparatus, a coding method, a decoding method, and a computer-readable recording medium that executes the methods, which can simplify the coding circuit and also simplify the processing for quantization by utilizing a position correlation of each component before performing color conversion and then executing color conversion.
According to one aspect of this invention, a first transformation unit transforms the information for each component by using a one-dimensional or two-dimensional position correlation of an image so that bias is generated in the amount of information, a second transformation unit transforms the information transformed by the first transformation unit to obtain a lightness signal and color signals, and a quantization unit performs quantization on the coefficients obtained through transformation by the first transformation unit and/or the second transformation unit.
According to another aspect of this invention, a separation unit separates information for each component into a DC component and AC components by predeter mined block, a transformation unit transforms the AC components separated by the separation unit to a lightness signal and color signals to obtain an AC lightness signal and AC color signals, and a quantization unit performs quantization on the AC lightness signal and/or the AC color signals obtained by the transformation unit.
According to still another aspect of this invention, a separation unit separates information for each component into a DC component and AC components by predetermined block, an AC component quantization unit performs quantization on the AC components separated by the separation unit. A first transformation unit transforms the DC component separated by the separation unit to a lightness signal and color signals to obtain a DC lightness signal and DC color signals, and a second transformation unit transforms the AC components quantized by the first quantization unit to a lightness signal and color signals to obtain an AC lightness signal and AC color signals. A quantization unit performs quantization on the AC lightness signal and/or the AC color signals transformed by the second transformation unit.
According to still another aspect of this invention, a separation unit separates information for each component into a DC component and AC components by predetermined block, a transformation unit transforms the DC component separated by the separation unit to a lightness signal and color signals to obtain a DC lightness signal and DC color signals, and a quantization unit performs quantization on the AC component separated by the separation unit.
Further, when an AC lightness signal and/or AC color signals are quantized, the quantization unit performs vector quantization on the signals.
Further, when an AC lightness signal and AC color signals are quantized, the quantization unit performs quantization on the ratio between the AC lightness signal and the AC color signal.
Further, the quantization unit performs vector quantization on a specific color that has an edge with the largest value of the AC components, and performs, for the other colors, quantization on the ratio between each AC component of the other colors and the vector of the specific color.
Further, the quantiza
Matsuura Nekka
Yagishita Takahiro
Yamazaki Yukiko
Mehta Bhavesh M.
Ricoh & Company, Ltd.
Sukhaphadhana Christopher
LandOfFree
Coding apparatus, decoding apparatus, coding 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 Coding apparatus, decoding apparatus, coding method,..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Coding apparatus, decoding apparatus, coding method,... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3255403