Image analysis – Image compression or coding – Quantization
Reexamination Certificate
1999-05-26
2003-02-11
Mancuso, Joseph (Department: 2623)
Image analysis
Image compression or coding
Quantization
C341S200000, C358S465000
Reexamination Certificate
active
06519366
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image processing apparatus, image processing method, and memory medium and, more particularly, to an image processing apparatus and image processing method for quantizing a multilevel image signal, and a memory medium.
2. Description of the Related Art
General image formation methods for halftone expression include error diffusion and mean density preserving method. These methods realize macroscopic halftone expression by expressing a density as a variation of an area where dot is on using a small number of gradation levels.
This processing increases the capacity of a line memory for holding quantization errors along with an increase in data resolution (an increase in resolution from 400 dpi to 600 dpi).
FIG. 3
is a block diagram showing an example of the arrangement of a gradation conversion section in the image formation apparatus. In
FIG. 3
, an error correction section
302
receives a signal DR′ obtained by dividing by a constant (17) a signal prepared by pre-processing an input multilevel image signal D, and binarization error data E generated in abinarization section
301
. The error correction section
302
generates an image signal DE error-corrected using a line memory.
FIG. 24
is a block diagram showing an example of the conventional arrangement of the error correction section
302
. The input binarization error data E is limited to a value from “−6 to +6” by limiter processing by a circuit (not shown) so as to avoid the influence of past binarization results called “false edge” when the input multilevel image signal D abruptly changes from a high density to a low density.
The binarization error data E limited to “−6 to +6” is halved by a division circuit
501
. The quotient (E/2) is supplied to a subtraction circuit
502
and line buffer
503
.
The subtraction circuit
502
calculates a difference EB (=E−E/2) between the binarization error data E and E/2 as an error component distributed to a pixel of interest, and supplies the difference to an addition circuit
504
. The addition circuit
504
adds an error component EA delayed by one line by the line buffer
503
for one line of 3-bit pixels, and an error component EB supplied from the subtraction circuit
502
, and supplies the sum (EA+EB) to an addition circuit
505
.
The addition circuit
505
adds the error component sum (EA+EB) and the image signal DR′ and outputs the sum as an image signal DE. That is, as shown in
FIG. 6
, the error correction section
302
shown in
FIG. 24
adds to the pixel value of the pixel of interest (DR′) the error component EA of a binarization error generated when a pixel A on one line is binarized, and the error component EB of a binarization error generated when a pixel B before one pixel is binarized.
The line buffer
503
holds 3-bit (−3 to +3) outputs from the division circuit
501
by one line. For example, the line buffer
503
has a capacity of 7,200 pixels×3 bits for an image resolution of 600 dpi and one line of 12 inch.
As the tones of images processed by image formation apparatuses such as a copying machine and printer become finer, the image resolution is projected to greatly increase. However, if the capacity of the line memory for holding quantization errors is increased with an increase in resolution, the cost of the line memory also increases. For example, a 4,800-pixel line memory is required to process an image at a resolution of 400 dpi, and a 7,200-pixel line memory is required to process an image at a resolution of 600 dpi. This simply increases the cost of the line memory to 1.5 times. If the resolution increases from 400 dpi to 800 dpi, the cost of the line memory doubles.
In this way, in the prior art, the capacity of the line memory must be increased in proportion to an increase in resolution.
SUMMARY OF THE INVENTION
The present invention has been made in consideration of the above situation, and has as its object to reduce the capacity of a memory necessary for image processing.
According to the present invention, there is provided an image processing apparatus for quantizing a multilevel image signal, comprising quantization means for quantizing a multilevel image signal for a pixel of interest, compression means for compressing an error component generated based on a quantization error generated by quantization, and storing the compressed error component in a memory, reconstruction means for reading out the compressed error component from the memory and reconstructing the error component, and correction means for correcting the multilevel image signal for the pixel of interest referred to by the quantization means on the basis of the reconstructed error component.
In the image processing apparatus, the compression means preferably uses, e.g., n (n≧2) pixels as a unit, generates synthesis data obtained by synthesizing error components belonging to each unit, and stores the synthesis data in a memory as a compressed error component for the unit.
In the image processing apparatus, the reconstruction means preferably decomposes each synthesis data as a compressed error component read out from the memory, thereby reconstructing an error component for each pixel.
In the image processing apparatus, the compression means preferably uses, e.g., n (n≧2) pixels as a unit, calculates a sum of error components belonging to each unit, and stores the sum in the memory as an error component compressed for the unit.
In the image processing apparatus, the reconstruction means preferably multiplies by, e.g., 1
each sum as a compressed error component read out from the memory, and outputs the product for a period of n pixels.
In the image processing apparatus, the memory is preferably, e.g., a line memory.
According to another aspect of the image processing apparatus of the present invention, there is provided an image processing apparatus for quantizing a multilevel image signal, comprising first arithmetic means for multiplying a signal value of a multilevel image signal by n (n≧2) to generate an n-multiple image signal, quantization means for quantizing the n-multiple image signal for a pixel of interest, second arithmetic means for multiplying by 1
an error component generated based on a quantization error generated by quantization to generate a 1
-multiple error component, storage means for using n pixels as a unit, calculating a sum of 1
-multiple error components belonging to each unit, and storing the sum in a memory as an error component corresponding to the unit, read means for reading out an error component corresponding to each unit from the memory and outputting the error component for a period of n pixels, and correction means for correcting the n-multiple image signal for the pixel of interest output from the first arithmetic means on the basis of an output from the read means, and supplying the n-multiple image signal to the quantization means.
According to the present invention, there is provided an image processing method of quantizing a multilevel image signal, comprising the quantization step of quantizing a multilevel image signal for a pixel of interest, the compression step of compressing an error component generated based on a quantization error generated by quantization, and storing the compressed error component in a memory, the reconstruction step of reading out the compressed error component from the memory and reconstructing the error component, and the correction step of correcting the multilevel image signal for the pixel of interest referred to in the quantization step on the basis of the reconstructed error component.
According to another aspect of the image processing method of the present invention, there is provided an image processing method of quantizing a multilevel image signal, comprising the first arithmetic step of multiplying a signal value of a multilevel image signal by n (n≧2) to generate an n-multiple image signal, the quantization st
Bhatnagar Anand P
Mancuso Joseph
LandOfFree
Image processing apparatus, image processing method, and... 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 apparatus, image processing method, and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing apparatus, image processing method, and... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3144150