Facsimile and static presentation processing – Static presentation processing – Attribute control
Reexamination Certificate
1999-01-21
2002-04-09
Rogers, Scott (Department: 2624)
Facsimile and static presentation processing
Static presentation processing
Attribute control
C358S451000, C382S252000
Reexamination Certificate
active
06369912
ABSTRACT:
This application is based on Application No. 10-026658 filed in Japan, the contents of which is hereby incorporated by reference.
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to image processing apparatuses and more particularly to an image processing apparatus for carrying out binarization using the error diffusion method.
2. Description of the Related Art
In the field of image processing, the technique of converting image data of continuous multiple gradation levels (multiple values) into binary data consisting of “0s” and “1s” is conventional known.
FIG. 15
is a block diagram showing a structure of an image processing apparatus adapting such a technique.
Referring to the figure, the image processing apparatus includes an MPU
501
for controlling the entire apparatus, an operation panel
502
for receiving user inputs, a readout sensor
503
formed of a photoelectric device such as a CCD and a driving system thereof, an A/D conversion circuit
504
for digitally converting an output from the readout sensor, a Log conversion circuit
505
for carrying out Log conversion of a digital signal, an MTF correction circuit
506
for carrying out sharpness correction, a gamma correction circuit
507
for carrying out gamma correction, a binarization circuit
508
for binarizing gamma-corrected data, and a printer (image recording device)
509
.
Readout sensor
503
scans a mixed format original of a gradation image and line copy image, for example, and produces a sampling analog signal. A/D conversion circuit
504
quantizes the sampling analog signal as gradation data where one pixel has a value of 8 bits (256 gradation levels), for example.
Log conversion circuit
505
calculates 8-bit gradation density data which has a log relationship with the gradation data from the gradation data. Sharpness correction circuit
506
corrects the sharpness of a gradation density data image using a digital filter such as a Laplacian filter.
Gamma correction circuit
507
corrects a difference in the gradation curve between readout sensor
503
and printer
509
to implement desired gamma characteristics as the entire image processing apparatus or to set gamma characteristics desired by users. Specifically, gamma correction circuit
507
has a lookup table (LUT) RAM of approximately 8 bits (
256
words). MPU
1
sets nonlinear gamma correction data in lookup table RAM and carries out gamma correction.
Binarization circuit
508
uses an area gradation binarization method such as the error diffusion binarization method to convert 8-bit gradation density data that is gamma-corrected into 1-bit binary data according to each light/dark level. The converted 1-bit binary data is printed on a recording medium by printer
509
. Printer
509
is an electronic photo printer or an ink jet printer, for example.
The error diffusion binarization method carried out in binarization circuit
508
calculates a density difference (binarization error) for each pixel between an input image density and an output image density and diffuses the calculation result, which is provided with particular weights, to peripheral pixels. This method is reported in R. W. Floyd, L. Steinberg “An adaptive algorithm for spatial gray scale” SID. 17, pp. 75-77 (1976).
FIG. 16
is a block diagram showing a specific structure of binarization circuit
508
in FIG.
15
. Here, binarization circuit
508
converts an 8-bit (256-gradation) input multi-valued image into a 1-bit (binary) image.
Referring to the figure, binarization circuit
508
includes an adder
601
, a comparator
602
, a selector
603
, a subtracter
604
, an error memory
605
, and an error weighting filter
606
.
Adder
601
adds a binarization error that is weight-averaged by error weighting filters
606
to a density value of a pixel of interest (*) of an input multi-valued image. Thus, the error is corrected. When the density value of an input image is 8 bits, an adder of approximately 10 bits with minus bits included is generally required as adder
601
.
Comparator
602
compares a fixed threshold Th and a density value of a pixel of interest of an error-corrected input multi-valued image. A binary output of the pixel of interest is provided according to the comparison result. Specifically, when the density value of a pixel of interest of an error-corrected input multi-valued image is at least threshold Th, “1” is output from comparator
602
. Otherwise, “0” is output. An 8-bit comparator is generally used as comparator
602
.
Selector
603
outputs a reference value of either High or Low that is set in advance by MPU
501
, for example, according to an output from comparator
602
.
Subtracter
604
calculates a difference (that is, a binarization error) between a reference value output from selector
603
and a density value of a pixel of interest of an error-corrected input multi-valued image. A 9-bit subtracter is generally used as subtracter
604
.
Error memory
605
has the structure of an FIFO and holds one to several lines of binarization errors. A memory where one word has a width of 9 bits is generally used as the FIFO memory.
Error weighting filter
606
calculates a weighted average value of binarization errors stored in error memory
605
. Specifically, binarization errors of pixels peripheral to a pixel of interest (*) are multiplied by the coefficients of {fraction (1/16)}-{fraction (3/16)} as shown in
FIG. 16
, and the multiplied errors are added together to provide a weighted average value of binarization errors. The weighted average value is added to a density value of a pixel of interest by adder
601
. An adder of 9-13 bits is generally required as error weighting filter
606
.
FIG. 17
shows an output image sample by the circuit in FIG.
16
.
However, the conventional image processing apparatus above has drawbacks described below.
(1) The effect of masking a nonuniform pitch, which causes overlapped or gapped pitches, of a printer is low because output patterns are irregular.
In the case of a laser beam printer, the nonuniform pitch of a printer mainly occurs when the inclination adjustment of a polygon mirror is insufficient or when the stability of paper feed is poor, and the nonuniform pitch occurs vertically to a sub scanning direction. In the case of an ink jet printer, the nonuniform pitch occurs when the accuracy of line feed of a printing head is insufficient or when the stability of paper feed is poor, and it also occurs vertically to a sub scanning direction.
(2) An irregular texture unique to a highlight portion of a photographic image is generated.
This problem is unique to the error diffusion method. An example is found in an irregular dot arrangement (texture) at a lower portion of the image (lower portion of a notebook image) shown in FIG.
17
.
(3) A hardware circuit for a feedback process cannot be formed by a synchronous circuit, and a higher speed is difficult to achieve as the number of gradation levels which can be represented is made larger.
Since the processing time of an adder generally becomes longer as the number of bits is larger, the processing speed becomes lower as the bit width of an input multi-valued image is made larger (that is, the number of gradation levels is made larger) in the error diffusion binarization method.
In order to solve the problems of (1) and (2) described above, a method has been proposed which periodically varies a binarization threshold in error diffusion binarization by a dither matrix having a size of S×S.
FIG. 18
is a block diagram showing a structure of an image processing apparatus adopting such a method.
This apparatus is different from the apparatus shown in
FIG. 16
in that a threshold that is input to comparator
703
is varied by a variable threshold matrix (dither matrix)
702
. In other words, this apparatus is formed so that the count values of a main scanning direction counter and a sub scanning direction counter are input to variable threshold matrix
702
and a threshold that is input to comparator
703
is varied as pixels are scan
Burns Doane , Swecker, Mathis LLP
Minolta Co. , Ltd.
Rogers Scott
LandOfFree
Image processing apparatus capable of applying line... 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 capable of applying line..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing apparatus capable of applying line... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2859874