 |
Description  |
|
|
CROSS REFERENCE TO RELATED APPLICATION
U.S. Pat. No. 4,748,678 entitled "Method of Storing and Retrieving Image
Data" and issued to Haruo TAKEDA et al discloses a technique wherein an
origianl image read out of an image file is printed out to add information
thereto, and the original image with the additional information is again
inputted to the system to discriminate a different part between the
original image and the newly inputted image and pick up the additional
information.
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image data filing system and more
particularly, to an image data filing system suitable for correcting an
image stored in the file and re-storing the corrected image.
2. Description of the Prior Art
Document image file systems (electronic file systems) utilizing large
capacity optical disks have recently become noticeable as the new means
for document management. Optical disks are large in capacity and capable
of recording image data so that document image information such as design
layouts, literatures, contracts and etc. can be stored therein. As an
example of systems of this kind, there is known, for example, a system
disclosed in a magazine "Nikkei Electronics" Mar. 28, 1983, pp. 105 to
120.
It becomes sometimes necessary to partially correct the image data, e.g.,
design layouts or the like, already stored in such an image file system
and re-store the corrected image data. According to the conventional
system, an already stored image is made corrected through the method
whereby the stored image is first printed out and applied to correction on
the printed paper, and the corrected image is read by an image input
device to store it in the image file as new image data in place of the
image data before the correction.
The conventional system, however, does not consider image quality
deterioration resulting from image printing and inputting operations, thus
posing a problem that as an image is frequently applied to correction, the
quality of such an image, particularly the image portion where correction
was not applied, gradually deteriorates. More specifically, since the
pixel densities of a printer and an image input device used in such a
system are different to each other in most cases, the size of an image
such as design layouts is reduced or magnified every correction operation
in order to make the sizes of the original image and the corrected image
coincide with each other. In addition, since there are digital/analog
conversion errors at the printer and image input device, line information
of the corrected image is degraded to thus result in distorted lines and
increased noises. If the same image data are applied to correction n
times, the original image information will have undergone the image input
device n+1 times and the printer n times. Consequently, as n becomes
large, the image quality deteriorates more remarkably.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide an image data filing
system capable of correcting an image already registered or stored in an
electronic file as many times as necessary without deteriorating the image
quality.
It is another object of the present invention to provide an image data
correcting method whereby a portion of image data stored in an electronic
file can be corrected easily.
The above objects are achieved by the present invention which provides an
image data filing system comprising: means for retrieving original image
data from a file, said original image data corresponding to image data
inputted as correction image data through an image input device; pattern
matching means for obtaining a difference between the inputted image data
and the retrieved image data; synthesizing means for forming a synthesized
and corrected image data using the difference image data obtained through
pattern matching and the retrieved image data; and means for storing the
synthesized image data in the file.
With the image data filing system constructed as above, a difference
between the correction image inputted from the image input device and the
original image retrieved from the file is obtained by pattern matching
means, the difference is synthesized with the original image by
synthesizing means to form a new corrected image. The new corrected image
includes the corrected portion and the non-corrected portion which were
both applied to the image input device only once, because the corrected
portion is additional information not present at printing-out of the
original image which constitutes the non-corrected portion applied to the
image input device only once and read from the file. Therefore, according
to the present invention, even if a same image is applied to correction n
times, information regarding both the corrected and non-corrected portions
has undergone the image input device only once. The resultant image
quality is evidently fine as compared to the conventional system wherein
image correction of n times results in application to the image input
device n+1 maximum times and to the printer n maximum times.
A further improved image data filing system according to the present
invention comprises: first means for printing out original information
onto a correction paper, the original information being obtained by
subjecting first image data read out from an image file to a first
conversion process; second means for reading the information containing
correction information on the correction paper as second image data; third
means for converting the second image data into third image data by
subjecting the second image data to a second conversion process, the third
image data being the second image data with the original information
removed therefrom; and fourth means for forming a corrected image data
using the first image data and the third image data.
An image data correcting method according to the present invention
comprises the steps of: printing out original information onto a
correction paper, the original information being obtained by subjecting
first image data read out from an image file to a first conversion
process; adding correction information to the correction paper and
inputting the information containing the correction information on the
correction paper as second image data; generating third image data by
subjecting the second image data to a second conversion process, the third
image data being the second image data with the original information
removed therefrom; and forming fourth image data using the first image
data and the third image data, the fourth image data being the third image
a portion of which was removed therefrom.
The first conversion process, typically a process of converting an original
image into dots, is employed to distinguish the original image information
represented by the first image data from the correction information to be
inputted by a user. In case of a dot conversion process, an original image
is printed on a correction paper in the form of dot information so that a
user can see the contents or positions of the already stored information
(e.g., characters, figures) from the correction paper and add the new
information and/or delete the contents of the already stored information.
The dot conversion process thins black pixels of the original image at
regular intervals to obtain dispersed black pixels. Therefore, if such
regularly dispersed black pixels are removed (second conversion process)
from the second image data obtained by reading the information on the
correction paper, the correction information inputted by the user can be
selectively extracted. Therefore, newly corrected image data can be
obtained through logical OR operation between the correction information
and the original image in the case of addition correction, and through
removal of the deletion information represented by the correction
information from the original image in the case of deletion correction.
The newly corrected image thus obtained is made of the image data having
undergone the input device only once so that the image quality is not
practically deteriorated even after several times of correction. Besides
the dot conversion process, the first conversion process in this invention
may include other modifications such as a thinning or broadening process
for line segments of an original image.
The foregoing and other objects, advantages, manner of operation and novel
feature of the present invention will be understood from the following
detailed description when read in connection with the accompanying
drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram showing an example of an image file system
embodying the present invention;
FIG. 2 is a schematic diagram showing the procedure starting from an
initial storage of an original image to the re-storage of a corrected
image according to a first embodiment of the present invention;
FIG. 3 is a flow chart for a program performing the procedure of retrieving
an original image;
FIG. 4 is a flow chart for a program performing the procedure of re-storing
the corrected image.
FIG. 5 is a flow chart for a program performing the procedure of locating
straight lines contained in the original image;
FIG. 6 illustrates how a straight line is located in the flow chart shown
in FIG. 5;
FIG. 7 illustrates image conversion associated with normalizing a
correction image;
FIGS. 8 and 9 show modifications of a rectangular frame used in image
correction;
FIG. 10 is a schematic diagram showing the procedure starting from the
initial storage of an original image to the re-storage of a corrected
image according to a second embodiment of the present invention;
FIG. 11 is a flow chart showing a print procedure using a correction paper
in the second embodiment;
FIG. 12 is a flow chart showing the synthesizing procedure of the
correction image with the original image in the second embodiment;
FIGS. 13 and 14 are a pixel layout and a program flow chart, respectively,
for use in explaining the dot conversion process for an original image;
FIG. 15 is a flow chart for a program performing the procedure of erasing
dots of the original image;
FIG. 16 is a flow chart for a program performing synthesizing the original
image with the partial correction image; and
FIG. 17 is a schematic diagram illustrating the procedure of deleting a
portion of a stored image according to the present invention.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
The invention will be detailed taking as an example an image file system
using optical disks.
FIG. 1 is a block diagram showing the overall arrangement of an image file
system. The system comprises an image input device 1 such as an image
scanner, television camera or the like, a printer 2 for printing image
data, an image file 3 for use in storing/reading image data, a display 4
for displaying image data, a data processor 5 which operates upon programs
stored therein, a keyboard 6 for inputting commands or the like, memories
7 to 12 for temporarily storing image data, and a bus 13 for
interconnecting these components.
FIG. 2 is a schematic diagram showing the procedure starting from the
initial storage of an original image to the re-storage of a corrected
image according to a first embodiment of the present invention, wherein
rectangular blocks represent image, those blocks with a peeled corner,
e.g., indicated by 21, represent printed paper and other blocks represent
image stored in memory.
Referring to FIG. 2, information 31 on a paper 21 is inputted into a memory
7 with the image input device 1, the original image 22 in the memory 7 is
stored in the image file 3. Information such as key words for image
retrieval is inputted with the keyboard 6 and stored in the image file 3
in one-to-one correspondence with the image 22.
If an image stored in the image file according to the above procedure
becomes necessary to be corrected, the program illustrated in FIG. 3 is
executed. At step 101, upon input of the number, key word or the like of
the image to be corrected, the image retrieved from the image file 3 is
stored in the memory 7. This original image 22 per se may be printed out
in the same manner as in the conventional system. However, to facilitate
the pattern matching process to be described later which is performed
during the correction processes, it is preferable to add at step 103
re-storage information such as a rectangular frame 33 for alignment
purpose, a key word 34 for image retrieval purpose, and the like, to the
original image 22. The rectangular frame 33 may be replaced by other marks
representative of the image position, and the key word 34 may be replaced
by other expression such as a bar code, a character in conformity with
character recognition of JIS standards, or the like. The line segments
constituting the rectangular frame 33, the key word 34 and the like are
sequentially stored in the memory 7 at predetermined positions using line
element patterns, character font patterns and the like. In the process of
writing the rectangular frame, all black pixels of the image 23 outside
the rectangular frame 33 may be converted into white pixels in order to
facilitate the pattern matching process. At step 105, the image thus
obtained as shown in FIG. 2 is printed out on a correction paper 24 by the
printer 2. In this embodiment, it is assumed that a user adds new
information 36 to the information 35 on the paper 24 to thereby obtain a
correction paper 25 whose image is intended to be re-stored in the image
file.
The process of re-storing a corrected image is performed in accordance with
the program illustrated in FIG. 4. First, at step 111, the image of the
corrected paper 25 is inputted in the memory 8 with the image input device
1 to thus obtain a correction image 26. The correction image 26 and the
original image 23 have differences therebetween in their contents caused
by the correction, and generally to some extent in their sizes and
inclinations. Namely, since the pixel densities of the printer 2 and the
image input device 1 are different in the strict sense of the word as
discussed previously, the size of the image 26 becomes smaller than that
of the image 23 if the printer 2 has a higher pixel density, and on the
contrary the size of the image 26 becomes larger than that of the image 23
if the image input device 1 has a higher pixel density. In addition,
generally there occurs an inclination of the image while a user
manipulates the image input device 1. The process of normalizing the size
and inclination of the correction image 26 relative to the size and
inclination of the original image 23 by using the rectangular frame 38,
and obtaining an image 27 will be described.
The normalization process of the size and inclination of the correction
image 26 at step 113 includes two steps. At one step, the size and
inclination of the image 26 are obtained by detecting four straight lines
constituting the rectangular frame 38. At the other step, the image 26
undergoes the coordinate conversion so as to coincide the rectangular
frame 38 of the correction image 26 with the rectangular frame 33 of the
original image 23, and the image 26 with corrected size and inclination is
stored in the memory 9. Various conventional methods may be used in
detecting straight lines constituting the rectangular frame 38, one
example of which is a known Hough conversion algorithm which is applied to
this embodiment.
FIG. 5 is a flow chart for a program detecting a left side vertical line
among the four straight lines of the rectangular frame 38, and FIG. 6
illustrates the manner of detecting the vertical line. In FIG. 6, the
formula of an objective straight line is expressed as
.gamma.=xsin.theta.+ycos.theta. where .gamma. is a distance from an origin
0, and .theta. is an inclination of the straight line. The characteristic
feature of the Hough conversion algorithm is the capability of detecting a
straight line irrespective of particular portions of document information
in the image. The outline of this algorithm is as follows: In the flow
chart shown in FIG. 5, at steps 130 to 131 parameters are initialized, at
steps 132 to 135 a candidate point, e.g., a black pixel P1 on a straight
line is searched, at steps 136 to 140 combinations of 65 and .theta. for
a straight line .gamma.=xsin.theta.+ycos.theta. passing through point P1
are obtained, and steps 132 to 141 are repeated to obtain a frequency
distribution. The function f(.gamma., .theta.) indicates the number of
candidate points on the straight line .gamma.=xsin.theta.+ycos.theta.. For
example, in FIG. 6, the value of f(.gamma., .theta.) of the straight line
(.gamma., .theta.) passing through points P1 to P4 is "4", whereas the
value of other straight lines (.gamma., .theta.) takes small value of 0 to
2. Therefore, the parameters .gamma. and .theta. which define a maximum
f(.gamma., .theta.) are determined at step 142 to thus enable to obtain a
straight line passing a maximum number of candidate points. The formulas
of the right, upper and lower straight lines of the rectangular frame 38
can be obtained in the similar manner.
The upper left, upper right, lower left and lower right corner points of
the rectangular frame 38 are then obtained from the intersections of the
four straight lines, and the four corner points are respectively defined
by (0, 0), (M1, N1), (M2, N2) and (M1+M2, N1+N2) in the new coordinate
system having the upper left corner point as its coordinate origin.
Next, the procedure of converting the correction image 26 into an image 27
by using the rectangular frame 38 will be described with reference to FIG.
7. The conversion from the image 26 into the image 27 can be expressed by
##EQU1##
where (0, 0), (m, 0), (0, n) and (m, n) are four corner points of the
rectangular frame of the original image 23, (x, y) is the coordinates of a
pixel in the image 26, and (X, Y) is the coordinates of a corresponding
pixel in the image 27. Using the transformed formula
##EQU2##
which is equivalent to the above formula, a point (x, y) corresponding to
each lattice point (X, Y), i.e., each point defined by an integer value X
and Y, is obtained. The density of a point (X, Y) is determined as the
density of a pixle nearest to the obtained point (x, y) to thus convert
the image 26 into the image 27. For example, in FIG. 7, point A is
converted into point a using the above formula, and the density of point
a' nearest to point a is used as the density of point A.
As the objective symbol for normalization of the size and inclination, the
rectangular frame constructed of four straight lines has been used in the
above embodiment. The normalization can be performed by other methods such
as through matching of characteristic points added at the four corners,
through matching of characteristic portions in an original document
without adding specific symbols, and the like. Further, in the above
embodiment, although the distortion caused by the difference of a pixel
density between the printer 2 and the image input device 1 is assumed
linear, the distortion will be non-linear in some cases. For example, such
non-linear distortion may appear if an image scanner of the type the line
sensor is driven by a motor is used as the image input device 1, because
the drive speeds at the start and during the constant drive operation
differs from each other. In such a case, instead of a rectangular frame, a
frame 33a composed of squares as shown in FIG. 8 or a frame 33b composed
of parallel lines like a bar code as shown in FIG. 9 may be added to the
original image 22. A change in distance between parallel lines of the
frame 33a or 33b is detected from the correction image 26. The detection
results are used to divide the correction image 26 into plural portions so
that each portion is subjected to the same conversion as described above,
thus realizing a highly precise normalization.
Returning back to FIG. 4, at step 115 the original image 22 corresponding
to the correction image 27 normalized by the above procedure is searched
from the image file 3 and written in the memory 7. If another image has
not been searched after the original image 22 was printed out, the above
search process is not needed because the image 23 is already stored in the
memory 7. The above search however becomes necessary for the case where a
plurality of images are sequentially printed out and corrected thereafter
to sequentially re-store the corrected images. The key word may be
inputted by a user to search an original image. However, the key word 34
such as a bar code, OCR-B fonts or the like added to the image 23 may be
used to search an original image. Since the key word 43 is being stored in
the memory 8 after the correction image was read and normalized, this key
word is recognized by the processor unit 5. A method of searching an image
from the image file using a given key word and storing the image in a
memory is known in the conventional system, so that the description
therefor is omitted herein.
Next, a difference image 27 is obtained through pattern matching between
the original image 22 stored in the memory 7 at the search process and the
correction image 27 stored in the memory 9 at the normalization process,
and the difference image 27 is stored in the memory 10. The above
procedure performed at step 117 will be described below. The difference
image 28 is formed by comparing pixels at the same coordinates of the
images 27 and 22 and generating a white pixel at the position where both
the pixels agree with each other and a black pixel at the position where
both the pixels do not agree. For instance, the information 44 is not
present on the image 22 so that the comparison result is a disagreement
and the information 44 per se becomes the information 46 on the image 28.
At this stage, the rectangular frame 42 and the key word 43 are erased by
changing the black pixels thereof to white pixels. Although most of the
information 41 coincides with the information 32, there is a local area
where the information 41 and the information 32 do not agree with each
other because they have undergone the printer 2 and the image input device
1. Thus, noise information 45 may be present on the image 28.
At step 119, the noise components are eliminated from the difference image
28 to obtain a difference image 29 which is then stored in the memory 11.
This process can be achieved by the following method by utilizing a
general feature that noises 45 have a narrower line width than that of the
correction information 46. In particular, black pixel areas are first
reduced in size to make the black pixel area with a narrower line width
disappear. Such a reduction process can be achieved by performing the
following formula for all coordinate values x and y:
f'(x, y)=f(x, y).LAMBDA.f(x+1, y).LAMBDA.f(x, y+1).LAMBDA.f(x+1, Y+1)
where f(x, y) is the density of a pixel at the coordinates (x, y) of the
difference image 29. This conversion formula means that the density of
each pixel in the difference image 28 is replaced by a logical AND density
of adjacent four pixels including the pixel now concerned. In case of a
binary image having white and black pixels alone, if at least one of the
adjacent pixels is white, the pixel now concerned is replaced by a white
pixel. In other words, the border of a black pixel area is narrowed by
changing a portion of black pixels to white pixels. Next, the size of the
black pixel area not disappeared by the reduction process is magnified to
restore the original size of the black pixel area. Such a magnification
process can be achieved by performing the following formula for all
coordinate values x and y:
f"(x, y)=f'(x, y) .nu.f'(x+1, y).nu.f'(x, y+1).nu.f'(x+1, y+1)
This conversion formula means that the density of each pixel of the
difference image 28 reduced in size by the reduction process is replaced
by a logical OR density of adjacent four pixels. In case of a binary
image, if at least one of adjacent four pixels is black, the pixel now
concerned is regarded as black. In other words, the border of the black
pixel area is magnified by changing a fraction of white pixels to black
pixels. In this embodiment, the densities of four adjacent pixels have
been used for the reduction and magnification processes. However, many
adjacent pixels, such as 9 pixels, 16 pixels may be used to make more
noise information disappear.
At step 121, the difference image 29 stored in the memory 11 and the
original image stored in the memory 7 are synthesized, and the synthesized
image is stored in the memory 12. In this synthesizing process, pixels at
corresponding positions of the image 29 and the image 22 are compared, and
the higher density is used as the pixel density of the synthesized image
30. The information 37 on the correction image 26 having an inferior image
quality due to application to the printer 2 and the image input device 1
is replaced by the information 32 on the original image 22 so that the
synthesized or corrected image 30 having no area with inferior image
quality can be stored in the memory 12. At the last step 123, the
corrected image 30 is registered or stored in the image file 3. In this
case, if an operator inputs a new registration key word with the keyboard
6, the corrected image 30 can be registered as a new image different from
the original image 22. Alternatively, if the key word used in searching
the original image 22 during the pattern matching process is again used,
the original image can be replaced by the synthesized image 30. Also, a
new key word with slight modification of a part of the key word of the
original image 22, such as with a new version number or a renewal history,
may be used to store the synthesized image as one of associated images
with the original image.
In the above embodiment, as the information of the difference image, only
the information present on the correction image and not present on the
original image, i.e., only the information added through correction has
been used. Contrarily, it is possible to use the information deleted
through correction as the information to be processed. In this case, in
the process of comparing pixels at corresponding positions of the images
27 and 22 at step 117, two types of difference image are obtained. The
first difference image is made of only those black pixels of the image 27
which disagree with the black pixels of the image 22, whereas the second
difference image is made of only those black pixels of the image 22 which
disagree with the black pixels of the image 27. Next, both the first and
second difference images are applied to the noise elimination process at
step 119. Lastly, in the synthesizing process at step 121, the first
difference image is added to the original image 22 and thereafter, the
second difference image is subtracted from the addition result. Namely,
the pixels of the original image 22 corresponding to the black pixels of
the second difference image are replaced by white pixels to obtain the
synthesized image 30.
Further, in the above embodiment, the difference image 29 has been obtained
through pattern matching between the original image 22 and the correction
image 26. This procedure may be replaced by the following simple method.
First, formed at step 103 are an image obtained by adding the rectangular
frame 33 to the original image and a rectangular image having only the
image information of the rectangular frame 33. Both the images are printed
out on two papers, respectively. An operator puts the paper with only the
rectangular frame on the paper with the original image 24 to which
correction information is added. The upper paper with the original image
and correction information is again read with the input image device 1.
The normalization process is performed similarly at step 113. With the
above procedure, the difference image 29 can be obtained without pattern
matching. The key word 34 may be added to the image with only the
rectangular frame in order to search the original image at the
synthesizing process.
Furthermore, the memories 7 to 12 have been used independently in the above
embodiment in consideration of simplifying the description. However, a
memory may be used in common in each process to reduce the number of
memories.
FIG. 10 is a schematic diagram showing the procedure of the initial storage
of an original image, correction of the stored image, and the re-storage
of a corrected image according to the second embodiment of the present
invention.
The image of information 230 (in this example, character A) on a paper 200
is inputted with the image input device 1 and stored in the image file 3
at a preset position as an image 201. As the stored image 201 is required
to be partially corrected, the image number, key word or the like are
designated and a search command is inputted from the keyboard 6 to read
the stored image from the image file or optical disk 3 and store it in the
memory 7. In this embodiment, the information 230 on the original image
201 stored in the memory 7 is converted into dots, and a rectangular frame
231 used for position alignment is added to the original image to form an
image 203 and store it in the memory 8. In addition to the rectangular
frame 231, re-storage information such as a bar code, a key word made of
OCR-B fonts conforming with JIS standards, or the like for image retrieval
may be added to the original image. Also, all black pixels outside the
frame area may be converted into white pixels during the write operation
of the rectangular frame 231 in order to facilitate to detect the frame.
The dot image 203 of the original image is printed out on a paper 204 which
is a correction paper on which correction information 232 is added to
obtain a correction image 205. The correction image 205 is read with the
image input device 1 and stored in the memory 9 as an image 206. The size
and inclination of information 230' of the image 206 differ from those of
the image 203. As discussed previously, if the pixel density of the
printer 2 is higher than that of the image input device 1, the information
230' and frame 231' of the image 206 become larger than those of the image
203. Contrarily, if the pixel density of the printer is higher than that
of the image input device 1, the information and frame of the image 206
become smaller than those of the image 203. The inclination of an image is
caused by a skew failure while the paper passes through the image input
device 1. Similar to the first embodiment, the size and inclination of the
image 206 are normalized using the rectangular frame 231' to obtain a
normalized image 207 having the same size and inclination as the image 203
and store the normalized image in the memory 10. Thereafter, the dots 230'
and rectangular frame 231' of the image 207 are erased to obtain a partial
correction image 208 which is then stored in the memory 11. Since the
image 208 contains only the correction information 232', a corrected image
210 can be obtained in the memory 12 by synthesizing the image 208 with
the image 201. The corrected image 210 is stored in the optical disk 3 as
a new image in place of the original image 201. The original image may be
left stored in the optical disk to use in forming another corrected image
therefrom.
FIG. 11 is a flow chart showing the print procedure for image correction
according to the second embodiment. In the flow chart, step 101 searches
the image 201 stored in the optical disk 3 and store it in the memory 7,
step 102 converts the read-out original image 201 into dots, step 103 adds
the re-storage information such as the rectangular frame, and step 105
prints out the image 203 with the frame to obtain the correction paper
204.
FIG. 12 is a flow chart showing the re-storage procedure of a corrected
image. In the flow chart, step 311 reads the image of the paper 205, which
is obtained by adding a correction on the correction paper 204, with the
image input device 1 to obtain the image 206, step 313 normalizes the size
and inclination of the image 206 to obtain the image 207, step 314 erases
the rectangular frame and dots of the image 207 to obtain the partial
correction image 208, step 315 searches the original image corresponding
to the normalized image 207, step 321 synthesizes the partial correction
image 208 with the original image 202 to obtain the corrected image 210,
and step 323 stores the synthesized image 210 in the optical disk 3. Among
these steps, steps 311, 313, 315, 321 and 323 correspond to steps 111,
113, 115, 121 and 123 shown in FIG. 4, respectively.
Next, the procedure of converting an image into dots at step 102 shown in
FIG. 11 will be detailed with reference to FIGS. 13 and 14.
The dot conversion process is a process to thin black pixels within an area
where black pixels are consecutively present. For example, all the black
pixels except specific dispersed two pixels as shown in FIG. 13 within a
4.times.4 pixel area are converted into white pixels. It is assumed here
that the coordinates of the upper left corner of the original image 201
and the dot image 203 are the coordinate origins, the density of a pixel
at the coordinates (x, y) on the original image is defined by OR(x, y),
and the density of a pixel at the coordinates (x, y) on the dot image is
defined by OB(x, y), where the x-axis is in the right direction, the
y-axis is in the down direction, and the interval between pixels is "1".
It is further assumed that the density of pixel is represented by a binary
value corresponding to white or black.
FIG. 14 shows an example of a flow chart for a program to converting the
original image into dots under the above-mentioned conditions. First, at
step 600 the densities OB(x, y) for all the pixels (x=1 to x-max, y=1 to
y-max) within the memory area 8 are initialized to white. At steps 602 to
608, the coordinates (x, y) are incremented to sequentially select the
pixels at each row starting from the first row of the image. At step 610
it is checked if the density OR(x, y) of a pixel of the original image at
the coordinate (x, y) is white or black. If it is white, the flow advances
to judgement step 618 where if it is judged that x is smaller than the
maximum value x-max, the flow returns to step 608 to increment the value
x. The density of the next pixel is checked and if the value x becomes the
maximum value x-max, then it is checked at step 620 if the value y is
smaller than the maximum value y-max. If it is y-max, the routine is
terminated. If y is smaller than y-max, the flow returns to step 604 to
increment the value y and check the density of the first pixel of the next
row. If the density OR(x, y) is black, the flow advances to step 612 where
the values of x and y coordinates are divided by 4 and the remainders are
substituted for variables Xmod and Ymod. The next step 614 is a step to
determine if the pixel with OB(x, y) corresponding to a pixel with OR(x,
y) is to be changed to white or black. In this embodiment, the density
OB(x, y) is set as black (step 116) is Xmod=1 and Ymod=0, or Xmod=3 and
Ymod=2, namely, if the pixel with OB(x, y) corresponds to one of the
pixels indicated by oblique lines in FIG. 13 within the divided area of
4.times.4 pixels of the image. If the pixel does not correspond to any one
of the two pixels, the pixel is not set as black but the flow advances to
step 618 where the coordinate value is checked. Thereafter, the coordinate
value is incremented or the routine is terminated as the case may be.
FIG. 15 is an example of a flow chart for a program erasing dots at step
314 shown in FIG. 12. It is assumed here that the density of a pixel at
the coordinates (x, y) of the dot image 207 is defined by OR(x, y), the
density of a pixel at the coordinates (x, y) of the erased dot image 208
is defined by OB(x, y), and the coordinate system is the same as that
shown in FIG. 13. In the flow chart shown in FIG. 15, similar to steps 600
to 608, 618 and 620 shown in FIG. 14, steps 700 to 708, 718 and 720
initializes the area (memory area 11) storing the erased dot image 208,
| | |
