 |
Description  |
|
|
BACKGROUND OF THE INVENTION
This invention relates to an image synthesizer and more particularly to an
image synthesizer of an image editing apparatus for editing, creating and
displaying a new pictorial image by processing two or more sets of image
data.
When an image editor of this type is used to synthesize a new image from
two or more original source images, the conventional procedure has been to
operate its keyboard, while watching the superposedmutually transparent
display of these source images, to adjust the two-dimensional positional
relationships among them. In connection with such a procedure, one method
of superposingly synthesizing two mutually transparent images A and B, for
example, is to read out pixel values of these images A and B respectively
stored in different image memories M.sub.a and M.sub.b by means of a data
processor to calculate their average values and to display such average
values calculated individually for each pixel. Another method is to read
out pixel values, for example, of the image A stored in a display memory
M.sub.v and those of the image B stored in an image memory M.sub.b by a
data processor to calculate their average values and to display such
average values after rewriting the display memory M.sub.v by these
calculated average values.
With either of these exemplary methods, data processing becomes extremely
time-consuming because averages of two or more sets of image data must be
calculated for all pixels in the superposed area. In the case of a
composite image requiring no transparency in superposition, the averaging
process is carried out only for pixels in boundary areas between different
images to make the boundary lines less conspicuous but the amount of
calculations to be carried out is still significantly large. By the second
of the two methods described above, furthermore, there is the disadvantage
of losing one of the source images because its image data stored in the
display memory is replaced by the newly calculated average values. In
order to make this image restorable, its data must be temporarily saved
and this means that an increased memory capacity is required.
SUMMARY OF THE INVENTION
It is therefore an object of the present invention to provide an image
synthesizer for an image editing apparatus with which a synthesized image
can be quickly displayed by superposing two or more source images one on
top of another such that they are all visible, being transparent to each
other.
It is another object of the present invention to provide an image
synthesizer for an image editing apparatus with which a synthesized image
can be quickly displayed by superposing one of the source images on top of
another such that their boundary lines appear inconspicuous.
In view of the above and other objects, an image synthesizer embodying the
present invention is characterized not only as having memory means for
storing image data of two or more source images and a display device for
displaying a pictorial image according to image data received thereby but
also as having an address generator which generates address signals and
transmits them sequentially to the memory means to cause image data
therein to be transmitted to the display device for a display. A data
processor for controlling the operations of all these components also
controls the sequence in which these address signals should be transmitted
such that each pixel on the display device displays a specified one of the
source images and that area densities of pixels displaying each of the
source images vary over the two-dimensional screen of a display device.
This makes it possible to display a synthesized image of two or more
source images which are mutually superposed and transparent to each other.
What is hereinafter referred to as a pattern memory may additionally be
provided to store data on a pattern according to which, when one of the
source images is superposed on top of another to form a composite image,
the boundary between the two can be made controllably less conspicuous.
BRIEF DESCRIPTION OF THE DRAWINGS
The accompanying drawings, which are incorporated in and form a part of the
specification, illustrate embodiments of the present invention and,
together with the description, serve to explain the principles of the
invention. In the drawings:
FIG. 1 is a block diagram of an image synthesizer embodying the present
invention,
FIGS. 2A and 2B are an example of source images and a superposingly
synthesized image obtained therefrom by the image synthesizer of FIG. 1,
FIGS. 3A and 3B are schematic drawings showing data arrangements in the
memory devices and a timing chart of processing by the image synthesizer
of FIG. 1,
FIG. 4 is a block diagram of another image synthesizer embodying the
present invention,
FIG. 5 is a drawing showing data arrangement in the memory devices of the
image synthesizer of FIG. 4,
FIGS. 6A and 6B are an example of source images and a superposingly
synthesized image obtained therefrom by the image synthesizer of FIG. 4,
FIGS. 7A and 7B are drawings showing data arrangements and a timing chart,
FIG. 8 is another example of source images and a superposingly synthesized
image obtained therefrom by the image synthesizer of FIG. 1,
FIG. 9 shows distribution of data in the image memories of FIG. 1,
FIGS. 10 and 11 are further examples of source images and superposingly
synthesized image obtained therefrom,
FIG. 12 shows an address pattern in connection with the example shown in
FIG. 11,
FIG. 13 is a block diagram of still another image synthesizer embodying the
present invention,
FIG. 14 is another example of source images and a superposingly synthesized
image obtained therefrom by the image synthesizer of FIG. 13,
FIGS. 15A, 15B and 15C are drawings showing arrangements of data in the
memories of FIG. 13,
FIG. 16 is a block diagram of still another image synthesizer embodying the
present invention,
FIG. 17 is a drawing for explaining a method of preparing a mask pattern,
FIGS. 18A-18D are examples of brush patterns for preparing a mask pattern,
FIG. 19 is a flow chart of a process for preparing a mask pattern,
FIG. 20 is a flow chart of another process for preparing a mask pattern,
FIGS. 21 and 22 are a drawing of a portion of pattern memory and a flow
chart for explaining still another process for preparing a mask pattern,
and
FIGS. 23A and 23B are drawings for showing a preprocessing procedure
embodying the present invention.
DETAILED DESCRIPTION OF THE INVENTION
In what follows, the present invention is described by way of examples. In
all these examples, the principles of the present invention are explained
for a situation where only two source images are synthesized but this is
not intended to limit the scope of the invention. Extension to situations
of syntheses of three or more source images should be readily
understandable from the disclosure below.
As shown by way of a block diagram in FIG. 1, an image synthesizer 10
according to one embodiment of the present invention is comprised of image
memories 11 and 12, a display device 13 and an address generator 15. Each
of the image memories 11 and 12 includes a semiconductor memory device and
peripheral circuits and stores a source image to be synthesized at
addresses corresponding to the two-dimensional coordinates thereon. The
display device 13 is of a known type for displaying a pictorial image on
its screen corresponding to image data outputted through a data bus 14
from the image memories 11 and 12. The address generator 15 is a device
which generates and transmits an address signal through an address bus 16
to the image memories 11 and 12 to specify image data to be outputted
therefrom. The image memories 11 and 12, the display device 13 and the
address generator 15 are all connected to and controlled by a data
processor 17 through a bus 18.
Let us assume that image data for original source images 21 and 22 shown in
FIG. 2A are stored respectively in the image memories 11 and 12 and that
it is desired to superpose them and to display their mutually transparent
image 23 on the display device 13. As shown in FIG. 2B, the superposingly
synthesized mutually transparent image 23 consists of vertically elongated
thin picture elements 24 and 25 respectively of the original source images
21 and 22 each of one pixel in width alternately arranged in horizontal
direction at the areal ratio of 1:1 such that the two original source
images 21 and 22 appear as if they were placed one on top of the other. A
method of generating this superposingly synthesized mutually transparent
image 23 is described next with reference to FIGS. 3A and 3B.
To start, the address generator 15 generates addresses of pixels in the
image memories 11 and 12 sequentially and alternately within one
horizontal display scan period. If the image memories 11 and 12 store
image data a0, a1, a2, a3,... of the image 21 and image data b0, b1, b2,
b3,... of the image 22, respectively, as shown in FIG. 3A, the address
generator 15 generates addresses in the image memories 11 and 12 where
image data a0, b1, a2, b3... are generated as shown in FIG. 3B and they
are transmitted as address signal to the image means 11 and 12 through the
address bus 16. The aforementioned data processor 17 retrieves image data
a0, b1, a2, b3,... for the image memories 11 and 12 sequentially according
to this address signal and transports them through the data bus 14 to the
display device 13. These data are sequentially displayed at the individual
pixels arranged on one horizontal line on the screen of the display device
13. With this processing repeated along each horizontal scan line, the
superposingly synthesized mutually transparent image 23 of FIGS. 2A and 2B
is displayed. In summary, image data from the two image memories Il and 12
are not averaged with respect to each pixel in the superposed area.
Instead, image data are alternately read from the image memories 11 and 12
according to address signal from the address generator 15 and they are
directly displayed on the display device 13. As a consequence, time
required for data processing can be significantly reduced.
FIG. 4 shows another image synthesizer 30 embodying the present invention
characterized as storing the image data of the source image 21 of FIG. 2A
in a display memory 31 and those of the other source image 22 of FIG. 2A
in an image memory 32. This image synthesizer 30 is further comprised of a
buffer memory 33 for temporarily storing a part of the image data stored
in the display memory 31 and a data transferring device 34 including an
address generator which functions as explained above. Under the control of
a data processor 39, address signal generated by the address generator of
the data transferring device 34 is received by the memories 31, 32 and 33
through an address bus 35 and image data are transferred among the
memories 31, 32 and 33 through a data bus 36 according to this address
signal. The image data thereby stored in the display memory 31 are
transmitted through a display bus 37 to be displayed by a display device
38 as a synthesized image.
Let us consider an exemplary horizontal scan line on the images 21 and 22
of FIG. 2A. At this moment, image data a0, a1, a2, a3,... of the source
image 21 and b0, b1, b2, b3,... of the other source image 22 are
sequentially stored in the display memory 31 and the image memory 32
respectively as shown in FIG. 5. Then, the data transferring device 34
causes the image data in the image memory 32 to be transferred to the
display memory 31 for every other pixel and the image data stored at the
address at the display memory 31 where these transferred data are written
are transferred to the buffer memory 33 to be temporarily stored therein.
Regarding this horizontal scan line considered above, therefore, every
other data b1, b3, b5,... are transferred from the image memory 32 to the
corresponding addresses of the display memory 31 and the data a1, a3,
a5,... originally at these addresses of the display memory 31 are
sequentially transferred to and saved in the buffer memory 33 as shown in
FIG. 5. This procedure is repeated for all other horizontal scan lines.
Thus, image data a1, b1, a2, b3,... for one frame stored in the display
memory 31 are outputted to the display device 38 through the display bus
37 and displayed on the screen of the display device 38 as a superposed
mutually transparent image 23 shown in FIG. 2.
When it is desired next to return from the superposed image 23 to the
original image 21, the data transferring device 34 sequentially reads out
the image data al, a3, a5,... and replaces every other data b1, b3, b5,...
in the display memory 31 thereby. The contents of the display memory 31
are thus returned to the original image data a0, a1, a2, a3... of the
source image 21. In summary, there is no averaging to be performed between
the two image data and the time for data processing can be reduced also by
this example. Another advantage of this example is that the capacity of
the buffer memory can be reduced because only those of the image data in
the display memory that are written over by image data from the image
memory are transferred to the buffer memory 33 to be saved therein.
FIGS. 6A and 6B show another example of using an image synthesizer
structured as shown in FIG. 1 to produce a superposingly synthesized
mutually transparent image of two original source images 21 and 22. Unlike
the example shown in FIGS. 2A and 2B, these two source images are
synthesized at the areal ratio of 3:2 such that the first source image 21
comes out more strongly. This is apparent by a comparison, for example,
between the synthesized pictures shown in FIG. 2A at 23 and in FIG. 6A at
43. The different is more clearly shown in FIG. 6B wherein 5 vertically
elongated picture elements 44 and 45 respectively from the original source
images 21 and 22 and both of the width of one pixel are arranged
horizontally and adjacent to each other at the ratio of 3:0 to form a unit
and such units are horizontally arranged one next to another.
To form the synthesized image 43 of FIG. 6A and 6B, the address generator
15 generates within one horizontal scan period addresses of the image
memories 11 and 12 at the ratio of 3:2. If the image data a0, a1, a2,
a3,... of the image 21 and b0, b1, b2, b3,... of the image 22 on this scan
line are stored respectively in the image memories 11 and 12 as shown in
FIG. 7A, the address generator 15 generates sequentially the addresses of
the image data a0, b1, a2, b3, a4; a5, b6,..., the semicolon indicating a
boundary between two of the aforementioned units of five elements such
that the desired ratio of 3:2 can be achieved. Thereafter, this series of
image data is displayed at pixels on one horizontal line on the screen of
the display device 13 in connection with FIG. 1. It now goes without
saying that this example also has the advantage of reduced processing time
because there is no averaging of data involved and that original source
images can also be superposed at different ratios other than 3:2 merely by
adjusting the address generator. The procedure shown above by way of FIGS.
7A and 7B can be achieved also by an image synthesizer structured as shown
in FIG. 4 and three or more original source images can be synthesized
similarly.
Moreover, the procedures according to the present invention are not limited
to alternate displaying of two source images in horizontal directions. As
shown below by way of examples, similar superposing effects can be
obtained by alternately displaying two source images not only horizontally
but also vertically. FIG. 8 shows the same two original source images 21
and 22 to be superposed differently to produce another synthesized image
48 by an image synthesizer structured as shown in FIG. 1, details of a
circled portion of the synthesized image 48 being also shown. FIG. 9 shows
image data for the source images 21 and 22 arranged in the image memories
11 and 12. In each, a horizontal row represents image data for one
horizontal scan line. If there are (n+1) pixels in the horizontal
direction on the screen of the display device, image data on a horizontal
line may, for example, be a00, a01, a02,... a0n in the image memory 21 and
b00, b01, b02,... b0n in the other image memory 22. Those on the lines
immediately therebelow are a10, a11, a12,... a1n and b10, b11, b12,...
b1n. To superpose the two source images 21 and 22 as shown in FIG. 8, the
address generator 15 generates the addresses of the image memories 11 and
12 alternately for each horizontal scan line. When the image data are
stored as shown in FIGS. 9A and 9B, the address generator 15 generates,
for example, the addresses of the image data a00, a01, a02,... a0n
sequentially during a certain horizontal scan period and those of b10,
b11, b12,... b1n during the next horizontal scan period. Thereafter, the
addresses of image data from the two image memories and 12 are alternately
generated. The image data corresponding to these addresses inputted
through the address bus 16 are outputted from the image memories 11 and 12
to the data bus 14 and the display device 13 displays on its screen as
shown in FIG. 8 the data inputted through the data bus 14.
FIG. 10 shows another example of obtaining a synthesized mutually
transparent image 49 by superposing the two source images 21 and 22 of
which the image data are stored as shown in FIG. 9. The address generator
generates addresses of the image memories 11 and 12 alternately for the
pixels during each horizontal scan period but by starting with an address
from a different image memory each time. In other words, if the address
generator 15 sequentially generates the addresses of the image data a00,
b01, a02, b03,... b0n during one horizontal scan period, the addresses to
be generated sequentially during the next horizontal scan period are those
of the image data b10, all, b12,... a1n. Addresses being thus generated
during each subsequent horizontal scan period, a synthesized image 49
superposed in a checkerboard fashion can be obtained as explained above in
connection with FIGS. 8 and 9. The superposed image 49 looks much like the
one shown in FIG. 2A at 23 and FIG. 8 at 48.
FIG. 11 shows still another example of obtaining a synthesized mutually
transparent image 50 by superposing the two source images 21 and 22 of
which the image data are stored as shown in FIG. 9. FIG. 12 shows an
example of a pattern of addresses to be generated by the address generator
15 for the purpose of this desired synthesis. When the pixels are arranged
in a matrix formation with horizontal rows and vertical columns, as in the
previous example shown in FIG. 9, FIG. 12 shows an address of which image
memory 21 or 22 should be generated for each pixel on each line, A and B
respectively indicating an address of the image memories 11 and 12. In
FIG. 12, i and j are dummy integers. This address pattern is stored within
the address generator 15.
More in detail, after an address pattern is determined as shown in FIG. 12,
the address generator 15 generates addresses of image data of desired
pixels on desired lines sequentially by making reference to this address
pattern. During a horizontal scan period for a (4i)th line, for example,
the address generator 15 sequentially generates the addresses of the image
data a00, b01, b02, b03, a04, b05, b06,... and during the next horizontal
scan period, those of the image data a10, a11, b12, b13, a14, a15, b16,
b17,... are sequentially generated. Thereafter, addresses are sequentially
generated for each horizontal scan similarly with reference to this
address pattern. As explained above, the image memories and 12 thereupon
receive address signal indicative of these generated sequence of addresses
through the address bus 16 and transmit corresponding image data through
the data bus 14. The display made accordingly on the display device 13 is
shown in FIG. 11 at 50.
The present invention has been described above only by way of image
synthesizers for superposing two source images one on top of the other
such that they are mutually transparent and the user can adjust their
positional relationships, but these applications are not intended to limit
the scope of the present invention. In what follows, image synthesizers
embodying the present invention for image editing apparatus with different
capabilities are described.
FIG. 13 shows the structure of an image synthesizer 30 according to still
another embodiment of the present invention. Components which have been
explained above with reference to FIG. 1 are assigned the same numerals as
before and not explained again. In FIG. 13, numeral 19 indicates a pattern
memory for storing mask patterns to be described in detail below. The
pattern memory 19 is comprised of a semiconductor memory device and
peripheral circuit as do the image memories 11 and 12, and is connected to
the address generator 15 through an address control bus.
Let us assume that an original source image 51 with a vase on a table as
shown in FIG. 14 is available and that it is desired to combine it with
another original source image 52 of a cup such that a synthesized image 54
of both the vase and the cup on the table becomes available. A mask
pattern as shown at 53 is used in this application with "1" indicating the
area in which the source images 51 and 52 are to be superposed, and is
stored in the aforementioned pattern memory 19. Data "0" are written at
pixel positions where superposition is not to take place. In FIG. 14,
numeral 55 indicates a portion of the mask pattern shown enlarged and
numeral 56 likewise indicates a corresponding portion of the synthesized
image 54 also shown enlarged.
Let us assume, as done above with reference to FIG. 9, that the image data
for the portions of the source images 51 and 52 in areas corresponding to
the enlarged areas 55 or 56 are arranged as shown in FIGS. 15A and 15B. In
other words, FIGS. 15A and 15B are portions of image data arrangement
stored respectively in the image memories 11 and 12. FIG. 15C shows the
arrangement of these image data inside this enlarged area 55 or 56 when
those of FIGS. 15A and 15B are superposed according to, or by using, the
mask pattern partly shown in FIG. 14 at 55 and 56.
Next, a procedure is explained in detail whereby the synthesized image 54
is produced by using image data for the source images 51 and 52 stored
respectively in the image memories 11 and 12 as well as the mask pattern
53 partly shown at 55 and 56 of FIG. 14 and stored as data in the pattern
memory 19. Regarding the mask pattern, since it is used to smooth out the
boundary between superposed and non-superposed areas, the densities of "1"
and "0" per unit area should be varied gradually from the superposed area
to the nonsuperposed area. For example, the mask pattern 53 (and 55) of
FIG. 14 shows that the addresses corresponding to the center part of the
cup are assigned "1", those corresponding areas not occupied by the cup
are assigned "0", and the frequency of occurrence decreases gradually from
the center to the outside of the cup area.
To start, the address generator 15 makes reference through the address
control bus to the pattern memory 19 regarding each pixel to determined
whether to access an address of the image memory 11 or 12, depending on
whether the mask pattern 53 on the pattern memory 19 corresponding to the
pixel is "0" or "1". If, for example, the image data arrangements
regarding a certain area are as shown in FIGS. 15A and 15B and the mask
pattern is as shown in FIG. 14, the address generator 15 sequentially
generates and outputs address signals corresponding to the image data b00,
b01, b02, b03, a04, b05, a06,... as shown in FIG. 15C. Upon receiving
address signals thus generated, the image memories 11 and 12 output the
corresponding image data to the data bus 14 and the display device 13
makes a display based on these outputted image data. The synthesized image
54 thus obtained has less sharp boundaries because the superposed image 52
and the background image 51 because of the use of the mask pattern 53.
FIG. 16 shows another image synthesizer 60 embodying the present invention
which, like the image synthesizer shown in and explained by way of FIG. 2
above, is characterized as being provided with a display memory 70. This
display memory 70, too, is comprised of a semiconductor memory device and
peripheral devices and serves to store image data which are transferred
through a display bus 69 to be displayed on a display device 65. An image
memory 61, a data transfer device 64 including an address generator and a
buffer memory 62 are also provided as shown in FIG. 2. Additionally, a
pattern memory 63, as explained above in connection with FIG. 13, is
connected to the data transfer device 64 through an address control bus
68.
Let us assume that the image data of the source images 51 and 52 shown in
FIG. 14 are respectively stored in the display memory 70 and the image
memory 61 in the matrix formations shown in FIGS. 15A and 15B, and that
data of the mask pattern 55 are stored in the pattern memory 63. Let us
further assume that it is desired to use this image synthesizer 60 to
display a synthesized image 54 with boundary lines properly made smooth
and then returned to the original condition. In this case, the data
transfer device 64 makes reference through the address control bus 68 to
the pattern memory 63 and, if the mask pattern 55 shows "1" regarding
certain pixels, the image data stored in the image memory 61 at the
corresponding addresses are caused to be sequentially transferred to their
corresponding addresses in the display memory 70. At the same time, those
image data in the display memory 70 at these corresponding addresses,
which are therefore to be replaced by the data transferred from the image
memory 61, are transferred to, and saved in, the buffer memory 42 as
explained above by way of FIG. 5. The image data of the image memory 41 at
addresses where the mask pattern 55 shows "0" are not transferred to the
display memory 70.
Let us now consider that the first rows in FIGS. 15A and 15B correspond to
a same horizontal scan line which also corresponds to the first row of the
mask pattern 55 shown in FIG. 14. For this horizontal line scan, since the
pattern data are (1, 1, 1, 1, 0, 1, 0), only the five image data b00, b01,
b02, b03, b05 corresponding to the data "1" are transferred to the
corresponding addresses of the image memory 70 containing a00, a01, a02,
a03, a05. At the same time, these image data a00, a01, a02, a03, a05 to be
overwritten are transferred to the buffer memory 62 and stored therein.
The addresses of the image memory 70 corresponding to this horizontal scan
line now contain the image data b00, b01, b02, b03, a04, b05, a06.
Thereafter, the display device 65 reads the contents of the image memory
70 as explained above in connection with FIG. 4 and a synthesized image as
shown in FIG. 14 is displayed. To return from this display of the
synthesized image to the original display of one of the source images (51
of FIG. 14), the image data temporarily saved in the buffer memory 42 are
sequentially returned to their addresses in the display memory 70
according to the positions of data "1" on the mask pattern.
Next, several methods of preparing a mask pattern are described. As shown
in FIG. 17, a brush pattern 72 is prepared in a known manner from an
original image 71 which is the subject of masking so as to produce a
desired mask pattern 73. FIGS. 18A and 18B are two examples of brush
patterns intended to have a black area at the center which gradually
becomes white in the edge areas. FIG. 18A is a pattern with a larger
central black area than the pattern shown in FIG. 18B. FIGS. 18C and 18D
are corresponding patterns with black areas represented by "1" and white
areas by "0". To prepare a mask pattern in a pattern memory (such as shown
in FIG. 13 at 19 and in FIG. 16 at 63), the memory contents are cleared
(S1 in FIG. 19), to start with. This may be accomplished, for example, by
writing "0" at all addresses. Next, the user specifies a coordinate system
with respect to the original image 71 of interest by means of a pointing
device such as a mouse (S2).
Next, one of available brush patterns such as shown in FIGS. 18A-18D is
selected with reference to the defined coordinate system (S3). If
different brush patterns are used, depending on the defined coordinate
system, affinity of the synthesized image to its surroundings (that is,
the degree of inconspicuousness) can be varied. In the case of the image
52 of a coffee cup, for example, the brush pattern of FIG. 15C may be used
near the bottom of the cup because it is less important to see its
boundary line with the table top but that of FIG. 15D may be preferred
near the top around its rim to show clearer boundary lines.
Next, the center point of the selected brush pattern is positioned at the
address of the pattern memory 19 or 63 corresponding to the specified
coordinate system and the logical sums are obtained between the data
already stored in the pattern memory 19 or 63 and the data on the selected
brush pattern (S4). The sums thus obtained are stored in the pattern
memory 19 or 63. This process is repeated until all relevant areas of the
image 71 have been covered (YES in S5). The mask pattern thus completed
may appear as shown in FIG. 14.
FIG. 20 shows another method of preparing a mask pattern. According to this
method, after the contents of the pattern memory are cleared (S11) as
explained above, a curve is specified to mark the boundary of the image
(such as that of a coffee cup) to be superposed (S12). Thereafter, data
"1" are written at addresses corresponding to this boundary curve. Next,
all data of the pattern memory inside this boundary curve are replaced by
"1" (S13). Finally, the center of a selected brush pattern is positioned
at the coordinate of a point on the boundary between data "1" and "0" and,
while the boundary is traced along pixels, logical sums of the data
already stored in the pattern memory and the brush pattern data are
calculated, the results being written in the pattern memory (S14).
A still another method of preparing a mask pattern is explained next by way
of FIGS. 21 and 22. FIG. 21 shows contents of a pattern memory and an
enlarged view of a portion thereof. First, as done in the previous
example, the memory contents are cleared (S21), data on the boundary curve
are replaced by "1" (S22) and those inside the curve are replaced by "1"
(S23). Next, a pixel where "1" is not written is picked for consideration
and its distance d to the aforementioned boundary curve is calculated
(S24). If the pixel under consideration is denoted by X as shown in FIG.
21, the distance d is defined by the formula d=(dx).sup.2 +(dy).sup.2
where dx and dy are the minimum numbers of pixels in the horizontal and
vertical directions between the pixel X and the boundary curve. In the
example shown in FIG. 21, dx=4, dy=3 and hence d=25.
A function f(x) of a dummy variable x is assumed to have been preselected
for the purpose of determining whether "1" should be written or not at a
pixel separated from the boundary by distance x, and the value f(d) is
obtained by using this function (S25). The function f(x) may be so
selected that "1" should be written if f(x) is greater than a certain
threshold value a. As an example, f(x) may be defined by f(x)=50-d+rnd(49)
where rnd(n) indicates a random number between 0 and n, and a=50. If f(d)
is greater than or equal to the threshold value (YES in S26), "1" is
written at this pixel (S27). If f(d) is less than the threshold value (NO
in S26), "0" is allowed to stay at this pixel. This entire process is
repeated for all pixels where "1" was not written in the beginning (S28).
The present invention further teaches a preprocessing method in order to
make the boundary between two mutually superimposed source images even
less conspicuous. Let us consider a situation, with the example of the
image 52 of a coffee cup, wherein there are pixels corresponding to data
"1" in the mask pattern 55 outside the area of the cup and, furthermore,
wherein their density is not similar to that in the boundary area inside
the cup. In such a situation, pixels outside the cup before the synthesis
still remain outside the cup after the synthesis.
By way of illustration, let us consider a portion of the image of 52 of the
cup, and assume that the image data for this portion are as shown in FIG.
23A wherein the numerals inside squares indicate the pixel values
(densities) at the corresponding positions. Stated briefly, this
preprocessing method is to synthesize two source images after rewriting
the pixel values of extended peripheral areas of the boundary by the pixel
value of the cup itself and its environs.
With reference to FIG. 23A, it is immediately noticed that pixel values of
areas other than those of the cup are "9". According to the preprocessing
method of the present invention, the pixel value of any pixel outside the
area of the cup and corresponding pixel value of "9" is replaced by that
of the pixel which is closest to it among the pixels inside the area of
the cup. If such preprocessing is carried out regarding every pixel which
is outside the area of the cup but of which the distance from the boundary
area is within the size of the brush pattern being used, the result is as
shown in FIG. 23B. With a preprocessing thus effected, pixel values become
same at the boundary both inside and outside the area of the cup. Thus, if
pixels outside the cup area remain after the synthesis, they are made less
conspicuous.
The foregoing description of preferred embodiments of the invention has
been presented for purposes of illustration and description. It is not
intended to be exhaustive or to limit the invention to the precise form
disclosed, and many modifications and variations are possible in light of
the above teaching. Such modifications and variations that may be apparent
to a person skilled in the art are intended to be included within the
scope of this invention.
* * * * *
|
|
 |
|
 |
Description |
|
