JP2685460B2 - Digital image interpolator - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial applications]   The present invention relates to a digital image interpolating device, and more
The digital image forming the interpolated image data to be provided to the input device.
The present invention relates to a tar image interpolation device. [Conventional technology]   Generally, this type of device can perform enlarged interpolation processing of images.
You. However, output to laser beam printer (LBP) etc.
When doing, synchronize the processing result with the high-speed pixel clock signal.
Need to output. Therefore, in the past,
The same pixel is read out or enlarged as many times as necessary.
Store the interpolated image data in the shared memory and then
It was read and transferred. [Problems to be solved by the invention]   However, in the case of the above conventional example, the former method expands.
Image increases as the rate increases to 2 times, 3 times, 4 times, ...
Becomes a mosaic, and the image quality is extremely poor.
Was. In addition, the latter method increases the resolution of the printer to high definition.
Process as the print size grows
The memory capacity for storing the finished image data becomes huge,
There was a problem from the point of view.   The present invention has been made in view of the above conventional example, and is simple.
Digitizer that can obtain high quality interpolated image with circuit configuration
It is an object of the present invention to provide an interpolating device for a digital image. [Means to solve the problem]   In order to achieve the above object, the digital image supplement of the present invention is added.
The interim device has the following configuration.   That is, a digital device that provides the interpolated image data to the image output device.
In the interpolator for digital images, integers of orthogonal addresses X and Y
An image memory that stores a plurality of image data at an address,
Arbitrary interpolation position to the real number address of the orthogonal address X, Y
Interpolation position setting means for setting
2m × 2n (m and n are
Image data selector to select (positive integer) image data
Stage and the selected image data from the set interpolation position.
Distance calculating means for obtaining the distance x, y to the
And a coefficient ω (y) uniquely determined based on the distance y, and Y
Intermediate interpolation image data I by 2n pieces of image data in the direction
A first interpolation image data forming means for forming (Y);
Noted distance x and coefficient ω uniquely determined based on the distance x
(X) and the 2m intermediate interpolated image data formed in the X direction.
Form the final interpolated image data I with the data I (Y)
A second interpolated image data forming means,
And the coefficient ω.
Based on (y) and the interpolation coefficient I (y),
(Y) = ω (y) · sin (πy) / πy
The obtained interpolation coefficient I (y) and the interpolation coefficient I (y)
2n number of the obtained products by taking each product with the image data corresponding to
The intermediate interpolated image data I (Y) is calculated according to the sum of products.
And the second interpolation image data forming means forms the distance
Based on the distance x and the coefficient ω (x), the interpolation coefficient I
(X) is expressed by the following equation, I (x) = ω (x) · sin (πx) / π
x and the calculated interpolation coefficient I (x)
The formed intermediate interpolation image corresponding to the interpolation coefficient I (x)
Take each product with the image data I (Y), and obtain the product of 2m
The final interpolated image data I is formed by the summation,
The number ω (y) depends on the value of n, and the coefficient ω
(X) is changed according to the value of m.
A digital image interpolating device. [Action]   With the above-described configuration, the present invention will first supplement the Y direction.
Inter-coefficient I (y) = ω (y) · sin (πy) / πy and 2n
Interpolated image data I (Y) from the sum of products of
And then for the X direction, the interpolation coefficient I (x) = ω
(X) .sin (πx) / πx and 2m pieces of interpolated image data I
(Y) to form the interpolated image data I from the sum
Make. In addition, when performing the interpolation,
Depending on the block size, that is, the value of the coefficient ω (y)
Depending on the value of n and the coefficient ω (x) depending on the value of m
Change. [Description of Example]   Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.
explain. [First embodiment]   FIG. 1 shows the digital image interpolation of the first embodiment according to the present invention.
It is a block block diagram of an apparatus. The first embodiment is a television
The NTSC video signal of the
Processing is performed, and the resulting interpolated image data is sent to the image printer.
Regarding what to output.   In FIG. 1, 1 is a video signal source, for example, a video signal source.
Generates the NTSC video signal VS used in the revision. 2
Is a luminance signal extraction circuit, for example, a well-known color subkey
Video signal VS from the video signal
Extract the frequency signal VLS. 3 is a sync signal extraction circuit,
For example, with a known circuit in a conventional television receiver, etc.
Video signal VS to video frame sync signal VFDS (vertical
Sync signal) and video scanning line sync signal VHDS (horizontal sync signal)
No.) is extracted. 4 is a sampling cycle generation circuit
Based on the extracted sync signals VFDS and VHDS.
Data sampler for A / D conversion of video luminance signal VLS
Generate a signal DSS (eg, 640 pixels / scan line). 5 is
This is an A / D conversion circuit that converts the input video luminance signal VLS into A / D
The converted luminance data LD is output. 6 is gamma change
This is a conversion circuit that uses the input brightness data LD to obtain an appropriate density level.
And output the density data ND. Generally, TV
The signal is matched to the display surface fluorescent material for image reception.
It is a signal with combined (1 / 2.2) γ characteristics. Then γ
The conversion circuit 6 linearly converts the luminance data LD having this characteristic.
And change the density level to suit the printer characteristics.
Convert to density data ND. 7 is a frame memory,
Store the input density data ND for one page. 11 is write
This is an address control circuit, and the density data to the frame memory 7
Control write address WAD and write timing of data ND
You. Specifically, count the data sampling signal DSS.
And at what pixel position (X coordinate) in the video scan line
Calculates whether or not it is compatible and calculates the video scanning line synchronization signal VHDS
Counting and the number of the video scanning line in the screen (Y coordinate)
Is calculated. With the above configuration, each density data
Data ND (image data GD) is the data sampling signal DSS
In succession on the integer address of the frame memory 7 in synchronization with
Written in different ways.   FIG. 6 is a description of the image data in the frame memory 7.
It is a figure which shows a memory mode. In the figure, a series of image data
GD (density data ND) corresponds to scan line (raster) for printing
It is stored as corresponding raster data. For example first
The 1st pixel, 2nd pixel, ..., Nth pixel of one raster are written in order.
Remembered. If necessary, the first raster is placed between the first and second rasters.
The blank area for the studio is stored. This way, X address, Y
Hardware control of addresses is simplified. Then the second la
1st pixel, 2nd pixel, ..., Nth pixel of the
Thus, finally, the first pixel, the second pixel of the M-th raster,
The Nth pixel is sequentially stored. In this case
Memory 7 is the number of pixels corresponding to, for example, a power of 2 for each raster
The image data GD of is continuously stored. Memorized in this way
The size of one image is 640 (pixels / raster) x 480
(Raster / image), memory space for this is 1024
× 512 = 512K bytes should be prepared.   In FIG. 1, 10 is an image printer, and prints
A print page synchronization message that indicates the top of the page when starting
No.PPDS is sent, and a series of print scans is performed when printing.
Line sync signal PSDS (sub-scan sync signal) and each print scan
In-line print pixel sync signal PGDS (main scan sync signal)
Send out. These sync signals are transmitted via the printer I / F circuit 9.
Then, it is input to the interpolation processing circuit 8 described later. 8 is interpolation processing
It is a circuit that outputs various synchronization signals from the image printer 10.
To set the interpolation position (real number coordinates).
For updating the integer read address of the frame memory 7
Generates synchronization signals (various carrier signals CAR) of
In addition, the sum of the products of the interpolation coefficient I and the original image data (interpolation image
Data HGD) will be generated. 12 is the read address control time
The various carrier signals CAR from the interpolation processing circuit 8
The integer read address RAD of the frame memory 7 in synchronization with
I will update. With the above configuration, on the frame memory 7
The image data GD is interpolated by the interpolated image data HGD
Output to the image printer 10. 13 is a sequence control unit
Control the operating procedure of the entire device.   FIG. 3 is an image by the sequence control circuit 13 of the first embodiment.
It is a flow chart of the image data processing procedure. Figure smell
Then, in step S1, a print instruction is issued from an operation unit (not shown).
It is determined whether or not If there is no print instruction,
At step S2, other processing is performed. Other processing
For example, check the status of the image printer 10 (power off, etc.).
Or the state of the video signal source 1 (whether signal is being sent or not)
Etc.) Check, etc. If there is nothing wrong with this check
Promptly return to step S1. In addition, if the step S1 is determined,
If there is a print instruction, the process proceeds to step S3, and the write address
Function to initialize density control circuit 11 and then write density data ND
To the frame memory 7 of a series of density data ND
Start writing. Frame memory 7 in step S4
It is determined whether or not the writing to is completed. If writing is not completed,
Then, in step S5, other processing is performed in the same manner as described above.
When the writing of all density data ND is completed,
Proceed to S6 to set the data write function of the write address control circuit 11.
Disable. In step S7, the image printer 10
It is determined whether or not the printing can be started. Print
If it is not ready to start, proceed to step S8,
And perform other processing. In a state where printing can be started again
If there is, proceed to step S9 to initialize the interpolation processing circuit 8.
You. Specifically, a differential analysis circuit (DDA) 14 and D described later.
Setting of each constant value (initial value, increment value) in DA15, various patterns
Clearing of groups of data (FIFO, latch, etc.), data flow not shown
Initialize the control counter and clear various gates
Do. In step S10, the read address control circuit 12 is initialized.
Then, the data read operation is enabled. Specifically
Is the image data read address RAD of the frame memory 7.
Initialize the counter etc. to hold the first image data
Make GD ready for reading. Print I / F in step S11
The circuit 9 is initialized. Specifically, with the image printer 10
Control information necessary for passing control commands and status signals between
Initialize the information I / O circuit and send / receive the information.
Other, various synchronization signals related to image interpolation processing and transfer
Initialization and initialization of gate circuits etc. related to these signals
Is performed. In this way, from the image printer 10 onwards
Read add in synchronization with the operation clock signal (various sync signals)
Frame control circuit 12 and interpolation processing circuit 8 operate,
Image data GD is sequentially read from the memory 7 and interpolation processing is performed.
The interpolated image data HGD generated and generated is printed by the printer I / F circuit 9
To the image printer 10 via. Meanwhile, the sequence
The control circuit 13 completes printing in step S12.
Wait for That is, whether printing is completed by the determination in step S12
If printing is not completed, determine in step S13.
Then, the other processing is performed in the same manner as described above. Eventually pre
When the printing is completed, the process proceeds to step S14 and the read address system is set.
The image data reading function of the control circuit 12 is disabled.   FIG. 2 shows the details of the interpolation processing circuit 8 of the first embodiment.
It is a lock block diagram. In the figure, 14 is DDA (Digital D
ifferential Analyzer) and print sub scanning direction
Perform digital differential analysis of the dress (Y coordinate). Specifically
Is the integer read address (Y coordinate) of the frame memory 7.
Generates the sub-scan carrier signal VCAR for updating.
, The real number address (interpolation
Position Y coordinate). 15 is a DDA, printer
Performs digital differential analysis of main scanning direction address (X coordinate)
U. Specifically, the integer read address of the frame memory 7
Main scan carrier signal HCAR for updating (X coordinate)
When generated, the real number between the integer read addresses
A dress (X coordinate of interpolation position) is generated. 16 is raw data
A group of scanning line buffers, which are sequentially read from the frame memory 7.
The projected image data GD can be stored in parallel for 7 scanning lines.
Can be. 17 is a sub-scanning direction interpolation table group,
Temporary about 8 image data GD lined up in one line in the scanning direction
Sub-scanning direction component interpolation coefficient I_YAnd the input image data
In order to obtain the distance between the set interpolation position and each image data
Each has a predetermined coefficient I based on the distance y._YGenerates the value of
I_YAnd the image data are taken. 18 is the sub-scanning direction
Interpolation adder group, which is the total of each product by interpolation table group 17
I is calculated as the temporary interpolation result in the sub-scanning direction._YAnd enter
Calculate the sum of products with force image data.   Reference numeral 19 is a main scanning direction interpolation data buffer group,
Scan direction interpolation Adder group 18 output temporary interpolation result I_YAnd enter
Accumulate 7 products in total in the main-scanning direction
Can be obtained. 20 is a main scanning direction interpolation table group
, 8 interpolation coefficients I arranged in one line in the main scanning direction_YAbout
Obtain the sum of the products of the final interpolation result I and the input image data
In order to adjust the distance between the set interpolation position and each interpolation image data.
Each has a predetermined coefficient I based on the distance x._XGenerates the value of
I_XAnd the above-mentioned temporary interpolation result.
Reference numeral 21 denotes a main scanning direction interpolation adder group, which is an interpolation table group 20.
The final interpolated image data V_out
Ask for.   FIG. 13 is a diagram showing tables 17 and 20 of the embodiment.
These are realized by ROM, and the upper side of the address input is
Address corresponding to the interpolation position information (Δx or Δy)
The lower side of the force corresponds to the value of the input pixel. Of the table
The contents are the interpolation coefficient of the interpolation position corresponding to each address.
The product of the corresponding pixel values is preset.   Next, details of the operation of the interpolation processing circuit 8 will be described. DDA14
When the page sync signal PPDS is input to
Change to the mode to select and output the contents of the period value register 14-2
I do. As a result, the adder 14-5 causes the initial value register 14-2
Add the initial value of and the increment value of increment value register 14-1
The addition result of is output. First scan line sync signal PSDS
When is input, the addition output of adder 14-5 is the current value register
14-3, and at the same time, the selector 14-4 displays the current value
The mode changes to select and output the contents of register 14-3.
Therefore, after that, every time the scanning line synchronization signal PSDS is input, the current
The contents of the value register 14-3 is the value obtained by adding the increment value to the previous value.
Will be updated by. Also during this update cycle the adder
Whenever an additional carrier occurs in 14-5, each time a carrier message is received
Signal VCAR is generated and output to the read address control circuit 12.
I do. From the above, the output of the adder 14-5 is
Except for the signal VCAR, the rest when the carry is normalized to 1.
Considered to be a minority part of. Therefore, video signal VS side run
If you consider the sampling interval distance to be unit distance 1,
The output of adder 14-5 from each sampling position
It represents a real number distance y measured in the scanning direction.   Similarly, scan line synchronization signal PSDS is input to DDA15.
And selector 15-4 selects the contents of initial value register 15-2
Change to output mode. As a result, the adder 14-5
Initial value and increment value register 15-1 of initial value register 15-2
The increment value of is added and the addition result is output. Then first
When the pixel synchronization signal PGDS of is input, the addition output of the adder 15-5
The force is taken into the present value register 15-3 and simultaneously selected.
15-4 selects and outputs the contents of the current value register 15-3
Change to mode. Therefore, after that, the pixel synchronization signal PGDS
Each time you enter, the current value register 15-3 will be set to the previous value.
It is updated by adding the increment value. Also this update service
If an addition carrier occurs in the adder 15-5 during the cycle,
A carrier signal HCAR is generated each time
Output to the control circuit 12. From the above, the adder 15-5
The output is normal to carry, except for the carrier signal HCAR.
It is considered to be the remaining minority part when converted. Therefore, bidet
Unit distance is the sampling interval of the signal direction
When it is considered as 1, the output of the adder 15-5 is each sampler.
Represents the real number distance x measured from the scanning position in the main scanning direction.
You.   FIG. 5 is a schematic diagram for explaining the operation mode of the DDA of the first embodiment.
FIG. Figure 5 shows an original image enlarged by 130% as an example.
It expresses the operation of DDA when performing interpolation processing. 130%
When enlarging, the value of 1 / 1.3 is displayed in binary as the increment value.
Use 15 bits after the decimal point. Hexagonal table for the minority part
It shows "313B". On the other hand, the initial value is the increment value "313B"
When the 8th carrier signal CAR is generated by adding
Set a value such that the current value (decimal part) becomes "0"
I do. The initial value in this case is "13B2". This
In Fig. 5, the first 10 clock signals are input from the printer 10 side.
The relationship that the first carrier signal CAR is generated when applied
Have been. Here, the clock signal from the image printer 10 side
As for the pixel synchronization signal PGDS, Fig. 5 shows the main scanning direction DDA15.
The following is an example of how this works for the image printer 10.
Input image while outputting 10 interpolated image data HGD
The data GD shows a relationship in which eight data are read.
If the clock signal is the scanning line synchronizing signal PSDS, it is shown in FIG.
Shows an example of operation in the sub-scanning direction DDA14.
Interpolation image data HG for 10 lines for image printer 10
While outputting D, input image data GD reads 8 lines
It shows the relationship of being issued. In addition, FULL15 bit
It is not necessary to add, and it is possible to carry out with the upper few bits.   In FIG. 2, the read address control circuit 12 is the same as the page.
Reset by PPDS, then DDA15
Of the carrier signal HCAR of the frame memory 7
Generate the lower read address (main scanning direction address)
The carrier signal VCAR from the DDA14 is counted and framed.
Higher-order read address of the memory 7 (address in the sub-scanning direction)
Generate Read address signal sent to frame memory 7
RAD is a combination of the upper and lower read addresses.
is there.   Raw data scan line buffer group 16 has 7 shift registers
(FIFO 16-1 to 16-7) are included in the frame memory
The image data GD for 7 scanning lines read from 7 is stored.
That is, the image data GD read from the frame memory 7 is
It is input to the FIFO 16-1 and shifted by 1024 pixels. Then FI
The output of FO16-1 becomes the input of FIFO16-2, and so on.
Then, the output of the FIFO16-6 becomes the input of the FIFO16-7. Soshi
Image data GD input to FIFO16-7 is 1024 pixels.
When shifted, the frame memory 7 synchronizes with this
The image data GD of the eighth scanning line is read out. This second
If the 8th scan line read raster is the current read raster,
Contents stored in any FIFO16-i (i = 1,2, ..., 7)
Is a read raster before i lines.   The sub-scanning direction interpolation table group 17 is composed of eight
Table (LUT) 17-1 to 17-8. Each LUT is a pair
Image data GD of the corresponding read raster and the current sample
Real number distance y (adder 14
The output of -5) is input and the density of the image data GD is
For the density of the interpolated image data HGD that should be generated between the positions
Contribution rate of how much to contribute from the sub scanning direction (sub
Between the partial interpolation coefficient Iy) in the scanning direction and the corresponding input pixel value
Output the product. For example, LUT17-1 is the current read raster
Image data of GD₀And interpolation of adder 14-5 output at present
The position y is used as an address input, and the LUT17
-1 is a predetermined interpolation coefficient Iy corresponding to the current interpolation position y
₀(Partial interpolation coefficient) value and the image data GD₀Value of (dark
Degree) product of Iy₀GD₀Is internally generated and output. same
In this way, the LUT17-8 is the image of the FIFO16-7 output 7 lines before.
Image data GD₇And the interpolation position y of the adder 14-5 is addressed
It is used as an input, so that the LUT17-8 can be
Interpolation position y obtained by adding 7 scanning line intervals to interposition y₇Corresponding to
Predetermined coefficient Iy₇Value (partial interpolation coefficient) and the image data GD₇
Value of product with value of Iy₇GD₇Is internally generated and output. This
Thus, the contents of the image data GD will cause the carrier signal HCAR to be generated.
The above-mentioned partial interpolation coefficient Iy_n
And the input image data GD are sequentially calculated. one
However, the contents of the read raster are synchronized with the generation of the carrier signal VCAR.
Then, the lines are updated one by one. And during this time,
Normally, the scan synchronization signal PSD is at a ratio of 1 or more (1.2, 1.3 etc.)
S occurs, and in synchronization with this, the interpolation position y in the sub-scanning direction
Will be updated. Then, the sub-scanning method of the image data GD
Expanded interpolation to the direction is performed.   In the above embodiment, the partial interpolation coefficient Iy_nIs currently calculated
Read raster and the preceding 7 line read raster
It is done at the same time, but in reality the interpolation position at the present time is approximately
Interpolation processing based on the image data on the upper and lower four lines
Sometimes you want to. In this case, for example, adder 14-5 output
4 stage shift register between the power and each LUT 17-1 to 17-8
Delay the phase of the current interpolation position by inserting
Just do it.   The sub-scanning direction interpolation adder group 18 is composed of LUT17-1 to LUT17-8.
Add all outputs and output the sum. Thus adder
The output of 18-7 is the image in the sub-scanning direction at the current interpolation position.
Temporary partial interpolation value GD based on image data GD_YIt is.   Next, the interpolation processing in the main scanning direction will be described. Main scanning direction supplement
Inter-process data buffer group 19 is synchronized with carrier signal HCAR
And sequentially store (latch) partial interpolation values in the sub-scanning direction.
You. This memory is from latch 19-1 to latch 19-2, latch
Synchronized with carrier signal HCAR from 19-2 to latch 19-3
And latch in the shift register format. Therefore, adder 18
If the time when the output of -7 becomes effective is the latch,
19-7 is the partial interpolation value GD 7 columns before_Y7Holding a rat
Chi 19-6 is 6 rows ago, ... Similarly, latch 19-1 is 1 row
Previous partial interpolation value GD_Y1Holding.   The main scanning direction interpolation table group 20 is composed of eight
Table (LUT) 20-1 to 20-8. Each LUT is a pair
Corresponding partial interpolation value GD_YAnd the current sampling position
Real distance x measured in the main scanning direction (adder 15-5 output)
Input the relevant partial interpolation value GD_YThe density of
Main scanning direction for the density of interpolation image data HGD to be created
Contribution ratio (in the main scanning direction)
Partial interpolation coefficient Ix) and the corresponding partial interpolation value GD_YOutputs the product of
You. For example, LUT20-1 is the current partial interpolation value GD_Y0And
The current interpolation position x of the calculator 15-5 output is used as the address input.
As a result, the LUT 20-1 is set to the current interpolation position.
The value of the predetermined coefficient Ix (partial interpolation coefficient) corresponding to x and the relevant part
Minute interpolation value GD_Y0The value (concentration) of
Output. Similarly, the LUT20-8 outputs the FIFO19-7 output.
Partial interpolation value GD 7 columns before_Y7And the interpolation position x of the adder 15-5
Is used as an address input, so that the LUT20-8
Interpolation position x obtained by adding 7 pixel intervals to the current interpolation position x₇To
Corresponding predetermined coefficient Ix₇Value of (partial interpolation coefficient) and the part
Interpolated value GD_Y7The value of the product of
You. Thus the partial interpolation value GD_Y7Is the carrier signal HCAR
The above-mentioned partial interpolation coefficient Ix
_nThe product of and is calculated sequentially. And during this time,
Normally, the pixel synchronization signal PGDS is set to 1 or more (1.2, 1.3, etc.).
Occurs, and the interpolation position x in the main scanning direction is updated in synchronization with this.
Be renewed. And by this, the main scanning direction of the image data GD
The expanded interpolation is performed.   In the above embodiment, the partial interpolation coefficient Ix_nIs currently calculated
Partial interpolation value GD_YAnd the preceding 7-part interpolation value GD_Y8 pixels
At the same time by using
Interpolated value I for 4 pixels before and after the interpolation position of_Y
There are also cases in which it is desired to interpolate. In this case, for example
There are 4 stages between the output of adder 15-5 and each LUT 20-1 to 20-8.
The current interpolation position by inserting a shift register
The phase of may be delayed.   The main scanning direction interpolation adder group 21 is composed of LUT20-1 to LUT20-8.
Add all outputs and output the sum. Thus adder
The output of 21-7 is the final interpolation value at the current interpolation position.
V_out(Interpolated image data HGD).   4A and 4B show the interpolation processing circuit 8 and the image printing.
Timing chart for explaining the mode of synchronization between data
It is. In FIG. 4A, the image printer 10 is
When the state becomes a, the sequence control circuit 13 reads out
Responsive enable signal is issued, and printing is in process.
From the image printer 10, the page sync signal PPDS and
The scanning line synchronization signal PSDS is output and the interpolation processing is synchronized with this.
Is performed.   FIG. 4 (B) is an enlarged table of one scanning line period in FIG. 4 (A).
FIG. In FIG. 4B, one scanning line period
Pixel sync signal PGDS is generated at a predetermined interval within
The carrier signal HCAR is output at a fixed ratio smaller than this. one
However, the interpolated image data HGD should be synchronized with the pixel sync signal PGDS.
Is output.   Next, the enlargement interpolation processing of the original image will be described in detail. Fig. 7
FIG. 7A is a diagram showing an original image of a person, and FIG.
The figure which shows the image which carried out the expansion interpolation process of the original image of FIG. 7 (A).
It is. The magnification of the image in the figure is 5.75 times in the main scanning direction.
Which is 4.25 times in the sub-scanning direction. Of such images
Enlargement uses the interpolation method of the present invention to increase the number of pixels in the original image.
It can be realized easily.   FIG. 8A is a diagram showing a partial area of the original image.
FIG. 8B shows the partial region of FIG. 8A in the main scanning direction.
75 times and 4.25 times in the sub-scanning direction. Fig. 8
The pixel pitch of (A) samples the video image signal VS.
Corresponding to the data sampling pitch when input
Fig. 8 (B) shows the sampling pitch as the minimum unit.
Is shown as a relatively enlarged partial area. Obedience
The partial area a of the original image corresponds to the partial area a ′ of the enlarged image.
Therefore, the interpolation position b assumed in the original image is the enlarged partial area.
It corresponds to the output pixel position b'in a '. Output there
The interpolation image data HGD is formed at the pixel position b '.
However, the value (density) of the interpolated image data HGD is determined.
Is a predetermined plurality of pixels surrounding the output pixel position b ′.
Each image from the pixel position b'using the value of the image data GD
Of the contribution rate (interpolation coefficient) determined by the distance to the data GD
Weighted, done by taking the sum of them
It is.   Next, when the value of the output pixel position b ′ in FIG.
I will explain the case. This corresponds to the interpolation position b in Fig. 8 (A)
It is equivalent to obtaining the real value. So Fig. 8
Rewrite (A) as shown in FIG. X mark in FIG.
The position is at point b. Here, the pixel position of the original image is marked with a circle.
The sampling pitch of the image data GD is mainly
The scanning and sub-scanning directions are equally spaced. Moreover, in FIG.
Think of the pixel spacing of an image as a unit distance (eg distance 1)
The coordinates in the main scanning direction are on the right with the interpolation position marked with × as the origin.
The direction is positive, the left direction is negative, and the coordinates in the sub-scanning direction are the same.
Set the interpolation position of the X mark as the origin to positive in the downward direction and upward in the upward direction.
Consider another coordinate system that is negative. Therefore new
In the horizontal coordinate system, the positions of the original image are ± Δx and ± Δy, respectively.
It is represented by applying. Δx and Δy are in the main scanning direction, respectively
And an original having a coordinate value that does not exceed the interpolation position b in the sub-scanning direction.
It is the distance from the pixel position c of the image.   Now, the main scanning direction between the pixel position of a certain original image and the interpolation position
When the distance of is set to x and the distance in the sub-scanning direction is set to y
In the main scanning direction (interpolation coefficient) I (x) and sub-scanning
The contribution ratio (interpolation coefficient) I (y) of each direction is calculated by the following equations.
To represent.   Here, we chose a multiple of 2 such that m = 4 and n = 4. next
This I (x), I (y) is the value that the input image data GD has
And multiply all input images within the range where this operation is used for interpolation.
Image data GD is calculated, and the interpolated value with the sum of them
I. That is, the value of the interpolation value I (interpolation image data HGD)
V_outThen, as shown in FIG. 9, the interpolation point b is surrounded by eight surrounding pixels.
Consider interpolating with a total of 64 pixel data of × 8 raster
And V_outIs Required by. This is an interpolator for 64 pixels
Find the value multiplied by a number and then sum the 64 values
Means However, if you try to realize this as it is
The hardware becomes huge and not practical. Also soft
If air is used, a sufficient processing speed cannot be obtained. So above
V_outThe expression is transformed into the following.   Thus, first the interpolation in the sub-scanning direction is performed, and then the
It is the same as above that interpolation in the main scanning direction is performed using the result.
Value. The first embodiment described above is realized according to this method.
ing.   FIG. 10 is a diagram showing details of the principle of interpolation in the main scanning direction.
You. In the figure, the symbol (V_i-Δ_x) Is the coordinate (i-Δx)
It is the value of the image data GD of the point. With these, interpolation image data
Data HGD value V₀Is calculated as follows.   Also, It is. The LUT instantly does the equivalent of this calculation.   FIGS. 11A to 11C are sub-scans using original image data.
Image data after interpolation in the main scanning direction after interpolation in the
FIG. 6 is a diagram showing how to increase the number of pixels (how to increase the number of pixels). No.
In Fig. 11 (A), 8 consecutive scan lines in the input image
Used, that is, at the same position in the sub-scanning direction on each scan line
Create a scan line using pixels to interpolate only in the sub-scan direction
I do. In FIG. 11 (B), the supplement is made only in this sub-scanning direction.
Interpolation is performed in the main scanning direction using the pixels on the intervening scanning lines.
Now. As a result, as shown in FIG. 11C, the pixels in one scanning line
The number of scan lines in an image
The output scan lines occur in a different manner. [Second Embodiment]   The second embodiment is the input pixel required to calculate the interpolation value.
Regarding generalization of area selection. That is, in the second embodiment
For example, each correction coefficient formula I (x),
I (y) The input pixel area is 8 × 8 in the first embodiment.
It is an extension of the above-mentioned one to 2m × 2n. [Third Embodiment]   The third embodiment relates to a modification of the configuration shown in FIG. That is, supplement
The LUT group of the inter-table groups 17 and 20 has a predetermined coefficient ω (x) or
Is a LUT that outputs only ω (y), its LUT output, and the pixel
You may comprise as a multiplier which calculates the product of a value.   This can significantly reduce the LUT capacity.
is there. By the way, the third embodiment trades image quality and circuit scale.
The size can be determined by turning off, and the present invention itself depends on the size.
It also shows that it does not exist. [Fourth Embodiment]   The fourth embodiment relates to a modification of the configuration shown in FIG. That is,
In one embodiment, DDA14,15 add and store only a fractional part.
It was configured to hold and output the carrier signal CAR,
Furthermore, the bit length of DDA can be extended to hold the integer part as well.
And specify the input pixel position with this integer value part
May be configured. [Fifth Embodiment]   The fifth embodiment relates to a modification of the configuration shown in FIG. That is,
In one embodiment, initialization of increment values and initial values of DDA14,15
Is performed by the control unit of the image printer 10.
However, in addition to that, connect the switch and short circuit signal pin.
You may do it by the method of making a short. This one aspect
As shown in FIG. [Sixth Embodiment]   The sixth embodiment relates to a modification of the configuration shown in FIG. That is,
In one embodiment, the DDA 14,15 sets the initial and current values
Hold each separately and select input to adder with selector
However, other than that, the cash register that holds the current value
For example, in the case of DDA in the main scanning direction, the scanning line synchronization signal PSD
In synchronization with S, and page synchronization if DDA in the sub-scanning direction
It may be configured to reinitialize in synchronization with the No. PPDS.
In this case, the initial value holding circuit and selector are unnecessary.
Interrupts the sequence control circuit by inputting a sync signal.
Signal, or reset the holding circuit with a sync signal.
Scanning is required. [Seventh embodiment]   The seventh embodiment relates to another aspect of the interpolation coefficient formula. That is,
It is preferable to interpolate at 2m × 2n points surrounding the interpolation position.
The interpolation coefficient formulas I (x) and I (y) are as follows.
Is also good.   In this way, ω (x) and ω (y) can be diversified. 7th real
The embodiment can set the value more easily than the second embodiment.
However, the light and shade stripes generated at the pixel intervals are strong. [Eighth Embodiment]   The eighth embodiment relates to another aspect of the interpolation coefficient formula. That is,
It is preferable to interpolate at 2m × 2n points surrounding the interpolation position.
The interpolation coefficient formulas I (x) and I (y) are as follows.
Is also good. In this way, ω (x) and ω (y) can be diversified. This
If so, the generation of light and shade stripes is weaker than in the first embodiment. A little more
However, the interpolation after obtaining the same number of interpolation points
The image sharpness drops. Therefore, depending on the application
There is room for division. [Ninth Embodiment]   The ninth embodiment relates to another aspect of the interpolation coefficient formula. That is,
It is preferable to interpolate at 2m × 2n points surrounding the interpolation position.
The interpolation coefficient formulas I (x) and I (y) are as follows.
Is also good. In this way, ω (x) and ω (y) can be diversified. This
If so, the generation of light and shade stripes is weaker than in the first embodiment. A little more
However, in the case of interpolation that finds the same number of interpolation points,
The sharpness of the statue drops. Therefore, depending on the usage
There is room for injury. [Tenth Embodiment]   The tenth embodiment relates to another mode of the interpolation coefficient formula. That is,
It is preferable to interpolate at 2m × 2n points surrounding the interpolation position.
The interpolation coefficient formulas I (x) and I (y) are as follows.
Is also good.   Where Io [] is the modified zero-order Bessel function of the first kind,
See also ω_a, Ω_bIs a configuration parameter, Select and use a value in the range. I_o[X] is, for example, And the like.   In this way, ω (x) and ω (y) can be diversified. First real
The generation of light and shade stripes is weaker than in the example. Also a little
However, the interpolation of the same number of interpolation points
Jap drops. For this reason, it can be used properly depending on the application
There is a ground. [Eleventh embodiment]   The eleventh embodiment relates to another mode of the interpolation coefficient formula. That is,
Interpolation coefficient is used to interpolate at 4x4 points surrounding the interpolation position.
Formulas I (x) and I (y) may be as follows: I (x) = (a + 2) x^Three-(a + 3) x^Two+1 (0 ≦ | x | <1) I (x) = ax^Three-5ax^Two+ 8ax-4a (1 ≦ | x | <2)   I (x) = 0 (2 ≦ | x |) Here, a is a real number constant. The same applies to I (y). [The invention's effect]   As described above, according to the present invention, first, the Y direction is related.
Interpolation coefficient I (y) = ω (y) · sin (πy) / π
Interpolated image data I based on the sum of products of y and 2n image data
(Y) and then for the X direction the interpolation factor I
(X) = ω (x) ・ sin (πx) / πx and 2m interpolation images
Interpolation image data I is formed from the sum of the image data I (Y)
For simple functional expressions and simple arithmetic processing such as
Therefore, the interpolated image data is formed, which is relatively easy.
The effect that an interpolated image can be formed with the circuit configuration
There is.   In addition, at the time of the interpolation, the value of the coefficient ω (y) is changed to the value of n.
And the coefficient ω (x) is changed according to the value of m.
Therefore, depending on the block size of the image data to be referenced,
You can set the optimum coefficient by
There is also an effect that a quality interpolation image can be obtained.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of a digital image interpolating device according to a first embodiment of the present invention, and FIG. 2 is a block diagram showing details of an interpolation processing circuit 8 according to the first embodiment. 3 is a flow chart of the image data processing procedure by the sequence control circuit 13 of the first embodiment, and FIGS. 4 (A) and 4 (B) are the interpolation processing circuit 8 and the image printer.
An operation timing chart explaining a mode of synchronization between 10 and 10, FIG. 5 is a schematic diagram explaining an operation mode of the DDA of the first embodiment, and FIG. 6 is a diagram showing one storage mode of image data in the frame memory 7, FIG. 7 (A) is a diagram showing an original image of a person, FIG. 7 (B) is a diagram showing an image obtained by enlarging and interpolating the original image of FIG. 7 (A), and FIG. 8 (A) is an original image. FIG. 8 (B) shows the partial area of FIG. 8 (A) in the main scanning direction.
5.75 times, 4.25 times in the sub-scanning direction, FIG. 9 shows details of FIG. 8 (A), FIG. 10 shows details of the main scanning direction interpolation principle, FIG. 11 ( A) to (C) are diagrams showing how to increase the image data (increase in the number of pixels) when the original image data is used to interpolate in the sub-scanning direction and then in the main scanning direction. FIG. 13 is a block diagram showing a part of the configuration of the DDA of the fifth embodiment, and FIG. 13 is a diagram showing tables 17 and 20 of the embodiment. In the figure, 1 ... Video signal source, 2 ... Luminance signal extraction circuit, 3 ...
Sync signal extraction circuit, 4 ... Sampling cycle generation circuit, 5
... A / D conversion circuit, 6 ... γ conversion circuit, 7 ... Frame memory 7, 8 ... Interpolation processing circuit, 9 ... Printer I / F circuit, 10 ... Image printer, 11 ... Write address control circuit, 12 ... Read address Control circuit, 13 ... Sequence control circuit.

Claims (1)

(57) [Claims] In a digital image interpolating device that provides interpolated image data to an image output device, an image memory that stores a plurality of image data at an integer address of orthogonal addresses X and Y, and an arbitrary interpolation at a real number address of the orthogonal addresses X and Y. Interpolation position setting means for setting a position, and image data selection means for selecting 2m × 2n (m, n is a positive integer) pieces of image data so as to surround the interpolation position with the set interpolation position as a center. A distance calculation means for calculating a distance x, y from the set interpolation position to the selected image data, the distance y, and a coefficient ω uniquely determined based on the distance y.
First interpolation image data forming means for forming intermediate interpolation image data I (Y) from (y) and 2n pieces of image data in the Y direction, the distance x, and a coefficient uniquely determined based on the distance x. ω
(X) and a second interpolated image data forming means for forming final interpolated image data I from the formed 2m intermediate interpolated image data I (Y) in the X direction. The data forming means obtains an interpolation coefficient I (y) based on the distance y and the coefficient ω (y) according to the following equation: I (y) = ω (y) · sin (πy) / πy Each product of the calculated interpolation coefficient I (y) and the image data corresponding to the interpolation coefficient I (y) is calculated,
The intermediate interpolation image data I (Y) is formed by the sum of the obtained 2n products, and the second interpolation image data forming means is based on the distance x and the coefficient ω (x), The interpolation coefficient I (x) is calculated according to the following formula: I (x) = ω (x) · sin (πx) / πx, and the calculated interpolation coefficient I (x) and the interpolation coefficient I (x) are obtained. The respective products of the formed intermediate interpolation image data I (Y) are taken, the final interpolation image data I is formed by the sum of the obtained 2m products, and the coefficient ω (y) is set to the value of the n. , And the coefficient ω (x) is changed according to the value of m.