JPH07210668A - Image information processing system - Google Patents

Abstract

PURPOSE:To obtain an artificially half-tone image in a form adapted to the sense of vision of human beings without the degradation of the image quality by performing the processing which cancels the rotation effect after the binary image conversion processing of image data subjected to the processing for the image rotation effect. CONSTITUTION:Input image data is written in a memory 11, and rotated image data is obtained from the memory 11 under the read control from a read control rotation control part 13 which receives rotation angle information O. The image data is subjected to interpolation processing by an interpolation part 14 and is inputted to a subtraction part 15, and the subtraction part 15 subtracts data stored in a dither pattern ROM 17 from interpolated data from the interpolation part 14 to perform the comparison, and one-bit output data is written in a memory 16. This writing to the memory 16 is controlled by a write control part 18, and data is read out from the memory 16 to be subjected to the reverse rotation processing by the control of a read control rotation control part 19 based on rotation angle information O and is outputted to, for example, a printer.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image information processing system, and more particularly to an image information processing system for solving the problem of image quality deterioration due to enlargement, reduction, rotation processing and the like.

[0002]

2. Description of the Related Art In an image information processing system for performing digital processing such as enlargement, reduction and rotation on an image signal,
Various problems occur depending on the treatment to be performed. For example, in the case of converting an analog signal into a digital signal for processing and performing a predetermined image processing and then converting into an analog signal, the sampling clock frequency fCLK is determined based on a well-known sampling theorem. fCLK 1
For signal components having frequency components of / 2 or more,
As shown in FIG. 2 (A), aliasing components occur, and beats and moire are generated in the finally obtained image signal, which deteriorates the image quality. Therefore, for the signal input to the A / D converter, the problem of the aliasing component is removed by using a low-pass filter (LPF) having a cutoff frequency of fCLK / 2, and the signal as shown in FIG. Is getting

By the way, in an image processing system such as an electronic still camera, there is a case where the image signal is subjected to processing such as reduction and enlargement for convenience. For example, when the reduction processing is performed, in FIG.
Each pixel is set to 1 for 768 × 240 pixels which is a video signal
Select every other, and 384 x 1 as shown in FIG.
A reduced image of 20 pixels is obtained. FIG. 6C shows the timing relationship of this reduced image processing, in which the original data is sampled based on the sampling clock fCLK,
By fetching and latching every other pixel data, the same pixel data is reduced to fCLK / 2 as the reduced data.
It is obtained based on the sub-sampling clock of the frequency of. Since this is substantially the sampling at the frequency of fCLK / 2, as shown in FIG.
A signal in the frequency band of CLK / 4 to fCLK / 2 causes a folding problem. Therefore, it is necessary to provide a low-pass filter having a cutoff frequency fCLK / 4 before the A / D converter.

More specifically, the reduction processing is, as shown in FIG. 24, a sampling interval (subsampling) larger than 1 with respect to the original pixel data having a pixel interval of 1 shown by a white circle. The reduction processing is to generate and read data of (interval). In this case, the sampling interval becomes larger than 1, and the cutoff frequency of the low-pass filter becomes 1 / (subsampling interval) · fCLK / 2.

Therefore, conventionally, as shown in FIG. 25, a plurality of analog low-pass filters (LPF1 to LPF5, ...) Having a cutoff frequency according to the reduction ratio are prepared, and low-pass filters corresponding to the reduction ratio are provided. Subsampling processing is performed after the output signal selected by the switch circuit is A / D converted. In the halftone dot dither, halftone dot processing in a diagonal direction of 45 degrees, which is advantageous for human visual quality, is performed, and dots are arranged in the 45 degree direction, and the apparent density is displayed by changing the size of each dot. is doing.

For example, as shown in FIG. 13, the diameter of the black circle in each square area is changed to display the shading, and as shown in FIG. 14, processing is performed on a pattern having an oblique angle (screen angle) of 45 degrees. . This type of dithering process may be performed on each of cyan, magenta, yellow, and black when performing color printing.

When halftone dot dither is applied to a block of length x and an interval X shown in FIG. 15A, processing is performed on a block inclined by θ = 45 degrees as shown in FIG. Is applied.

FIG. 16 is an illustration of an example of dither processing. Each unit of the dither matrix of the block composed of 8 × 8 pixel units is provided with a threshold value that is arranged in a spiral shape from the center and has a sequentially increasing value, and each threshold value is set in advance. Comparison value 25
When each unit value is larger than 25, white is given, and when it is smaller, black is given, and finally, a substantially circular black circle as shown in FIG. Therefore, the comparison value is 25
If it is set to a larger value, the black circle in the figure (B) becomes larger,
If the comparison value is set smaller than 25, the black circle becomes smaller.

[0009]

As described above, in the conventional image information processing system, in order to solve the problem of aliasing, an analog low-pass filter having a cutoff frequency according to the reduction ratio is provided in the preceding stage of the A / D converter. A plurality of low pass filters corresponding to the reduction ratio are used by switching. However, preparing a plurality of low-pass filters having different cut-off frequencies causes a problem in cost, and an analog low-pass filter having a variable cut-off frequency cannot be easily obtained.

Further, in FIG. 17A, a dither pattern (indicated by a dotted line) used when performing dither processing for each pixel indicated by a solid line is required, but when performing processing for generating each block of the dither pattern. , The figure (B)
As shown in FIG. 5, the threshold value may be spread over a plurality of threshold values, making it impossible to compare with the comparison value.

Therefore, conventionally, as compared with FIG.
In the case of rotating by the angle θ, the adjacent unit is shifted in the vertical and horizontal directions so that the angle formed by the line connecting the center of the pixel unit and the center of the adjacent pixel unit becomes θ as shown in FIG. There is. In this case, the distance X between adjacent pixel units is X = x / cos θ, and the interval X increases as θ increases.
At this time, if θ is 45 degrees, as shown in FIG. 7C, X = √2x, and the shaded areas in the figure cannot be processed, and only every other image can be processed. And the image quality deteriorates.

[0012] Therefore, an object of the present invention is to provide an image information processing system capable of obtaining a pseudo halftone image in a mode more suitable for human vision without deterioration of image quality.

[0013]

In order to solve the above-mentioned problems, the image information processing system according to the present invention has a first rotation effect for performing processing for obtaining a predetermined image rotation effect on the supplied image data. Means, a binary image processing means for performing a predetermined binary image processing on the output data of the first rotation effect means, and the first rotation effect on the output data of the binary image processing means. Second rotation effect means for performing processing for obtaining an image rotation effect that cancels the image rotation effect in the means.

[0014]

According to the present invention, the binary image processing is performed on the image data that has been processed to obtain the predetermined image rotation effect, and the image rotation effect that cancels the above-described rotation effect is obtained on the obtained binary image data. I am giving it.

[0015]

Embodiments of the present invention will now be described with reference to the drawings. FIG. 1 is a configuration block diagram showing an example of an image information processing system related to the present invention. The image signal has a low-pass filter (LPF) 1 having a cut-off frequency of fCLK / 2 removed components of fCLK / 2 or more, converted into a digital signal by the A / D converter 2, and then input to the digital filter 3. , Sub-sampling part 4
Is subsampled by. The digital filter 3 is a digital filter whose cutoff frequency changes according to the reduction ratio, and has, for example, a well-known one-dimensional IIR filter configuration as shown in FIG. 2A, and is added to the multipliers 31 and 32. 33, 34, adder 35, 1-clock delay unit 3
6 and.

The transfer function of the digital filter shown in FIG. 2A is H (Z) = (1-a) / 2 (1 + Z-1) / (1-a)
Z-1) and by changing the coefficient a (0≤a <1),
As shown in FIG. 2B, the frequency characteristic changes and the cutoff frequency can be changed. That is, the coefficient a
The cutoff frequency can be set smaller as is increased.

Although the above description is for the reduced image processing, the same thing occurs during the image rotation processing. For example, consider a rotation process of rotating an NTSC field image (longitudinal rectangular pixel of 1: 2.4) 90 degrees to the left as shown in FIG. 3A. Then, the process of obtaining the pixel indicated by x from the original pixel of the white circle is performed. In this case, the pixel interval in the horizontal direction is 1 in the original image, whereas the image after rotation by 90 degrees is 2.4, which necessitates thinning processing. Therefore, the aliasing phenomenon occurs at (fCLK / 2) · (1 / 2.4), and this aliasing frequency changes according to the rotation angle, so the cutoff frequency of the low-pass filter also needs to be changed accordingly.

It is also possible to prepare a cut-off frequency determined in accordance with the rotation angle of the rotation processing in a table format and select a cut-off frequency suitable for the processing.

According to the above, it is possible to solve the problem of aliasing that occurs when the sub-sampling interval is larger than the original interval, as well as the image reduction and rotation processing.

Therefore, the above example can be applied to the case of correcting the pincushion type distortion as shown in FIG. 4A which is one of the optical distortions. The solid line in FIG. 4A is the actual image data obtained from the optical system and the photoelectric conversion unit, and the optical axis position O
The farther away from, the actual image is obtained at a position farther from the original position. Therefore, reduction processing for shrinking the image data further inward as it moves away from the optical axis position O is required. That is, as shown in FIG. 3B, a thinning process is required at the time of pincushion distortion correction, and it is necessary to solve the above-mentioned folding back problem. This problem can be solved by the present invention.

FIG. 5 is a block diagram of another configuration of the image information processing system related to the present invention. In FIG. 5, components designated by the same reference numerals as those in FIG. 1 indicate components having similar configurations and functions. Digital filter (LPF) 3
The cutoff frequency is variable according to the reduction ratio supplied from the CPU 9. The output signal from the digital filter 3 is recorded in the memory 5. Recording / reading control to / from the memory 5 is performed by an address signal Add from the address control unit 10 based on information such as a reduction rate and an enlargement rate from the CPU 9. When reading from the memory 5, the sub-sampling function of FIG. 1 is executed by lengthening the read interval. Interpolator 6
Performs an interpolation process for generating data at a position other than the original pixel data that originally exists for the image data read from the memory 5, using the interpolation coefficient k from the address control unit 10. .

As will be described later, the digital emphasizing section 7 performs emphasizing processing for compensating for the deterioration of the frequency characteristic when enlarging the image data. Thus, the digitally emphasized image data is converted into an analog signal by the D / A converter 8 and output.

Enlarged image processing will be described with reference to FIG. 2 with respect to the normal frequency characteristic shown in FIG.
When the double enlargement process is performed, the frequency characteristic deteriorates as shown by the solid line in FIG. The enlargement process is shown in Figure (C).
Since the pixel data indicated by black circles is obtained by interpolation between pixel data of the white circle pixel data shown in (1) and is arranged at the original rate, the frequency characteristic deterioration occurs as described above. .

More generally, as shown in FIG. 7, pixel data indicated by x, which is a narrower interval than original pixel data arranged at horizontal intervals and vertical intervals indicated by white circles, is obtained by interpolation. Then, the sampling interval becomes smaller than 1.

FIG. 8 is another basic configuration block diagram of the image information processing system related to the present invention. In the figure, FIG.
And the same reference numeral have the same configuration function.

As described above, the conventional system can perform the enlargement process only uniformly on the entire screen, and cannot deal with the case where the enlargement ratio changes depending on the place of the screen such as distortion correction.
The present embodiment solves this problem.

The digital emphasizing section 7 is a digital filter for compensating for the deterioration of the frequency characteristic.
The digital filter forming the emphasizing unit 7 is, for example, as shown in FIG.
Filter processing is performed on 3 × 3 block data a to i for a pixel block as shown in FIG.
The output signal P is P = kxa + ky (b + c + d + e) + kz (f + g +
h + i). Here, kx + 4ky + 4kz = 1

In order to obtain a to i from the input image data, as shown in (B), 1H delay sections 711 and 712,
A circuit including the one-clock delay units 713 to 718 can be used. The data a to h thus obtained are filtered by the digital filter shown in (C). This digital filter includes adders 721 to 72.
3 and multipliers 724 to 726.
The filter characteristics are changed by setting the coefficients kx, ky, and kz supplied to 25 and 726 to appropriate values. Here, assuming that kx = 1 to 5, ky = (1−Kz) / 4 Kz = 0. Therefore, for kx = 1 to 5 as shown in FIG. , The figure (B)
As shown in, the characteristic has a peak at f CLK / 4. To change the peak frequency, the number of vertical and horizontal pixels in the block of FIG. 9A may be changed.

This example can also be applied to (sub) sampling that occurs during rotation processing. When performing 90-degree rotation processing as shown in FIG. 11A (same as FIG. 3A), the vertical (sub) sampling interval is 1 /
Since it is 2.4, which is smaller than 1, the band frequency characteristic is deteriorated. This sampling interval changes according to the rotation angle,
Blurring occurs and the image quality deteriorates. The deterioration of the frequency characteristic is emphasized and compensated by the digital emphasizing unit 7 in FIG.

The same thing occurs when correcting barrel distortion caused by the optical system. The solid line in FIG. 12A is the image data obtained via the optical system, and the image data after compensation is shown by the dotted line. As it goes from the optical axis point O to the periphery (as the distance from the center increases), it shifts inward from the original position and is coupled to the optical system. The distortion is corrected by changing the enlargement ratio according to the distance from the center. In this correction, the sampling interval becomes smaller than 1 as shown in FIG.

Next, as an embodiment of the present invention, in order to obtain a pseudo halftone image in a mode that is more suitable for human vision, for an image that has been subjected to the above-described processing or an image that has not been subjected to such processing. The image information processing system will be described below. This embodiment relates to halftone dot dither.

In the present embodiment, the rotation process is performed using the memory, the dither process is performed after the interpolation, and the process is returned by the reverse rotation process. The processing configuration block diagram for that is shown in FIG.
9 is shown. The input image data (6 bits) is written to the memory 11 under the control of the write control unit 12, and read out from the read control rotation control unit 13 which receives the rotation angle θ information, and then the memory 1 is read.
The rotation image data (6 bits) is obtained from 1. The image data read from the memory 11 is interpolated by the interpolation unit 14 and then input to the subtraction unit 15. The subtraction unit 15 calculates the dither pattern R from the interpolation data from the interpolation unit 14.
The data (6 bits) stored in the OM 17 is subtracted and compared, and 1-bit output data is written in the memory 16. Writing to the memory 16 is controlled by the write controller 18, and is controlled by the read control rotation controller 19 on the basis of the rotation angle information θ to read from the memory 16 to perform the reverse rotation process, for example, as a printer output. It

FIG. 20 shows 4 of image data using a memory.
In the figure for explaining the rotation processing of 5 degrees, the memory output is indicated by a black circle with respect to the memory input indicated by a white circle. Therefore, the image shown in FIG.
The image is inclined as shown in FIG. The image rotated in this way by 45 degrees is subjected to dither processing with a dither pattern in which θ is 0 degree (FIG. 21A), and then 4
By performing the rotation processing in the reverse direction of 5 degrees, the diagonal dither matrix pattern processing as shown in FIG.

[0034]

As described above, according to the present invention,
It is possible to obtain a pseudo halftone image in a mode that is more suitable for human vision without deterioration of image quality.

[Brief description of drawings]

FIG. 1 is a configuration block diagram showing an example of an image information processing system related to the present invention.

FIG. 2 is a block diagram of a digital filter in the system shown in FIG.

FIG. 3 is an explanatory diagram of image rotation processing in FIG. 1.

FIG. 4 is a diagram showing an example of application to correction of pincushion type distortion which is one of optical distortions.

FIG. 5 is a configuration block diagram showing another example of the image information processing system related to the present invention.

FIG. 6 is an explanatory diagram of enlarged image processing.

FIG. 7 is an explanatory diagram of enlarged image processing.

FIG. 8 is a block diagram showing a basic configuration of still another example of the image information processing system related to the present invention.

9 is a diagram showing the configuration of a digital filter that constitutes the emphasis unit 7 in FIG.

10 is a frequency characteristic diagram of the digital filter shown in FIG.

FIG. 11 is a diagram for explaining a 90-degree rotation process.

FIG. 12 is a diagram for explaining an operation when correcting barrel distortion caused by the optical system.

FIG. 13 is a diagram for explaining dither processing.

FIG. 14 is an oblique angle (screen angle) in dither processing
It is a figure for demonstrating the process with respect to a pattern of 45 degrees.

FIG. 15 is a diagram for explaining halftone dot dither processing.

FIG. 16 is an explanatory diagram of an example of dither processing.

FIG. 17 is a diagram for explaining a problem of conventional dither processing.

FIG. 18 is an explanatory diagram of the operation of the conventional dither processing.

FIG. 19 is a configuration block diagram of an embodiment of the image information processing system according to the present invention.

FIG. 20 is a diagram for explaining 45-degree rotation processing of image data using a memory according to an embodiment of the present invention.

FIG. 21 is a diagram in which the image rotated by 45 degrees in FIG. 20 is subjected to dither processing with a dither pattern in which θ is 0 degrees.

FIG. 22 is a diagram for describing a folding component during image processing.

FIG. 23 is an operation explanatory diagram of reduction processing.

FIG. 24 is a more detailed explanatory diagram of reduction processing.

FIG. 25 is a configuration diagram for solving a conventional folding back problem.

[Explanation of symbols]

 1 Low-pass filter (LPF) 2 A / D converter 3 Digital low-pass filter 4 Sub-sampling part 5,11,16 Memory 6,14 Interpolation part 7 Digital enhancement part 8 D / A converter 9 CPU 10 Address control part 12,18 Write control Section 13, 19 read control rotation control section 15 subtraction section 17 dither pattern ROM

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical display location G09G 5/36 520 K 9471-5G H04N 1/405 3/22 B 5/765 5/781 H04N 1 / 40 104 7734-5C 5/781 510 Z 5/91 L

Claims (1)

[Claims] 1. A first rotation effect means for performing a processing for obtaining a predetermined image rotation effect on the supplied image data, and a predetermined binary image processing for output data of the first rotation effect means. The binary image processing means for performing the above, and the first data concerning the output data of the binary image processing means.
Second rotation effect means for performing processing for obtaining an image rotation effect that cancels the image rotation effect in the rotation effect means of
An image information processing system comprising: