JPH05130455A - Picture quality improving device - Google Patents

Abstract

PURPOSE:To provide a picture quality improving device suitable for a video equipment, capable of adding an edge emphasis component to a slope of an input signal in a natural form without giving a sense of incongruity to a viewer and adjusting the degree of edge emphasis, and easy to employ a digital circuit. CONSTITUTION:A waveform synthesizer 1-3 implements amplitude synthesis, phase synthesis and phase nonlinear processing based on outputs Sb, Sc of an orthogonal high pass filter 1-1 and an in-phase high pass filter 1-2 to obtain a signal Se resulting from making a waveform changing part of the signal Sc steep. A subtractor 1-4 subtracts the signal Sc from the signal Se to obtain an edge emphasis component Sg. The edge emphasis component Sg is added to an edge of the input signal Sa by an adder 1-5 to obtain an output signal Sh subject to sure edge emphasis.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION This invention relates to a television (TV).
The present invention relates to an image quality improving device suitable for various video devices such as a receiver and a video tape recorder (VTR), and various image processing devices for handling image data. Then, the present invention improves the image quality by sharpening the changing portion of the waveform, that is, the waveform edge, but in a natural form without giving discomfort to the viewer.
An object of the present invention is to provide an image quality improving device capable of improving the sharpness and resolution of a reproduced image.

[0002]

2. Description of the Related Art Conventionally, in the contour correction used for improving the image quality, the contour correction component is obtained by quadratic differential processing,
This correction component is added to the original signal in an appropriate amount. In the contour correction by this method, the secondary differential waveform, which is the contour correction component, has a peak considerably outside the midpoint of the waveform change portion (edge portion) of the original signal. Therefore, if this second-order differential waveform is added to the original signal, preshoot or overshoot may occur, which does not have the expected image quality improvement effect, and black-and-white edging is possible at the edge portion of the reproduced image. This may result in unnatural contour correction.

[0003]

SUMMARY OF THE INVENTION The problem to be solved by the present invention is to perform pre-processing by edge enhancement by smoothly adding a waveform step at the midpoint position of the waveform change portion (edge portion) of the original signal. It is possible to prevent unnatural contour correction due to shoots and overshoots, improve the sharpness and resolution in a natural manner without giving a sense of discomfort to the viewer, and an image quality improving device suitable for IC integration. In order to do so, what kind of measures should be taken?

[0004]

In order to solve the above problems, the present invention provides a quadrature high-pass filter which is supplied with a first signal as an input signal and outputs a second signal, and An in-phase high-pass filter that is supplied with a first signal and outputs a third signal; and the second and third signals are supplied, and
A waveform synthesizer for outputting a fourth signal having a steepened waveform edge of the third signal by non-linear processing using an amplitude value and a phase value obtained by vector-synthesizing two signals; , Third and fourth signals are provided,
The third signal, which is a part of the first signal, is converted into the fourth signal.
And an adder / subtractor that obtains an output signal in which the waveform edge of the first signal is emphasized by replacing the signal of FIG.

[0005]

BEST MODE FOR CARRYING OUT THE INVENTION The present invention is intended to improve the image quality by sharpening the waveform changing portion (edge portion) of the original signal by performing nonlinear processing (nonlinear phase conversion processing) on the phase. FIG. 1 shows the first and second embodiments of the image quality improving apparatus of the present invention (both using the non-linear processing corresponding to claim 2). 2 and 5 to 10 are diagrams for explaining the operation thereof, and FIGS. 3 and 4 are diagrams showing a specific configuration example of the waveform synthesizer. In the operation explanatory diagram, a simplified simulated expression method is also used for convenience. Further, even when a digital circuit is taken as a specific circuit example, the signal waveform is shown as an analog waveform in order to make the explanation of the operation thereof easy to understand.

In the first embodiment shown in FIG. 1A, 1
-1 is a quadrature high-pass filter, 1-2 is an in-phase high-pass filter, 1-
3 is a waveform synthesizer, 1-4 is a subtracter, and 1-5 is an adder. These are the basic configurations of the present invention. For convenience of explanation, signal delay due to processing time of each circuit itself,
Also, a delay circuit or the like normally used only for simply correcting the delay will be omitted except where it is necessary for explanation. As an example of the input signal handled by this image quality improving apparatus, a television luminance signal, a color signal, an RGB signal, etc. are assumed. Therefore, for an input signal having a frequency component with an upper limit of 4 MHz, nonlinear processing is performed on the slope portion of the signal waveform by the following processing, and a component exceeding the frequency band of 4 MHz is added to sharpen the edge to enhance the edge. Is going.

First, in order to explain the basic operation of the present invention, the input signal Sa coming from the line L1 is shown in FIG.
A signal of one cycle of a cosine wave as shown in (a) is used. This is an example of a luminance signal band-limited by the upper limit frequency f2 (about 4 MHz), and has a waveform with an amplitude of 1. The horizontal axis is represented by the phase θ for convenience of explanation.

[0008]

[Equation 1]

It has the same meaning as. As the value of the frequency of the cosine wave at this time, consider about 2 MHz for the time being. The input signal Sa is first supplied to the quadrature high-pass filter 1-1.
The characteristics of this orthogonal high-pass filter are shown in FIGS. 2 (a) and 2 (c).
The characteristics are as shown in. In the frequency characteristic shown in FIG. 2A, the imaginary part is represented by the following equation,

[0010]

[Equation 2]

The real part becomes 0. The impulse response has a differential characteristic as shown in FIG. 2C, and has a phase difference of π / 2, which is orthogonal to the signal Sa. In order to realize such a quadrature high-pass filter 1-1, a transversal filter of an origin symmetric type, which uses an analog circuit or a digital circuit and has the impulse response of FIG. 2C as a coefficient value, is used. Orthogonal high-pass filter 1-1
Output waveform Sb becomes a sine wave as shown in FIG. On the other hand, the input signal Sa is also supplied to the in-phase high-pass filter 1-2. The characteristics of the in-phase high-pass filter are shown in FIGS. 2 (b) and 2 (d).
The characteristics are as shown in. In the frequency characteristic shown in FIG. 2B, the real part is represented by the following equation,

[0012]

[Equation 3]

The imaginary part is 0. The impulse response has a high-pass filtering characteristic as shown in FIG. 2D and is in phase with the signal Sa. To realize such an in-phase high-pass filter 1-2, an analog circuit or a digital circuit is used to form a transformer having a coefficient value symmetrical with respect to the axis of time = 0, as in the impulse response of FIG. A Versal filter or the like is used. The output waveform Sc of the in-phase high-pass filter 1-2 has the same cosine wave as the signal Sa as shown in FIG. 5C. The quadrature high-pass filter 1-1 and the in-phase high-pass filter 1-2 have the same amplitude characteristic G (f), and the characteristic is expressed by the following equation,

[0014]

[Equation 4]

The quadrature high-pass filter 1-1 and the in-phase high-pass filter 1
-2 has a characteristic that the phases are different from each other by π / 2 and that they are orthogonal to each other. Since the characteristic shown in FIG. 2A is the characteristic of a high-pass filter having a value in the imaginary part, the one having the characteristic is called an orthogonal high-pass filter. Also, FIG.
The characteristic shown in (b) is the characteristic of a high-pass filter having a value in the real part, so that the one having that characteristic is called an in-phase high-pass filter. The characteristics shown in FIGS. 2A and 2B are
The VSB of the television is set so that the value increases linearly in the range of f = 0 to f1 (1.25 MHz) and becomes a constant value in the range of f = f1 to f2 (4 MHz).
This is because the (residual sideband) transmission characteristics were taken into consideration, but as a result, this setting was suitable for the present invention. Figure 1
Returning to (a), the output signals Sb and Sc of the quadrature high-pass filter 1-1 and the in-phase high-pass filter 1-2 are supplied to the waveform synthesizer 1-3 at the next stage. Here, first, as shown in FIG. 2F, the two signals Sb and Sc are arranged on the vertical axis because Sb is a quadrature component and on the horizontal axis because Sc is an in-phase component. Then, by vector composition of the quadrature component and the in-phase component shown in the following equation,

[0016]

[Equation 5]

As the combined amplitude, the square root Sd of the sum of squares of the two signals Sb and Sc is obtained. The signal S is shown in FIG.
It is a wave form diagram of d. The signal Sd is 1 in this example,
Naturally, it will not be negative. From FIG. 2 (f), the combined phase θi is

[0018]

[Equation 6]

It is calculated by The value of the phase θi is usually −π
It is calculated in the range of / 2 to π / 2. Further, the polarities of θ on the horizontal axis of FIG. 5 and the combined phase θi have an inverse relationship. That is, θ
i = −θ. Next, when the absolute value of this phase θi is obtained, it becomes a value in the range of 0 to π / 2, as shown in FIG. 5 (e). Furthermore, the absolute phase value is converted into the maximum value π
When the power of α (here, α = 4) is used with reference to / 2,
As in (f), the waveform is such that peaks are concentrated at positions of ± π / 2 on the horizontal axis. The position of ± π / 2 on the horizontal axis is shown in FIG.
It is almost at the midpoint position of the waveform changing portion of the input signal Sa shown in (a). The operation of the waveform synthesizer 1-3 that obtains the output Se using the phase value θi, the combined amplitude Sd, and the signal Sc that perform such non-linear processing is represented by the following expression.

[0020]

[Equation 7]

A waveform diagram of the output Se is shown in FIG. As shown in the figure, the signal Se has a waveform in which the rising and falling portions of the signal Sc are steepened. As shown in FIG. 5 (f), the new phase value θo obtained by the non-linear processing has a region close to 0 near the original phase value θi of 0.
In the vicinity of ± π / 2 of the original phase value θi, the new phase value θo
Rapidly concentrates on π / 2. Therefore, as shown in FIG. 5 (g), the signal Se based on this new phase value θo
Indicates that the phase value θo changes abruptly as the signal value around θ on the horizontal axis approaches 0 (value near 1 in this case), and θ on the horizontal axis approaches ± π / 2. , Its signal value also changes rapidly. Therefore, the signal Se has a waveform in which the rising and falling portions of the signal Sc are steepened (the phase value θi at the position of the substantially midpoint of the rising and falling portions is ± π / 2). At this time, since the zero cross point of the signal Se is not displaced from the zero cross point of the signal Sc, the phase information of the original signal Sc (that is, the signal S
The phase information of a) is preserved even after the non-linear processing. Zero cross points are indicated by circles in FIG. A specific configuration method of the amplitude synthesizer 1-3 will be described using the examples of FIGS. 3 and 4. First, FIG. 3 shows some configuration examples of basic element blocks of the waveform synthesizer. FIG. 3A shows the combined amplitude Sd and the combined phase θi.
This is an example of a configuration for obtaining the above, and is functionally composed of four blocks (3-1 to 3-4). In block 3-1, the phase angle θi of the equation 6 is obtained from the input signals Sb and Sc, and in block 3-2, the contribution of the orthogonal component as shown in the following equation from the input signal Sb and the phase angle θi,

[0022]

[Equation 8]

## EQU3 ## In block 3-3, the input signal Sc
And the phase angle θi, the contribution of the in-phase component as shown in the following equation,

[0024]

[Equation 9]

Then, the adder of the block 3-4 adds and synthesizes the signals from the blocks 3-2 and 3-3 to obtain the output Sd as shown in the equation (5). When these are configured by digital circuits, the blocks 3-1 to 3-3 preliminarily calculate output values for all expected input data and write them in a ROM or the like, and refer to this to output. It is realized by a table lookup method or the like.

FIG. 3B shows another example for obtaining the combined amplitude Sd, which is functionally composed of four blocks. Blocks 3-5 and 3-6 are multipliers,
The squared values Sb ² and Sc ² of the input signals Sb and Sc are obtained, respectively. The squared values are combined by the adder in block 3-7, the square root is obtained in block 3-8, and S is output.
d is obtained. Such a structure can be realized by either a digital circuit or an analog circuit, but in the case of a digital circuit, the block 3-8 also uses a ROM or the like. FIG. 3C shows the phases θi and Sc from the signals Sd and Sc.
2 is a circuit configuration for obtaining a signal indicating the polarity of. Block 3
At -9, from the input signals Sd and Sc,

[0027]

[Equation 10]

The phase θi is obtained. Since the value in the parentheses in the formula is a positive value, the obtained value of θi is a positive value of π / 2 or less. A block 3-10 outputs a signal sgn (Sc) representing a polarity such as -1, 0, 1 corresponding to the negative, zero, and positive values of the signal Sc. 3 in FIG.
Two examples will be shown in which the waveform synthesizer is configured by combining the above components. FIG. 4A is composed of five blocks. The signals Sb and Sc are input to the block 4-1, and the inside of the block 4-1 has a configuration as shown in FIG.
The signal Sd is obtained. In block 4-2, the signals Sd and S
c is input, and the inside of the block 4-2 has a structure as shown in FIG. 3C, and the combined phase .theta.i is obtained. Block 4
Sd and .theta.i are input to -3, and a nonlinear process relating to the phase is performed as in the following equation.

[0029]

[Equation 11]

Block 4-4 is block 3-10 of FIG.
Similarly, the signal indicating the polarity of the signal Sc is generated.
Block 4-5 is a multiplier and blocks 4-3 and 4
The signal Se from -4 is taken to obtain the signal Se. The processing of FIG. 4 (a) can be summarized as an expression as follows.

[0031]

[Equation 12]

Next, the waveform synthesizer shown in FIG.
It consists of two blocks (4-6 to 4-12). Each of the first blocks 4-6 to 4-9 in FIG.
Is the same as the blocks 3-1 to 3-4 of FIG.
i is determined and added to blocks 4-10. In block 4-10, as in block 4-3, non-linear processing such as the following equation is performed.

[0033]

[Equation 13]

The value of the phase .theta.i from block 4-6 is -.pi.
Since it can take a value in the range of / 2 to π / 2, the absolute value conversion process is required. Block 4-11 is block 3-
It outputs the polarity signal of the signal Sc.
The last block 4-12 is the same multiplier as the block 4-5, and takes the product of the signals added from the block 4-10 and the block 4-11 to obtain the output Se. Figure 4
The processing of the waveform synthesizer shown in (b) can be summarized into the following equation.

[0035]

[Equation 14]

As described above, in the waveform synthesizer of the block 1-3, the signal Se in which the waveform changing portion of the signal Sc is made steeper is obtained by the signals Sb and Sc, but the nonlinear processing (shown in FIG. 4 is performed. Processing in blocks 4-3 and 4-10)
, The real number parameter α is used. This non-linear processing is performed on the phase, and is represented by the following equation with the input phase θi and the output phase θo.

[0037]

[Equation 15]

In this equation, the value of the parameter α is 1, 2,
FIG. 2 (g) shows the phase change (obtained in the range where .theta.i is positive) when the values are 4 and 8. When α = 1, θo = θ
i, and the input phase value is output as is. α
As becomes larger, the output becomes difficult to output (the output phase θo does not easily increase from 0 even if the input phase becomes larger), and the value of the output phase θo concentrates on π / 2. This means that the processing is such that the 0 phase region is expanded as α increases. That is, as α becomes larger, the new phase value θo due to the non-linear processing has a range of values close to 0 near 0 of the original phase value θi, and a new phase value θo near ± π / 2 of the original phase value θi. A large phase value θo means that it is rapidly concentrated in π / 2. Here, returning to FIG. 1A, the next block 1-4 to which the signal Se from the waveform synthesizer 1-3 and the signal Sc from the in-phase high-pass filter 1-2 are supplied is the signal Se from the signal Se. It is a subtractor that functions to subtract the signal Sc.
In block 1-4, a signal Sg represented by the following equation is obtained.

[0039]

[Equation 16]

This signal Sg is an edge enhancement component, and its waveform is shown in FIG. 5 (h). As can be seen from the above Expression 16, this signal Sg is obtained from the difference between the signals Se and Sc, and in order to perform the process of replacing the signal Sc with the edge enhancement, instead of the signal Sc, the signal Sg between these signals is changed. You are seeking the difference between. The last block 1-5 in FIG. 1A is an adder, which calculates an output signal Sh by adding and synthesizing the edge emphasis signal Sg to the input signal Sa as in the following equation.

[0041]

[Equation 17]

FIG. 5 (i) is a waveform diagram of the signal Sh output from the output line L2. Compared to the input signal Sa to the present apparatus shown in FIG. 5 (a), the signal Sh has a steep change portion of the waveform and is edge-emphasized. In addition, the degree of emphasis is natural and smooth without any unnatural preshoot or overshoot being added. The sharpening of this smooth edge (that is, the waveform sloped portion) is a characteristic of this device. Is. In FIG. 5, in order to make it easier to understand how the phase change of the signal is manipulated, the horizontal axis represents the phase value θ.
The operation of each block will now be seen using the operation waveform diagram with the horizontal axis as the time axis. FIG. 7 shows an example in which the input signal Sa is a sine wave of 2 MHz, and can be regarded as an example in which FIG. 5 is rewritten on the time axis. However, the waveform synthesizer 1-3
The parameter α has a value of 4. 7A to 7G show the following signals, respectively. (A) Input signal Sa (b) Output signal Sb of quadrature high-pass filter 1-1 (c) Output signal Sc of in-phase high-pass filter 1-2 (d) Obtained in waveform synthesizer 1-3 Synthetic amplitude Sd (e) Output signal Se of waveform synthesizer 1-3 (f) Edge-enhanced signal Sg (h) obtained as output of subtractor 1-4 Edge-enhanced signal Sh that is the final output signal

FIG. 8 shows an example in which the input signal Sa has a step waveform. This is an example of a luminance signal band-limited by the upper limit frequency f2 (4 MHz), and the waveform part whose amplitude changes from 1 to 0 is processed by a low-pass filter having a 100% roll-off characteristic at a frequency of 0 to 4 MHz. It is a waveform obtained by doing. Other conditions are the same as in the case of FIG. Each of the waveform diagrams in FIGS. 8A to 8G has a one-to-one correspondence with FIGS. 7A to 7G. As is clear from FIGS. 7 and 8, the final output signal Sh has the waveform edge sharpened significantly as compared with the input signal Sa.

More specifically, the output waveform Sh is compared with the input waveform S
Compared with a, the output waveform Sh has a smooth waveform step at a substantially midpoint of the waveform changing portion (edge portion) of the input waveform Sa, the slope of the waveform changing portion is steepened, and the edge is accurately emphasized. You can see that it has a different waveform. When the output signal Sh is reproduced, a contour-corrected image can be obtained.
In addition, the edge enhancement processing of this image quality improving apparatus does not become the edge enhancement processing that exceeds the amplitude of the original signal such as preshoot and overshoot as in the conventional contour correction, but the edge enhancement processing within the amplitude of the original signal. is there. Therefore, even if a device incorporating this image quality improving device is configured with a digital circuit, the overflow problem does not occur, and the device can perform good image quality improvement.

In this way, the signal Sh output from the line L2 is subjected to edge enhancement, and as a result, a new sideband component is formed, and a spectrum beyond the band originally possessed by the input signal Sa is newly added. Become a signal. The addition of this new spectrum equivalently gives the viewer the impression that the resolution of the original signal has improved, and serves to improve the sharpness of the image. The steepness of the edge portion of the signal Sh (the degree of edge enhancement) is determined by the parameter α of the waveform synthesizer 1-3.
Is a fixed value, it depends on the frequency characteristic of the edge portion in the original signal before edge enhancement. If the rising portion and the falling portion (edge portion) of the original signal have a steeper slope, a strong degree of edge enhancement is performed. On the other hand, for a gentle slope, a weak degree of edge enhancement is performed. Further, as shown in FIG. 5, the zero cross point of the signal Sh (the zero cross point is indicated by a circle in FIG. 5) is the signal S.
Since it does not deviate from the zero crossing point of a, the phase information of the original signal Sa is preserved even after the non-linear processing.

As described above, the edge enhancement component Sg added to the input signal depends on the frequency characteristic of the input signal and has a complete correlation including the input signal and phase information. The sharpness and the resolution can be improved in a natural manner without giving a feeling of strangeness to the viewer.

Next, two examples will clarify what the waveform edge looks like when the value of the parameter α used in the nonlinear processing in the waveform synthesizer 1-3 is changed. FIG. 9 is a waveform diagram of the output signal Sh when the step waveform used in FIG. 8 is the input signal Sa. In FIGS. 9A to 9D, the output signal Sh is obtained by setting the value of α as follows. (A) Sh when α = 1 (b) Sh when α = 2 (c) Sh when α = 4 (d) Sh when α = 8 FIG. 10 shows the same frequency 2 MHz as FIG. 7. FIG. 6 is a waveform diagram of an output signal Sh when a sine wave is used as an input signal Sa. Figure 10
The values of α in (a) to (d) are shown in FIGS. 9 (a) to (d), respectively.
It is set according to.

As is clear from FIGS. 9 and 10, α = 1
In the case of, the edge enhancement processing is not performed, but as α becomes larger than 1, the degree of edge enhancement progresses, and it can be seen that the edge enhancement effect is saturated at a certain value or more. From experience, it seems that α = about 2 to 4 is suitable, but the value of α can be adjusted according to taste to optimize the degree of edge enhancement.

The edge enhancement degree is adjusted by changing the signal levels of the outputs Sb and Sc of the two high-pass filters 1-1 and 1-2 at the same time and equivalently changing the magnitude of the signal Sg. Is also possible. Further, as shown in FIG. 6 and the following equation,

[0050]

[Equation 18]

It is also possible to optimize the setting by changing the frequency characteristic of the edge emphasis component by switching and using several kinds of high-pass filters having different characteristics. The high-pass filters (1) to (4) described in Eq. 18 represent four combinations of quadrature and in-phase filters. In any case, it is important to select the optimum frequency characteristic according to the purpose. As described above, by changing the value of the parameter α and the characteristics of the high-pass filter, the edge emphasis is independent of the degree of edge emphasis depending on the frequency characteristics of the edge part in the original signal before the edge emphasis described above. You can adjust the degree of emphasis.

FIG. 1B shows the second embodiment. This embodiment can prevent a phase shift even for an input signal sampled at a low sampling frequency, and is suitable for a digital circuit configuration. Blocks 1-1 to 1-5 in the same figure are shown in FIG.
It has exactly the same function as the one with the same block number shown in. Newly added in the second embodiment are a phase adjuster 1-6, an amplifier 1-7, a high-pass filter 1-8, and an adder 1-9. For this additional block,
The patent application by the present inventor (reference number H03000791, image quality improving device on filing date September 30, 1991) has been described in FIG. 22, so a detailed description will be omitted, but a brief description will be given. It looks like this: The amplifier 1-7 amplifies the value of the waveform edge enhancement signal Sg with the amplification factor β (≧ 1) to obtain the output Si, makes the value of the amplification factor β variable, and sets the optimum edge enhancement effect. Of the

As shown in FIG. 2E, the high-pass filter 1-8 outputs the high frequency component of the input signal Sa and the frequency component (> f2) outside the frequency band included in the input signal from the signal Sh. It is extracted and multiplied by γ (≧ 0) to obtain the edge emphasis component again. Then, the edge emphasis component is added to the adder 1-
This is for adding and synthesizing the signal Sh in 9 to give the output signal Sk obtained from the line L3 a further edge enhancing effect.

The functions of blocks 1-7 to 1-9 are necessary when the nonlinear processing of the waveform synthesizer cannot be performed very strongly, for example, the signal to be handled is sampled and discretized. This is the case. Especially, it is effective when you want to suppress the phase shift due to nonlinear processing within a certain range, which tends to occur when the sampling frequency cannot be set to a too high value, and can be used to assist the function of the waveform synthesizer. ..

First, the phase adjuster 1-6, in response to the phenomenon of phase shift due to non-linear processing that occurs when the signal is sampled and discretized as described above, the in-phase component Sc.
Find the zero crossing point of. Next, using the plurality of sampled values picked up in the order closer to the zero crossing point and the plurality of sampled values of the output Se of the waveform synthesizer 1-3 corresponding to these in time, the former plurality of sampled values are used. By reflecting the amplitude ratio of the sampled values in the latter plurality of sampled values and recombining, the information of the zero crossing points can be preserved and the phase shift can be reduced. By performing such processing, it becomes possible to set the value of the sampling frequency to a value lower than usual. Thus, when the signal is sampled and the circuit is digitally constructed, the blocks 1-6 to 1-9 function effectively. Of course, this embodiment also has the same effect as the first embodiment.

Next, the third embodiment will be described. The overall block structure is the same as that of the first embodiment shown in FIG. The third embodiment corresponds to claim 3, and the difference from the first and second embodiments is the method of nonlinear processing in the waveform synthesizer 1-3. In the first and second embodiments, as shown in FIG. 2 (g), the phase nonlinear processing is continuously performed from the input phase θi = 0. On the other hand, in the third embodiment, as shown in FIG. 11 (f), the output phase θo is 0 until the input phase θi reaches the predetermined value, and when the input phase θi becomes the predetermined value or more, the input phase θi This is a non-linear process that sets the output phase θo according to the value of.

The operation of the third embodiment will be described with reference to FIGS. 11 to 17, focusing on the non-linear processing. As an example of the input signal Sa coming from the line L1, one cycle of a cosine wave as shown in FIG. 13A is used. This is an example of a luminance signal band-limited by the upper limit frequency f2 (about 4 MHz), and has a waveform with an amplitude of 1. The horizontal axis is represented by the phase θ shown in Expression 1 for convenience of description.

Quadrature high-pass filter 1-1, in-phase high-pass filter 1
The characteristics shown in FIGS. 6A and 6B are used as the characteristics of −2. That is, the characteristics of the orthogonal high-pass filter 1-1 are as shown in FIG.
The characteristics are as shown in (a) and (c), and the characteristics of the in-phase high-pass filter 1-2 are as shown in FIGS. 11 (b) and (d). Output waveform Sb of quadrature high-pass filter 1-1
Is a sine wave having an amplitude of 0.5 as shown in FIG. 13B because of the balance with the gain of the filter. In-phase high-pass filter 1
The output waveform Sc of −2 is the signal Sa as shown in FIG.
It becomes a cosine wave with the same shape as and an amplitude of 0.5. Waveform synthesizer 1-
3 obtains the combined amplitude Sd from the signals Sb and Sc as in the first embodiment. The composite phase θi obtained by the waveform synthesizer 1-3 is given by the following equation from FIG. 11 (e).

[0059]

[Formula 19]

The value of the phase θi is usually calculated in the range of -π / 2 to π / 2 and becomes as shown in FIG. 13 (e). Horizontal axis θ
, And the polarity of θi is in the opposite relationship. Further, the absolute value of the combined phase value θi is 0 to απ / 2 (where 0 ≦
When α <1) is set to 0 and απ / 2 to π / 2 has a linear characteristic according to the input phase value,
As in (f), the phase waveform is such that peaks are concentrated on the horizontal axis π / 2. The operation of the waveform synthesizer 1-3 accompanied by such phase non-linear processing is performed by combining the synthesized phase value θi and the synthesized amplitude Sd.
When expressed by an equation using
Is required.

[0061]

[Equation 20]

A waveform diagram of the output Se is shown in FIG.
As shown in the figure, the signal Se has a waveform in which the rising and falling portions of the signal Sc are steepened.

As shown in FIG. 13 (f), the new phase value θo obtained by the non-linear processing extends from the original phase value θi of 0 to a predetermined value in a region where the phase value θo = 0 and the original phase value θo increases. In the vicinity of ± π / 2 of θi, the new phase value θo suddenly concentrates at π / 2. Therefore, as shown in FIG. 13G, in the signal Se based on the new phase value θo, the region of the signal value where the horizontal axis θ of the signal Sc is 0 (0.5 in this case) is widened. Then, as the horizontal axis θ approaches ± π / 2,
Since the phase value θo changes abruptly, the signal value of the signal Se also changes abruptly. Therefore, the signal Se has a waveform in which the rising and falling portions of the signal Sc are steepened (the phase value θi at the position of the substantially midpoint of the rising and falling portions is ± π / 2). At this time, since the zero cross point of the signal Se is not displaced from the zero cross point of the signal Sc, the phase information of the original signal Sc (that is, the signal S
The phase information of a) is preserved even after the non-linear processing. The zero cross points are indicated by circles in FIG.

A concrete construction method of this waveform synthesizer is shown in FIG.
This will be described using the example of No. 2. The symbols in FIG. 12 and FIG.
4 and the same reference numerals do not mean the same block. FIG. 12 shows a first configuration example of the waveform synthesizer. FIG. 12 is functionally divided into five blocks (3-1
~ 3-5). In block 3-1, the phase angle θi of equation 19 is obtained from the input signals Sb and Sc, and in block 3-2, the contribution of the orthogonal component as shown in the following equation from the input signal Sb and the phase angle θi,

[0065]

[Equation 21]

Is seeking. In block 3-3, the contribution of the in-phase component as shown in the following equation from the input signal Sc and the phase angle θi

[0067]

[Equation 22]

Seeking. The adder of the block 3-4 adds and synthesizes the signals from the blocks 3-2 and 3-3 to obtain the output Sd as shown in Expression 5. In the next block 3-5, the signal Sc, the phase value θi from the block 3-1, and
The combined amplitude value Sd from the block 3-4 is supplied, and the output Se is obtained by performing the calculation as shown in Expression 20. When these are configured by digital circuits, blocks 3-1 to 3-1
3-3 and 3-5 can be realized by a table lookup method or the like in which output values for all expected input data are calculated in advance and written in a ROM or the like and the output is obtained by referring to this. .. The block 3-4 can be made by a well-known general-purpose circuit.

As described above, the waveform synthesizer 1-3 performs the processing shown in Expression 20 on the basis of the signals Sb and Sc to obtain the signal Se in which the waveform changing portion of the signal Sc is made steeper. Non-linear processing performed on the phase (block 3-
In 5), the real number parameter α is used. When the input phase is θi and the output phase is θo, the nonlinear processing of the phase is expressed by the following equation.

[0070]

[Equation 23]

In this equation, the value of the parameter α is 0, 0.2
FIG. 11 (f) shows how the phase changes nonlinearly at 5, 0.5, and 0.75. Thus, when α = 0, θ
The phase value input at o = θi is output as is,
As α increases, the output becomes harder to output, and the output concentrates on π / 2. This means that the processing is such that the 0 phase region is expanded as α increases. That is, as α increases, the new phase value θo due to the non-linear processing expands the region where the phase value becomes 0, and near ± π / 2 of the original phase value θi, the new phase value θo rapidly increases by ± π. It means to concentrate on / 2.

Returning now to FIG. 1, the waveform synthesizer 1-3
Signal Se from the same and signal Sc from the in-phase high-pass filter 1-2
The subtractor 1-4 to which is supplied obtains the edge enhancement component Sg in the same manner as in the first embodiment. FIG. 13H is a waveform diagram of Sg which is the edge emphasis component. As can be seen from Equation 16, this signal Sg is obtained from the difference between Se and Sc, and the edge-enhanced signal S is used instead of the signal Sc.
This means that the difference between these signals is obtained in order to perform the process of exchanging e. The next block 1-5 is an adder which, like the first embodiment, adds and synthesizes the edge enhancement signal Sg with the input signal Sa to obtain the output signal Sh. FIG. 13I is a waveform diagram of the signal Sh obtained from the output line L2. The signal Sh is the input signal S of FIG.
The change portion of the waveform is steeper than that of a, and the effect of edge enhancement can be seen. Further, the degree of emphasis is not particularly accompanied by an unnatural preshoot or overshoot, and is characterized in that the edge, that is, the waveform slanted portion, has a smooth and natural shape. Further, as shown in FIG. 13, the zero crossing point of the signal Sh (the zero crossing point is indicated by a circle in FIG. 13) is not displaced from the zero crossing point of the signal Sa, so that the phase of the original signal Sa is changed. Information is preserved after non-linear processing.

In FIG. 13, the horizontal axis is represented by the phase value θ in order to make it easier to understand how the phase change of the signal is manipulated. Next, the operation will be described with a waveform diagram with the horizontal axis as the time axis. To do. FIG. 14 is an example of the output signal Sh when the input signal Sa has a step waveform. The signal Sa has an upper limit frequency f2 (4
MHz) is an example of a luminance signal band-limited, and a waveform part in which the amplitude of the signal changes from 1 to 0
A signal waveform Sh obtained by performing edge enhancement processing by processing with a low-pass filter having a 100% roll-off characteristic at MHz is shown. The difference between FIGS. 14A to 14D is the value of the parameter α of the waveform synthesizer 1-3, and the signal Sh is obtained under the following settings. (A) α = 0 (b) α = 0.25 (c) α = 0.5 (d) α = 0.75 In FIG. 14 (a), the nonlinear processing function of the waveform synthesizer 1-3 is stopped. This is the case, and it can be seen that the falling waveform portion becomes steeper as the value of α is increased in comparison with this.

FIG. 15 shows an example of processing when the input signal Sa has a pulse waveform. The signal Sa has an upper limit frequency f2 (4 MH
It is an example of a 2T pulse whose band is limited in z). The difference between the four waveforms is the value of the parameter α as in FIG.
5 (a) to (d) correspond one-to-one with FIGS. 14 (a) to (d). FIG. 16 shows an example of processing when the input signal Sa is a sine wave of 2 MHz. 16A to 16D are also shown in FIG.
There is a one-to-one correspondence with (a) to (d), and the parameter α
It is an output example of Sh when the value of is changed in the range of 0 to 0.75. Also in FIGS. 15 and 16, similar to FIG.
The falling waveform portion becomes steeper as the value of α is increased. The inventor considers that α = 0.5 is a suitable value from experience, but adjusts the value of α according to his / her own preference, and determines the degree of edge enhancement to be the optimum value for each user. can do.

In FIG. 17, the value of the parameter α is fixed to 0.5 and the value of the frequency f of the input signal Sa which is a sine wave is set to 1M.
Hz ~ 3.58MHz (3.58MHz = color sub-carrier frequency fsc)
7 shows the waveform of the output signal Sh when the value is changed in the range of. (A) f = 1 MHz, α = 0.5 (b) f = 2 MHz, α = 0.5 (c) f = 3 MHz, α = 0.5 (d) f = 3.58 MHz, α = 0.5 Thus, as the value of the frequency f increases, the degree of sharpening of the edge portion advances. The frequency characteristics of the in-phase and quadrature high-pass filters 1-2 and 1-1 are shown in FIG.
It is based on the fact that it is linearly increased as in (a) and (b).

The waveform synthesizer 1-3 described above is shown in FIG.
(F), or nonlinear processing for the phase as shown in Expression 23 is performed to widen the zero phase region and perform conversion so as to concentrate the phase change on the axis of ± π / 2, thus edge enhancement is attempted. , FIG. 11 shows another example that brings about a similar effect.
Shown in (g) and (h). In FIG. 11 (g), the phase θo
Is zero when the absolute value of the phase .theta.i is .alpha..pi. / 2 (where 0.ltoreq..alpha. <1) and the characteristic of the sine square curve is up to .pi. / 2 for .theta.i beyond that.

Further, FIG. 11 (h) has the same gist, but shows a phase conversion characteristic created by connecting several set points by linear approximation. The object of the present invention can be achieved by using a circuit that realizes these characteristics as a non-linear phase converter of the above-mentioned waveform synthesizer. The same figure (g),
The curve shown in (h) has the same value of α, and the curve shown in FIG.
Compared to the curve of (f), the region close to the zero phase extends (that is, the phase θo does not increase easily even if the phase θi is increased).
As the concentration on / 2 progresses, the degree of edge enhancement appears strongly in the non-linear processing according to FIGS.

A problem in performing the non-linear processing in the digital circuit is that the zero crossing of the waveform, that is, the phase distortion tends to occur, depending on the sampling frequency. In this embodiment, since the combined amplitude value and the combined phase value of the signals from the quadrature and in-phase high-pass filters are used together, the phase distortion is small. However, if the value of the parameter α is increased, the phase distortion phenomenon cannot be avoided. In order to avoid this kind of phase distortion, the phase adjuster may be used as described in the second embodiment. The third embodiment naturally has the same effect as that of the first embodiment.

As described above, the edge enhancement processing according to the present invention has a perfect correlation with the input signal including the phase information, and the edge enhancement component is smoothly added in a natural form. The device can give a sense of improvement in sharpness and resolution in a natural manner without giving a feeling of strangeness to a viewer.

Further, each of the illustrated embodiments can be easily constructed with a well-known circuit using a commercially available IC or the like, so that the entire apparatus can be realized at low cost. Further, in the above embodiment, the luminance signal of the television is used as the input signal example,
The image quality improving device of the present invention can be applied to color signals, RGB signals, and the like. Further, even for digitized image data, contour correction processing and edge enhancement processing equivalent to those of the present invention can be realized by software processing using a computer, and the present invention is applicable to processing of image data by software. It can be applied. Furthermore, the present invention is also effective for improving waveform deterioration of a general digital transmission communication system.

[0081]

As described above, the image quality improving device of the present invention has the following effects. (A) Edge enhancement is performed by accurately adding a waveform step to the midpoint of the inclined portion of the input signal, and as a result, an output signal to which a frequency component outside the frequency band of the input signal is added is obtained, The outline is emphasized. (B) It is an edge enhancement process within the amplitude of the original signal rather than the edge enhancement over the amplitude of the original signal such as preshoot and overshoot as in the conventional contour correction. Therefore,
Even if a device incorporating this image quality improving device is configured by a digital circuit, the overflow problem does not occur, and the device can perform good image quality improvement. (C) The degree of edge enhancement with respect to the input signal depends on the content frequency of the input signal. Further, despite the non-linear processing, the degree of edge enhancement has a complete correlation including the input signal and phase information. .. Further, the edge emphasis component smoothly makes the waveform change portion steep. Therefore, the image quality improving device can give the viewer a feeling of improvement in sharpness and resolution in a natural manner without giving a sense of discomfort to the viewer. (D) Despite the non-linear processing, the original signal phase information is preserved, so that the image quality improving device is suitable for the digital circuit configuration. is there. (E) By adjusting the non-linear processing parameter of the waveform synthesizer, the degree of edge enhancement can be adjusted separately from the degree of edge enhancement in (c) above. Therefore, the viewer can obtain various image quality by adjusting the degree of edge enhancement according to his / her preference, and can set the optimum image quality among them. (F) Each component of the image quality improving device of the present invention can be realized with a simple circuit configuration by using a general-purpose component such as a commercially available IC. Therefore, this image quality improving device can be easily manufactured at low cost and has a wide range of applications, so it is industrially effective,
Be beneficial.

[Brief description of drawings]

FIG. 1 is a block configuration diagram of an embodiment.

FIG. 2 is an operation explanatory diagram of the first embodiment.

FIG. 3 is a diagram showing an example of specific constituent elements of the waveform synthesizer of the first embodiment.

FIG. 4 is a diagram showing a specific configuration example of the waveform synthesizer of the first embodiment.

FIG. 5 is an operation explanatory diagram of the first embodiment.

FIG. 6 is an operation explanatory diagram of the first embodiment.

FIG. 7 is an operation explanatory diagram of the first embodiment.

FIG. 8 is an operation explanatory diagram of the first embodiment.

FIG. 9 is an operation explanatory diagram of the first embodiment.

FIG. 10 is an operation explanatory diagram of the first embodiment.

FIG. 11 is an operation explanatory diagram of the third embodiment.

FIG. 12 is a diagram showing a specific configuration example of the waveform synthesizer of the third embodiment.

FIG. 13 is an operation explanatory diagram of the third embodiment.

FIG. 14 is an operation explanatory diagram of the third embodiment.

FIG. 15 is an operation explanatory diagram of the third embodiment.

FIG. 16 is an operation explanatory diagram of the third embodiment.

FIG. 17 is an operation explanatory diagram of the third embodiment.

[Explanation of symbols]

 1-1 Quadrature high-pass filter 1-2 In-phase high-pass filter 1-3 Waveform synthesizer 1-4 Subtractor 1-5 Adder

[Procedure amendment]

[Submission date] February 17, 1992

[Procedure Amendment 1]

[Document name to be amended] Statement

[Correction target item name] 0034

[Correction method] Change

[Correction content]

The value of the phase θi from block 4-6 is -π.
Since it can take a value in the range of / 2 to π / 2, the absolute value conversion process is required. Block 4-11 is block 3-
It is the same as 10 and outputs the polarity signal of the signal Sc. The last block 4-12 is the same multiplier as the block 4-5, and takes the product of the signals added from the block 4-10 and the block 4-11 to obtain the output Se. The processing of the waveform synthesizer shown in FIG. 4 (b) can be summarized as an equation as follows.

[Procedure Amendment 2]

[Document name to be amended] Statement

[Correction target item name] 0061

[Correction method] Change

[Correction content]

[0061]

[Equation 20]

[Procedure 3]

[Document name to be amended] Statement

[Correction target item name] 0063

[Correction method] Change

[Correction content]

As shown in FIG. 13 (f), the new phase value θ ₀ by the non-linear processing extends from the original phase value θ i of 0 to a predetermined value, and the region of the phase value θ ₀ = 0 expands. In the vicinity of ± π / 2 of the phase value θi, the new phase value θ ₀ suddenly concentrates on ± π / 2. Therefore, as shown in FIG. 13G, in the signal Se based on the new phase value θ ₀ , the region of the signal value where the horizontal axis θ of the signal Sc is 0 (0.5 in this case) is widened. Then, as the horizontal axis θ approaches ± π / 2,
Since the phase value θ ₀ changes abruptly, the signal value of the signal Se also changes abruptly. Therefore, the signal Se has a waveform in which the rising edge and the falling edge of the signal Sc are steepened (the phase value θi at the substantially midpoint of the rising edge and the falling edge is ± π / 2). At this time, since the zero cross point of the signal Se is not displaced from the zero cross point of the signal Sc, the phase information of the original signal Sc (that is, the signal S
The phase information of a) is preserved even after the non-linear processing. The zero cross points are indicated by circles in FIG.

[Procedure amendment 4]

[Document name to be amended] Statement

[Correction target item name] 0064

[Correction method] Change

[Correction content]

A concrete construction method of this waveform synthesizer is shown in FIG.
This will be described using the example of No. 2. The symbols in FIG. 12 and FIG.
4 and the same reference numerals do not mean the same block. FIG. 12 shows a configuration example of the waveform synthesizer. FIG. 12 is functionally divided into five blocks (3-1 to 3-
It consists of 5). In block 3-1, the input signal Sb
And Sc, the phase angle θi of the equation 19 is obtained, and the block 3
At -2, the contribution of the orthogonal component as shown in the following equation from the input signal Sb and the phase angle θi is ──────────────────────────── ───────────────────────────

[Procedure amendment]

[Submission date] October 29, 1992

[Procedure Amendment 1]

[Document name to be amended] Statement

[Correction target item name] 0010

[Correction method] Change

[Correction content]

[0010]

[Equation 2]

[Procedure Amendment 2]

[Document name to be amended] Statement

[Correction target item name] 0050

[Correction method] Change

[Correction content]

[0050]

[Equation 18]

[Procedure 3]

[Document name to be corrected] Drawing

[Name of item to be corrected] Figure 2

[Correction method] Change

[Correction content]

[Fig. 2]

[Procedure amendment 4]

[Document name to be corrected] Drawing

[Name of item to be corrected] Figure 6

[Correction method] Change

[Correction content]

[Figure 6]

[Procedure Amendment 5]

[Document name to be corrected] Drawing

[Name of item to be corrected] Fig. 11

[Correction method] Change

[Correction content]

FIG. 11

Claims (3)

[Claims] 1. A first signal, which is an input signal, is supplied,
A quadrature high-pass filter that outputs a second signal, the first signal is supplied, an in-phase high-pass filter that outputs a third signal, and the second and third signals are supplied. A waveform synthesizer that outputs a fourth signal having a steep waveform edge of the third signal by nonlinear processing using an amplitude value and a phase value obtained by vector-synthesizing two signals; 1, 3, and 4 signals are supplied, and the waveform edge of the first signal is emphasized by replacing the third signal, which is a part of the first signal, with the fourth signal. And an adder / subtractor for obtaining an output signal. 2. The waveform synthesizer uses a circuit for obtaining the phase value as a positive value of π / 2 or less, and a parameter α for the phase value smaller than π / 2 as a reference value. A circuit that obtains a new phase value by performing a non-linear phase conversion process with a small value, and a circuit that obtains the fourth signal from the new phase value, the amplitude value, and the polarity of the third signal. The image quality improving apparatus according to claim 1, wherein the image quality improving apparatus comprises: 3. The waveform synthesizer includes a circuit for obtaining the phase value as a value of −π / 2 or more and π / 2 or less, and an absolute value of the phase value as an input, and the input up to a predetermined value. Is an output value of zero, and for values above that, a non-linear phase conversion process is performed using the parameter α so that an output value corresponding to the input value is obtained, and a new phase value with a widened zero phase region 2. The image quality improving apparatus according to claim 1, further comprising: a circuit for obtaining the fourth signal and a circuit for obtaining the fourth signal based on the new phase value and the amplitude value.