JPH0818829A - Noise eliminating device - Google Patents

Abstract

PURPOSE:To provide a noise eliminating device eliminating a noise in response to the quality of a video signal by adjusting a feedback quantity depending on the motion of the video signal. CONSTITUTION:This device is provided with a frame delay means 11 delaying a signal in response to an input video signal, a 1st subtractor means 12 obtaining a difference signal between an input video signal and the output signal of the frame delay means 11, an orthogonal transform means 14 applying orthogonal transform to the difference signal obtained in this way, a nonlinear processing means 15 applying nonlinear processing to the output of the orthogonal transform means 14, an orthogonal inverse transform means 16 applying orthogonal inverse transform to the output, an attenuation means 18, a 2nd subtractor means 19 to obtain a difference signal between the input video signal and the output signal of the attenuation means 18. Further, a motion detection means 21 detecting a motion based on the output of the orthogonal transformation means 14 and an adaptive control means 22 applying at least one of a threshold level of nonlinear processing of the nonlinear processing means 15 and attenuation of the attenuation means 18.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a noise eliminating device for effectively eliminating noise contained in a video signal.

[0002]

2. Description of the Related Art With the recent development of semiconductor memory, a frame memory can be used at a low cost.
Dimensional processing is becoming popular. As for a noise eliminator used in a home VTR or a TV receiver, many devices using a frame memory have been devised. Among them, a Hadamard Transform, which is a method of orthogonal transformation, is used as a noise eliminator that utilizes the difference in three-dimensional statistical properties of video signals and random noise.
A frame recursive noise eliminator using the above has been proposed (for example, Journal of the Television Society, Vol. 37, No. 37).
12, 1983, pp56-62).

Video signals without noise include horizontal, vertical,
Correlation is large in any direction along the time axis, while random noise has a property that the correlation is small in the horizontal, vertical, or any direction along the time axis. The noise removing device using Hadamard transform is a method for removing noise by more effectively using the difference in the three-dimensional correlation between the video signal and the random noise. A frame recursive noise eliminator using Hadamard transform has a lower resolution of a moving image portion than a simple frame recursive noise eliminator without Hadamard transform under the condition that the amount of improvement in S / N ratio is the same. There is an advantage that there are few.

A conventional frame recursive denoising apparatus using Hadamard transform will be described below. FIG.
Figure 1 shows the configuration of a conventional noise eliminator using Hadamard transform. In FIG. 18, reference numeral 1 is a first subtractor, which subtracts an input video signal and a signal obtained by delaying an output signal obtained by removing noise from the input video signal by one to several frames to obtain a frame difference signal. Reference numeral 2 delays the noise-removed output signal by one to several frames. A serial-parallel converter 3 converts a time-series data string into a time-parallel data string so as to match the order of the Hadamard transform. Reference numeral 4 is a Hadamard converter which performs Hadamard conversion on parallel data strings. 5_1
5 to 5_k are non-linear processing units that perform non-linear processing on the data Hadamard-transformed by the Hadamard converter 4. Reference numeral 6 denotes an operation that is the reverse of the conversion operation performed by the Hadamard transformer 4, that is, Hadamard inverse transformation is performed on the data that has been subjected to the nonlinear processing by the Hadamard inverse transformer. A parallel-serial converter 7 converts a parallel data string that has undergone Hadamard inverse conversion into a serial data string. 8
The second subtractor subtracts the output of the parallel-serial converter 7 from the input video signal to obtain an output signal from which noise has been removed.

The operation of the noise eliminator configured as described above will be described below. First, the first subtractor 1 calculates the difference between the signal delayed by N (N = 1, 2, ...) Frames by the frame memory 2 and the input signal. Since the random noise and the motion component in the video signal have a small correlation in the time axis direction, they are extracted as a frame difference signal according to the noise and the signal amplitude.

The serial-parallel converter 3 temporally parallels the temporally serial frame difference signals output from the first subtractor 1 at m sample points in the horizontal direction and n lines in the vertical direction (m and n are natural numbers). Convert to different data. The serial-parallel converter 3 is composed of, for example, (n-1) line memories and (m-1) * n latches. Here, m = 4 samples, n =
The case of two lines will be described as an example.

The data of the temporally parallel pixel blocks generated by the serial / parallel converter 3 is shown in the form of a matrix (Equation 1).

[0008]

[Equation 1]

Here, the data of the pixel block composed of x00 to x13 will be described. Based on x00,
x01, x02, and x03 are data located on the screen by 1 sample, 2 samples, and 3 samples to the right, respectively. When x10 is used as a reference, x11, x12, and x13 are 1 on the screen, respectively.
The data is located to the right of samples, 2 samples, and 3 samples. Further, x10 to x13 are data which are located one line below on the screen with respect to x00 to x03.

The Hadamard transform unit 4 performs the Hadamard transform operation represented by (Equation 2) on the data of the temporally parallel pixel blocks of 4 sample points in the horizontal direction and 2 lines in the vertical direction. Expand to the frequency component of. However, yij (0 ≦ i ≦ 1, 0 ≦ j ≦ 3) is data after Hadamard transform.

[0011]

[Equation 2]

Since the random noise has a small correlation,
It is evenly distributed to each frequency component of yij in (Equation 2).

The non-linear processing units 5_1 to 5_k extract noises evenly distributed in each frequency component by Hadamard transform. FIG. 19 shows an input / output characteristic diagram of the nonlinear processing units 5_1 to 5_k showing an example of the input / output relationship of the nonlinear processing units 5_1 to 5_k. Here, the output of the non-linear processing unit for yij which is the output of the Hadamard converter 4 is y'ij. In FIG. 19, the horizontal axis represents the input yij and the vertical axis represents the output y'ij. As can be seen from FIG. 19, when the input yij whose absolute value is the value “A” or more is input, the output y′ij is zero.

After that, the noise components extracted by the non-linear processing units 5_1 to 5_k are subjected to the calculation represented by (Equation 3) in the Hadamard inverse transformer 6 to restore the data to real space domain components again.

[0015]

(Equation 3)

Further, after the noise component x'ij returned to the real space domain is converted into time-series data by the parallel-serial converter 7, the second subtractor 8 converts the noise-containing input signal from the input signal. By performing the subtraction, the conventional noise removing device using the Hadamard transform realizes the noise removing action.

[0017]

However, in the above-mentioned non-linear processing units 5_1 to 5_k, the non-linear processing threshold "A" is fixed. For example, when the motion component of the video signal is equal to or less than the value "A". Is a non-linear processing unit 5_1
5_k erroneously extracts a motion component as noise and feeds it back, and thus has a problem of causing a phenomenon such as an afterimage or tailing in a moving image portion.

Further, for example, the amplitude of noise is large and the value is "A".
If it exceeds the above, the non-linear processing unit 5 cannot completely extract the noise component and returns only a part of the noise, resulting in a problem that the noise removing effect cannot be sufficiently obtained. Had.

The present invention solves the above-mentioned conventional problems, and provides a noise eliminator which does not cause deterioration such as an afterimage or tailing in a moving image portion by adjusting a feedback amount according to a movement amount of a video signal. The purpose is to provide.

[0020]

In order to achieve this object, a noise removing device of the present invention is a delay means for delaying a signal in response to an input video signal, the input video signal and an output signal of the delay means. A first subtraction means for obtaining a difference signal between the first and second subtraction means, an orthogonal transformation means for performing an orthogonal transformation on the difference signal obtained by the first subtraction means, and a non-linear processing for the output of the orthogonal transformation means. And an inverse orthogonal transform means for performing an orthogonal inverse transform, which is an inverse transform of the orthogonal transform in the orthogonal transform means, on the output of the nonlinear transform means, and an output of the orthogonal inverse transform means is attenuated. An attenuating means, a second subtracting means for obtaining a difference signal between the input video signal and an output signal of the attenuating means, a motion amount detecting means for detecting a motion amount using an output of the orthogonal transforming means, and the motion amount Based on the output of the detecting means composed of the adaptive control means for adaptively controlling at least one of attenuation at threshold and the damping means of the non-linear processing in the nonlinear processing means.

In order to achieve the same object, the noise eliminating apparatus of the present invention provides a delay means for delaying a signal in response to an input video signal and a differential signal between the input video signal and the output signal of the delay means. First subtraction means for obtaining, orthogonal transformation means for performing orthogonal transformation on the difference signal obtained by the first subtraction means, and nonlinear processing means for performing non-linear processing on the output of the orthogonal transformation means An orthogonal inverse transform means for performing an orthogonal inverse transform, which is an inverse transform of the orthogonal transform in the orthogonal transform means, on the output of the non-linear processing means, an attenuating means for attenuating the output of the orthogonal inverse transform means, Second subtracting means for obtaining a difference signal between an input video signal and an output signal of the attenuating means, a motion amount detecting means for detecting a motion amount using the output of the orthogonal transforming means, and an output of the motion amount detecting means. And an adaptive control means for adaptively controlling at least one of a non-linear processing threshold in the non-linear processing means and an attenuation amount in the attenuating means based on the output of the isolated point removing means. Composed of.

In order to achieve the same object, the noise eliminating apparatus of the present invention provides a delay means for delaying a signal responsive to an input video signal and a differential signal between the input video signal and the output signal of the delay means. First subtraction means for obtaining, orthogonal transformation means for performing orthogonal transformation on the difference signal obtained by the first subtraction means, and nonlinear processing means for performing non-linear processing on the output of the orthogonal transformation means An orthogonal inverse transform means for performing an orthogonal inverse transform, which is an inverse transform of the orthogonal transform in the orthogonal transform means, on the output of the non-linear processing means, an attenuating means for attenuating the output of the orthogonal inverse transform means, Second subtraction means for obtaining a difference signal between the input video signal and the output signal of the attenuating means, a plurality of movement amount detection means for detecting a movement amount using the difference signal, and a plurality of movement amount detections hand Motion amount calculation means for calculating a motion amount from the output of the motion compensation device, and adaptive control for adaptively controlling at least one of the threshold value of the non-linear processing in the non-linear processing means and the attenuation amount in the attenuation means based on the output of the motion amount calculation means. And means.

In order to achieve the same object, the noise eliminator of the present invention provides a delay means for delaying a signal in response to an input video signal and a differential signal between the input video signal and the output signal of the delay means. First subtraction means for obtaining, orthogonal transformation means for performing orthogonal transformation on the difference signal obtained by the first subtraction means, and nonlinear processing means for performing non-linear processing on the output of the orthogonal transformation means An orthogonal inverse transform means for performing an orthogonal inverse transform, which is an inverse transform of the orthogonal transform in the orthogonal transform means, on the output of the non-linear processing means, an attenuating means for attenuating the output of the orthogonal inverse transform means, Second subtraction means for obtaining a difference signal between the input video signal and the output signal of the attenuating means, at least one movement amount detection means for detecting a movement amount in the horizontal direction using the difference signal, Based on the output of at least one isolated point removing means for removing the isolated point for each amount detecting means, the motion amount calculating means for calculating the motion amount from the output of the at least one isolated point removing means, and the output of the isolated point removing means. And adaptive control means for adaptively controlling at least one of the threshold value of the non-linear processing in the non-linear processing means and the attenuation amount in the attenuating means.

In order to achieve the same object, the noise removing apparatus of the present invention provides a delay means for delaying a signal in response to an input video signal and a differential signal between the input video signal and the output signal of the delay means. First subtraction means for obtaining, orthogonal transformation means for performing orthogonal transformation on the difference signal obtained by the first subtraction means, and nonlinear processing means for performing non-linear processing on the output of the orthogonal transformation means An orthogonal inverse transform means for performing an orthogonal inverse transform, which is an inverse transform of the orthogonal transform in the orthogonal transform means, on the output of the non-linear processing means, an attenuating means for attenuating the output of the orthogonal inverse transform means, Second subtraction means for obtaining a difference signal between the input video signal and the output signal of the attenuating means, a motion amount detecting means for detecting a motion amount using the output of the orthogonal transforming means, and an output of the orthogonal transforming means An edge amount detecting means for detecting an edge component of image information using the threshold value, a non-linear processing threshold in the non-linear processing means based on the output of the motion amount detecting means and the output of the edge amount detecting means, and the attenuation in the attenuating means. And an adaptive control means for adaptively controlling at least one of the quantities.

[0025]

According to the present invention, the adaptive control means controls at least one of the threshold value of the non-linear processing of the non-linear processing means and the attenuation amount of the attenuating means by the above-mentioned configuration, thereby suppressing the deterioration of the moving area image and the moving area image. It works to efficiently remove noise in both still area images.

[0026]

An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 shows a block diagram of a noise eliminating apparatus according to a first embodiment of the present invention. In FIG. 1, 10 is an input terminal to which a video signal including noise is added. Reference numeral 11 denotes a frame delay means, which is connected to a second subtraction means 19 which will be described later, and outputs an output signal obtained by removing a noise component from an input video signal output by the second subtraction means 19 for N frames (N = 1, 1).
2, ...) It delays. Reference numeral 12 is a first subtraction means, which is connected to the input terminal 10 and the frame delay means 11 and is an N output from the input video signal and the frame delay means 11.
This is to obtain a differential signal (frame differential signal) from the signal delayed by the number of frames. Reference numeral 13 denotes a serial / parallel conversion means, which is connected to the first subtraction means 12 and converts a temporally serial signal into temporally parallel data to generate pixel block data. Here, the size of the pixel block data is m samples in the horizontal direction and n lines in the vertical direction (m and n are natural numbers). Further, in the present embodiment, for the sake of explanation, it is assumed that m = 4 and n = 2 as an example. Here, a configuration example of the serial-parallel conversion means 13 will be described.
FIG. 2 is a block diagram of the serial / parallel conversion means 13, wherein the serial / parallel conversion means 13 is (m−1) × n 1-sample delay means 20.
1 to 206 and (n-1) line delay means 207.

In FIG. 1, reference numeral 14 is an orthogonal transforming means, which is connected to the serial / parallel transforming means 13 and is connected to the serial / parallel transforming means 13.
The orthogonal transformation is applied to the data of the pixel block generated in (1). In the present embodiment, the orthogonal transform in the orthogonal transform means 14 will be described by taking the Hadamard transform as an example. The Hadamard transform has an advantage that the circuit configuration is simple and the circuit can be commonly used for the transform and the inverse transform. 15_1 ~
Reference numeral 15_k denotes k (k = m × n) non-linear processing means, which is connected to the orthogonal transformation means 14 and performs non-linear processing on the data after the orthogonal transformation to extract a noise component. Reference numeral 16 is an orthogonal inverse transforming means, which is a non-linear processing means 15_.
It is connected to 1 to 15_k and performs inverse orthogonal transform on the data extracted as the noise component.

Reference numeral 17 denotes a parallel / serial conversion means, which is connected to the orthogonal inverse conversion means 16 and performs orthogonal inverse conversion on the temporally parallel data (pixel block data) of m samples in the horizontal direction and n lines in the vertical direction. , By converting the average value of the sample points which are overlapped between different blocks (integration operation), the signals are converted into a serial signal in terms of time. Here, a configuration example of the parallel-serial conversion means 17 will be described. FIG. 3 is a block diagram of the parallel / serial conversion means 17, wherein the parallel / serial conversion means 17 is (m−1) × n 1-sample delay means 211 to 216.
And (n−1) line delay means 217 and (m × n−)
1) It is composed of one addition means 218 to 224 and an attenuation means 225 having a gain of 1 / j. In the example, m
= 4, n = 2, the above-mentioned values are (m-1) * n = 6, n-1 = 1, and m * n-1 =
7, 1 / j = 1/8.

In FIG. 1, reference numeral 18 denotes an attenuating means, which is connected to the parallel / serial converting means 17 and reduces the gain of the signal integrated by the parallel / serial converting means 17. 19 is a second subtraction means, which is connected to the input terminal 10 and the attenuation means 18,
The noise component is removed from the input video signal by subtracting the output signal of the attenuator 18, which is a noise component, from the input video signal. The output of the second subtraction means 19 is connected to the frame delay means 11. An output terminal 20 is connected to the second subtraction means 19 and outputs a video signal from which noise is removed.

Reference numeral 21 is a motion amount detecting means, which is the orthogonal transformation means 1.
4 is used to detect the absolute value of the motion amount of the signal component included in the frame difference signal output from the first subtraction means 12 using the output of the orthogonal transformation means 14.

Reference numeral 22 is an adaptive control means, which is a motion amount detection means 2
1. Non-linear processing means 15_1 to 15_k, damping means 18
Connected to. Here, a configuration example of the adaptive control means 22 will be described. FIG. 4 is a block diagram of the adaptive control means 22.
Reference numeral 31 is a first control means, which is connected to the motion amount detecting means 21 and the non-linear processing means 15_1 to 15_k, and based on the motion amount output from the motion amount detecting means 21, the non-linear processing means 15
The second control means 232 controls the threshold value for the non-linear processing of _1 to 15_k. The second control means 232 is connected to the motion amount detecting means 21 and the damping means 18, and sets the motion amount output by the motion amount detecting means 21. Based on the feedback coefficient a in the damping means 18,
Is to control.

The operation of the noise eliminator of the first embodiment constructed as above will be described below. A video signal containing noise is input from the input terminal 10. The first subtraction means 12 calculates the difference between the N frame (N = 1, 2, ...) Delayed signal output by the frame delay means 11 and the input video signal, and outputs a frame difference signal.
The first subtraction unit 12 detects irregular noise having a small correlation between frames and a motion component of a video signal. The output of the first subtraction means 12 is 0 in the static region containing no noise.

The time-serial signals of noise and motion components detected by the first subtracting means 12 are converted into serial-parallel converting means 1.
3, 3 (m-1) 1-sample delay units 201 to 206 composed of latches and (n-1) line-delay units 207 composed of line memories are used to m samples in the horizontal direction. It is converted into n lines of time-parallel data in the vertical direction. Now, as an example, description will be made assuming that m = 4 and n = 2. The time-parallel blocks generated by the serial-parallel conversion means 13 are shown in (Formula 4) in the form of a matrix.

[0035]

[Equation 4]

Here, by using FIG. 5, k pieces (k = m × n =
The pixel block composed of the data x00 to x03 and x10 to x13 of 8) will be described. FIG. 5 is a block diagram of a 4 × 2 pixel block. When x00 is used as a reference, x01, x02,
x03 is the data located on the screen by 1 sample, 2 samples, and 3 samples to the right, and when x10 is used as a reference, x11, x12, and x13 are 1 sample each on the screen,
The data is located to the right of two samples and three samples. Further, x10 to x13 are data which are located one line below on the screen with respect to x00 to x03.

4 × 2 converted by the serial / parallel conversion means 13
Pixel block data xij (0 ≦ i ≦ 1, 0 ≦ j ≦
3) is subjected to Hadamard transform by the orthogonal transform means 14. The data after Hadamard transformation is yij (0 ≦ i ≦ 1, 0 ≦ j ≦
As 3), the conversion formula is shown in (Equation 5).

[0038]

(Equation 5)

By subjecting the frame difference data xij to Hadamard transform, an irregular noise component having a small correlation in the horizontal and vertical directions is, for example, white noise having a flat frequency characteristic. Almost evenly distributed over y13, each level is about 1/8 before Hadamard transform. On the other hand, the motion component of the video signal has frequency characteristics, and if this is Hadamard transformed, y00 ~
Focus on a specific component of the eight components of y13.

Next, the operation of the motion amount detecting means 21, the adaptive control means 22 and the non-linear processing means 15_1 to 15_k will be described with reference to FIGS.

First, the motion amount detecting means 21 detects the amount of motion of the signal component contained in the frame difference signal output from the first subtracting means 12 using the output of the orthogonal transforming means 14. For example, The motion amount may be detected by performing arithmetic calculation or absolute value calculation using a plurality of signals (y00 to y13) output by the orthogonal transformation means 14. Here, as a simple example, the value obtained by performing the absolute value calculation of the y00 signal (sum of 8 pixels) output from the orthogonal transformation means 14 is used as the motion amount.

If the motion amount output from the motion amount detecting means 21 is large, it indicates that the frame difference signal output from the first subtracting means 12 includes not only a noise component but also a motion component of the signal. When the motion amount output from the motion amount detecting means 21 is small, it indicates that the frame difference signal output from the first subtracting means 12 contains only a noise component.

The motion amount obtained by the motion amount detecting means 21 is
It is output to the first control means 231 and the second control means 232 of the adaptive control means 22.

The first control means 231 of the adaptive control means 22
Is based on the motion amount obtained by the motion amount detecting means 21,
Y00 to y in the non-linear processing means 15_1 to 15_k
A threshold value for non-linear processing on the data of the pixel block consisting of 13 8 components is generated. Here, FIG. 6 shows an example of the input / output relationship of the non-linear processing means 15_1 to 15_k.
FIG. 6 is an input / output characteristic diagram of the non-linear processing means 15_1 to 15_k. When the absolute value of the input is less than or equal to the threshold value (indicated by the value "A" in the figure), the linear output characteristic is output, and the absolute value of the input is the threshold value. In the above case, the characteristic has the opposite slope, and when the absolute value of the input is more than twice the threshold value, the output is set to "0".
A so-called non-linear operation with a maximum threshold value is performed. Here, when the output of the motion amount detecting means 21 is large, the frame difference data is regarded as the motion component of the video signal, and the threshold value in the non-linear processing means 15_1 to 15_k is reduced to reduce the feedback amount of the signal of the motion component. To reduce the deterioration of moving images. On the other hand, when the output of the motion amount detecting means 21 is small, the frame difference data is regarded as a noise component and the threshold value in the non-linear processing means 15_1 to 15_k is increased, that is, the noise feedback amount is increased, and the noise removing effect is obtained. To increase.

Here, an example of the input / output relationship in the first control means 231 is shown in FIG. FIG. 7 shows the first control means 2.
31 is an input / output characteristic diagram showing the relationship between the value of the motion amount detecting means 21 and the threshold values of the non-linear processing means 15_1 to 15_k.

The outputs non-linearly processed by the non-linear processing means 15_1 to 15_k are made to correspond to the input yij and y '.
ij. The orthogonal inverse transform means 16 performs a 4 × 2 order Hadamard inverse transform shown in (Equation 6) on y′ij. Here, x′ij is the output of the orthogonal inverse transform means 16.

[0047]

(Equation 6)

As can be seen from (Equation 6), the coefficient 1/8
The Hadamard inverse transformation formula is the same as the Hadamard transformation formula except for the presence or absence of.

Next, the data subjected to the Hadamard inverse transformation by the orthogonal inverse transformation means 16 is the data x'ij of 4 × 2 temporally parallel pixel blocks in the parallel / serial transformation means 17 shown in FIG.
(0 ≦ i ≦ 1, 0 ≦ j ≦ 3) is converted into eight temporally serial data, and at sample points where different pixel blocks overlap, the blocks are averaged.

Then, the parallel / serial converted data is multiplied by the feedback coefficient generated by the adaptive control means 22 in the attenuation means 18.

Next, the operations of the motion amount detecting means 21, the adaptive control means 22 and the damping means 18 will be described with reference to FIGS. 1 and 4.

Second control means 232 of adaptive control means 22
Controls the gain, that is, the feedback coefficient a (0 ≦ a <1), by the attenuator 18, based on the amount of motion obtained by the motion amount detector 21. Here, when the output of the motion amount detecting means 21 is large, the frame difference data is regarded as the motion component of the video signal, and the feedback coefficient a in the attenuating means 18 is reduced to reduce the feedback amount of the signal of the motion component. Make it smaller to prevent deterioration of moving images. On the other hand, if the output of the motion amount detecting means 21 is small, the frame difference data is regarded as a noise component, and the feedback coefficient a is increased by the attenuating means 18, that is, the noise feedback amount is increased, and the noise removing effect is increased. To do.

Here, an example of the input / output relationship in the second control means 232 is shown in FIG. FIG. 8 shows the second control means 2.
An input / output characteristic diagram of 32 shows the relationship between the value of the motion amount detecting means 21 and the feedback coefficient of the damping means 18.

Finally, the second subtraction means 19 subtracts the data extracted as noise from the input video signal, so that a noise-free video signal with little deterioration of the moving image is obtained at the output terminal 20. be able to.

As described above, according to this embodiment, the input terminal 10 to which the video signal is added and the frame delay means for delaying the output of the second subtraction means by N frames (N = 1, 2 ...). 11, input video signal and frame delay means 11
From the output signal from the first subtraction means 12 and serial-parallel conversion means 13 that converts temporally serial frame difference data into temporally parallel pixel block data.
, Orthogonal transformation means 14 for performing orthogonal transformation, and non-linear processing means 15 for performing non-linear processing on the orthogonally transformed data.
_1 to 15_k, an orthogonal inverse transform means 16 that performs an inverse orthogonal transform on the non-linearly processed data, and data of pixel blocks that are temporally parallel to the orthogonally inversely converted data are serially temporally arranged. After being converted into data, parallel-serial conversion means 17 for averaging overlapping data among a plurality of blocks, attenuating means 18 for multiplying temporally serialized data by a feedback coefficient, and extracted from the input video signal. The second subtraction means 19 for subtracting the noise component, and the motion amount detection means for detecting the motion amount of the signal component included in the frame difference signal output from the first subtraction means 12 by using the outputs of the orthogonal transformation means 14. 21 and the non-linear processing means 15_1 to 1
By providing the adaptive control means 22 for controlling the threshold value of 5_k and the feedback coefficient in the attenuation means 18 according to the motion amount which is the output of the motion amount detecting means 21, the moving area and the stationary state are reduced in a state where the deterioration of the moving image is small. Noise components in both areas can be effectively removed.

In the present invention, the nonlinear processing means 15
The input / output characteristics in _1 to 15_k are not limited to those shown in FIG. 6, and for example, the slope of the characteristic represented by a straight line may not be 1, and may have the characteristic represented by a non-straight line. . The input / output relationship of the first control means 231 and the input / output characteristics of the second control means 232 are not limited to those shown in FIGS. 7 and 8.

Further, in the present embodiment, the data of the pixel block is continuously defined by m samples in the horizontal direction and n lines in the vertical direction (m and n are natural numbers). , N lines at intervals of q lines in the vertical direction (where m, n, p, and q are natural numbers) may be defined. However, in this case, the q line delay means may be used as the line delay means used in the serial / parallel conversion means or the parallel / serial conversion means, and the p sample delay means may be used as the 1 sample delay means.

The second embodiment of the present invention will be described below with reference to the drawings. FIG. 9 shows a block diagram of a noise eliminating apparatus according to the second embodiment of the present invention.

First, the configuration of the second embodiment is shown in FIG.
The first embodiment will be described in comparison with the first embodiment. In FIG. 9, 30
Is an input terminal, 31 is a frame delaying means, 32 is a first subtracting means, 33 is a serial-parallel converting means, and 34 is an orthogonal converting means, which are similar to 10 to 14 in FIG. Three
5_1 to 35_k are non-linear processing means, which are 15_1 in FIG.
~ 15_k. Reference numeral 36 is an orthogonal inverse conversion means, 37 is a parallel / serial conversion means, 38 is an attenuation means, 39 is a second subtraction means, 40 is an output terminal, 41 is a motion amount detection means, which are the same as 16 to 21 in FIG. It is a thing.

Reference numeral 42 is an isolated point removing means for removing isolated points from the output of the motion amount detecting means 41. An example of the configuration will be described. FIG. 10 is a block diagram of the isolated point removing means 42. Reference numeral 241 is an r sample delay means for delaying input data by r sample time (r is a natural number), and 242 is an s sample delay means. Data is delayed by s sample time (s is a natural number),
Numeral 243 is an intermediate value selecting means for selecting an intermediate value data from a plurality of input data and outputting the absolute value thereof.

That is, the isolated point removing means 42 uses data without delay, data delayed by r sample time by the r sample delay means 241, and data delayed by (r + s) sample time by the s sample delay means 242. By selecting and outputting an intermediate value by the intermediate value selecting means 243, a filter that performs isolated point removal using the data after r samples and the data before s samples is configured.

Reference numeral 43 denotes an adaptive control means, which operates in the same manner as the adaptive control means 22 of FIG. 1 in the first embodiment and has the same configuration as that shown in FIG. The adaptive control means 43
The isolated point removing means 42 and the non-linear processing means 35_1 to 35_
k, connected to the damping means 38, and the first control means 231
Are non-linear processing means 35_1 to 35_ based on the motion amount detection signal from which the isolated point removal means 42 outputs.
The second control means 242 controls the threshold value for the non-linear processing of k, and the second control means 242 uses the isolated point removed motion amount detection signal output from the isolated point removal means 42 to attenuate the attenuation means 38.
Is to control the feedback coefficient a.

The operation of the noise eliminator of the second embodiment constructed as above will be described below in comparison with the operation of the first embodiment. An input video signal containing noise is applied to the input terminal 30, the first subtraction means 32 takes the difference between the input video signal and the output signal of the frame delay means 31, and the serial / parallel conversion means 33 performs temporal parallel processing. Data of a pixel block composed of m × n pieces of data is generated, the m × n Hadamard transform is performed by the orthogonal transformation means 34, the nonlinear processing is performed by the nonlinear processing means 35_1 to 35_k, and the orthogonal inverse transformation means 36 performs m. The inverse n × n Hadamard transform is performed, the time-parallel m × n data is serially converted by the parallel / serial conversion means 37, the attenuation means 38 multiplies the feedback coefficient, and then the second subtraction means 39. The basic operation of obtaining the noise-removed video signal at the output terminal 40 by subtracting the data extracted as noise at the input video signal is the same as in the first embodiment.

In the first embodiment, the adaptive control means 22 is controlled by using the motion amount detecting means 21, but in the second embodiment, the isolated point data is removed from the output of the motion amount detecting means 41. The adaptive control means 43 is controlled via the isolated point removal means 42 having a function.

The motion amount detecting means 41 is an orthogonal transformation means 34.
Is used to detect the motion amount of the signal component included in the frame difference signal output by the first subtraction means 32. The motion amount of the signal component depends on the video signal, and the video signal itself is in the horizontal direction. , The vertical direction or the time axis direction has a correlation, and it is rare that the amount of motion is large (or small) in only one pixel, for example. Therefore, in such a case, by considering the corresponding pixel as a singular point (isolated point) and performing a process of replacing it with the amount of movement of surrounding pixels (isolated point removal process), the amount of movement that is extremely natural from the perspective of the entire display image. Can be obtained. Therefore, the isolated point removing means 42 is provided between the motion amount detecting means 41 and the adaptive control means 43.
Via the motion amount detected by the motion amount detecting means 41, the isolated point removing means 42 removes the isolated point data of the motion amount,
The adaptive control means 43 is a non-linear processing means 35_1 to 35_k.
The threshold value for the non-linear processing performed in step 1 and the feedback coefficient for the attenuation processing performed by the attenuation means 38 are adaptively controlled according to the natural motion amount output by the isolated point removal means 42.

In the first embodiment, there was a slight sense of discomfort due to the difference in the noise removal amount between the isolated point of the motion amount and the surroundings, but in the second embodiment, the discomfort was hardly seen,
A noise-removed video signal with little deterioration of the moving image can be obtained from the output terminal 40.

As described above, according to the second embodiment, the input terminal 30 to which the video signal is added and the frame for delaying the output of the second subtracting means by N frames (N = 1, 2 ...). The delay means 31 and the first subtraction means 32 for taking the difference between the input video signal and the output signal from the frame delay means 31.
, Serial-parallel conversion means 33 for converting temporally serial frame difference data into temporally parallel pixel block data, orthogonal conversion means 34 for performing orthogonal conversion, and non-linear for orthogonally converted data. Non-linear processing means 35_1 to 35_k for performing processing, orthogonal inverse transformation means 36 for performing orthogonal inverse transformation on the non-linearly processed data, and data of pixel blocks that are temporally parallel to the orthogonally inverse-transformed data. After being converted into temporally serial data, parallel-serial conversion means 37 for averaging overlapping data among a plurality of blocks, and attenuation means 38 for multiplying the temporally serialized data by a feedback coefficient, The frame difference which is the output of the first subtraction means 32 is calculated by using the outputs of the second subtraction means 39 for subtracting the extracted noise component from the input video signal and the orthogonal transformation means 34. A motion amount detecting means 41 for detecting a motion amount of a signal component included in a signal, an isolated point removing means 42 for removing isolated data from the motion amount detected by the motion amount detecting means 41, and a non-linear processing means 35_1 to 35_k. Of the threshold value and the feedback coefficient in the attenuation means 38 to the isolated point removal means 4
By providing the adaptive control means 43 that controls according to the amount of movement of the isolated point removed which is the output of No. 2, noise components in both the moving region and the still region are effectively removed in a state where the moving image is less deteriorated. it can.

It should be noted that r sample delay means 241 of the isolated point removing means 42 and r and s of the s sample delay means 242.
With respect to the sample time delay amount shown in (4), if r = s, the isolated point removing means 42 constitutes a so-called intermediate value filter, but the isolated point removing means 42 does not necessarily have to be limited to the filter r = s. Further, a plurality of isolated point removing means having the configuration used in this embodiment may be used in series to form one isolated point removing means 42.

In this embodiment, the isolated point removing means 42 removes the isolated points in the horizontal direction.
By replacing the r, s sample delay means with the r, s line delay means, it is possible to remove isolated points in the vertical direction.
If it is replaced with an s frame delay means, it is possible to remove isolated points in the time axis direction. Furthermore, it goes without saying that the performance of isolated point removal can be improved by combining isolated point removal in the horizontal direction, vertical direction, and time axis direction.

The third embodiment of the present invention will be described below with reference to the drawings. FIG. 11 shows a block diagram of a noise removing apparatus according to a third embodiment of the present invention.

First, the configuration of the third embodiment is shown in FIG.
The first embodiment will be described in comparison with the first embodiment. In FIG. 11, 5
Reference numeral 0 is an input terminal, 51 is a frame delay means, 52 is a first subtraction means, 53 is a serial / parallel conversion means, and 54 is an orthogonal conversion means, which are similar to 10 to 14 in FIG.
55_1 to 55_k are non-linear processing means, which are 15_ in FIG.
It is the same as 1 to 15_k. Reference numeral 56 is an orthogonal inverse conversion means, 57 is a parallel-series conversion means, 58 is an attenuation means, 59 is a second subtraction means, and 60 is an output terminal.
It is the same as 6 to 20.

61_1 and 61_2 are motion amount detecting means, which are connected to the serial / parallel converting means 53, and the serial / parallel converting means 53.
Is used to detect the amount of horizontal movement of the signal component included in the frame difference signal output by the first subtraction means 52.

Reference numeral 62 denotes a motion amount calculating means, and the motion amount detecting means 6
1_1 and 61_2, and the motion amount detection means 61_
The motion amount is calculated using a plurality of horizontal motion amounts that are outputs of 1 and 61_2.

Reference numeral 63 denotes an adaptive control means, which operates in the same manner as the adaptive control means 22 of FIG. 1 in the first embodiment and has the same configuration as that shown in FIG. The adaptive control means 63
Motion amount calculation means 62, non-linear processing means 55_1 to 55_
k, is connected to the damping means 58, and is the first control means 231.
Is for controlling the threshold value for the non-linear processing of the non-linear processing means 55_1 to 55_k based on the motion amount calculation signal output from the motion amount calculating means 62, and the second control means 2
Reference numeral 42 controls the feedback coefficient a of the attenuation means 58 based on the motion amount calculation signal output from the motion amount calculation means 62.

The operation of the noise eliminator of the third embodiment configured as described above will be described below in comparison with the operation of the first embodiment. An input video signal containing noise is added to the input terminal 50, the first subtraction means 52 takes the difference between the input video signal and the output signal of the frame delay means 51, and the serial / parallel conversion means 53 performs temporal parallel processing. Data of a pixel block composed of m × n pieces of data is generated, the orthogonal transformation means 54 performs Hadamard transformation of m × n order, the nonlinear processing means 55_1 to 55_k perform nonlinear processing, and the orthogonal inverse transformation means 56 performs m. A Hadamard inverse transformation of the order × n is performed, the time-parallel m × n pieces of data are serially converted by the parallel / serial conversion means 57, the attenuation means 58 multiplies the feedback coefficient, and then the second subtraction means 59. The basic operation of obtaining the noise-removed video signal at the output terminal 60 by subtracting the data extracted as noise at the input video signal is the same as in the first embodiment.

In the first embodiment, the adaptive control means 22 is controlled by using the motion amount detecting means 21, but in the third embodiment, a plurality of motion amount detecting means 61_1 are output from the output of the serial / parallel converting means 53. , 61_2 are used to detect the amount of movement in the horizontal direction, the amount of movement calculation means 62 calculates the amount of movement, and the output is used to control the adaptive control means 63.

The motion amount detecting means 61_1 and 61_2 use the output of the serial / parallel converting means 53 to generate the first subtracting means 32.
The motion amount calculating means 62 detects horizontal motion amounts of the signal components included in the frame difference signal
Calculates the total amount of motion using the plurality of horizontal motion amounts obtained here. The motion amount calculation means 62, for example, with respect to a plurality of input data, the maximum value, the minimum value,
The average value or the intermediate value is calculated (selected) and the absolute value is output. The adaptive control means 63 is a motion amount calculation means 62.
Non-linear processing means 55_1 using the motion amount calculated in
˜55_k threshold and damping means 5 for non-linear processing
The feedback coefficient for the attenuation processing performed in 8 is adaptively controlled.

In the first embodiment, only one moving amount detecting portion is configured to be output by the moving amount detecting means 21 using the output of the orthogonal transforming means 14, but in the third embodiment, it is constructed. Using the output of the serial / parallel conversion means 53, a plurality of horizontal movement amounts are detected by the movement amount detecting means 61_1 and 61_2, and the movement amount calculating means 62 calculates the movement amount from the obtained plurality of outputs. Because of the above configuration, the third embodiment is configured so that a more correct amount of movement can be obtained. Therefore, in the third embodiment, a more natural video signal and a noise-removed video signal with less deterioration of the moving image can be obtained from the output terminal 60.

As described above, according to the third embodiment, the input terminal 50 to which the video signal is added and the frame for delaying the output of the second subtraction means by N frames (N = 1, 2 ...). The delay means 51 and the first subtraction means 52 for taking the difference between the input video signal and the output signal from the frame delay means 51.
A serial-parallel conversion means 53 for converting the temporally serial frame difference data into temporally parallel pixel block data, an orthogonal conversion means 54 for performing an orthogonal conversion, and a nonlinear for the orthogonally converted data. Non-linear processing means 55_1 to 55_k for performing processing, orthogonal inverse transformation means 56 for performing orthogonal inverse transformation on the non-linearly processed data, and pixel block data that is temporally parallel to the orthogonally inverse transformed data. After being converted into serial data in time, parallel-serial conversion means 57 for averaging the overlapping data among a plurality of blocks, and attenuation means 58 for multiplying the time-serialized data by a feedback coefficient. A frame which is the output of the first subtraction means 52 using the outputs of the second subtraction means 59 for subtracting the noise component extracted from the input video signal and the serial-parallel conversion means 53. A plurality of motion amount detection unit 61_1 for detecting a horizontal motion amount of the signal component included in the divided signal,
61_2 and a motion amount calculating means 6 for calculating a total motion amount from a plurality of outputs of the motion amount detecting means 61_1 and 61_2.
2 and the adaptive control means 63 for controlling the threshold values of the non-linear processing means 55_1 to 55_k and the feedback coefficient of the attenuation means 58 according to the motion amount output from the motion amount calculating means 62, thereby reducing the deterioration of the moving image. In this state, noise components in both the moving area and the stationary area can be effectively removed.

In the third embodiment, the case where the motion amount detecting means 61_1 and 61_2 are two has been described, but the present invention is not limited to two, and a plurality of similar ones may be used. It goes without saying that the effects will increase if configured. Also, the motion amount detecting means 61_1, 61_2
Has described that the amount of movement in the horizontal direction is detected, but the amount of movement is not limited to the horizontal direction, and the amount of movement in the vertical direction may be detected. Good.

The fourth embodiment of the present invention will be described below with reference to the drawings. FIG. 12 shows a block diagram of a noise elimination apparatus according to the fourth embodiment of the present invention.

First, the configuration of the fourth embodiment is shown in FIG.
The third embodiment will be described in comparison with the first embodiment. In FIG.
Reference numeral 70 is an input terminal, 71 is a frame delaying means, 72 is a first subtracting means, 73 is a serial-parallel converting means, and 74 is an orthogonal converting means, which are similar to 50 to 54 in FIG. 75_1 to 75_k are non-linear processing means, which are similar to 55_1 to 55_k in FIG. Reference numeral 76 is an orthogonal inverse conversion means, 77 is a parallel-series conversion means, 78 is an attenuation means,
Reference numeral 79 is a second subtracting means, and 80 is an output terminal, which are similar to 56 to 60 in FIG.

81_1 and 81_2 are motion amount detecting means connected to the serial / parallel converting means 73, and the serial / parallel converting means 73.
Of the horizontal direction of the signal component included in the frame difference signal output from the first subtraction means 72, respectively.
It is the same as 2.

Reference numerals 82_1 and 82_2 are isolated point removing means, which are connected to the motion amount detecting means 81_1 and 81_2, respectively.
This is to remove isolated points from the outputs of the motion amount detecting means 81_1 and 81_2, and its operation and configuration are the same as those of the isolated point removing means 42 described in the second embodiment with reference to FIG. , The data delayed by r sample time by the r sample delay means 241 and the data delayed by r + s sample time by the s sample delay means 242 are used to select and output an intermediate value by the intermediate value selection means 243. As a result, a filter that performs isolated point removal using the data after r samples and the data before s samples is configured.

Reference numeral 83 denotes a motion amount calculating means, which is an isolated point removing means 8
2_1 and 82_2 and isolated point removing means 82_
The motion amount is calculated by using the plurality of horizontal motion amounts from which the isolated points are removed, which are the outputs of 1 and 82_2.

Reference numeral 84 denotes an adaptive control means, which is the adaptive control means 22 of FIG. 1 in the first embodiment or that of FIG. 1 in the third embodiment.
It operates in the same manner as the adaptive control means 63 of No. 1 and has the same configuration as that shown in FIG. The adaptive control means 84 includes a motion amount calculation means 83 and non-linear processing means 75_1 to 75_.
k, connected to the damping means 78, and the first control means 231
Are non-linear processing means 75_1 to 75 based on the motion amount calculation signal from which the isolated point is removed, which is output from the motion amount calculating means 83.
Controlling the threshold for the non-linear processing of _k,
The second control means 242 is based on the motion amount calculation signal from which the isolated point is removed, which is output from the motion amount calculation means 83.
The feedback coefficient a of 8 is controlled.

The operation of the noise eliminator of the fourth embodiment constructed as above will be described below in comparison with that of the third embodiment. An input video signal containing noise is added to the input terminal 70, the first subtraction means 72 takes the difference between the input video signal and the output signal of the frame delay means 71, and the serial / parallel conversion means 73 performs temporal parallelization. Data of a pixel block composed of m × n pieces of data is generated, the orthogonal transformation means 74 performs Hadamard transformation of m × n order, the nonlinear processing means 75_1 to 75_k perform nonlinear processing, and the orthogonal inverse transformation means 76 performs m processing. The inverse n × Hadamard transform is performed, the parallel-serial converter 77 converts time-parallel m × n data into serial data, the attenuator 78 multiplies the feedback coefficient, and the second subtractor 79. The basic operation of obtaining the noise-removed video signal at the output terminal 80 by subtracting the data extracted as noise in the input video signal is the same as in the third embodiment.

In the third embodiment, the serial / parallel conversion means 53
Output from the plurality of motion amount detecting means 61_1, 61_2
The amount of movement in the horizontal direction is detected using
Although the amount of movement is calculated in 2 and the adaptive control means 63 is controlled using this output, in the fourth embodiment, a plurality of movement amount detection means 81_1 are output from the output of the serial / parallel conversion means 73.
The isolated point removing means 82_ having a function of removing the isolated point data with respect to the horizontal movement amount detected by using 81_2.
The amount of movement calculation is performed by the amount of movement calculation means 83 via 1, 82_2, respectively, and the adaptive control means 84 is used using this output.
Control.

First, the motion amount detecting means 81_1, 81_2
Then, by using the output of the serial / parallel conversion means 73, the horizontal movement amount of the signal component included in the frame difference signal output from the first subtraction means 72 is detected, and the isolated point removal means 82_1 and 82_2 move. Quantity detecting means 81_1, 81
The isolated amount data in the horizontal direction is removed by using the horizontal motion amount output by each _2, and the motion amount calculation means 83
Then, the total amount of movement is calculated using the plurality of horizontal movement amounts from which the isolated points have been removed by the isolated point removal means 82_1 and 82_2, and the adaptive control means 84 calculates it by the movement amount calculation means 83. Using the amount of motion with isolated points removed,
The threshold value for the non-linear processing performed by the non-linear processing means 75_1 to 75_k and the feedback coefficient for the damping processing performed by the damping means 78 are adaptively controlled. The movement amount calculation means 83
Is the maximum value for multiple input data,
The minimum value, average value, or intermediate value is calculated (selected) and the absolute value is output.

In the second embodiment, the motion amount detecting unit 41 outputs only one motion amount detecting unit by using the output of the orthogonal transforming unit 14, and further uses the isolated point removing unit 42. In the third embodiment, the motion amount detecting means 61_1 and 61_2 are used to obtain a plurality of horizontal motion amounts by using the output of the serial / parallel conversion means 53. The motion amount calculating means 62 calculates and obtains the motion amount from the detected plural outputs. However, in the fourth embodiment, a plurality of horizontal motion amounts are detected by using the output of the serial / parallel conversion means 73 so that the structure having both the features of the second and third embodiments is obtained. The isolated point data of the horizontal motion amount detected by the means 81_1 and 81_2 are removed by the isolated point removing means 82_1 and 82_2, and the obtained horizontal movements in which the isolated points are removed are removed. Since the movement amount calculation means 83 calculates the movement amount from the amount, the fourth embodiment is configured so that a more correct movement amount can be obtained. Therefore, in the fourth embodiment, a more natural video signal and a noise-removed video signal with less deterioration of a moving image can be obtained from the output terminal 80.

As described above, according to the fourth embodiment, the input terminal 70 to which the video signal is added and the frame for delaying the output of the second subtraction means by N frames (N = 1, 2 ...) The delay means 71 and the first subtraction means 72 for taking the difference between the input video signal and the output signal from the frame delay means 71.
And a serial / parallel conversion means 7 for converting temporally serial frame difference data into temporally parallel pixel block data.
3, orthogonal transformation means 74 for performing orthogonal transformation, and non-linear processing means 7 for performing non-linear processing on the orthogonally transformed data.
5_1 to 75_k, an orthogonal inverse transform unit 76 that performs an inverse inverse transform on the non-linearly processed data, and data of pixel blocks that are temporally parallel to the orthogonally inversely converted data. After being converted into data, a parallel / serial conversion means 77 for averaging the overlapping data among a plurality of blocks, an attenuating means 78 for multiplying the temporally serialized data by a feedback coefficient, and an input video signal are extracted. The horizontal subtraction amount of the signal component included in the frame difference signal which is the output of the first subtraction unit 72 is detected by using the outputs of the second subtraction unit 79 for subtracting the noise component and the serial-parallel conversion unit 73. A plurality of motion amount detecting means 81_1, 8
1_2, the isolated point removing means 82_1 and 82_2 for removing the isolated point data of the horizontal motion amount, and the motion amount calculating means 83 for calculating a total motion amount from a plurality of outputs of the isolated point removing means 82_1 and 82_2. , Non-linear processing means 75_
By providing the adaptive control means 84 for controlling the threshold value of 1 to 75_k and the feedback coefficient of the attenuation means 78 according to the motion amount of the isolated point removal which is the output of the motion amount calculation means 83,
It is possible to effectively remove the noise component in both the moving area and the still area while the moving image is less deteriorated.

In the fourth embodiment, the case where the movement amount detecting means 81_1 and 81_2 are two has been described, but the present invention is not limited to two, and a plurality of similar ones may be used. It goes without saying that the effects will increase if configured. In that case, the isolated point removing means may be configured by using the same number as the motion amount detecting means.

Further, the motion amount detecting means 81_1, 81_2
Has described that the amount of movement in the horizontal direction is detected, but the amount of movement is not limited to the horizontal direction, and the amount of movement in the vertical direction may be detected. Good.

The fifth embodiment of the present invention will be described below with reference to the drawings. FIG. 13 shows a block diagram of a noise eliminating apparatus according to a fifth embodiment of the present invention.

First, the configuration of the fifth embodiment is shown in FIG.
The second embodiment will be described in comparison with the second embodiment. In FIG. 13, 9
Reference numeral 0 is an input terminal, 91 is a frame delay means, 92 is a first subtraction means, 93 is a serial / parallel conversion means, and 94 is an orthogonal conversion means, which are the same as 30 to 34 in FIG.
95_1 to 95_k are non-linear processing means, and 35_ in FIG.
It is the same as 1 to 35_k. Reference numeral 96 is an orthogonal inverse transforming means, 97 is a parallel / serial transforming means, 98 is an attenuating means, 99 is a second subtracting means, 100 is an output terminal, 101 is a motion amount detecting means, and 102 is an isolated point removing means. Of 3
It is the same as 6 to 42.

An edge detecting means 103 is connected to the orthogonal transforming means 94 and the adaptive control means 104. Here, a configuration example of the edge detection unit 103 will be described. FIG. 14 is a block diagram of the edge detecting means 103. 251 is a horizontal edge detecting means for detecting edge information of a signal in the horizontal direction from a signal output by the orthogonal transforming means 94, and 252 is a vertical edge detecting means. The edge information of the signal in the vertical direction is detected from the signal output by the orthogonal transformation means 94.
Reference numeral 53 is an edge signal generation means, which comprehensively generates an edge signal from the edge information output by the horizontal edge detection means 251 and the vertical edge detection means 252. The edge signal generation means 253 calculates (selects) a maximum value, a minimum value, and an average value for a plurality of input data, and outputs the absolute value thereof.

Reference numeral 104 denotes an adaptive control means, which is an isolated point removing means 102, an edge detecting means 103, and a non-linear processing means 95_.
1 to 95_k, connected to the attenuation means 98. Here, a configuration example of the adaptive control means 104 will be described. FIG. 15 is a block diagram of the adaptive control means 104. 261 is a first control means, which is an isolated point removal means 102, an edge detection means 103,
For non-linear processing of the non-linear processing means 95_1-95_k based on the amount of motion of the isolated point removed which is connected to the non-linear processing means 95_1-95_k and the edge signal output by the edge detection means 103 262 is a second control means for controlling the threshold value of
It is connected to the isolated point removing means 102, the edge detecting means 103, and the attenuating means 98, and the attenuating means 98 is based on the isolated point removed motion amount output by the isolated point removing means 102 and the edge signal output by the edge detecting means 103. The feedback coefficient a of is controlled.

The operation of the noise eliminator of the fifth embodiment constructed as above will be described below by comparing its operation with that of the second embodiment. An input video signal containing noise is added to the input terminal 90, the first subtraction means 92 takes the difference between the input video signal and the output signal of the frame delay means 91, and the serial / parallel conversion means 93 performs temporal parallel processing. Data of a pixel block consisting of m × n pieces of data is generated, the orthogonal transformation means 94 performs Hadamard transformation of m × n order, the nonlinear processing means 95_1 to 95_k perform nonlinear processing, and the orthogonal inverse transformation means 96 performs m. The inverse n × n Hadamard transform is performed, the time-parallel m × n data is serially converted by the parallel / serial conversion means 97, the attenuation means 98 multiplies the feedback coefficient, and then the second subtraction means 99. The basic operation of obtaining the noise-removed video signal at the output terminal 100 by subtracting the data extracted as noise at the input video signal is the same as in the second embodiment.

In the second embodiment, the adaptive control means 43 is controlled via the isolated point removing means 42 having the function of removing isolated point data from the output of the motion amount detecting means 41.
In the fifth embodiment, the isolated point removing means 1 having a function of removing isolated point data from the output of the motion amount detecting means 101.
While controlling the adaptive control means 104 via 02,
The adaptive control means 104 is controlled by using the edge signal from the output of the orthogonal transformation means 94 through the edge detection means 103.

The motion amount detecting means 101 is orthogonal transformation means 9
4 is used to detect the motion amount of the signal component included in the frame difference signal output by the first subtraction unit 92.
The motion amount of the signal component depends on the video signal, and the video signal itself has a correlation in the horizontal direction, the vertical direction, or the time axis direction. For example, when the motion amount of only one pixel is large (or small), It is rare. Therefore,
In such a case, the pixel is considered to be a singular point (isolated point), and a process of replacing it with the amount of motion of surrounding pixels (isolated point removal process) is performed to obtain an extremely natural amount of motion as seen from the entire display image. be able to. Therefore, the isolated point removing unit 102 uses the motion amount output from the motion amount detecting unit 101 via the isolated point removing unit 102 between the motion amount detecting unit 101 and the adaptive control unit 104, and the isolated point removing unit 102 obtains the isolated point data of the motion amount. It is removed and output to the adaptive control means 104.

Further, in the edge detecting means 103,
As shown in FIG. 14, using the output of the orthogonal transformation means 94, the edge information of the signal component included in the frame difference signal output by the first subtraction means 92 is detected using the horizontal edge detection means 251 and the vertical edge detection means 252. The detected edge signal is generated by the edge signal generation means 253. The edge detection method in the horizontal edge detection unit 251 and the vertical edge detection unit 252 may obtain edge information by performing combination processing such as arithmetic operation and absolute value operation using the output of the orthogonal transformation unit 94, for example. As a simple example, the horizontal edge detecting means 251 uses the y02 component, which is one of the outputs of the orthogonal transforming means 94, as the value obtained by performing the absolute value calculation as horizontal edge information, and the vertical edge detecting means 252 is orthogonal. The y10 component, which is one of the outputs of the converting means 94, may be used to output the value obtained by the absolute value calculation as vertical edge information.

In the adaptive control means 104, as shown in FIG. 15, the first control means 261 controls the isolated point removal means 1
02, the threshold value for nonlinear processing of the nonlinear processing means 95_1 to 95_k is controlled based on the isolated point removed motion amount and the edge signal output by the edge detection means 103, and the second control means 262 controls the threshold value. The amount of movement of the isolated point removed by the isolated point removing means 102 and the edge detecting means 10
The feedback coefficient a of the attenuating means 98 is controlled based on the edge signal output by the signal A3.

Here, an example of the input / output relationship in the first control means 261 is shown in FIG. FIG. 16 is an input / output characteristic diagram of the first control means 261 and shows isolated point removal means when the value of the edge detection means 103 is, for example, below the threshold “B” or above the threshold “B”. Non-linear processing means 95_1 to 9
5 shows the relationship of thresholds for non-linear processing of 5_k.

FIG. 17 shows an example of the input / output relationship in the second control means 262. FIG. 17 is an input / output characteristic diagram of the second control means 262, which shows isolated point removal means when the value of the edge detection means 103 is, for example, a threshold “C” or less and a threshold “C” or more. The relationship between the amount of movement of the isolated point removed by the output 102 and the feedback coefficient of the attenuator 98 is shown.

In the second embodiment, the adaptive control means 43 is controlled via the isolated point removing means 42 having the function of removing the isolated point data from the output of the motion amount detecting means 41.
In the fifth embodiment, the isolated point removing means 1 having a function of removing isolated point data from the output of the motion amount detecting means 101.
While controlling the adaptive control means 104 via 02,
Since the adaptive control means 104 is controlled by using the edge signal from the output of the orthogonal transformation means 94 via the edge detection means 103, when the video signal has an edge component, for example, by reducing the feedback amount, the image quality due to the edge component is improved. It is possible to obtain from the output terminal 60 a noise-removed video signal with less deterioration (tailing phenomenon due to edge components of moving images, collapse of high-frequency components of moving images, etc.).

As described above, according to the fifth embodiment, the input terminal 90 to which the video signal is added and the frame for delaying the output of the second subtracting means by N frames (N = 1, 2 ...). The delay means 91 and the first subtraction means 92 for taking the difference between the input video signal and the output signal from the frame delay means 91.
And a serial-parallel conversion means 9 for converting temporally serial frame difference data into temporally parallel pixel block data.
3, orthogonal transformation means 94 for performing orthogonal transformation, and non-linear processing means 9 for performing non-linear processing on the orthogonally transformed data.
5_1 to 95_k, orthogonal inverse transform means 96 that performs orthogonal inverse transform on the non-linearly processed data, and pixel block data that is temporally parallel to the orthogonal inverse transformed data and is temporally serial data. After the conversion into the data, the parallel-serial conversion means 97 for averaging the overlapping data among the plurality of blocks, the attenuating means 98 for multiplying the temporally serialized data by the feedback coefficient, and the input video signal are extracted. A motion amount for detecting the horizontal motion amount of the signal component included in the frame difference signal which is the output of the first subtraction unit 92 using the outputs of the second subtraction unit 99 for subtracting the noise component and the orthogonal transformation unit 94. An edge is detected by using the output of the detection unit 101, the isolated point removal unit 102 that removes isolated data from the motion amount detected by the motion amount detection unit 101, and the output of the orthogonal transformation unit 94. An edge detection unit 103 for, nonlinear processing means 55_1~55_k threshold and attenuating means 58
Degradation of a moving image is provided by providing an adaptive control unit 104 for controlling the feedback coefficient of the above according to the amount of motion of the isolated point removed which is the output of the isolated point removal unit 102 and the edge information which is the output of the edge detection unit 103. Motion area,
It is possible to effectively remove the noise component in both the stationary regions.

The input / output relationship in the first control means 261 and the input / output characteristics in the second control means 262 of the adaptive control means 104 are not limited to those shown in FIGS. 16 and 17.

Also, by inserting an isolated point removing means that performs the same operation as the isolated point removing means 102 between the edge detecting means 103 and the adaptive control means 104, adaptive control is performed using the edge information from which the isolated point has been removed. It goes without saying that the adaptive control by the means 104 can effectively remove the noise components in both the moving region and the still region in a state where the moving image is less deteriorated.

[0109]

As is apparent from the above description, the noise removing apparatus of the present invention sets parameters for noise component extraction such as a threshold value of non-linear processing and a feedback coefficient according to the motion amount or edge information of the input video signal. By adaptively controlling the moving area image, it is possible to effectively remove the noise component in both the moving area and the stationary area while the deterioration of the moving area image is small.

[Brief description of drawings]

FIG. 1 is a block diagram of a noise eliminator according to a first embodiment of the present invention.

FIG. 2 is a block diagram of serial-parallel conversion means.

FIG. 3 is a block diagram of parallel-serial conversion means.

FIG. 4 is a block diagram of adaptive control means.

FIG. 5 is a block diagram of a pixel block

FIG. 6 is an input / output characteristic diagram of non-linear processing means 15_1 to 15_k.

FIG. 7 is an input / output characteristic diagram of the first control means 231.

FIG. 8 is an input / output characteristic diagram of the second control means 232.

FIG. 9 is a block diagram of a noise eliminator according to a second embodiment of the present invention.

FIG. 10 is a block diagram of an isolated point removing means 42.

FIG. 11 is a block diagram of a noise eliminator according to a third embodiment of the present invention.

FIG. 12 is a block diagram of a noise eliminator according to a fourth embodiment of the present invention.

FIG. 13 is a block diagram of a noise eliminator according to a fifth embodiment of the present invention.

FIG. 14 is a block diagram of edge detection means 103.

FIG. 15 is a block diagram of adaptive control means 104.

FIG. 16 is an input / output characteristic diagram of the first control means 261.

FIG. 17 is an input / output characteristic diagram of the second control means 262.

FIG. 18 is a block diagram of a conventional noise eliminator.

FIG. 19 is an input / output characteristic diagram of nonlinear processing units 5_1 to 5_k.

[Explanation of symbols]

10, 30, 50, 70, 90 Input terminal 11, 31, 51, 71, 91 Frame delay means 12, 32, 52, 72, 92 First subtraction means 13, 33, 53, 73, 93 Serial parallel conversion means 14, 34, 54, 74, 94 Orthogonal transformation means 15_1 to 15_k, 35_1 to 35_k Non-linear processing means 55_1 to 55_k, 75_1 to 75_k Non-linear processing means 95_1 to 95_k Non-linear processing means 16, 36, 56, 76, 96 Inverse orthogonal transformation Means 17, 37, 57, 77, 97 Parallel / serial conversion means 18, 38, 58, 78, 98 Attenuation means 19, 39, 59, 79, 99 Second subtraction means 20, 40, 60, 80, 100 Output terminals 21, 41, 101 Motion amount detecting means 22, 43, 63, 84 Adaptive control means 42, 82_1, 82_2, 102 Isolation Point removing means 61_1, 61_2, 81_1, 81_2 Motion amount detecting means 62, 83 Motion amount calculating means 103 Edge detecting means 104 Adaptive control means 201-206, 211-216 1 Sample delay means 207, 217 Line delay means 218-224 Addition Means 225 Attenuation means 231 First control means 232 Second control means 241 r Sample delay means 242 s Sample delay means 251 Horizontal edge detection means 252 Vertical edge detection means 253 Edge signal generation means 261 First control means 262 Second Control means

Claims (13)

[Claims] 1. A delay means for delaying a signal in response to an input video signal, a first subtraction means for obtaining a differential signal between the input video signal and an output signal of the delay means, and the first subtraction means. Orthogonal transformation means for performing orthogonal transformation on the obtained difference signal, non-linear processing means for performing non-linear processing on the output of the orthogonal transformation means, and the orthogonal transformation for the output of the non-linear processing means Means for performing an orthogonal inverse transform which is an inverse transform to the orthogonal transform in the means, an attenuating means for attenuating the output of the orthogonal inverse transforming means, and a differential signal between the input video signal and the output signal of the attenuating means. Second subtraction means for obtaining, a motion amount detection means for detecting a motion amount using the output of the orthogonal transformation means, and a threshold value of the non-linear processing in the non-linear processing means based on the output of the motion amount detection means And adaptive control means for adaptively controlling at least one of the attenuation amounts in the attenuating means, the output signal of the second subtracting means being delayed by the delay means, and the output signal of the second subtracting means. Is extracted as a signal from which noise has been removed. 2. The adaptive control means comprises first control means for controlling the threshold value of the non-linear processing in the non-linear processing means based on the output of the motion detecting means, and attenuation amount of the attenuating means based on the output of the motion detecting means. 2. The noise elimination device according to claim 1, further comprising a second control means for controlling the. 3. A delay means for delaying a signal in response to an input video signal, a first subtraction means for obtaining a difference signal between the input video signal and an output signal of the delay means, and the first subtraction means. Orthogonal transformation means for performing orthogonal transformation on the obtained difference signal, non-linear processing means for performing non-linear processing on the output of the orthogonal transformation means, and the orthogonal transformation for the output of the non-linear processing means Means for performing an orthogonal inverse transform which is an inverse transform to the orthogonal transform in the means, an attenuating means for attenuating the output of the orthogonal inverse transforming means, and a differential signal between the input video signal and the output signal of the attenuating means. Second subtraction means for obtaining, motion amount detection means for detecting a motion amount using the output of the orthogonal transformation means, isolated point removal means for removing isolated points from the output of the motion amount detection means, and the isolated point Removal Adaptive control means for adaptively controlling at least one of the threshold value of the non-linear processing in the non-linear processing means and the attenuation amount in the attenuating means based on the output of the stage, and the output signal of the second subtracting means is the delay A noise removing device, wherein the output signal of the second subtracting means delayed by the means is taken out as a noise-free signal. 4. The isolated point removing means recognizes that the motion information output by the motion detecting means is isolated information by comparing it with the motion information of surrounding pixels in the horizontal direction, the vertical direction, or the time axis direction. 4. The noise elimination device according to claim 3, wherein the noise elimination device is composed of a filter that operates so as to calculate a correction value in the case of occurrence. 5. The adaptive control means comprises a first control means for controlling a threshold value of the non-linear processing in the non-linear processing means based on the output of the isolated point removing means, and an attenuating means based on the output of the isolated point removing means. The noise removing device according to claim 3, further comprising a second control unit that controls the amount of attenuation. 6. A delay means for delaying a signal in response to an input video signal, a first subtraction means for obtaining a differential signal between the input video signal and an output signal of the delay means, and the first subtraction means. Orthogonal transformation means for performing orthogonal transformation on the obtained difference signal, non-linear processing means for performing non-linear processing on the output of the orthogonal transformation means, and the orthogonal transformation for the output of the non-linear processing means Means for performing an orthogonal inverse transform which is an inverse transform to the orthogonal transform in the means, an attenuating means for attenuating the output of the orthogonal inverse transforming means, and a differential signal between the input video signal and the output signal of the attenuating means. Second subtraction means for obtaining, a plurality of motion amount detecting means for detecting a motion amount using the difference signal, and a motion amount calculating means for calculating a motion amount based on outputs of the plurality of motion amount detecting means And before An output signal of the second subtraction means, which has adaptive control means for adaptively controlling at least one of the threshold value of the non-linear processing in the non-linear processing means and the attenuation amount in the attenuation means based on the output of the motion amount calculation means. Is delayed by the delay means, and the output signal of the second subtraction means is taken out as a noise-free signal. 7. The adaptive control means controls the threshold value of the non-linear processing in the non-linear processing means based on the output of the motion amount detecting means, and the damping means based on the output of the motion amount detecting means. The noise removing device according to claim 6, further comprising a second control unit that controls the amount of attenuation. 8. A delay means for delaying a signal in response to an input video signal, a first subtraction means for obtaining a difference signal between the input video signal and an output signal of the delay means, and the first subtraction means. Orthogonal transformation means for performing orthogonal transformation on the obtained difference signal, non-linear processing means for performing non-linear processing on the output of the orthogonal transformation means, and the orthogonal transformation for the output of the non-linear processing means Means for performing an orthogonal inverse transform which is an inverse transform to the orthogonal transform in the means, an attenuating means for attenuating the output of the orthogonal inverse transforming means, and a differential signal between the input video signal and the output signal of the attenuating means. Second subtraction means for obtaining, at least one motion amount detecting means for detecting a horizontal motion amount using the difference signal, and at least one isolated point for removing an isolated point for each of the motion amount detecting means Removing means, a motion amount calculating means for calculating a motion amount based on the output of the at least one isolated point removing means, and a threshold value of the non-linear processing in the non-linear processing means and the attenuation based on the output of the isolated point removing means. Means for adaptively controlling at least one of the attenuation amounts in the means, the output signal of the second subtraction means is delayed by the delay means, and the output signal of the second subtraction means is noisy. A noise eliminator characterized in that it is taken out as a removed signal. 9. The isolated point removing means compares the motion information output by the motion amount detecting means with the motion information of surrounding pixels in the horizontal direction or the time axis direction, and recognizes that the information is isolated. 9. The noise eliminator according to claim 8, comprising a filter that operates so as to calculate a correction value. 10. The adaptive control means comprises a first control means for controlling a threshold value of the non-linear processing in the non-linear processing means based on the output of the isolated point removing means, and an attenuating means based on the output of the isolated point removing means. 9. The noise removing device according to claim 8, further comprising a second control unit that controls the amount of attenuation of the noise. 11. A delay means for delaying a signal in response to an input video signal, a first subtraction means for obtaining a difference signal between the input video signal and an output signal of the delay means, and the first subtraction means. Orthogonal transformation means for performing orthogonal transformation on the obtained difference signal, non-linear processing means for performing non-linear processing on the output of the orthogonal transformation means, and the orthogonal transformation for the output of the non-linear processing means Means for performing an orthogonal inverse transform which is an inverse transform to the orthogonal transform in the means, an attenuating means for attenuating the output of the orthogonal inverse transforming means, and a differential signal between the input video signal and the output signal of the attenuating means. Second subtraction means for obtaining, motion amount detecting means for detecting a motion amount using the output of the orthogonal transforming means, and edge amount detecting means for detecting an edge component of image information using the output of the orthogonal transforming means And adaptive control means for adaptively controlling at least one of the threshold value of the non-linear processing in the non-linear processing means and the attenuation amount in the attenuating means based on the output of the motion amount detecting means and the output of the edge amount detecting means. Then, the output signal of the second subtraction means is delayed by the delay means, and the output signal of the second subtraction means is taken out as a signal from which noise has been removed. 12. The edge amount detecting means detects a horizontal edge component of the image information in the horizontal direction by using the output of the orthogonal transforming means, and the vertical edge of the image information by using the output of the orthogonal transforming means. The noise according to claim 11, further comprising vertical edge amount detecting means for detecting an edge component in a direction, and edge amount calculating means for calculating an edge amount from an output of the horizontal edge amount detecting means and an output of the vertical edge amount detecting means. Removal device. 13. The first control means for controlling the threshold value of the non-linear processing in the non-linear processing means based on the output of the motion amount detecting means and the output of the edge amount detecting means.
The noise elimination device according to claim 11, further comprising: second control means for controlling the attenuation amount of the attenuation means based on the output of the motion amount detection means and the output of the edge amount detection means.