JPH05328310A - Noise reducder for muse decoding processing - Google Patents

Abstract

PURPOSE:To improve an SN ratio and to prevent the generation of blurr in the case of in a moving image part laminance applying noise reducing processing to a video signal based upon a MUSE system. CONSTITUTION:Output laminance signals Y and chroma signals C from a still image processing means 8 and a moving image processing means 11 are mixed by a mixing processing means 13 in accordance with a movement signal outputted from a moving area detecting means 7. A low band area of an output brigtness signal Y from the means 13 is substituted for a low band area of an output brightness signal outputted from an input processing means 3 by a low band substituting means 14. An output luminance signal Y from the means 14 and an output chroma signal C from the means 13 are adaptively noise-reduced by an adaptive noise reducer 26 capable of changing the noise reducing quantity of the moving signal so that the noise reducing quantity of a still image part is increased and that of a moving image part is reduced.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a MUSE decoder used in a high-definition receiver and, more particularly, to noise reduction processing for the MUSE decoder.

[0002]

2. Description of the Related Art M is one of the high definition broadcasting systems.
The USE method was developed by NHK. This MUSE
The method is “NHK technology research” 1987. As reported in the paper "Development of MUSE system" by Yuichi Ninomiya in Volume 39, No. 2, it transmits wide band high-definition signals by band compression using subsampling. In order to receive the transmission signal, a MUSE decoder for returning the band-compressed HDTV signal to the original broadband HDTV signal is required.

This MUSE decoder is described, for example, in "Hi-Vision MUSE Satellite Broadcasting, Hi-Vision Technology, September 12, 1991, NHK Engineering Service", which will be described with reference to FIG. However, 1 is the MUSE signal input terminal, 2 is the LPF
(Low-pass filter) 3, input processing means, 4 A / D
(Analog / digital) conversion means, 5 de-emphasis processing means, 6 inverse transmission gamma processing means, 7 motion area detection means, 8 still image processing means, 9 inter-frame interpolation means, 10 inter-field interpolation Insertion means, 11 is a moving image processing means, 12 is an intra-field interpolation means, 13 is a mix processing means, 14 is a low band replacement means, 15 is a control signal generation means, 16 is an audio signal processing means, and 17 is TCI (Time).
Compressed Integration) decoder, 18 is output processing means, 19 is inverse matrix means, 20 is gamma processing means, 21 is D / A (digital / analog) conversion means, 22 is LPF, 23, 24, 2
Reference numeral 5 is an output terminal.

In the figure, the MUSE signal input from the input terminal 1 is supplied to the input processing means 3 through the LPF 2. In the input processing means 3, this MUSE signal is
After being converted into a digital signal by the D / D conversion means 4, it is input-processed by the de-emphasis processing means 5, the reverse transmission gamma processing means 6 and the like, and the motion area detection means 7, the still image processing means 8, the moving image processing means 11 and It is supplied to the low frequency replacement means 14. Further, the digital MUSE signal output from the A / D conversion means 4 is supplied to the control signal generation means 15 and the audio signal processing means 16, and the control signal generation means 15 generates the control signal necessary for each processing in the MUSE decoder. Processing such as demodulation of voice data is performed.

In the moving area detecting means 7, the digital MU is used.
A moving portion (moving portion) of an image in the SE signal is detected, and a signal indicating the presence or absence of this moving portion (hereinafter referred to as a motion signal) is formed and output. Further, in the still image processing means 8, the time-division multiplexed luminance signal Y and the chroma signal C of the digital MUSE signal are imaged in four fields by the inter-frame interpolation means 9, the inter-field interpolation means 10, etc., respectively. The signal is interpolated, demodulated into the original wideband video signal, and supplied to the mix processing means 13. In the moving image processing means 11, the original broadband video signal is demodulated by the field interpolation means 12 and the like into the luminance signal Y and the chroma signal C, which are similarly time-division multiplexed, and are supplied to the mix processing means 13.

In the mix processing means 13, the still image processing signal from the still image processing means 8 and the moving image processing means for each of the luminance signal Y and the chroma signal C according to the motion signal from the motion area detecting means 7. The video processed signal from 11 is mixed. That is, the moving image processing signal is adaptively mixed in a region where the video signal has a motion and the still image processing signal is increased in other regions. Mix processing means 1
3, the luminance signal Y and the chroma signal C are separately output, and only the luminance signal Y is supplied to the low frequency band replacement means 14,
The low-frequency component of the luminance signal Y is replaced with the low-frequency component of the luminance signal of the video signal in the digital signal output from the input processing means 3, and a new luminance signal Y is obtained.
It is supplied to the TCI decoder 17 together with the chroma signal C.

Here, the chroma signal C is a line-sequential signal in which two color difference signals R-Y and B-Y are alternately time-division multiplexed for each line (scanning line). The color difference signals RY and BY are separated from the chroma signal C, line interpolation is performed, and the continuous color difference signals RY and BY are supplied to the output processing means 18 together with the luminance signal. In the output processing means 18, first, the luminance signal Y and the two color difference signals RY and BY are supplied to the inverse matrix means (MTX) 19 to generate three primary color signals R, G and B. Is subjected to gamma correction by the gamma processing means 20, and then converted into an analog signal by the D / A conversion means 21 and output. These analog primary color signals R, G and B are output terminals 2 via LPF 22 respectively.
It is output from 3, 24, and 25.

In such a conventional MUSE decoder,
For noise reduce, the still image processing means 8 is provided with a noise reducer, and the difference between the current video field, the video field one frame before and the video field two frames before is detected as noise, and this detection output It is known that noise reduction is performed by performing a coring process on the signal and subtracting it from the video signal of the current field.

[0009]

By the way, in the above-mentioned conventional example, the noise reduction is performed by the inter-frame interpolation processing of the still image processing. This is due to the high inter-frame correlation of the still image and the noise reduction. It utilizes that there is no correlation. However, noise-reduced still images are mixed with moving images that have not been subjected to such processing, and luminance signals that have not been subjected to noise-reduction processing are replaced in the low frequency region, so noise-reduction processing is performed. The image part is only in the high frequency region of the still image, and there is no noise reduction effect in the flat (low frequency region) still image part of the image. If noise occurs in the low frequency region of the replaced luminance signal, In a flat still image part, there was a problem that noise was very noticeable.

It is an object of the present invention to provide a noise reducer which solves such a problem and makes noise less noticeable even in a flat still image portion of an image.

[0011]

In order to achieve the above object, the present invention provides an adaptive noise reducer at a stage subsequent to the low frequency substituting processing means, and reduces the noise according to the movement detected by the moving area detection processing means. Control the amount.

[0012]

[Function] The adaptive noise reducer adaptively controls the noise reducer to reduce the noise reduce amount in the moving image area and increase the noise reduce amount in the still image area with respect to the low-pass substituting video signal. Apply processing. As a result, noise reduction will be applied to the part that was not subjected to noise reduction processing in the conventional example, and in the still image area, the noise reduction effect will be large and the noise in this part will not be conspicuous. , In the moving image area, the noise reducing effect is small and the image is not blurred.

[0013]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of a MUSE decoder according to the present invention. Reference numeral 26 is an adaptive noise reducer, and parts corresponding to those in FIG.

In the figure, the conventional MUS shown in FIG. 7 is used.
Similar to the E decoder, the noise-reduced luminance signal Y from the still image processing means 8 and the luminance signal Y from the moving image processing means 11, and the chroma signal C from the still image processing means 8 and the moving image processing means 11 are also included. The chroma signal C from
The low frequency component of the luminance signal output from the mix processing unit 13 is adaptively mixed by the mix processing unit 13 according to the motion signal from the motion region detecting unit 7, and further input by the low frequency substituting unit 14. The low-frequency component of the luminance signal in the output digital signal of the means 3 is replaced, and the adaptive noise reducer 26 is provided at the subsequent stage of the low-frequency replacement means 14, and the low-frequency replacement means 1 is provided.
The luminance signal Y outputted from the No. 4 and the chroma signal C outputted from the mix processing means 13 are supplied.

The adaptive noise reducer 26 operates according to the motion signal output from the motion area detecting means 7,
Noise reduction processing is adaptively performed so that the noise reduction amount is small in a region where the image motion is large and the noise reduction amount is large in a region where the image motion is small. In this case, the luminance signal Y and the chroma signal C are again time-division multiplexed and input to the adaptive noise reducer 26, and noise reduction processing is performed in a time division manner.

The luminance signal Y and the chroma signal C, which have been adaptively noise-reduced in the entire screen and in the entire frequency range by the adaptive noise reducer 26, are separated again, and by means of the TCI decoder 17 and the subsequent means shown in FIG. The same processing as that of the MUSE decoder is performed, and the original wideband high-definition video signal is output from the output terminals 23, 24 and 25.

As described above, the adaptive noise reducer 2
In No. 6, the noise reduction process is performed adaptively so that the noise reduction amount in the moving image region is small and the noise reduction amount is large in the still image region for the Y signal subjected to the low frequency substituting process. Noise reduction is performed by the adaptive noise reducer 26 even on a portion not subjected to noise reduction processing other than the high frequency area of the still image area due to the processing and low frequency substituting processing, and noise is reduced in the still image area. Noise is not noticeable in this part because the reduce effect is large, and in the moving image area, the noise reduce effect is small and the video signal is not blurred.

FIG. 2 is a block diagram showing a specific example of the adaptive noise reducer 26 shown in FIG. 1, in which 201 is an input terminal of a video signal and 202 is a moving area detecting means 7 (FIG. 1).
The input terminal of the motion signal output from the device, 203 is a multiplier that changes the constant K, 204 is an adder, 205 is a multiplier that changes the constant K, and 206 is a 1-frame delay in the case of a luminance signal Y, and a color difference In the case of the signal C, a frame delay means for obtaining a delay of 2 frames, and 207 is an output terminal.

In FIG. 2, the adaptive noise reducer 26 time-divisionally noise-reduces the luminance signal Y and the chroma signal C as described above. Where first
Assuming that the luminance signal Y is input from the input terminal 201, it is supplied to the multiplier 203 and multiplied by a constant K (0 <K <1). This constant K changes according to a motion signal from the motion area detection means 7 (FIG. 1) input from the input terminal 202, and is large in a moving image area with motion and small in a still image area without motion. For example, K = 0.9 in a completely moving region and K = 0.1 in a still image region without moving. For example, the constant K is digitally determined according to the amount of motion in the image. The constant K may be changed in steps, or the constant K may be changed in an analog range of 0 <K <1 in accordance with the amount of movement.

The output luminance signal of the multiplier 203 is supplied to the adder 204. The adder 204 adds the output signal of the multiplier 205 and supplies it to the TCI decoder 26 of FIG.
06 and is delayed by one frame, and the multiplier 205
Is multiplied by a constant (1-K) and supplied to the adder 204. Like the constant K of the multiplier 203, the value K of this constant (1-K) is also controlled according to the motion signal from the input terminal 202.

Here, in the loop consisting of the adder 204, the frame delay means 206, and the multiplier 205, the K-multiplied input luminance signal Y is further multiplied by (1-K) and then the next K-multiplied input luminance signal. Since it is added to Y, and such addition is sequentially performed, the output signal of the multiplier 205 is a noise-reduced signal. When the constant K is small, this noise-reducing effect is large. Therefore, in the still image area where there is no motion, the constant K is small as described above, and in this case, the luminance signal obtained at the output terminal 207 is sufficiently suppressed in noise. Noise is sufficiently suppressed in the area and becomes inconspicuous. On the other hand, in the moving image area where there is a motion, since the constant K is large as described above, the luminance signal obtained at the output terminal 207 is mainly the input signal Y multiplied by K, and therefore the noise reducing effect is small. In this case, there is no deterioration in resolution due to the noise reduce process, and in this case, even if there is noise, it is hardly noticeable, so that a high-quality image without blur can be obtained.

The above is the case of the luminance signal, but when the chroma signal C is input from the input terminal 201, the two input signals of the adder 204 have a correlation as the delay amount of the frame delay means 206. Therefore, it is set to 2 frames. Other than this, it is the same as the case of the above luminance signal. Therefore, the frame delay means 206 has a delay means for one frame and a delay means for two frames, and switches the delay means used for the luminance signal and the chroma signal.

FIG. 3 is a block diagram showing another specific example of the adaptive noise reducer 26 in FIG.
Reference numerals 1 and 302 are subtractors, and 303 is a multiplier. The parts corresponding to those in FIG.

In the figure, the input luminance video signal Y from the input terminal 201 is supplied to the subtractor 301 and also to the subtractor 302, and the frame delay means 20.
The output signal of 6 is subtracted. These difference signals are multiplied by the multiplier 3
The constant K is multiplied by 03, supplied to the subtractor 301, and subtracted from the input luminance signal Y. The output luminance signal of the subtractor 301 is supplied from the output terminal 207 to the TCI decoder 17 of FIG. In this case, the delay amount of the frame delay means 206 is one frame as in the specific example of FIG. 2, but when the input signal from the input terminal 201 is the chroma signal C, it is two frames. Needless to say.

Here, the constant K in the multiplier 303 changes in the range of 0 <K ≦ 0.5 according to the motion signal from the input terminal 202. Then, the multiplier 303 is set such that the constant K becomes a value closer to 0 as the amount of motion is larger in the moving image area where there is a motion, and the constant K becomes a value closer to 0.5 in the still image area where there is no motion. Is controlled. As a result, the same effect as the specific example shown in FIG. 2 can be obtained.

The specific example of the adaptive noise reducer 26 shown in FIG. 1 has been described above.
The adaptive noise reducer 26 is not limited to this specific example, and may have any configuration as long as it adaptively performs noise reduce processing according to the amount of motion of the image.

As described above, according to this embodiment, a still image and a moving image are adaptively mixed, and a noise reduction process is adaptively applied to a luminance signal and a chroma signal that have undergone replacement processing. As a result, noise reduction processing can be performed according to the still image area and the moving image area of the image, and it is possible to obtain an image signal with good S / N and less blurring of the image.

FIG. 4 is a block diagram showing another embodiment of the MUSE decoder according to the present invention, in which 27 is a delay means, parts corresponding to those in FIG. ..

In FIG. 4, the motion signal output from the motion area detecting means 7 is delayed by the delay means 27 and then supplied to the adaptive noise reducer 26. Other than this, it is the same as the embodiment shown in FIG. The luminance signal Y supplied to the adaptive noise reducer 26 with respect to the motion signal output from the motion area detecting unit 7 by the mix processing in the mix processing unit 13 according to the motion signal and the processing in the low frequency band substituting unit 14. , The chroma signal C is delayed, but the delay means 27 delays the motion signal by this delay so as to eliminate the time difference between the luminance signal Y, the chroma signal C and the motion signal supplied to the adaptive noise reducer 26. ing.

As a result, the motion signal indicating the motion amount of the video signal is supplied to the adaptive noise reducer 26 in the same phase as the luminance signal Y and the chroma signal C, and the motion of the image is performed by the adaptive noise reducer 26. The noise reduce process according to is more accurately performed than the embodiment shown in FIG.

FIG. 5 is a block diagram showing another embodiment of the MUSE decoder according to the present invention.
Reference numeral 6b is an adaptive noise reducer, and the portions corresponding to those in FIG.

In the figure, two adaptive noise reducers 26a and 26b are used, and the luminance signal Y from the low frequency band replacement means 14 is supplied to the adaptive noise reducer 26a.
The adaptive noise reducer 26b is mixed with the mix processing means 13
Are supplied with the chroma signals C from each. Other than this, it is the same as the embodiment shown in FIG. The adaptive noise reducers 26 a and 26 b adaptively perform noise reduction processing for each of the luminance signal Y and the chroma signal C in accordance with the motion signal from the motion area detection means 7, and supply the noise reduction processing to the TCI decoder 17.

As a result, the luminance signal Y in which the still image and the moving image are adaptively mixed, and the low frequency region is subjected to the replacement processing, and the chroma signal in which the still image and the moving image are adaptively mixed are respectively provided. Noise reduction processing can be performed according to the above.

The adaptive noise reducers 26a, 2
When, for example, the configuration shown in FIG. 2 is used as 6b, the frame delay means 206 in each of the adaptive noise reducers 26a and 26b is composed of one delay means, unlike the embodiment shown in FIG. The delay amount of the frame delay unit 206 of the adaptive noise reducer 26a is one frame, and the delay amount of the frame delay unit 206 of the adaptive noise reducer 26b is two frames.

As described above, also in this embodiment, similarly to the embodiment shown in FIG. 1, a still image and a moving image are adaptively mixed and further applied to a video signal subjected to a low frequency replacement process. Since the noise reduce process can be performed effectively, the noise reduce process can be performed according to the still image region and the moving image region of the image, and the S / N is good and the image signal with less blur of the image in the moving image region can be obtained. It becomes possible to obtain. Further, in this embodiment, since the noise reduction processing can be separately applied to the luminance signal and the chroma signal, the adaptive noise reducer 26 for the luminance signal Y is obtained.
The noise reduction amount K at a and the noise reduction amount K at the adaptive noise reducer 26b for the chroma signal C can be appropriately made different, and each signal can be subjected to noise reduction processing according to the noise. Become.

As in the embodiment shown in FIG. 4, the delay means delays the motion signals supplied to the adaptive noise reducers 26a and 26b with respect to the motion signals supplied to the mix processing 13. Needless to say, good things.

FIG. 6 is a block diagram showing still another embodiment of the MUSE decoder according to the present invention, in which 26 'is an adaptive noise reducer, and the parts corresponding to those in FIG. The description is omitted.

In the figure, an adaptive noise reducer 26 'is provided between the TCI decoder 17 and the output processing means 18. The other parts are similar to those of the embodiment shown in FIG. The luminance signal Y from the low frequency substituting means 14 and the chroma signal C from the mix processing means 13 are supplied to the TCI decoder 17, and the luminance signal Y generated here and two color difference signals RY and BY are obtained. Are supplied to the adaptive noise reducer 26 '. As the adaptive noise reducer 26 ', for example, the one configured as shown in FIG. 2 is provided for each of the luminance signal Y, the two color difference signals RY, BY, and the movement from the movement area detecting means 7 is performed. TC depending on the signal
The luminance signal Y and the color difference signal R output from the I decoder 17
Noise reduction processing similar to that in the above-described embodiment is performed for each of −Y and BY. These noise-reduced signals are supplied to the output processing means 18.

As described above, also in this embodiment, as shown in FIG.
In the same manner as in the embodiment shown in Fig. 4, the still image and the moving image are adaptively mixed, and the noise reduction process can be adaptively applied to the video signal subjected to the low-frequency replacement processing, so that the still image It is possible to perform noise reduction processing according to the image area and the moving image area, and it is possible to obtain a video signal with good S / N and less blurring of the image in the moving image area.

Also in this embodiment, the motion signal supplied to the adaptive noise reducer 26 'may be delayed by the delay means as in the embodiment shown in FIG. 4, or the adaptive noise reducer 26' may be used. In ′, the noise reduction amount may be different between the luminance signal Y and the color difference signals RY and BY.

[0041]

As described above, according to the present invention,
Noise reduction processing according to the good motion of the video can be performed over the entire video signal, the S / N in the still image area where noise is easily noticeable is significantly improved, and the deterioration of resolution in the video area can be prevented and the video can be prevented. It is possible to eliminate the blurring of a part and obtain a high quality image as a whole.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of a MUSE decoder according to the present invention.

FIG. 2 is a configuration diagram showing a specific example of an adaptive noise reducer in FIG.

3 is a configuration diagram showing another specific example of the adaptive noise reducer in FIG. 1. FIG.

FIG. 4 is a block diagram showing another embodiment of the MUSE decoder according to the present invention.

FIG. 5 is a block diagram showing another embodiment of the MUSE decoder according to the present invention.

FIG. 6 is a block diagram showing another embodiment of the MUSE decoder according to the present invention.

FIG. 7 is a block diagram showing an example of a conventional MUSE decoder.

[Explanation of symbols]

1 MUSE signal input terminal 3 input processing means 7 motion area detection means 8 still image processing means 11 moving image processing means 13 mix processing means 14 low-frequency replacement means 17 TCI decoder 18 output processing means 23, 24, 25 output terminals 26, 26a , 26b, 26 'Adaptive noise reducer 27 Delay means

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Masato Sugiyama, inventor 292 Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa Inside the Hitachi Media Visual Media Laboratory (72) Inventor Kentaro Teranishi, 292 Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa Hitachi, Ltd., Visual Media Laboratory (72) Inventor, Hatsuji Kimura, 292 Yoshida-cho, Totsuka-ku, Yokohama, Kanagawa Prefecture, Ltd., Visual Media Laboratory, Hitachi, Ltd. (72) Inventor Kei Hasegawa, 4 Kanda Surugadai, Chiyoda-ku, Tokyo 6-chome Hitachi, Ltd. (72) Inventor Yuichi Ninomiya 1-10-11 Kinuta, Setagaya-ku, Tokyo Inside the Japan Broadcasting Corporation Broadcasting Technology Laboratory

Claims (4)

[Claims] 1. A still picture region processing means for performing inter-field interpolation processing in units of at least 4 fields on each of a luminance signal and a line-sequential color difference signal of an input MUSE signal, and decoding the original wide band still picture signal. A moving image region processing means for performing field interpolation processing for each of the luminance signal and the line-sequential color difference signal of the input MUSE signal, and decoding the original wideband moving image signal, and a luminance signal from the still region processing means. , A still image area signal for each line-sequential color difference signal, a luminance signal from the moving area processing means, and a moving image area signal for each line-sequential color difference signal.
A mixing unit that adaptively mixes the luminance signal in the SE signal according to the amount of movement between frames or fields, and the low-frequency component of the luminance signal output from the mixing unit are included in the input MUSE signal. The still image is provided with low-pass substituting means for substituting the low-frequency component of the luminance signal or mixing at a predetermined ratio, and line-sequential decoding means for converting the line-sequential color difference signal from the mixing means into a simultaneous color-difference signal. The MU that decodes the MUSE signal obtained by band-compressing the high-definition signal into the original wide-band high-definition signal by performing noise reduction processing by the area processing means.
A noise reducer for MUSE decoding processing, characterized in that, in the SE decoder, a noise reducer utilizing correlation between frames or fields is provided after the low-pass replacing means. 2. The noise reducer for MUSE decoding processing according to claim 1, wherein a correlation coefficient of the noise reducer is adaptively controlled according to the motion amount. 3. The noise reducer according to claim 1, wherein the noise reducer has a first noise reducer whose correlation coefficient is adaptively controlled according to a motion amount between one frame or one field, and between one frame. Alternatively, it comprises a second noise reducer whose correlation coefficient is adaptively controlled according to the amount of movement between two frames, and the luminance signal output from the low-pass replacement processing means is supplied to the first noise reducer. The noise reducer for MUSE decoding processing, wherein the line-sequential color difference signal output from the mixing means is supplied to the line-sequential decoding means via the second noise reducer. 4. The noise reducer according to claim 1, wherein the noise reducer is a first noise reducer whose correlation coefficient is adaptively controlled according to a motion amount between one frame or one field, and between one frame. Alternatively, it comprises a second noise reducer whose correlation coefficient is adaptively controlled according to the amount of movement between two frames, and the first noise reducer is provided at a stage subsequent to the low-pass replacement processing means, and the second noise reducer is provided. A noise reducer is provided in the subsequent stage of the line-sequential decoding means.
Noise reducer for decoding processing.