JP2646965B2 - MUSE decoder motion detection circuit - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION The present invention relates to a band-compressed MUSE.
More specifically, the present invention relates to a MUSE decoder for demodulating a signal into an original wideband high-definition television signal and reproducing the signal, and more particularly, to a MUSE decoder which shares a circuit of an edge detection process which is a part of a motion detection process and a video / video signal reproduction process. The present invention relates to a motion detection circuit.

[0002]

2. Description of the Related Art MUSE (Muitiple Sub-Nyquist Sampling) for applying a sub-Nyquist sample that goes around the original high-definition television signal in four fields as a method of band-compressing a wideband high-definition television signal to a practical level for transmission.
Encoding) method.

[0003] The MUSE system is a system developed by NHK (Japan Broadcasting Corporation), and various documents (eg, December 1, 1990)
The detailed description is omitted since it is described in "MUSE-High-Vision Transmission System" by Yuichi Ninomiya published by IEICE.

In the MUSE system, a high-definition television signal (hereinafter referred to as a "baseband signal") having a band of up to about 22 MHz for a luminance signal (hereinafter referred to as a "Y signal") and about 7 MHz for a chrominance signal (hereinafter referred to as a "C signal"). With a bandwidth of 27 MHZ
Approximately 8 MHz for transmission on one channel of satellite broadcasting
Bandwidth compression processing. The MUSE decoder demodulates the band-compressed signal (hereinafter, referred to as “MUSE signal”) into an original baseband signal.

FIG. 6 is a block diagram showing a configuration of a conventional MUSE decoder.

The input terminal 1 is supplied with a MUSE signal.
After the band is limited by the LPF 2 having a cutoff frequency of 8.1 MHz, the digital signal is sampled by the A / D converter 3 using a clock signal of 16.2 MHz. The output signal of the A / D converter 3 is input to an audio decoding unit 4, a synchronization processing unit 5, and a de-emphasis processing unit 6.

The audio decoder 4 separates and decodes the audio signal from the input signal, and outputs an audio 4-channel independent data signal 7.

In the synchronization processing section 5, a synchronization signal 8 and a control signal 9 are separated from an input signal and supplied to each signal processing section.

[0009] The de-emphasis processing section 6 performs processing reverse to the emphasis in the MUSE encoder (transmission side processing). The signal that has been subjected to the de-emphasis processing is subjected to inverse processing of the transmission gamma on the encoder side by the inverse gamma correction processing unit 10, and becomes a linear signal 11.

The signal 11 after the inverse gamma correction processing is input to a motion detection processing section 12, a field interpolation processing section 13, and an inter-frame interpolation processing section 14.

The motion detection processing section 12 detects a moving image area using the signal of the frame difference, and generates a motion signal 15 which is the basis of the mixing ratio of the still image area and the moving image area in the MIX processing section 18. The detailed operation of the motion detection processing will be described later.

The field interpolation section 13 interpolates the moving image area in each field, and demodulates data at a rate of 16 MHz into data at a rate of 32 MHz. The field interpolation processing of the moving image area will also be described later. The signal interpolated in the field is converted by the sampling frequency converter 17 to 32
The MHZ rate is converted into a 48 MHz rate sampling signal, which is input to the MIX processing unit 18.

The interpolating section 14 performs interpolation by interpolating missing samples in the still image area with the samples of the previous frame. 16M by frame interpolation
The data at the Hz rate is data at the 32 MHz rate. The Y signal after the frame interpolation processing is passed through an LPF 19 having a cutoff frequency of 12 MHz in order to prevent aliasing interference occurring at the time of sampling frequency conversion in the next processing.

After passing through the LPF, the sampling frequency is converted from a 32 MHz rate to a 24 MHz rate in order to perform field interpolation. 24 MHZ at sampling frequency converter 20
Data converted to z-rate is interpolated by field interpolator 2
In step 1, the missing sample is interpolated with the sample of the previous field to produce a sampling signal of a 48 MHz rate.
It is input to the processing unit 18.

The MIX processing unit 18 converts the moving image signal from the sampling frequency conversion unit 17 and the still image signal from the inter-field interpolation processing unit 21 into a motion signal 15 from the motion detection processing unit 12.
Mix according to.

The output signal of the MIX processing unit 18 is obtained by separating a line-sequentially scanned color signal into an RY signal 23 and a BY signal 24 by a TCI decoding unit 22 and calculating a Y signal 25 by an inverse matrix unit 26. The signal is converted into three signals of an R signal, a G signal, and a B signal. These three signals are converted into analog signals by a D / A converter 27, and output as an HD signal of an R signal 29, a G signal 30, and a B signal 31 through an LPF 28 having a cutoff frequency of about 21 MHz.

Next, the motion detection processing and the field interpolation processing will be described. Various documents (for example, Electronics Life 19 issued by Japan Broadcasting and Publishing Association) describe the method of motion detection and field interpolation of the conventional MUSE method.
January 1990, p.38-48 "How HDTV receivers work")
Introduced in

FIG. 7 is a block diagram showing the configuration of a conventional motion detection process. The signal 11 subjected to the inverse gamma correction processing is input to three blocks of an edge detection processing unit 37, a two-frame difference detection processing unit 32, and a one-frame difference detection processing unit 33 in the motion detection processing unit 12.

The edge detection processing section 37 detects a signal of an edge portion corresponding to the contour of the object. FIG. 8 is a schematic diagram of edge detection. Since the input signal has a 16 MHz rate, there are points having data (hereinafter, referred to as "sample points") and points having no data (hereinafter, referred to as "non-sample points") at every 32 MHz on the same line. Is performed in two different ways.

FIGS. 8A and 8B show the detection at the sample point and the non-sample point of the horizontal edge detection, and FIG. 9 shows the detection at the sample point and the non-sample point of the vertical edge detection, respectively. (A) and (B) show. Since the four methods are similar, a description will be given using the horizontal edge detection at the sample points shown in FIG. 8A as a representative. The six image data of FIG.
When A6, | A1-A2 |, | A3-A4 |, | A5
-A6 | is determined, and the maximum value of the three values is set as the value of the horizontal edge of A4.

[0020] That is, the horizontal edge values of _{A 4 = MAX {| A1-} A2 |, | A3-
A4 |, | A5-A6 |｝.

8 (B) and FIGS. 9 (A) and 9 (B), the values of the vertical edge and the horizontal edge for each pixel are similarly obtained.

The horizontal edge value of B7 = MAX ｛| B1−B2 |, | B3−B4
|, | B5-B6 |｝ Vertical edge value of C4 = MAXＭＡ | C1-C2 |, | C3-C4
|, | C4-C5 |｝ Vertical edge value of D6 = MAX ｛| D1-D2 |, | D2-D3
|, | D4-D5 |｝

The two-frame difference detection processing unit 32 obtains a difference signal between two frames, and performs motion detection corresponding to a slow motion (2 frames = 1/15 second).

The one-frame difference detection processing unit 33 obtains a difference signal between one frame and a motion faster than the two-frame difference (1
The motion detection corresponding to (frame = 1/30 second) is performed.

The detected three types of signals, the edge signal 16, the two-frame detection signal 34, and the one-frame detection signal 35 are input to a moving image area detection processing unit 36. The moving image area detection processing unit 36 adjusts the sensitivities of the one-frame difference detection signal 35 and the two-frame difference detection signal based on the edge signal 16, takes the maximum value of the two signals for each pixel, and outputs the result as the motion signal 15. I do.

FIGS. 10A and 10B are schematic diagrams showing an example of a conventional field interpolation process. FIG. 10A shows a process of interpolating a sample point (interpolated point E7), and FIG. 10B shows a process of interpolating a non-sampled point (interpolated point F7) from surrounding sample points. Since the two methods are similar, FIG.
The field interpolation process at the sample points shown in FIG. The 13 image data shown in FIG.
When E13, the processing of the function f1 is performed on all data to generate E7 data, and interpolation of sample points is performed. In the same manner as in FIG. 10B, interpolation of non-sample points is performed.

As described above, the motion detection processing unit 12 and the edge detection processing unit 37 and the field interpolation unit 13 perform processing using the same data. Comparing the image data of FIGS. 8A and 10A, A1 and E4, A2 and E9, A3 and E
2, A4 and E7, A5 and E5, A6 and E10 are the same data. 8 (B) and 10 (B), FIG. 9 (A) and FIG. 10 (A), FIG.
The same applies to (B) and FIG. 10 (B).

[0028]

As described above, the conventional MU
The SE decoder has independent processing systems even though the motion detection processing unit and the field interpolation processing unit perform similar processing. For this reason, the circuit scale becomes large and it is difficult to reduce the circuit size.

Therefore, in the present invention, the same function circuit is shared between the edge detection processing unit as a part of the motion detection processing and the field interpolation processing unit as a part of the video / video signal reproduction processing circuit. The purpose of the present invention is to reduce the circuit scale and increase the accuracy of edge detection.

[0030]

To accomplish the above object means to provide a process, the motion detection circuit of MUSE decoder for reproducing and demodulating the MUSE signal band-compressed original broadband to high definition television signal, the frame The difference signal
Motion detection processing means for detecting a moving image area using the
In the field for performing interpolation of the image data in the field
And interpolation means.
From the difference value of the image data in a predetermined range within the
Edge detection means for obtaining the edge signal;
Characterized in that share each other between said field interpolation unit section, and constitutes the edge detection means so as to provide only one of the motion <br/>-out detection processing means and interpolation means in said field And a MUSE decoder motion detection circuit.

The present invention also provides a MUSE decoder motion detection circuit, wherein the one or more circuit blocks are at least line memories for image data.

[0032]

In the MUSE decoder motion detection circuit according to the present invention, the same functional parts as the edge detection means included in the motion detection processing means and the field interpolation means for performing field interpolation for reproducing a moving picture signal are performed. , The circuit scale can be reduced.

Further, an edge signal can be obtained from image data in a wide range in a field and image data at an equal distance from the surroundings without increasing the circuit scale.
This makes it possible to improve the detection accuracy of the edge signal required for each pixel.

Further, by making the shared portion a line memory for image data, the circuit required for edge detection can be reduced to about 30%.

[0035]

Next, first and second embodiments of the present invention will be described sequentially with reference to the drawings.

First, a first embodiment will be described with reference to FIGS. 1, 2 and 3.

FIG. 1 is a block diagram of a MUSE decoder of the present invention, and FIG. 2 is a block diagram of a motion detection processing section of the MUSE decoder of the present invention. Both block diagrams are the same as the block diagrams showing the conventional example of FIGS. 6 and 7 except that the edge signal 16 from the field interpolation unit 13 is input to the motion detection processing unit 12. Therefore, the description of the same parts as in the related art will be omitted, and different parts will be mainly described.

As described in the conventional example, the motion detection processing unit
The edge detection processing unit 37 in 12 and the field interpolation unit 13 perform processing using the same data. Therefore, in the present invention, as shown in FIG. 2, the edge detection processing unit 37 shown in FIG. 7 of the conventional example is deleted, and the edge signal 16 is converted using the image data used in the processing by the field interpolation unit 13. The calculation process for obtaining is performed.

FIG. 3 is a schematic diagram showing edge detection in the field interpolation unit 13 according to the first embodiment of the present invention. FIG. 3A shows sample points and FIG. 3B shows non-sample points. )
Shown in Since the two methods are similar, the edge detection processing at the sample points shown in FIG. When the 13 pieces of image data in FIG. 3A are E1 to E13, | E4−E9 |, | E2−E7 |, and | E5−E10 | are obtained, and the maximum value of the three values is set to E7. Horizontal edge value, | E4−
E5 |, | E6-E7 |, | E7-E8 | are obtained, and the maximum value of the three values is set as the value of the vertical edge of E7.

That is, the horizontal edge value of E7 = MAX ｛| E4-E9 |, | E2-E7
|, | E5-E10 |｝ Vertical edge value of E7 = MAXＭＡ | E4-E5 |, | E6-E7
|, | E7−E8 |｝.

Similarly, the values of the horizontal edge and the vertical edge for each pixel are obtained for FIG.

The horizontal edge value of F7 = MAX ｛| F1-F6 |, | F4-F10
|, | F2-F8 |｝ Vertical edge value of F7 = MAX ｛| F3-F4 |, | F4-F5
|, | F6-F8 |｝

As described above, the edge signal 16 detected by using the image data of the field interpolation unit 13 is
The signal 34 from the two-frame difference detection processing unit 32 and the one-frame difference detection processing unit
It is configured to generate the motion signal 15 in combination with the signal 35 from 33.

Therefore, according to the first embodiment, the circuit of the edge detection processing unit 37 can be shared with the field interpolation unit 13, so that the circuit scale can be reduced. The shared part in the embodiment corresponds to, for example, a line memory for image data.

Next, a second embodiment will be described with reference to FIGS.

As can be seen from a schematic diagram showing an example of the conventional field interpolation (FIG. 10) and the edge detection processing (FIGS. 8 and 9), the field interpolation is more than the edge detection processing. Using the data. By sharing the circuit of the edge detection processing unit with the field interpolation unit, in the first embodiment, the conventional edge detection processing was performed using the data of the field interpolation unit 13 (FIG. 3). In the second embodiment, the edge detection processing is performed using more data than in the conventional case.

FIGS. 4 and 5 are schematic diagrams showing the edge detection processing according to the second embodiment. FIGS. 4A and 4B show the detection of a horizontal edge.
FIGS. 5A and 5B show the detection of the vertical edge. Since the four methods are similar, a description will be given using the horizontal edge detection at the sample points shown in FIG. When the seven image data in FIG. 4A are A1 to A7, | A1−A2 |, | A
3-A4 |, | A5-A6 |, | A4-A7 |, and the maximum value of the four values is taken as the value of the horizontal edge of A4.

That is, the horizontal edge value of A4 = MAX ｛| A1−A2 |, | A3−A4
|, | A5-A6 |, | A4-A7 | 7.

Similarly, the values of the horizontal edge and the vertical edge for each pixel are obtained for FIGS. 4B, 5A and 5B.

Horizontal edge value of B7 = MAXＭＡ | B1−B2 |, | B3−B4
|, | B5-B6 |, | B2-B8 |, | B6-B9 |｝ Vertical edge value of C4 = MAX ｛| C1-C2 |, | C3-C4
|, | C4-C5 |, | C6-C7 | 7 Vertical edge value of D6 = MAX ｛| D1-D2 |, | D2-D3
|, | D4-D5 |, | D7-D8 |, | D8-D9 |

According to the present invention, when such edge detection processing is performed, the edge detection processing (see FIGS. 8 and 9), which has been deviated from the data of the previous line in the conventional example, is reduced to a uniform distance around the periphery. It can be detected using certain data.
Therefore, the accuracy of the edge signal for each pixel can be improved.

Therefore, according to the above-described second embodiment, the circuit of the edge detection processing unit 37 is shared with the field interpolation unit 13, and the edge detection processing using more data than before is performed. The accuracy of the edge signal can be improved. The shared part in the embodiment is, for example, a line memory for image data.

The present invention can be applied not only to the MUSE decoder but also to other image processing devices as long as the data interpolation and edge detection are performed using data in the same field. .

[0054]

As described above, the present invention is configured so that the circuit of the edge detection processing unit, which is a part of the motion detection processing, and the circuit of the field interpolation processing unit are shared. The circuit scale can be reduced, and the circuit required for edge detection processing can be reduced to about 30%.

Further, when performing the edge detection processing, the number of data in the vertical line direction can be increased, so that it is possible to perform detection using data at an equal distance from the surroundings, and to detect an edge signal for each pixel. Accuracy can be improved. As a result, the accuracy of the motion signal output from the motion detection processing unit is improved.

[Brief description of the drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of one embodiment of the present invention.

FIGS. 3A and 3B are schematic diagrams showing edge detection using field interpolation at sample points and non-sample points according to the present invention, respectively.

FIGS. 4A and 4B are schematic diagrams showing horizontal edge detection at sample points and non-sample points in the present invention, respectively.

FIGS. 5A and 5B are schematic diagrams showing vertical edge detection at sample points and non-sample points according to the present invention, respectively.

FIG. 6 is a block diagram showing a configuration of a conventional MUSE decoder.

FIG. 7 is a block diagram showing a motion detection processing unit in a conventional MUSE decoder.

FIGS. 8A and 8B are schematic diagrams showing horizontal edge detection at sample points and non-sample points in a conventional MUSE decoder, respectively.

FIGS. 9A and 9B are schematic diagrams showing vertical edge detection at sample points and non-sample points in a conventional MUSE decoder, respectively.

FIGS. 10A and 10B are schematic diagrams showing field interpolation processing at sample points and non-sample points in a conventional MUSE decoder, respectively.

[Explanation of symbols]

 1 MUSE signal input terminal 3 A / D converter 6 De-emphasis processing unit 12 Motion detection processing unit 13 Field interpolation unit 15 Motion signal 16 Edge signal 17 Sampling frequency conversion unit 18 MIX processing unit 27 D / A converter 32 2 Frame difference detection processing section 33 1 frame difference detection processing section 36 Video area detection processing section 37 Edge detection processing section

Claims (3)

(57) [Claims] 1. A motion detecting circuit of a MUSE decoder for demodulating and reproducing a band-compressed MUSE signal into an original wide-band high-definition television signal and detecting a moving image area using a frame difference signal.
Detection processing means and a field for interpolating the image data in the field
Interpolating means for performing a motion detection process within a predetermined range in the field.
Edge signal from the difference value of the image data
The edge detection unit, the shared each other in a motion detecting processing means and the field interpolation means, only the edge detection means on one of the <br/> motion detection means and the interpolation means in said field A MUSE decoder motion detection circuit characterized by being configured to be provided. 2. A subsump that makes a cycle with a period of 4 fields.
The MUSE signal band-compressed by the
MUSE Deco demodulates and reproduces high-definition television signals
In the motion detection circuit of the coder, the motion for detecting the moving image area using the frame difference signal
An image in a predetermined range within the field is provided to the detection processing means.
Edge detection means for detecting an edge from a difference value of image data
Field for interpolating image data in the field without
Image used by the field interpolation means in the field interpolation means
And outputs an edge signal detected by using the data, the motion
Detection processing means for outputting from the field interpolation means;
Configured to perform motion detection based on the edge signal
A MUSE decoder motion detection circuit, characterized in that: 3. The motion detection processing means and the inside of the field
Interpolation means and at least the line memory for image data
The MUSE data according to claim 1, wherein the MUSE data is shared.
Coder motion detection circuit.