JPH05130564A - Motion detecting circuit and video recording and reproducting device - Google Patents

Abstract

PURPOSE:To correctly detect motion even in slow motion reproduction, etc. CONSTITUTION:In the slow motion reproduction, a memory delay line 110 performs a one-field delay process for each field having a different field number, inputted at each repetitive period, from a last input field at each repetitive period of the same field and subtracters 142 and 143 detect motion signals. The motion signal of one field detected at each period is judged as a correct signal and outputted to the outside, and also stored in memory delay lines 151, 161, and 171. Then the motion signals stored in the memory delay lines 151, 161, and 171 are repeatedly outputted as correct motion signals for motion signals other than the motion signal of the one field detected at each period.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to a V for recording a video signal.
The present invention relates to a recording / reproducing apparatus such as a TR (Video Tape Recorder), an optical disc, a magneto-optical disc, and the like, and more particularly to a motion detection technique for detecting a motion amount of a video during reproduction.

[0002]

2. Description of the Related Art When a video signal recorded on a VTR or the like is played back, the screen is thinned to adjust the playback time or to produce special effects such as slow motion, still images, and double speed playback. , Special playback may be repeated to repeat the same screen.

On the other hand, in the case of the interlace type video signal,
One picture (frame) is composed of two fields having an interlacing relationship with each other.

Therefore, when the recorded interlace system video signal is thinned out or special reproduction is repeated to repeat the same screen, a field video signal is created from the reproduced video signal and the field video is reproduced. It is necessary to maintain order. This is because if the field order is not maintained, the center of gravity of the image moves and vertical resolution deteriorates.

Therefore, conventionally, a video signal of a different field is created from the reproduced video signal by using an interpolation filter technique.

On the other hand, when a video signal having a different field is created from the reproduced video signal by using the technique of the interpolation filter, that is, when the field conversion is performed, only the data in the same field is reproduced. Field conversion used (hereinafter,
In addition to performing "field-to-field conversion"), when the reproduced image is a still image, the data of a plurality of fields in which the scanning lines are interlaced is used. It is known that the image quality during special reproduction can be improved by performing field conversion (hereinafter referred to as "field-to-field field conversion"). Also, when the reproduced video is a still image, the luminance signal and color signal in the video signal must be separated.
Rather than two-dimensional Y / C separation using data in the same field, three-dimensional Y / C separation using multiple field data using inter-frame correlation It is desirable to carry out by C separation.

However, such a field in the field
In order to select whether to perform the field conversion, the field conversion between fields, the two-dimensional Y / C separation, or the three-dimensional Y / C separation, the field conversion is performed. It is necessary to judge whether the reproduced image is a still image or not by detecting the motion of the image from the data of the field to be recorded and the data of several fields immediately before that field. Become.

A circuit described in Japanese Patent Publication No. 13790/1993 is known as a motion detecting circuit for detecting the motion of an image.

FIG. 5 shows the configuration of this motion detection circuit.

In FIG. 5, 401 is a video signal input terminal, 402 and 403 are connected to a motion adaptive Y / C separation circuit and a motion adaptive field conversion circuit, respectively, and output the motion amount detected by this circuit. Output terminals 411, 4
Reference numeral 12 is a delay circuit that delays the input video signal by one frame, 413 and 414 are 262 horizontal scanning periods (hereinafter, the horizontal scanning period is indicated by “H”), 1H delay circuits, and 421 and 422 are respectively. Subtractors 431, 432 are non-linear absolute value circuits, 433 is a synthesis circuit, 434 is a low-pass filter (hereinafter referred to as "LPF") that passes only signals in a frequency region lower than an arbitrary frequency, and 435 is an edge. A detection circuit, 436 is an isolated point removal circuit, 437 is a motion synthesis circuit, and 438 is a motion expansion filter.

In this device, the video signal input from the input terminal 401 is appropriately delayed by the delay circuits 411 and 412,
Subtractor 421 from the input video signal and the output of delay circuit 412
Thus, the motion signal of the difference between the two frames is subjected to subtraction processing of the input video signal and the output of the delay circuit 411 by the subtractor 422,
The LPF 434 removes the color signal component and calculates the motion signal of the difference between the frames.

The motion signals of the difference between the two frames and the difference between the one frame are subjected to the nonlinear processing and the absolute value processing by the nonlinear absolute value circuits 431 and 432, respectively, according to the edge information detected by the edge detection circuit 435. And the synthesis circuit 43
It is weighted by 3 and synthesized. The influence of noise or the like is removed from the motion signal output from the synthesis circuit 433 by the isolated point removal circuit 436.

The motion signal α output from the isolated point removal circuit 436 is input to the motion synthesis circuit 437 together with the motion signal β delayed by 262H in the delay circuit 413 and the motion signal γ further delayed by 1H in the delay circuit 414. , Coefficient k1,
After being synthesized while being weighted by k2 and k3, they are output from the output terminal 402 and control the motion adaptive Y / C separation circuit. The motion adaptive Y / C separation circuit is a circuit that selectively executes two-dimensional Y / C separation and three-dimensional Y / C separation according to the amount of motion.

Further, the motion expansion filter 438 spatially blunts the motion combined signal output from the motion combining circuit 437 and outputs it from the output terminal 403.
Control the field conversion circuit. The motion adaptive field conversion circuit is a circuit that selectively executes field conversion within a field and field conversion between fields according to the amount of motion.

As described above, according to the circuit disclosed in Japanese Examined Patent Publication No. 3-13790, it is possible to detect the motion of the video signal with high accuracy, and the motion adaptive Y / C separation circuit and the motion adaptive fission circuit are provided. It is possible to prevent erroneous control of the field conversion circuit and improve the image quality during special reproduction.

Now, referring to FIG. 6, the motion adaptive type field.
An example of field conversion between fields and field conversion between fields of the field conversion circuit will be shown.

In FIG. 6, 501 is data of an arbitrary line of an arbitrary input field, 502 is data one line before the data 501, and 503 is a field to which the data 501, 502 belongs. Field data immediately before the field and is the data interlaced to the data 501 and 502.

If the reproduced image is a moving image and the input field and the even field of the output field match each other, the data is reproduced.
The data 501 and 502 are added in a ratio of 3: 1 and output, and when the even field of the input field and the output field do not match, the data 501 and 502 are added in a ratio of 1: 3 and output. This is the in-field field conversion.

On the other hand, when the reproduced image is a still image and the odd and even numbers of the input field number and the output field number match, the data 501 and the data 503 are added one by one and output. , 502 and 503 if the even field of the input field number and the output field number do not match.
Are added one by one and output. This is the field-to-field conversion.

As described above, in order to perform field conversion between fields in a still image, the field data to be field-converted and the field and the scanning line are interleaved. The data of the previous field to be laced is required.

When the reproduced video is repeated by slow motion reproduction or the like, the data of the immediately preceding field required for field conversion between fields is field converted. In some cases, the data is exactly the same as that of the field to be processed, in which case correct field-to-field field conversion cannot be performed. For this reason, in such a case, the field data for which field conversion is to be performed and the field data for which the scanning line is interlaced immediately before the repetition of the image are performed. Is stored in a memory means, and the data between the fields is used by using the data.
I'm trying to convert the field.

[0022]

In FIG. 7, <A> indicates the normal reproduction of the recorded moving image.

In the figure, 601, 602, 603, 604,
Reference numeral 605 is a time-sequential 1-field video signal reproduced in order from the tape.

<B> is a slow motion mode of the same image at 1/6 times.
In this example, each video signal is repeatedly reproduced 6 times. The video signals 611 to 616 have the same picture and the same field number as the video signal 601, and the video signals 621 to 625 have the same picture and the same field as the video signal 602.

Now, referring to FIG. 7B, for example, data 62
4 is subjected to field conversion, the above-described conventional motion detection circuit detects the motion between the two frames by the data 616 and the data 622, and the result 2 is detected.
It is determined that the reproduced video signal is a moving image based on the motion detection result of the frame difference. Therefore, the motion adaptive field conversion circuit uses only the data 624 and the field in the field.
Perform the field conversion.

However, when the data 625 is converted into a field, the conventional motion detection circuit uses the same field as the data 621 and the data 623 to convert the difference between two frames.
In order to detect the movement of one frame difference, the data 625
Will be judged as a still image. Therefore, in the motion adaptive field conversion circuit, the data 625, the data 625 and the scanning line have an interlaced relationship, and the reproduced image stored in the memory means is repeated. The field conversion between fields is performed by using the data 616 which is the data just before the start of. As a result, the playback video signal that is a moving image
Since the field conversion between fields is performed, an afterimage is generated in the data after the field conversion, and the reproduced image is significantly deteriorated.

That is, the same file such as slow motion
If the circuit is repeatedly reproduced for a certain period of time, the circuit described in Japanese Patent Publication No. 3-13790 cannot correctly detect the motion of the reproduced video signal, and thus the motion adaptive Y / C separation circuit or the motion adaptive type. There is a problem that the control of the field conversion circuit is erroneous.

Therefore, it is an object of the present invention to provide a motion detecting circuit which can correctly detect the motion of a reproduced video signal even when the same field is repeatedly reproduced for a certain period. And

[0029]

[Means for Solving the Problems] To achieve the above object,
The present invention is a motion detection circuit for inputting a video signal that constitutes one frame from two fields and detecting the motion of the input video signal, and the field to which the input video signal belongs is A field jump detection means for detecting whether or not the field input immediately before the field is repeated, and a field which is not the repetition of the immediately preceding field depending on the detection content of the field jump detection means. A video signal delay means capable of delaying the video signal by at least one frame, which delays only the video signal of the field by one field at the timing when the field that does not repeat the immediately preceding field is input; The file you entered immediately before
A motion detecting means for detecting at least a motion of a field video signal which is not repeated in the field with respect to the field video signal delayed by one frame by the video signal delaying means. Motion detection circuit.

Further, in order to achieve the above object, the present invention provides:
A motion detection circuit for inputting a video signal constituting one frame from a field and detecting a motion of the input video signal, the video signal delaying means delaying the input video signal, and one field. Minute motion video signal motion memory and the input video signal field
The field jump detection means for detecting whether or not the field is the repetition of the field input immediately before the field concerned, and the field which is not the repetition of the field input immediately before.
When a field video signal is input, the field n (where n is an arbitrary natural number) frame before the input field delayed by the video signal delay means is input.
The motion detection means for detecting the motion between the input field video signals with respect to the field video signal and updating the content of the motion memory; and the motion detection means immediately before the field jump detection means. For a field that is not a repeat of the field input to, the motion detected by the motion detection means is output as the motion of the video signal of the field, and the field that is a repeat of the field input immediately before is output. 2), the second motion detection circuit is provided with a selecting means for outputting the motion stored in the motion memory as the motion of the video signal of the field.

[0031]

According to the first motion detecting circuit of the present invention, the video delay means, depending on the detection content of the field jump detecting means, does not repeat the immediately preceding field.
Only the video signal of the field is delayed by one field at the timing when a field that is not a repetition of the immediately preceding field is input. Therefore, only the non-repeating field of the immediately preceding field can be delayed by one field for each period of the non-repeating field of the immediately preceding field, resulting in the immediately preceding field. When inputting a field that is not repeated,
The field before the frame is delayed by just one frame. Therefore, even when the same field is continuously input a plurality of times, such as when playing back in slow motion, the video signal delay means for one frame is used to immediately precede the field. For a field that is not repeated, the movement of the field with respect to the field one frame before can be detected. Similarly, by facilitating the video signal delay means for n frames, the fields 1 to n frames before the field concerned are not repeated for fields which are not repeated. The movement with respect to can be detected.

Further, according to the second motion detecting circuit of the present invention, the motion detecting means, when the video signal of the field which is not the repetition of the field inputted immediately before is inputted, The video signal of the input field delayed by the signal delay means with respect to the video signal of the field n (where n is any natural number) frame before the input field. The motion is detected during the interval and the contents of the motion memory are updated. Then, according to the content detected by the field jump detection means, for a field which is not repeated just before the field input, the motion detected by the motion detection means is regarded as the motion of the video signal of the field. With respect to the field which is output and which is the repetition of the field input immediately before, since the motion stored in the motion memory is output as the motion of the video signal of the field, the slow motion reproduction is performed. When the same field is input several times in succession, such as, etc., not only for fields that do not repeat the field that was input immediately before, but also for the field that was input immediately before.
It is possible to output the correct movement even for the repeated field of the field.

[0033]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the motion detecting circuit according to the present invention will be described below by taking application to a VTR as an example.

First, the configuration of the VTR according to this embodiment is shown in FIG.
Shown in.

FIG. 4 shows, as an example, a video reproducing system of a VTR for recording / reproducing according to the D2 format referred to in the SMPTE standard. According to the D2 format, the video signal is recorded on the tape as digital data.

In the figure, numeral 800 is a tape, numeral 801 is a magnetic head, numeral 802 is a decod-desha full concealment for rearranging data according to the standard and error correction and correction of data, and numeral 110 is at least 4 files. -Dedicated memory delay line, and 130 is a field jump detection circuit. The decoder / desha full conceal 802 outputs the reproduced video data to the memory delay line 110 and the reproduced field number to 130. The field number is each field
It is a sequential number of the field recorded on the tape at the recording position of the field together with the image data of the field.

Further, 100 is a motion detection circuit, 810 is a motion adaptive Y / C separation circuit, 820 is a motion adaptive field conversion circuit, 830 is a D / A conversion circuit, and 813 and 82.
2, 826 are multiplexers. Motion adaptive Y / C
The separation circuit 810 includes a two-dimensional Y / C separation circuit 811 and a three-dimensional Y / C separation circuit 812, controls the multiplexer 813 according to the amount of motion detected by the motion detection circuit 100, and controls the two-dimensional Y / C separation circuit. 811 or the output of the three-dimensional Y / C separation circuit 812 is output. The motion adaptive field conversion circuit 820 includes a luminance signal in-field conversion circuit 821, a luminance signal inter-field conversion circuit 823, a color difference signal in-field conversion circuit 824, and a color difference signal field. A conversion circuit 825 is provided, and multiplexers 822 and 826 are provided according to the amount of motion detected by the motion detection circuit 100.
To output the outputs of the inter-field conversion circuits 823 and 825 or the in-field conversion circuits 821 and 824. The D / A conversion circuit converts the video data into an analog signal and outputs it.

Next, details of the memory delay line 110, the motion detection circuit 100, and the field jump detection circuit 130 will be described.

FIG. 1 shows the internal structure of the motion detecting circuit 100 together with the memory delay line 110 and the field jump detecting circuit 130.

In FIG. 1, 101 is an input terminal for the video data reproduced from the tape 800, 102 is an input terminal for inputting the field number of the video data input to the input terminal 101, 103 is a VTR or the like. Input terminal for inputting the playback mode of the video signal, and 104 is a motion adaptive Y /
An output terminal for outputting to the C separation circuit 810, and an output terminal 105 for outputting the detected motion signal to the motion adaptive field conversion circuit 820.

Reference numeral 110 denotes a memory delay line for appropriately delaying the video data input from the input terminal 101 and outputting data necessary for motion detection according to the reproduction status of the video data.
Reference numeral 1 denotes an output terminal for directly outputting the reproduced video data input from the memory delay line 110 through the input terminal 101, 11
Reference numeral 2 denotes an output terminal for outputting the video data one frame before the video data output from the output terminal 111, or the video data corresponding thereto from the memory delay line 110, 113.
Is the two frames of the video data output from the output terminal 111.
This is an output terminal for outputting the video data before the video or the video data corresponding thereto from the memory delay line 110.

Reference numeral 120 denotes an edge detection circuit for detecting an edge portion of a picture pattern of the video data, and 130 denotes a video data field.
A field jump detection circuit for inputting a field number and a reproduction mode to detect the occurrence of repetitive video reproduction, etc.
Reference numeral 41 is an LPF that passes only signals in a frequency region lower than an arbitrary frequency.

142 and 143 are subtractors 144 and 145.
Is an absolute value circuit that performs absolute value processing on the input data,
Motion signal conversion circuits 820 and 148 that convert the input signal according to the edge information detected by the edge detection circuit 120 generate motion signals with a difference between one frame and motion signals with a difference between two frames. A synthesizing circuit for synthesizing by weighting with an arbitrary coefficient value, 149 is a noise removing circuit, 151, 1
Reference numeral 52 is a delay line for one clock when the motion signal is sampled at a frequency four times as high as the subcarrier,
4 is a coefficient unit having an arbitrary coefficient value α, and 155 is a delay line 15
A synthesis circuit that synthesizes the output of 1 and the outputs of the coefficient multipliers 153 and 154 with arbitrary weighting, 161 is a delay line of 1H, 162
Is a synthesizing circuit for synthesizing the output of the synthesizing circuit 155 and the output of the delay line 161 with arbitrary weighting, and 171 is about 262H.
, 172 is a coefficient unit having an arbitrary coefficient value β, 17
Reference numeral 3 is a motion synthesizing circuit for synthesizing the output of the noise removing circuit 149 and the output of the coefficient multiplier 172 with arbitrary weighting.

The operations of the memory delay line 110, the motion detection circuit 100, and the field jump detection circuit 130 will be described below.

The video data reproduced from the tape 800 is input to the reproduction video data input terminal 101, and the file is reproduced.
The field number is input to the field number input terminal 102.

The field number of the video data input to the input terminal 102 is the field jump detection circuit 13.
Input to 0. In the field jump detection circuit 130, the field number of the current field input from the input terminal 102 and the value 1 input from the input terminal 102 and stored in the field jump detection circuit 130.
By comparing with the field number of the previous field, it is shown that the reproduction mode input from the input terminal 103 is the forward reproduction, and the field number of the previous field. On the other hand, if the field number of the current field is advanced by one, it is judged to be normal reproduction, and if it is advanced by two or more, fast-forward reproduction is carried out.
Judge as motion reproduction.

Further, it is shown that the reproduction mode input from the input terminal 103 is the reverse reproduction, and the field number of the current field is different from the field number of the previous field. If − is returned by one or more, it is judged as reverse fast forward reproduction including −1 × speed reproduction, and if it is not returned, slow motion reproduction including still, and the judgment result is stored in the memory delay line 110 and synthesized. Circuits 155, 1
62, coefficient unit 172, motion synthesizing circuit 173, etc., respectively.

The video data input to the input terminal 101 is input to the memory delay line 110. In the reproduced video data, the scanning lines are interlaced with each other for each field. Therefore, in order to detect the motion, a signal with a difference of 2 fields (1 frame) or 4 fields is used. A signal with a field (2 frame) difference is required. Therefore, the memory delay line 1
In 10, field jump information is input from the field jump detection circuit 130 so that correct motion detection can be performed even when the field sequence is out of order during special reproduction, and is internally appropriate. The output terminals 111, 11 of the memory delay line 110 after performing various delay processes.
2, 113, the current field image data, 1 respectively
Video data corresponding to the video data before the frame is output.

The delay process of the memory delay line 110 will be described with reference to FIG.

In FIG. 2, (a) is the video data input from the input terminal 101, and (d), (b) and (e).
(C) is a field that stores video data for one field.
It is a rud memory.

Also, 201 to 296 are each 1
This is field image data, and each numeral in the figure indicates a field number of the image data. Also, in the figure,
Time is changing from top to bottom. That is, in the figure,
Each video data in the same row shows the input video data at the same time point and the delayed video data stored in each field memory, and the data in the row below the input video data at the time point after one field equivalent time has elapsed. The video data and the delayed video data stored in each field memory are shown.

Now, in the figure, (A) shows the case of normal reproduction, and (a) 201 to 205 is video data which is temporally continuous.

These video data 201 to 205 are output from the output terminal 111 of FIG. 1 as the video data of the current field. 2A and 2B store video data for performing a subtraction process on the signals of FIGS. 2A and 2A to detect a motion signal having a difference between frames. (A) for normal playback
The video data of (a) is stored in the field memory (d) (b).
The video data 211 is output from the output terminal 112 of FIG. 1 at the same timing as the video data 201, and the video data 212 at the same timing as the video data 202.

FIGS. 2A and 2C store video data for detecting a motion signal having a difference between two frames by performing subtraction processing on the signals of FIGS. The image data of b) is sequentially shifted to the field memories (e) and (c), and further delayed by 2 fields. Video data 22
1 is output from the output terminal 113 in FIG. 1 at the same timing as the video data 201 and 211, and the video data 222 is output at the same timing as the video data 202 and 212.

By the above processing, (d) is set to 1 by the memory delay line 110 for the input video data at an arbitrary time point.
Field-delayed video data is stored, (b) stores video data delayed by one frame (2 fields) by the memory delay line 110, and (e) shows memory delay line 110.
The video data delayed by 3 fields is stored in (c).
Means that the video data delayed by 2 frames (4 fields) by the memory delay line 110 is stored.

Next, FIG. 2B shows the case of slow motion reproduction, and (B) (a) is the video data inputted from the input terminal 101, which is temporally from 250 to 257. Continuous video data, 251 to 254 are video data of the same field, and 255 to 257 are video data of the same field. These video data are directly output from the output terminal 111 of FIG. 1 as the video data of the current field as in the normal reproduction.

FIGS. 2B and 2B are image data for detecting a motion signal having a difference between one frame by subtracting the image data of FIGS. 2B and 2A. However, during special reproduction, just by sequentially shifting the video data of (B) and (a) in the field memory as in the normal reproduction, the video data data necessary for correct motion detection processing is created. It is not possible.

Accordingly, the field jump detection circuit 1
When the field jump information from 30 indicates special reproduction such as slow motion, memory delay processing according to the reproduction mode is required.

Therefore, for example, in the case of slow motion reproduction as shown in FIG. 2B, the video data of (B) and (a) are first sequentially input to the field memory (d). 1 field delay. And next,
Of the video data in the field memory (d), only the video data 281 and 285 immediately before the field number is changed are written in the field memory (b) in the next stage. On the other hand, the field memory (b) holds the written video data until the data is written next from (d).

Similarly, the field memories (b) (e)
Also, of the video data, only the video data 260, 262, 290 and 292 immediately before the field number is changed are written in the next field memory. The field memories (e) and (c) also hold the written video data until the next writing of data.

Of the video data in the field memory, the video data 281 and 28 immediately before the field number changes.
Writing of 5, 260, 262, 290, and 292 may be performed by writing to the next field memory at the timing of the input video data (a) 251, 255 immediately after the field number is changed.

Now, from the output terminal 112 of FIG.
The contents stored in the field memory (b) are output.
Similarly, the contents of the fourth stage field memory (c) are output from the output terminal 113.

By performing the above processing, (B)
Immediately after the field of the input video data of (a) is changed, that is, immediately after the field number of the video data of (B) (a) is changed, an image of a different field is stored in each field memory. The data can be arranged in the order of input, and correct video data between one frame and two frames can be obtained.

However, the values are correct 1 except immediately after the field number is changed in the input video data of (B) and (a).
Image data having a difference between frames and a difference between two frames cannot be obtained. Therefore, correct motion detection cannot be performed. Therefore, the amount of movement immediately after the field is changed is set as the amount of movement until the next field number is changed by the processing described later.

The video data input from the reproduced video terminal 101 is not a signal for interlacing scanning lines for each field, but a method for interlacing the same location for each field. With video data, motion detection can be performed with video data having a difference between one field and three fields, but even in this case, the same operation as the memory delay process described above is required during special reproduction.

Now, referring to FIG. 1, the memory delay line 110
The output terminals 111, 112, and 113 output the video data of the current field, the video data corresponding to the video data of the previous frame, and the video data corresponding to the video data of the previous frame, respectively. .. The image data of the current field output from the output terminal 111 is the edge detection circuit 12
0 and the subtractors 142 and 143. Also, the video data delayed by one frame output from the output terminal 112 is input to the subtractor 142, and the subtractor 14
In step 2, subtraction processing is performed on the image data of the current field output from the output terminal 111, and a motion signal having a difference between 1 frames is detected.

Similarly, the 2 output from the output terminal 113
The video data delayed by the frame is input to the subtractor 143, and the subtractor 143 performs subtraction processing on the video data of the current field output from the output terminal 111 to obtain 2 frames.
The motion signal of the difference between the frames is detected. However, when the video data input to the reproduction video data input terminal 101 is a composite signal in which a color signal and a luminance signal are multiplexed, such as in the NTSC system, the phase of the subcarrier carrying the color difference signal for each frame is Since it has been inverted, the motion signal of the difference between one frame includes an erroneous color difference signal component difference. Therefore, the output of the subtractor 142 is input to the LPF 141,
The color signal component is removed in F141.

If the video data input to the reproduced video data input terminal 101 is non-multiplexed video data such as a component signal, the LPF 141 is unnecessary.

In the video data of the current field output from the output terminal 111 and input to the edge detection circuit 120, this circuit detects the edge portion of the pattern of the video data and outputs this edge information. 1-frame difference output from the LPF 141 and the subtractor 143, and 2
The motion signals of the difference between frames are the absolute value circuits 144 and 145.
After being subjected to absolute value processing respectively, are input to the motion signal conversion circuits 146 and 147, respectively.

The motion signal conversion circuits 146 and 147 convert the input motion signals of the difference between one frame and the difference between two frames according to the edge information detected by the edge detection circuit 120. That is, when detecting the motion of the video data, the edge part of the pattern is often erroneously detected as a motion, and the erroneous motion detection of the edge part has a bad influence on the motion adaptive field conversion circuit 820 and the like. large. Therefore, with respect to a place determined to be an edge portion by the edge detection circuit 120, processing for reducing the amount of motion detected according to the strength of the edge detected by the edge detection circuit 120 is performed even if motion is detected. To do. By the conversion processing in the motion signal conversion circuits 146 and 147, the adverse effect when the edge portion of the pattern of the video data is erroneously detected as a motion is reduced.

Differences between one frame and two frames after motion signal conversion processing is performed by the motion signal conversion circuits 146 and 147.
The motion signals of the inter-frame difference are both input to the combining circuit 148 and are combined with arbitrary weighting. In this embodiment, the output of the motion signal conversion circuit 146 and the output of the motion signal conversion circuit 147 are added in a one-to-one ratio to perform synthesis by an addition averaging method. The motion signal combined by the combining circuit 148 is input to the noise removing circuit 149.

Next, the noise removal circuit 149 in the motion detection circuit 100 will be described.

FIG. 3 shows the configuration of the noise elimination circuit 149.

In FIG. 3, reference numeral 301 denotes the synthesis circuit 1 of FIG.
Input terminals to which the motion signals output from 48 are input, 3
Reference numeral 02 is an output terminal for outputting a motion signal after removing the influence of noise and the like, 311, 312 are delay lines of 1H, and 321 is a combining circuit. The motion signal input from the input terminal 301 in this circuit is input to the combining circuit 321 together with the signal delayed by 1H by the delay line 311 and the signal further delayed by 1H by the delay line 312.

The synthesizing circuit 321 selects one of the three input motion signals, or multiplies each motion signal by an arbitrary coefficient value and adds them. That is, in this way, the correlation of the detected motion signal in the vertical direction is checked, and the protruding ones are removed as noise or the like as compared with the upper and lower motion signals. The video data reproduced from the tape 800 is composed of a continuous waveform in the scanning line direction, that is, in the horizontal direction on the screen. When noise or the like is added to the video data waveform, the influence of the noise or the like is displayed on the screen. This is because it is spread in the horizontal direction, and its influence does not spread in the vertical direction on the screen. By performing such processing, it is possible to reduce the influence of noise and the like.
The output signal of the synthesizing circuit 321 is output from the output terminal 302 as a motion signal after removal of noise and the like and is output from the coefficient unit 172.
Is input to the motion synthesizing circuit 173 together with the output of.

In FIG. 1, the motion synthesizing circuit 173 outputs the field output from the field jump detecting circuit 130.
According to the field jump information, when the slow motion is not reproduced (when the field number is sequentially changed), the output of the noise removing circuit 149 and the coefficient unit 1
The output of 72 is combined with the output motion signal,
In the case of motion reproduction (when the same field number is continuous), the output of the coefficient unit 172 is output as it is. In the present embodiment, when the slow motion reproduction is not performed, the larger one of the output of the noise elimination circuit 149 and the output of the coefficient unit 172 is selected and output.

The synthesizing circuit 155 includes a motion synthesizing circuit 173.
The delay circuit 151 delays the motion signal output by the coefficient multiplier 153 and the motion signal output from the motion synthesizing circuit 173 by one clock with a clock having a frequency four times that of the subcarrier. The motion signal
The motion signal output from the motion synthesizing circuit 173 is delayed by the delay circuits 151 and 152 for two clocks with a clock having a frequency four times as high as that of the subcarrier, and then the motion signal subjected to the attenuation processing by the coefficient unit 154 is input. It However, in this embodiment, it is assumed that video data obtained by sampling an analog video signal at a frequency four times as high as that of the subcarrier is used for recording.

In this synthesizing circuit, according to the field jump information output from the field jump detecting circuit 130, in the case of slow motion reproduction, the delay circuit 151.
Is output as is. If the motion is not the slow motion reproduction, the three motion signals are combined and output. In this embodiment, the largest one of the three motion signals is selected and output as the combining process in the case of non-slow motion reproduction. If not slow motion playback,
The delay circuits 151 and 152, the coefficient units 153 and 154,
The synthesizing circuit 155 operates as a horizontal filter that widens the detected motion while attenuating it in the horizontal direction.

The motion signal output from the synthesis circuit 155 is output from the output terminal 104 connected to the motion adaptive Y / C separation circuit 810 to control the motion adaptive circuit 810, and at the same time, delayed by 1H by the delay circuit 161. It is input to the synthesis circuit 162 together with the motion signal. This synthesis circuit 162
Then, according to the field jump information output from the field jump detection circuit 130, in the case of slow motion reproduction, the output of the synthesis circuit 155 and the delay circuit 161.
The output of is alternately output for each field, and the slow mode
If not motion reproduction, the two motion signals are combined and output. However, in this embodiment, the
In the case of non-motion reproduction, the maximum one of the two motion signals is selected and output.

The motion signal output from the synthesizing circuit 162 is output from the output terminal 105 connected to the motion adaptive field converting circuit 820 to control the motion adaptive field converting circuit 820 and to the delay circuit 171. Is entered.

On the other hand, in the delay circuit 171, the output of the coefficient multiplier 172 is equal to the output of the noise elimination circuit 149 by one filter.
The input motion signal is delayed so as to be delayed (262H or 263H), and the delayed motion signal is input to the coefficient unit 172. The motion signal input to the coefficient unit 172 is slow-moved according to the field jump information output from the field jump detection circuit 130.
In the case of motion playback, the input motion signal is
If not slow motion reproduction, it is output after being attenuated by an arbitrary coefficient value. The motion signal output from the coefficient unit 172 is input to the motion synthesis circuit 173, as described above.

As described above, according to the present embodiment, according to the field jump information output from the field jump detection circuit 130, during the same slow motion reproduction,
Motion synthesis circuit 173, synthesis circuit 155, synthesis circuit 16
2. Controlling the coefficient unit 172, delay line 151, delay line 16
1. The delay line 171 is made to function as a field memory for storing one field of motion signal.
The output of the field memory composed of 1, 161, and 171 is fed back to the motion synthesizing circuit 173 as it is.

That is, for example, at the time of slow motion reproduction as shown in FIG. 2B, the motion detection is correctly performed for the video data 251 immediately after the field number is changed, and the detected motion signal is delayed by the delay circuit. 151, 15
2, a filter composed of the coefficient units 153 and 154, the synthesizing circuit 155, and a filter composed of the delay circuit 161 and the synthesizing circuit 162, which spatially expands and outputs the same, and to the field memory composed of the delay lines 151, 161, 171. Store the detected motion signal.

For video data other than immediately after the field number that cannot be correctly detected for motion detection, the video data immediately after the field number has been changed is obtained, and the delay lines 151, 161, 17 are obtained.
The motion signal stored in the field memory consisting of 1 is repeatedly used as it is.

Therefore, a correct motion detection result can always be obtained, and erroneous control of the motion adaptive Y / C separation circuit 810 and the motion adaptive field conversion circuit 820 can be prevented.

Further, by adopting a circuit configuration for attenuating and feeding back the detected motion signal during normal reproduction as in the present embodiment, the detected motion signal is not only directed in the horizontal and vertical directions, but also Depending on the coefficient value of the coefficient unit 172, it is possible to widen by several fields in the time axis direction, and the detection omission of the motion in the image data in which the picture pattern changes in two field cycles is greatly missed. Can be reduced.

In this embodiment, the D2 format V
Although the motion detection circuit has been described by taking the application to TR as an example, the motion detection circuit according to the present embodiment is not limited to the digital VT of another system.
In addition to the R and analog VTRs, the present invention can be similarly applied to other video recording / reproducing devices such as an optical disk device and a magneto-optical disk device. In addition to NTSC signals, PAL,
The same can be applied to video signals of other systems such as SECAM.

[0088]

As described above, according to the motion detecting circuit of the present invention, the motion of the reproduced video signal is correctly detected even when the same field is repeatedly reproduced for a certain period. It is possible to provide a motion detection circuit capable of doing the above.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a motion detection circuit according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram showing an operation of a memory delay line according to an embodiment of the present invention.

FIG. 3 is a block diagram showing a configuration of a noise elimination circuit according to an embodiment of the present invention.

FIG. 4 is a block diagram showing a part of the configuration of a VTR according to an embodiment of the present invention.

FIG. 5 is a block diagram showing a configuration of a conventional motion detection circuit.

FIG. 6 is an explanatory diagram showing a motion adaptive field conversion process.

FIG. 7 is an explanatory diagram showing a conventional motion detection operation.

[Explanation of symbols]

 101 video data input terminal 102 field number input terminal 103 reproduction mode signal input terminal 104, 105 motion signal output terminal 110 memory delay line 120 edge detection circuit 130 field jump detection circuit 141 low-pass filter 142, 143 Subtractor 144, 145 Absolute value circuit 146, 147 Motion signal conversion circuit 148 Synthesis circuit 149 Noise removal circuit 151, 152 Delay circuit 161, 171 Delay circuit 153, 154 Coefficient unit 172 Coefficient unit 155, 162 Motion combination circuit 173 Motion combination circuit 800 tape 810 motion adaptive Y / C separation circuit 820 motion adaptive field conversion circuit

Front page continued (72) Inventor Miyoko Yoshikoshi 292 Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa Inside Hitachi Media Video Research Laboratory (72) Inventor Kotaro Okiguchi 2-15-12 Nishishinbashi, Minato-ku, Tokyo Stock Company Hitachi, Ltd. Home Appliance Business Headquarters

Claims (9)

[Claims] 1. A motion detection circuit for inputting a video signal constituting one frame from two fields and detecting the movement of the input video signal, wherein the field to which the input video signal belongs is The field
The field jump detection means for detecting whether or not the field input just before the field repeats, and the field jump detection means for detecting the field which is not repeated just before the field jump detection means. Immediately before inputting the video signal, a video signal delay means capable of delaying the video signal by at least one frame, which delays the video signal by one field at the timing when a field that is not a repetition of the previous field is input. The video signal of the field which is not repeated in the field input to the
The motion of the field delayed by
A motion detection circuit comprising at least motion detection means for detecting. 2. A motion detection circuit for inputting a video signal constituting one frame from two fields and detecting the movement of the input video signal, wherein the video signal delay means delays the input video signal. And a motion memory for storing the motion of one field of the video signal, and a field to which the input video signal belongs,
Field jump detection means for detecting whether or not the field input immediately before the field concerned is repeated, and when the video signal of the field which is not the field repetition input immediately before is inputted. , N from the input field delayed by the video signal delay means,
(Where n is an arbitrary natural number) a motion detecting means for detecting a motion between the input field video signal and the field video signal before the frame and updating the contents of the motion memory. Depending on the content detected by the field jump detection means, for a field which is not repeated just before the field input, the motion detected by the motion detection means is output as the motion of the video signal of the field. The field that is the repetition of the field entered immediately before.
And a selection means for outputting the motion stored in the motion memory as the motion of the video signal of the field. 3. A motion detection circuit for inputting a video signal constituting one frame from two fields and detecting a motion of the input video signal, wherein a motion of the video signal for one field is detected. A motion memory to be stored, a field jump detection means for detecting whether or not the field to which the input video signal belongs is a repetition of the field input immediately before the field, and field jump detection. Depending on the detection contents of the means,
Only the video signal of the field that is not repeated
A video signal delay means capable of delaying a video signal by at least one frame, which is delayed by one field at a timing when a field which is not a repetition of the immediately preceding field is inputted, and a field inputted immediately before being inputted. 1 for video signal delay means for field video signals that are not repeated
The motion of the field delayed by the frame relative to the video signal is at least detected, and the motion detection means for updating the content of the motion memory and the content detected by the field jump detection means are input immediately before. For a field that does not repeat the field, the motion detected by the motion detecting means is output as the motion of the video signal of the field, and the field that is the repetition of the field input immediately before is output.
And a selecting means for outputting the motion stored in the motion memory as the motion of the video signal of the field. 4. A motion detection circuit for inputting a video signal constituting one frame from two fields and detecting the movement of the input video signal, the video signal delaying means delaying the input video signal. And a motion memory for delaying the movement of one field of the video signal by one field, and the field to which the input video signal belongs repeats the field input immediately before the field. Field jump detection means for detecting whether or not the video signal of the field of the input field is input when the video signal of the field which is not the repetition of the field input immediately before is input. The motion of the field delayed by the delay means with respect to the video signal in the field one frame before the input field;
A motion detecting means for obtaining a motion with respect to the video signal of the field before the frame, synthesizing and outputting the motion,
Depending on the detection content of the field jump detection means, for the field that is the repetition of the field input immediately before, the delayed motion is output by the motion memory, and the field input immediately before is not repeated. Regarding the field, a combining circuit for combining and outputting the motion output by the motion detecting means and the motion delayed by the motion memory at a predetermined ratio, and the detection content of the field jump detecting means, For a field that does not repeat the field input immediately before, a filtering process is performed to spread the motion output by the synthesizing circuit to the surrounding space on the image plane, and the field is processed.
A field that is output as a motion of a video signal of a field and is input to the motion memory, and a field that is a repetition of the field that is input immediately before is the motion output by the synthesizing circuit as the motion of the video signal of the field. A motion detection circuit comprising a filter circuit for outputting and inputting to the motion memory. 5. The motion detection circuit according to claim 4, wherein the video signal is digital data obtained by sampling an analog video signal at a predetermined frequency, and the filter circuit detects the field jump. Depending on the content detected by the means, the field that is not repeated is the field that was input immediately before.
For the field, a filter that spreads the motion of the video signal in the horizontal direction on the image plane by convolving the data of the video signal with the value of the data that has a front-back relationship in the horizontal direction on the field image plane. A motion detection circuit which performs processing and outputs as a motion of a video signal of the field. 6. The motion detection circuit according to claim 4, wherein the video signal is digital data obtained by sampling an analog video signal at a predetermined frequency, and the filter circuit detects the field jump. Depending on the content detected by the means, the field that is not repeated is the field that was input immediately before.
For the field, the video signal data is convolved with the value of the data that is vertically above or below the field in the field image plane, so that the motion of the video signal is changed in the vertical direction on the image plane. A motion detection circuit which performs a filtering process for expanding the video signal and outputs it as a motion of the video signal of the field. 7. A recording medium for recording a video signal constituting one frame from two fields, a reproducing means for reproducing the video signal stored in the recording medium, and a video signal reproduced by the reproducing means. 7. The motion detection circuit according to claim 1, wherein the motion detection circuit detects the motion of the input video signal, and display means for displaying an image represented by the video signal reproduced by the reproducing means. A field for converting the oddness of the field by using the video signal in the same field according to the motion output from the motion detection circuit so that the oddness field is alternately displayed on the display means. Video recording / reproducing having a motion adaptive field converting means for switching between intra-field conversion processing and inter-field conversion processing for converting field oddness using a video signal of an adjacent field. apparatus. 8. A recording medium for recording a composite video signal in which a luminance component signal and a chrominance component signal are multiplexed, which constitutes one frame from two fields, and a video signal stored in the recording medium. 7. A motion detecting circuit according to claim 1, 2, 3, 4, 5 or 6, wherein the reproducing means for reproducing the video signal and the video signal reproduced by the reproducing means are inputted and the motion of the inputted video signal is detected. Field adjacent to the two-dimensional Y / C separation processing for separating the luminance component and the color component of the video signal reproduced by the reproducing means using the video signal in the same field in accordance with the output motion of Motion-adaptive Y / C separation means for switching between the three-dimensional Y / C separation processing for separating the luminance component and the color component of the video signal reproduced by the reproducing means using the video signal of Y / C separator There video recording and reproducing apparatus characterized in that it comprises a display means for displaying an image showing a luminance component signal and the color component signals separated. 9. A video recording / reproducing apparatus according to claim 7, wherein the recording is a magnetic tape, and more particularly, a video tape recorder.