JPH01229586A - Multiplexing signal extracting device - Google Patents

Abstract

PURPOSE:To surely extract an additional signal by deciding the contents of a movement adaptive process performed at the transmission side with decision of the maximum likelihood not with detection of movement. CONSTITUTION:In case of a still picture, the average signals of 1st and 2nd fields are repetitively transmitted in every 1st and 2nd fields. In case of a moving picture, a movement adaptive process is carried out at the transmission side so that the signals of the 1st or 2nd field are transmitted. At the reception side the average signal is obtained between the upper and lower lines for each field together with a difference signal secured between the line average signal of one of both fields and the signal of the other field. Then the amplitudes of these three signals are compared with each other so that the contents of the movement adaptive process carried out at the transmission side can be decided. Thus it is possible to decide the contents of the movement adaptive process of the transmission side with no deterioration in transmission efficiency nor the influence of transmission noise.

Description

[Detailed Description of the Invention] [Objective of the Invention] (Field of Industrial Application) This invention maintains compatibility with an NTSC color television broadcasting system, while also providing a signal that is separate from the original color television signal in this system. The present invention relates to multiplexed signal extraction @l suitable for extracting a multiplexed signal on the receiving side when transmitting a color television signal multiplexed with an original color television signal.

(Prior Art) The NTSC system, which is one of the color television broadcasting systems, can be said to be an excellent system that is compatible with black and white television broadcasting and has sufficient performance as a color television broadcasting system. This can be seen from the results of implementation in Japan, the United States, and other countries.

Incidentally, over its long history, the image quality of the NTSC system has been significantly improved since its initial implementation as a result of constant efforts by both the transmitting and receiving sides.

However, in the NTSC system, there is a desire for further improvement in image quality due to the recent spread of large screen displays.

IDT is a method for improving the image quality of the NTSC system.
V (1Broved Definition Te
There is a method called 1 edition). This method aims to improve image quality by making full use of the transmitted NTSC color television signal (hereinafter referred to as NTSC signal) on the receiving side.

This IDTV could not be implemented under conventional analog technology, but has become possible due to recent advances in digital technology.

According to this IDTV, compared to the conventional analog system,
Image quality can be significantly improved.

However, since this IDTV is based on the NTSC system, the upper limit of the image quality that can be improved is limited by the NTSC standard. Here, the upper limit items for the system include (1) screen aspect ratio (2) horizontal resolution of 330 Tv lines.

The aspect ratio of (1) is currently 4:3, but it is known that users prefer ratios such as 5:3 or 6:3 (Broadcasting System published by Japan Broadcasting Publishing Association (Editor) (See page 80 of 2 Japan Broadcasting Corporation)).

In addition, the high-definition television broadcasting system (H1ghDef'
1nition Te1evision), 16
:9 aspect ratio may be adopted (CCI
RReport 801-2).

Regarding the horizontal resolution (2), in the NTSC system, 4
．． Since it is specified as 2MHz, the limit is 330Tv. On the other hand, considering the number of effective scanning lines (480 lines), the vertical resolution can be 450 TV lines even if you take into account margins such as opacity. Therefore, at this stage, it is desired to improve the horizontal resolution in terms of horizontal and vertical balance.

As an example of a system that improves the above two items and maintains compatibility with current television receivers, for example, 5 Josep
h L, LoClcero “A Compatible
le High-Definitlon televi
sion System (SLSC) with Ch.
roinance and Aspect Rati
o lnpuruvements'”5WPTE Jo
urnal, May 1985. Below, these 5
The LSC method will be described.

FIG. 21 shows a spectrum diagram of the 5LSC method.

In FIG. 21, signals of 0 to 4.2 MHz are signals for maintaining compatibility with current television receivers. The signal from 4.9 to 10.1 MFLz is an additional signal used to expand the aspect ratio and the resolution of one luminance and one chromaticity. Therefore, in this 5LSC system, the signal for one station is transmitted using the band for two channels, and one channel basically transmits a signal similar to the current television broadcast signal, and the other channel transmits a signal that is basically similar to the current television broadcast signal. , and sends additional signals to improve image quality.

According to such a configuration, channels received by current television receivers do not include additional signals, so
It is considered that there is high compatibility with respect to interference. However, 1
It is not efficient because it occupies two channels per station. Particularly in situations where channel allocation is near the limit, such as in Japan, implementation is expected to be difficult. In addition, when considering intra-station and inter-station transmission, current television broadcasting equipment is
Since it does not have a band that extends to 0 MHz, it is necessary to invest in all new equipment.

From the above, it is preferable to perform transmission within the band of one channel. Moreover, if additional signals can be multiplexed around the baseband of 4.2 MHz, compatibility with current television and broadcast equipment such as video tape recorders and transmitters can be achieved.

As one method of multiplexing additional signals to around 4.2MTlz of the baseband, T, Puklnukl st.

”Extended Deflnlslon TV
Fully CoBable with E
xlsting 5standards” I
EEE Tr, onCou*unicatlon
Vol, C0M-32NO, 8, August 19
There is a method according to 84.

In the NTSC system, in the case of still images, this method uses luminance detail components (approximately 4 to
This is a 6 MHz signal (hereinafter referred to as a brightness high frequency signal) that is multiplexed. Here, the unused area is
In the vertical-time direction spectrum diagram of FIG. The area of the third quadrant is used. In addition, in the figure, C
is a color difference signal.

By the way, this method is applicable only to still images and not to moving images. In the case of video, the spectrum spreads in the time direction and the original NTS
This is because the C signal and the additional signal (luminance high frequency signal Y, . . . ) overlap, making it impossible to separate both signals on the receiving side.

Since the brightness high frequency signal YH is effective for still images, even if the above method can transmit the additional signal only for still images, the objective of improving the resolution of still images can be achieved.

However, when transmitting a signal for enlarging the aspect ratio as an additional signal, the additional signal must be sent not only for still images but also for moving images. Therefore, the above-mentioned additional signal multiplexing method, which can transmit additional signals only for still images, cannot be used to transmit additional signals for expanding the aspect ratio.

In order to solve this problem, it is conceivable to limit the motion component of the current NTSC signal, but doing so is likely to cause unnatural motion as a side effect, and the existing television reception Compatibility with the machine will be impaired.

Furthermore, since the luminance high frequency signal YH has a much lower level than the low frequency component in the case of a general natural image, even if it is multiplexed, there is little interference with current television receivers. On the other hand, when transmitting an additional signal for expanding the aspect ratio, it is necessary to transmit from high-level low-frequency components to high-frequency components, which poses a problem of interference with the current NTSC signal. In order to solve this problem, it is possible to lower the transmission level of the additional signal, but if this is done, a new essential problem arises in that the signal-to-noise ratio during reception and reproduction deteriorates.

As mentioned above, in order to multiplex an additional signal to expand the aspect ratio within 4.2MHz of the baseband, it is necessary to: ■ Be able to transmit the additional signal regardless of whether it is a moving image or a still image ■ Current televisions It is necessary to satisfy the following conditions: there is little interference to the John receiver, and the signal-to-noise ratio (S/N ratio) during reception and reproduction of the additional signal can be secured. However, at present, no method has been developed that can satisfy these two conditions.

(Issue 8 to be solved by the invention) This invention has been made to deal with problem (2) of the above two problems.

Specifically, when motion adaptive processing is performed on the transmitting side in order to transmit additional signals for still images and moving images, the receiving side reliably determines the content of the motion adaptive processing on the transmitting side and calculates the amount of information from the received signal. It is an object of the present invention to provide a device that can ensure extraction of additional signals.

In other words, when motion adaptive processing is performed on the transmitting side of a television broadcasting system, and motion adaptive processing is performed on the receiving side in synchronization with the processing on the transmitting side, on the receiving side, as well as on the transmitting side, the image is converted from the image signal. If motion is detected, noise will be superimposed during transmission, making it impossible to synchronize motion adaptive processing with the transmitting side. This is because normally, when performing motion detection, motion detection is performed such that when the difference value between frames of a predetermined image signal is greater than or equal to a threshold value, it is used as a moving image, and when it is less than the threshold value, it is used as a still image. This is because if the S/N ratio of the image signal decreases due to transmission noise or the like, malfunctions in motion detection will increase.

This makes it impossible to accurately determine the content of the motion adaptive processing on the transmitting side, and it becomes impossible to accurately extract the additional signal.

In order to determine the content of motion adaptive processing on the transmitting side without being affected by transmission noise, a signal indicating motion may be transmitted separately from the image signal.

However, when performing motion adaptive processing on a pixel-by-pixel basis, the same bandwidth as that required for image signal transmission is required to transmit signals indicating motion, which reduces transmission efficiency. Significantly decreased.

Therefore, this invention provides the following advantages:
The present invention aims to provide a multiplexed signal extraction device that can determine the content of motion adaptive processing on the transmitting side without being affected by transmission noise, and can ensure extraction of additional signals.

[Structure of the invention] (Means for solving the problem) In order to achieve the above object, the present invention provides the following in the case of still images:
1st. The average signal of the second field is the average signal of the first field. It is repeatedly transmitted for each second field, and in the case of a moving image, the first field.
It is assumed that motion adaptive processing is performed on the transmitting side so that the signal of one field is transmitted among the second fields, and on the receiving side, the average signal of the upper and lower two lines of each field is calculated, and the signal of one field is transmitted. Find the difference signal between the line average signal of the field and the signal of the other field,
By comparing the amplitudes of these three signals,
The content of the motion adaptive processing on the transmitting side is determined.

(Function) As described above, the present invention uses maximum likelihood judgment to determine the content of motion adaptation processing on the transmitting side, without using motion detection, so there is no reduction in transmission efficiency. Of course, the determination can be made without being affected by transmission noise, and the additional signal can be reliably extracted.

(Embodiments) Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a circuit diagram showing the configuration of an embodiment of the present invention. Before explaining FIG. 1, an example of an additional signal multiplexing television signal broadcasting system to which the present invention is applied will be explained with reference to FIGS. 2 to 19.

In FIG. 2, 11 is an input terminal for color television signals. This signal has an aspect ratio of 16:9, a number of scanning lines of 525, and a frame frequency of 59.94Hz (hereinafter referred to as
It is a progressive scanning (non-interlaced) signal of 60 Hz). In the drawings, such non-interlaced signals are expressed as 525/60. This signal consists of a luminance signal represented by Y and chromaticity signals represented by I and Q.

The signal supplied to the input terminal 11 is sent to the screen division filter 1.
2. This screen division filter 12 divides the input signal into a portion corresponding to the center portion F1 of the screen F in FIG. 3 and a portion corresponding to the side portion F2. Here, the aspect ratio of the screen center portion F1 is set to 4:3.

Screen center portion Fl output from the screen division filter 12
The signal (hereinafter referred to as the center signal) is sent to the time expansion circuit 1.
3, and the time axis is expanded by 5/4 times. on the other hand,
The signal of the screen side portion F2 (hereinafter referred to as side signal) is expanded four times by the time expansion circuit 14. Figure 4 shows the state of time expansion. Of the effective horizontal scanning period of 53 μs in terms of interlacing, 42 IJs is allocated to the screen center portion Fl, and 11 μs is allocated to the screen side portion F2. This relationship is (42 + 11): 42 , assuming an overscan of about 6%,
It can also support an aspect ratio of 16:9. Hereinafter, the explanation will be made using parameters based on the relationship shown in Figure 4.
If you do not want to allow overscanning, you can slightly change the parameters described below.

The center signal is time-expanded by a factor of 5/4, resulting in a band of 0 to 10 MHz. Among the time-expanded center signals, the luminance signal Y is supplied to the luminance high frequency separation circuit 15. This brightness high frequency separation circuit 15 receives input signals from 8 to 8.
It is separated into a 10 MHz brightness high band signal YH and a 0 to 3 MHz brightness low band signal YL.

The luminance high frequency signal YH has its level suppressed by a level conversion circuit 15, and then is converted by a YH enhancer 17 into a signal suitable for multiplexing. On the other hand, the luminance low-frequency signal YL is supplied to the NTSC encoder 19 after being subjected to blurring processing in the motion adaptive blurring processing circuit 18 to make it a signal suitable for multiplexing with the luminance high-frequency signal YH and the side signal. . At this time, the spectrum of the luminance low-frequency component YL is limited to a region as shown in FIG. 5(a).

Among the center signals output from the time expansion circuit 13, the chromaticity signals I and Q are converted to NTSC by the color difference band limiting filter 20.
After being limited to a band that meets the standard, it is supplied to the NTSC encoder 19. And this NTSC encoder 1
9, the luminance low frequency signal YL is converted into an NTSC color television signal, and then the addition circuit 21
is supplied to

On the other hand, the side signal is time-expanded four times by the time expansion circuit 14, and is made into a signal with a band of 2.2 MHz.

This time-expanded side signal is sent to the time-division color multiplexing circuit 22.
is supplied to This time-division color multiplexing circuit 22 receives color difference signals 1. After band-limiting Q to 0.25 MHz, line sequential multiplexing is performed. Furthermore, by time-division multiplexing this line-sequential multiplexed signal and the luminance signal Y, the signal shown in FIG. 6 is obtained. In this case, the color difference signal I.

The amplitude of Q is set to 1.33 (110,75) times the normal value. This makes it possible to improve the S/N ratio on the receiving side.

The output of the time-division color multiplexing circuit 22 is output by the band compression circuit 23 at 525/4 r per 1/30 [second] in the vertical direction.
c, p, h], and the horizontal direction is compressed to a band of 1 [MHz]. The level of this compressed output is suppressed by the level conversion circuit 24, and then converted by the side information encoder 25 into a signal suitable for multiplexing. The spectrum of this converted signal is located in the shaded area in FIGS. 5(a) and (b), and as shown in FIG. 5(a), it is separated from the center signal in the horizontal and vertical spectral regions. in position.

The output of the side information encoder 25 is supplied to the adder circuit 21 and added to the output of the NTSC encoder 19.

This addition output is sent to the YH encoder 1 by the addition circuit 26.
It is added to the luminance high frequency signal YH output from 7. This addition output becomes a transmission signal.

FIG. 7 is a circuit diagram showing an example of a specific configuration of the motion adaptive blurring processing circuit 18.

In this FIG. 7, a luminance low frequency signal YL having a band of 0 to 8 MHz is supplied to the input terminal 1b from the luminance high frequency separation circuit 15 shown in FIG. 2 above.

This luminance low-pass signal YL is supplied to a horizontal and vertical two-dimensional low-pass filter (hereinafter referred to as LPF) 2b which has the shaded area in FIG. Ingredients are removed. This diagonal high-frequency component has little contribution to visual perception, so even if it is deleted, it has almost no effect on deterioration of image quality.

The output of the two-dimensional LPF 2b is supplied to the horizontal HPF 3b, and high frequency components of 4 to 8 MHz are extracted.

This high frequency component of 4 to 8 MHz is subjected to motion adaptive blurring processing. That is, this 4-8 MHz component is supplied to a still image processing circuit 4b and a non-interlace/interlace (hereinafter referred to as NINT/INT) conversion circuit 6b. As shown in FIG. 8, the still image processing circuit 4b has a frame delay circuit IC that delays an input signal by one frame, and an adder circuit 2C calculates the sum of the input and output signals of this frame delay circuit 1C. , the output signal of this adder circuit 2C is halved by a 1/2 coefficient circuit 2C. As a result, an average output of image signals for two frames (in this case, image signals at 1/30 second intervals) is obtained from the 1/2 coefficient circuit 3C.

This average output is supplied to the NINT/INT conversion circuit 5b and converted into an interlaced signal. This NINT
The /INT conversion circuit 5b has a line distribution circuit 1d, as shown in FIG.
The signal for one frame is divided into odd line and even line signals.

Then, the signal on one line is transferred to the field delay circuit 2d.
By selectively selecting this signal and the signal on the other line by the switch circuit 3d in accordance with the field switching signal, an interlaced signal is obtained. Here, the output signal of the switch circuit 3d is
Although the signal is in the form of an interlaced signal, since the information for one frame is divided into two fields and transmitted, there is no motion component between the two fields.

Therefore, there is no generation of vertical high-frequency components due to so-called interlace folding that occurs when there is movement, so there is no worry of crosstalk to additional signals.

The above NINT/ to which the output of the above horizontal HPF 3b is supplied
The INT conversion circuit 6b performs scanning line conversion by deleting even-numbered line signals among the 525 scanning line signals in a certain frame, and deleting odd-numbered line signals in the next consecutive frame. . Therefore, in this case, the motion component is preserved between two consecutive frames.

The output of the NINT/INT conversion processing circuit 6b is supplied to the switching circuit 8b after being subjected to predetermined moving image processing by the moving image processing circuit 7b. This switching circuit 8b is NINT/
Either one of the outputs of the variable force circuit 5b and the logical circuit 7b of the INT variation circuit 5b is selected, and the control thereof is performed by the motion return detection circuit 9b. The motion aliasing detection circuit 9b controls the switching circuit 8b so that the output of the video processing circuit 7b is selected when the output signal of the NINT/INT conversion circuit 6b has an aliasing component due to motion, and there is no aliasing component. At this time, the switching circuit 8b is controlled so that the output of the NINT/INT processing circuit 5b is selected.

Note that details of the moving image processing circuit 7b and the motion aliasing detection circuit 9b will be described later.

Among the 0 to 8 MHz components output from the two-dimensional LPF 2b, the 0 to 4 MHz components are extracted by subtracting the input and output of the horizontal HPF 3b by the subtraction circuit 10b, and are extracted by the NINT/INT conversion circuit 11b. Converted to interlaced signal. This conversion process is
The conversion process is the same as that of the NINT/INT conversion circuit 6b described above.

The output of the NINT/INT conversion circuit 11b and the output of the switching circuit 8b are added by an adding circuit 12b.

The reason for using such adaptive operation is as follows.

Regarding motion, since the same signal is transmitted by field repetition, two consecutive fields within one frame (1/30 second) can be treated as a still image. but,
In such a method, movement becomes a deterioration factor. This will be explained with reference to FIG. The figure shows a rectangular pattern moving horizontally (from left to right) at a constant speed. FIG. 10(a) shows the original signal, in which smooth motion is expressed from field n to field n+3. FIG. 10(b) shows the case of field repetition described above. According to FIG. 10(b), it can be seen that the edge portion swings left and right with respect to the center of gravity of the movement. Such discontinuity in display with respect to continuity of movement is visually recognized as jerky and unnatural movement called motion jerkiness. For example, it is also reported in Makoto Miyahara, "Relationship between visual characteristics and image quality for moving images and its application to TV signal band compression," J NHK Technical Research, 1970, Vol. 27, No. 4, pp. 141 to 171. As described above, the above-mentioned field repetition has a narrow tolerance in terms of visual characteristics in terms of smoothness of motion.

Therefore, in this embodiment, when the width of movement between two fields becomes large, as shown in FIG. 11(a), the luminance signal Y is Remove horizontal high frequency components. That is, in the motion area, locally field thinning transmission is performed. Therefore, the motion area is 30
The area shown by the broken line in Figure 11(a), which causes Hz flicker, is the so-called uncovered background, which is the area left after movement, and is not a very useful area in terms of image quality, so it is visually There is almost no deterioration. As shown in FIG. 11(b), on the receiving side, the horizontal high-frequency components of the luminance signal Y can be deleted and reproduction can be performed using a field in which only the additional information is superimposed.

FIG. 12 is a circuit diagram showing an example of a specific configuration of the moving image processing circuit 7b of FIG. 7.

In FIG. 13, the output signal of the NINT/INT conversion circuit 6b of FIG. 7 is supplied to an IH delay circuit 1e and an adder circuit 2e. The adder circuit 2e is the IH delay circuit 1
Add the input and output signals of e. The amplitude of this addition output is halved by the 1/2 coefficient circuit 3e, and becomes an average signal of two lines worth of signals.

The two-line average signal outputted from the 1/2 coefficient circuit 3e and the output of the field delay circuit 10e, which will be described later, are compared in terms of absolute value of amplitude by an absolute value comparison circuit 4e. As a result, among the two consecutive fields of signals,
A more significant signal is determined. This determination signal is supplied to the median filter 5e, and isolated point noise components are removed. The judgment signal from which the noise component has been removed and the signal delayed by one line by the line delay circuit 6e are combined by an OR circuit 7e to form a continuous judgment signal for two lines. Here, when the output of the field delay circuit 10e has a larger amplitude than the line average signal and is determined to be a significant signal, the switch circuit 9e is turned off and the information of the second field is deleted. That is, the area of the n+1 field indicated by the broken line in FIG. 11(a) is deleted. This is a determination to perform field thinning by deleting information in the second field of the interlaced structure.

Next, in exactly the same way as the above process, the field delay circuit 10
The two-line average signal of the output signal of e is sent to the line delay circuit 11.
It is obtained by the e1 adder circuit 12e and the 1/2 coefficient circuit 13e. Then, this two-line average signal and the field delay circuit 10
From the input signal e, continuous determination signals for two lines are obtained by the absolute value comparison circuit 14e, the median filter 15e, the 1-line delay circuit 16e, and the 2-OR circuit 17e. When the input signal of the field delay circuit 10e has a larger amplitude than the two-line average signal and is determined to be a significant signal, the switch circuit 19e is turned off and field thinning is performed to delete the information of the first field. . This means that the broken line portion of the n field in FIG. 11(a) has been deleted.

Since the above video processing is performed in units of one frame, only one field period in which the first field signal appears at the output of the field delay circuit 10e and the second field signal appears at the input, the switch circuit 8e , 18e are turned on, and the operations described above are performed.

Through the above processing, when there is significant information in either the first field or the second field within one frame, the luminance signal Y is deleted from the two adjacent upper and lower lines of the other field, and the luminance signal Y is locally Field thinning status is obtained. This signal is processed as field thinning in the motion area by prohibiting field repetition, so that the smoothness of the motion is not impaired.

FIG. 13 is a circuit diagram showing a specific configuration of the motion aliasing detection circuit 9b of FIG. 7.

In this Fig. 14, N I NT/INT in Fig. 8
The interlaced signal output from the conversion circuit 6b is processed using a field delay circuit 1f so that the signals of the first and second fields are simultaneously processed into horizontal and vertical two-dimensional HPFs 2f and 3f.
is supplied to These two-dimensional HPFs 2f and 3f have a pass band defined by the shaded area in FIG. 4. The component of this region is the seventh
Since it has been deleted in the horizontal and vertical two-dimensional HPF 2b in the figure, it does not originally exist. However, in the case of FIG. 7, after this deletion, the NINT/INT conversion circuit 6b
Scan line conversion is performed at As a result, in the case of a moving image, aliasing components occur in the shaded area in FIG. 4 due to this scanning line conversion. This aliasing component is extracted by two-dimensional I (PF2f, 3f.Here, two 2-dimensional HPFs with the same characteristics are used to extract this aliasing component.
When processing a signal in subsequent signal processing, the first . This is because it is more convenient to use the signal structure based on the second field. That is, there is a line offset relationship between the first and second fields, and the center of gravity of the signal is shifted by one line, so two two-dimensional HPFs 2 f are applied to each field.

3f is used. Here, the two-dimensional HPF
The signal structure of the output of 2f is the signal structure of the first field,
In order to use the signal structure of the output of the two-dimensional HPF 3f as the signal structure of the second field, the two-dimensional HPF 2f, 3
Of the outputs of f, only the output of the two-dimensional HPF 3f is delayed for one field period by a field delay circuit 4f, and then both are alternately selected by a switch circuit 5f.

The feature of the operation in FIG. 13 is that the first .
The second field-to-field operation is performed, and no inter-frame operation is performed.

The output of the switch circuit 5f has an interlaced structure. After the absolute value of this signal is taken by an absolute value circuit 6f, it is converted into an aliasing component detection signal by a nonlinear circuit 7f. This converted output is supplied to a switching circuit 8b in FIG. 7 after isolated noise components are removed by a median filter 8f.

FIG. 14 is a circuit diagram showing an example of a specific configuration of the band compression circuit 23, level conversion circuit 24, and side information encoder 25 shown in FIG. 2. Note that in FIG. 14, the band compression circuit 23 and the side information encoder 25 are confused, so for convenience of explanation, the level conversion circuit 24 is shown on the human-powered side, but in principle, the band compression circuit 23 and the side information encoder 25 are It can be inserted anywhere in the signal processing path.

As shown in Figure 5 above, if a signal is transmitted using the area allocated for side information (the shaded area in the figure), the permissible amount of information is 525 vertically per 1/30 second. /4 [c, p, hl x horizontal 1 [MHz
It is the information amount of l.

The circuit shown in FIG. 14 converts an input signal with a frame frequency of 60 Hz into a signal every 1/30 seconds using a frame thinning circuit 5g.

Because frames are thinned out on the sending side in this way,
On the receiving side, reproduction is performed using frame interpolation. In this case, to reduce the unnaturalness of the movement,
The sum of the input signal and the output signal of the frame delay circuit 1g is taken by the adder circuit 2g, and this is halved by the 1/2 coefficient circuit 3g.
Multiply this to obtain the average output of the signal for two frames. 2D L
PF4g limits the side signal to the spectrum characteristics shown in FIG. 15(a). During transmission, 1/2 of the information is processed by the line thinning circuits 13d and 20d for interlace conversion, so the amount of information is 525/4 [c, p, hl xl [
M7]. The signal obtained in this way is
The information is supplied to the field thinning circuit 5g, and the information is thinned out for each frame, resulting in a frame frequency of 30H2.

The vertical LPF 6g extracts a signal having the spectrum shown in FIG. 15(b) from the signal having the spectrum shown in FIG. 15(a) output from the field thinning circuit 5g. The adder circuit 7g subtracts its output signal from the input signal of the vertical LPF 6g.

By passing this subtraction output through the horizontal LPF8g, the first
A signal having the spectrum shown in FIG. 5(C) can be obtained. The vertical frequency shifter 9g performs a 4-line inversion process on this signal in the vertical direction to produce the signal shown in FIG. 15(d).
Obtain a signal with the spectrum shown in .

The line thinning circuit 10g performs line thinning processing on the output signal of the vertical frequency shifter 9g to obtain a signal with the number of scanning lines being 262.5 and one horizontal period being time expanded to 64 μs. The frame frequency of this signal is 3
07, the band is from 0 to 0.5 MHz. The line distribution circuit 11g divides the input signal, and the number of scanning lines is 131.
Get two signals. The time compression/time division multiplexing circuit 12g is
The two signals output from the line distribution circuit 11g are
By time-compressing the signal to 2 and time-division multiplexing the compressed output, a signal having the number of scanning lines of 131×2 in one horizontal period is obtained. As a result, a signal with 131 scanning lines, a frame frequency of 30 Hz, and a band of 0 to IMHz can be obtained.

The line interpolation filter 13g converts the signal with 131 scanning lines output from the time compression/time division multiplexing circuit 14g into a signal with 525 scanning lines.

At this time, the signal peak value becomes 1/4 and is distributed over 525 scanning lines. This signal is transmitted every 1/30 seconds, 1/6
Since the output is only for 0 seconds, the line distribution circuit 14
Distribute to odd and even lines using g. Then, after the signal of one line is delayed by one field by the field delay circuit 15g, both signals are selectively selected by the switch circuit 16g according to the field switching signal, so that the number of scanning lines is 262. 5. Obtain a signal with a field frequency of 607 and a band of 0 to IMHz.

The output of the vertical LPF 6g is further supplied to a line thinning circuit 17g. This line thinning circuit 17g thins out the number of scanning lines of the input signal to 131, and doubles the time. As a result, the number of scanning lines is 131 and the frame frequency is 30Hz.

Signals in the band 0 to IMHz can be obtained. This signal is converted to 525 scanning lines by a line interpolation filter 18g, and its peak value is reduced to 1/4. Therefore, no total signal energy is converted. Thereafter, this signal is transmitted to the scanning line number 2 by a line distribution circuit 19g, a field delay circuit 20g, and a switch circuit 21g.
62.5 lines, field frequency 60Hz, band 0 to IM
This becomes the TIz signal.

The outputs of the switch circuits 16g and 21g are sent to the multiplier circuit 22g.
, 23g. The frequency of this subcarrier for orthogonal modulation is (6/7) fsc = 195 fH - 3,07 MHz, where fH is set to the horizontal synchronization frequency, and the phase is inverted for each field. This phase inversion is performed by a phase shift circuit 24g and a switch circuit 25g.

Note that 26g is a phase shift circuit for giving a phase difference of π/2 to the subcarriers supplied to the multiplication circuits 22g and 23g.

The outputs of the multiplier circuits 22g and 23g are supplied to an adder circuit 27g, where the in-phase and quadrature modulation components are summed. A horizontal band pass filter (hereinafter, a band pass filter will be referred to as BPF) removes unnecessary components other than 2 to 4 MHz from this addition output. As a result, a multiplexed signal having a spectrum indicated by diagonal lines in FIG. 4 is obtained. Note that the multiplication circuit 25d
.

The signal input to 26d is a signal having a spectrum shown in a solid line frame in FIG.

According to such a configuration, there is no need to perform field thinning, and information can be transmitted for each field, thereby eliminating deterioration of motion. However, since oblique components are largely removed, the resolution characteristics deteriorate.

FIG. 17 shows Y, which performs multiplexing processing of the luminance high-frequency signal YH,
An example of a specific configuration of the encoder 17 is shown.

In order to transmit the luminance high frequency signal YH, the area marked YH in FIG. 5, that is, vertical (525/2) ± (1/1B)
c, p, h, ] horizontal 1 to 2 [M) lz], and only the static area portion is multiplexed.

As an effect of reducing noise power at the cost of reducing the amount of transmitted information, the level conversion circuit 16 converts 8 to 10 PvI in order to reduce the transmission level and reduce interference when receiving an NTSC television receiver. Decrease the amplitude of the Hz frequency component.

Regarding the transmission of the horizontal luminance high-frequency signal, even if the vertical spectrum is limited to a certain extent, the effect will not deteriorate much, so the vertical LPF 1h is used to limit the vertical spectrum to 525/8 [c, p, h]. Since we are considering transmitting only still images, the switch circuit 2h extracts one field's worth of information in two frames.

Since this information can be expressed by 131 scanning lines, the line thinning circuit 3h thins out the number of scanning lines to 1/4. Then, the even and odd lines of the signal having 131 scanning lines are transferred to an offset line distribution circuit 4h.
frame delay circuit 5h for only one line information.
After being delayed, the information on both lines is switched at a frame period by a switch circuit 6h and output. As a result, a signal with 65 scanning lines and a frame frequency of 30 Hz is obtained.

The vertical interpolation filter 7h converts this signal into a signal with 525 scanning lines. The offset line distribution circuit 8h is
This signal is divided into odd line and even line signals. The signals of these two lines are processed by a field delay circuit 9h and a field changeover switch 10h, so that the number of scanning lines is 26.
It is converted into a signal with 2.5 lines and a frame frequency of 60 Hz. This signal is processed by the next multiplier circuit 11h so that the frequency is (
12/7) Multiplex modulation is performed using subcarriers whose phase is inverted for each field at f sc [Hzl.

As a result, the output signal of the field switch circuit 10h is converted to the YH region in FIG. 4. This conversion signal is
The signal is supplied to the BPF 13e only when the image is a still image via a switch 12h controlled by a still image determination signal. And this BP
Only the 1-2 MHz components are extracted by F13h. This extracted output is supplied to the adder circuit 26 in FIG. 1 and multiplexed with the side signal and the like.

FIG. 18 shows processing blocks on the receiving side.

In FIG. 18, 41 is an input terminal to which a received signal is supplied. The received signal supplied to this input terminal 41 is processed by an NTSC decoder 42 into a luminance signal Y and a chromaticity signal 1. It is decoded by Q. In this embodiment, since all the additional signals are included in the luminance region, the luminance signal output of the NTSC decoder 42 includes the additional signals. Therefore, the luminance signal output of the NTSC decoder 42 is Y,
The signal is supplied to both the decoder 43 and the side information decoder 44 and decoded. However, in FIG. 20, in order to simplify the explanation, the NTSC decoders 42, Y,
The decoder 43 and side information decoder 44 are shown separately, and since the center signal is processed while containing the additional signal, some interference remains in the additional signal. Therefore, an actual television receiver includes an NTSC decoder 42, a YH decoder 43, and a side information decoder 4.
4 signal separation circuits operate as one, each decoder 42
．． The configuration is such that unnecessary signals are not input to 43 and 44.

The 4-5 MHz brightness high-band signal YH decoded by the YH decoder 43 is added to the brightness low-band signal YL of the center signal in an adder circuit 45 . As a result, a luminance signal Y having a band of 0 to 5 MHz is obtained.

This luminance signal Y is time-compressed by a factor of 415 in the time compression circuit 46, so that it has a band of 0 to 6.25M&. This time compressed signal is processed by a non-interlaced conversion circuit 47 with a number of scanning lines of 525, a frame frequency of 60 Hz, and a band of 0 to 1.
It is converted to a 3MHz signal.

On the other hand, the side signal decoded by the side information decoder 44 is time-compressed by a factor of 1/4 by the time compression circuit 48. As a result, the number of scanning lines is 525, the frame frequency is 60,
A signal in the band 0 to 8 MHz is obtained.

This time-compressed side signal and the center signal output from the interlace conversion circuit 47 are combined by a screen synthesis circuit 49 and converted into a signal having a large aspect ratio of 16:9. Luminance signal Y, chromaticity signal 1. output from this screen synthesis circuit 49. Q is R, G by the inverse matrix circuit 50.

The signal is converted into a B primary color signal and used for display.

FIG. 19 shows an example of a specific configuration of the side information decoder 44.

In the figure, an additional signal multiplexed composite signal output from an NTSC decoder 42 is supplied to the Y/C separation circuit 11. This Y/C separation circuit 11 converts the input signal into a luminance signal Y
and chromaticity signal C. Of these, the luminance signal Y is supplied to the additional signal extraction filter 21. This additional signal extraction filter 21 extracts an additional signal included in the input luminance signal Y from the input luminance signal Y. The extracted additional signal is supplied to the orthogonal synchronous demodulation circuit 31 and reproduced into a baseband signal using a subcarrier having a frequency of 6/7fSC (subcarrier frequency). This demodulated output is converted into a television signal as an additional signal by a band compression decoding circuit 41.

Note that the subcarrier for demodulation is generated by the subcarrier recovery circuit 51 according to the additional signal multiplex decoded signal.

Now, returning to FIG. 1, one embodiment of the present invention will be described in detail. Note that FIG. 1 is a concrete example of FIG. 19. In FIG. 1, the field delay circuit 2
j9, line delay circuits 3j, 4j, addition circuits 5j, 6j
, 1/2 coefficient circuits 7j, 8j, 10j, absolute value circuit 11
j1 median filter 12j1 minimum value judgment circuit 13j1
Switch circuits 14j, 15j, 17j, 18j, vertical H
It is configured by PF19j.

In FIG. 1, the horizontal HPF 1j extracts components of 2 to 4 MHz from the input signal of the side information decoder 44. It is now assumed that the signal being input to the horizontal HPF 1j is an n+1 field signal. In a signal with 525 scanning lines made from two consecutive fields of n and n+1 in one frame, the information of adjacent scanning lines is originally (525/8) [c, p, hl, narrow band Since the components are expressed by 525 scanning lines, the correlation is very high and they can be considered to be almost identical signals.

As shown in FIG. 11(b), in the video area, n
, n+1 fields, only an additional signal whose phase is inverted for each field is multiplexed into one field, and the luminance signal Y is deleted. On the other hand, in the still image area, in addition to the additional signal that is inverted for each field, there is a luminance signal Y representing the same image.

The output signal of HPFlj is sent to field delay circuit 2j, 1-line delay circuits 3j and 4j, and addition circuit 5j.

6j, 1/2 coefficient circuits 7j, and 8j are supplied to upper and lower line averaging circuits. From this upper and lower line average circuit,
In one field of the moving image area, only the additional signal appears, and in the other field, the additional signal with the luminance signal Y superimposed thereon appears. Further, when the inter-field inversion average is obtained by the addition circuit 9j and the 1/2 coefficient circuit 10j, in the case of a still image,
The luminance signal Y is canceled and only the additional signal is obtained. On the other hand, in the case of a moving image, the luminance signal Y is not canceled out.

Therefore, by selecting the minimum output from the three outputs of the average output of the upper and lower two lines in each field of n and n+1, and the field inversion average output, it is possible to obtain a signal that does not include the luminance signal and consists only of the additional signal. can.

That is, the additional signals included in the three modes of average of the upper and lower two lines in each field of n and n+1 and field inversion average output are all the same, and the luminance signal Y is excluded in at least one mode depending on the image content. There is.

This algorithm is determined by the minimum value determination circuit 13. In this case, the signals of the above three modes are subjected to the absolute value of the amplitude in the absolute value circuit 11j, and after removing single pulse noise in the median filter 12'', are supplied to the minimum judgment circuit 13j, Used for minimum value determination. In addition,
The minimum value determination circuit 13j has two modes in which the same minimum value is taken.
If there are more than one, it is determined that the mode is still image mode.

The above is a determination algorithm for the three modes, and since this algorithm performs a type of maximum likelihood determination, it is characterized by fewer malfunctions due to transmission noise and the like.

In order to separate the additional signal and the luminance signal Y, 525 scanning line signals are required. In the still image mode, this signal may be generated from signals of fields n and n+1. Therefore, in the still image mode, the switch circuit 14j
, 15j select the signals of fields n and n+1, respectively.

In the n+1 field processing, the above-mentioned addition circuit 6j and 1/2 coefficient circuit 8j are processed from 262.5 scanning line signals.
The average of the upper and lower lines is taken at , and the phase of the average is inverted by an inverting circuit 16j to generate 262.5 scanning line signals. This signal is selected by the switch circuit 15j, and the original 262.5 scanning line signals are selected by the switching circuit 14j, resulting in a total of 525 scanning line signals. Similarly,
In the n-field processing, the addition circuits 5j11/
The average of the upper and lower lines is taken by the two-coefficient circuit 7J, and the phase of this is inverted by the inversion circuit 16j, and the original scanning line signal is sent to the switch circuit 14j.

Select with 15j.

The connection states of switch circuits 14j and 15j are controlled by the output of minimum value determination circuit 13j. A valid signal is only for one field period of one frame (two fields, 1/30 seconds). Therefore, this effect signal appears at 1/
Switch circuit 17j, 1 only for 60 seconds, 1 field period
8j is turned on and this valid signal is supplied to the vertical HPF 19j.

This vertical HPF 19 j is (525/2) ± (52
5/8) [Has a passband of c, p, hl and extracts an additional signal from the input signal. The extracted additional signals are orthogonally demodulated by multiplication circuits 20j and 21j. Horizontal LPF22j
extracts the 0-IMHz component from this demodulated output. As a result, the number of scanning lines is 262.5, and the frame frequency is 30)
A signal of 1z is obtained. This signal is the line thinning circuit 2
3j into a signal having 131 scanning lines.

The vertical spectrum of the signal output from the horizontal LPF22j is limited to the 525/8 [c, p, hl bands, so even if it is converted to a signal with 131 scanning lines by line thinning, the information will not be preserved. be done.

The signal shown in FIG. 15(d) is transmitted by time-division multiplexing two signals with 131 scanning lines, so the original number of 131 scanning lines is transferred to the time division circuit 24j1 and the time expansion circuit 25j. After returning to the two signals, the line superimposing circuit 26j converts it into a signal with 262.5 scanning lines. The spectrum of this signal is shown in FIG. 15(d).

After that, this signal is processed by the line interpolation circuit 27j to convert the number of scanning lines to 5.
Convert to 25 signals. Next, when this signal is shifted by (525/8)[c, p, hl] by the vertical frequency shifter 28j, a signal having the spectrum shown in FIG. 15(c) is obtained.

On the other hand, the signal on the multiplication circuit 21j side has 131 scanning lines and has a spectrum as shown in FIG. 15(b). This signal is converted into a signal with 525 scanning lines by a line interpolation circuit 29j, and then added to the output of the vertical frequency shifter 28j by an adder circuit 30j. This results in
The adder circuit 30j outputs a signal having the spectrum shown in FIG. 15(a). However, this signal is 1/3
Since the signal appears once every 0 seconds, the frame delay circuits 31j and 32j have a delay amount of 1/60 seconds.

An adder circuit 33j and a 1/2 coefficient circuit 33 generate a frame interpolation signal, and a switch circuit 35j selectively selects this signal and the output of the frame delay circuit 31j every 1/60 seconds,
A signal with 525 scanning lines and a frame frequency of 60 Hz is obtained.

By the way, as shown in FIG. 20, the frame interpolation signal is simply the average of the signals of the previous and subsequent frames, and therefore has an element that causes unnatural motion. Therefore, the motion detection circuit 36j detects the amount of motion of the image, and uses the detection output to control the high frequency component of the frame interpolation signal. That is, when there is movement, the unnaturalness of the movement is eliminated by suppressing the high frequency components of the frame interpolation signal.

Note that the high-frequency component of the field interpolation signal output from the 1/2 coefficient circuit 34j is extracted by the horizontal LPF 37j to which this frame interpolation signal is supplied and the addition circuit 38j that performs subtraction processing on the input/output signal of this horizontal LPF 37j. . The amplitude of this high frequency component is controlled by the multiplication circuit 39j according to the motion detection output. This control output is added to the low frequency component output from the horizontal LPF 37j in the adder circuit 40j.

The output of the switch circuit 35j is a level conversion circuit 41 having characteristics opposite to those of the level conversion circuit 24 on the transmitting side shown in FIG.
It is converted to the original signal at j. This signal is sent to color decoder 4
2j, luminance signal Y1 chromaticity signal 1. It is decoded by Q.

According to this embodiment described in detail above, it is possible to determine the content of motion adaptive processing on the transmitting side without reducing transmission efficiency, and without being affected by transmission noise.
It is possible to contribute to accurate extraction of side signals.

. This is because, in this embodiment, the content of the motion adaptive processing on the transmitting side is determined by maximum likelihood determination rather than by image motion detection. In this case, even if the judgment is incorrect, this
This corresponds to a case where there is little difference between a moving image and a still image, so no failure occurs.

Although one embodiment of the present invention has been described above in detail, it goes without saying that the present invention is not limited to this embodiment.

[Effects of the Invention] As described above, according to the present invention, it is possible to accurately determine the content of motion adaptive processing on the transmitting side without being affected by transmission noise and without causing a decrease in transmission efficiency. I can do it,
Additional signal extraction on the receiving side can be ensured.

[Brief explanation of the drawing]

FIG. 1 is a circuit diagram showing the configuration of an embodiment of the present invention, and FIG.
The figure is a circuit diagram showing the configuration of a transmission system to which the present invention is applied, Figures 3 to 6 are diagrams for explaining the operation of Figure 2, and Figure 7 is a motion adaptive processor shown in Figure 1. FIG. 8 is a circuit diagram showing an example of a specific configuration of the still image processing circuit shown in FIG. 7, and FIG.
A circuit diagram showing an example of a specific configuration of the NINT/INT conversion circuit shown in the figure, Figures 10 and 11 are diagrams for explaining the operation of Figure 7, and Figure 12 is the video processing shown in Figure 7 A circuit diagram showing an example of a specific configuration of the circuit, FIG.
FIG. 14 is a circuit diagram showing an example of a specific configuration of the motion aliasing detection circuit shown in FIG. 15 and 16 are diagrams for explaining the operation of FIG. 14, and FIG. 17 is the Y shown in FIG. 2. A circuit diagram showing an example of a specific configuration of an encoder, FIG. 18 is a circuit diagram showing a configuration of an additional signal multiplexing television signal receiving device, and FIG. 19 is an example of a specific configuration of a side information decoder shown in FIG. 18. FIG. 20 is a diagram for explaining the operation of FIG. 1, and FIGS. 21 and 22 are diagrams for explaining different examples of the conventional transmission system. 1j...Horizontal HPF, 2j...Field delay circuit, 3j, 4j...Line delay circuit, 5j, 6j, 9j
. 30j, 33j, 38j, 40j...addition circuit, 7j
, 8j, 10j, 34j...1/2 coefficient circuit, 11j
...Absolute value circuit, 12j...Median filter, 1
3j...minimum value determination circuit, 14j, 15j. 17j, 18j, 35j... switch circuit, 16j.
... Inverting circuit, 19j... Vertical HPF, 20j. 21j, 39j, . . . multiplication circuit, 22j, 37j, .
・Horizontal LPF, 23j...Line thinning circuit, 24j
...Time division circuit, 25j...Time expansion circuit, 26
j... Line superimposition circuit, 27j, 29j... Line interpolation circuit, 28j... Frequency shift circuit, 31j. 32j...Frame delay circuit, 36j...Motion detection circuit, 41j...Level conversion circuit, 42j...Color decoder. Applicant's agent Patent attorney Takehiko Suzue F2 F+ F2 Figure 3 (53LIS) Figure 4 Figure 8 Fee 1 = t7J1 sentence - Figure 9 Figure 16 (b) Figure

Claims (1)

[Claims] It has a 2:1 interlace scanning structure, has a horizontal frequency of fx [MHz] or more, and a vertical frequency of fy [cph].
] In the luminance signal from which the oblique high-frequency components have been removed,
If the component whose horizontal frequency is fx [MHz] or more and vertical frequency is less than fy [cph] is a still image, the average signal of both fields is calculated in the first and second fields of each frame. A first color television signal that is transmitted repeatedly, and in the case of a moving image, is transmitted so as to thin out the signal in one field of each frame, and a first color television signal that has a frequency that is an integral multiple of the horizontal synchronization frequency. , a second television signal modulated using a signal whose phase is inverted for each field as a carrier wave, and a second television signal that is multiplexed with the first television signal using the oblique high-frequency component deletion area. For each frame of the television signal received by the television signal receiving device that receives the television signal,
a first line averaging means for obtaining an average signal of two lines above and below the first field; and for each frame of said television signal received;
a second line averaging means for obtaining an average signal of two lines above and below the second field; a difference calculating means for obtaining a difference signal between the output of the line averaging means and the signal of the second field; By comparing the magnitude of the amplitude of the output of the second line averaging means and the output of the difference calculating means, if the image is a still image, a moving image, or a moving image, the above-mentioned addition of the above-mentioned first and second fields is determined. A determining means for determining which field is the field in which only the signal is present; and if the determining means determines that it is a still image, the signals of the first and second fields are selected, and the additional signal is selected from the first field. If the additional signal exists in the second field, select the signal of the first field and the output of the first line averaging means, and if the additional signal exists in the second field, select the signal of the second field. A multiplexed signal extracting device comprising: a selection means for selecting a signal and the output of the second line averaging means; and a filter means for selecting the second television signal from the output of the selection means. .