JPH0548669B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to a high-definition television receiver, and particularly to a high-definition television receiver suitable for connection with video equipment such as a video camera and a VTR.

[Background of the invention]

By significantly compressing the bandwidth of high-definition television signals,
The MUSE method has been proposed, which allows transmission via satellite broadcasting in the 12GHg band. (Ninomiya and 2 others "Satellite one-channel transmission system for high-definition television (MUSE)" Television Society technical report TEBS page 95-237 ~
(Page 42) As shown in Figure 1, this transmits a sample point that is 1/4 of the sample point of the original signal, and transmits a complete picture in 4 fields. , and then dot-interlaced and transmitted. The theoretical compression ratio of this method is 1/4.

Demodulation of this signal is performed as follows.

For still images, the received signal has a frame memory and is demodulated using information on all sample points in the current field and the past three fields.

At this time, since the point X has not been transmitted, it is reproduced by two-dimensional interpolation from sample points around it.

For moving image parts, interpolation and demodulation are performed using only the transmitted sample points of the field.

As described above, since the demodulation methods are different between the still image portion and the moving image portion, motion detection is required on the receiving side. Additionally, since images in videos are blurred, if the camera pans slowly, the entire screen will move in parallel, making the blur more noticeable. For this reason, so-called motion correction is required to eliminate the positional shift between the image of the past frame and the image of the current frame.

Due to these problems, the following two types of control signals are required in advance on the sending side. (Refer to the above document) (1) Motion vector: A signal that indicates how much the position has shifted between frames. (2) Motion detection mode control signal: Controls whether motion detection is performed using a 1-frame difference or a 2-frame difference. This kind of processing on the sending side increases the bandwidth.
This means that it can be compressed to around 8MHg, making satellite broadcasting in the 12GHg band possible.

Even when receiving transmission signals of this MUSE method, the home-use
We need a VCR that allows us to record at home, shift the time, and watch it whenever we want, just like a VCR. You will also need a home video camera that can be connected to this receiver and VTR to enjoy high-quality images. However, incorporating such video equipment into a system causes the following disadvantages.

Due to its mechanism, VTRs cause time axis fluctuations. Due to this variation in the time axis, the still image portion also appears to move, and there is a high possibility that the motion detection of the receiver will determine that it is a moving image and the image will be blurred.

In addition, signal dropouts called dropouts occur in VTRs, but if this occurs during the period in which motion vectors are inserted, motion vector information may not be obtained or may be erroneous, causing motion compensation to malfunction. be.

Furthermore, when considering home video cameras,
In order to convert the signal to MUSE, it is necessary to insert the two control signals mentioned above, but for this purpose, a frame memory, high-speed adder/subtractor, etc.
It is necessary to use large amounts of memory and high-speed logic elements. Therefore, it becomes difficult to reduce the size, power consumption, and cost of the device.

On the other hand, when assembling a system with a VTR or video camera after demodulating the original high-definition television signal,
This requires a high-speed AD converter and DA converter of approximately 64.8 MHg, which is four times the sampling frequency of MUSE (approximately 16.2 MHg), which poses an obstacle to reducing the cost of the receiver.

[Purpose of the invention]

An object of the present invention is to provide a high-definition television receiver that can provide overall high-quality images at low cost when combined into a system with a VTR and a video camera.

[Summary of the invention]

As mentioned above, MUSE performs inter-field offset sampling and further performs dot interlacing to compress the band. Therefore,
By performing motion compensation and demodulating the dot interlace by changing the interpolation process according to the motion and putting it into a state of offset sampling between fields, even if it is displayed as is, almost no aliased components of the signal will be perceptible due to the visual integration effect. Moreover, when the inter-field offset sampling state is restored to the original high-definition television signal by two-dimensional interpolation,
Changes in blur in still image portions due to motion detection malfunctions are extremely small and pose almost no problem. Therefore, for example, a MUSE signal that has been offset sampled between 4 fields can be interpolated between frames to perform offset sampling between 2 fields.
By demodulating to the state of the MUSE signal and recording it on a VTR etc. in the state of offset sampling between two fields, time axis fluctuations as mentioned above,
Problems caused by dropouts can be reduced. In addition, by introducing a configuration in which the MUSE signal played from the VTR and the signal from the video camera in the state of offset sampling between two fields is guided to a high-definition television receiver and demodulated to the original wideband television signal, the inter-frame There is no need for motion vector correction processing performed by interpolation processing or motion detection between two frames, and a simple motion detection circuit for demodulating the MUSE signal of offset sampling between two fields is sufficient.

Therefore, for example, when considering a high-definition television video camera for home use, a frame memory or a High-speed arithmetic elements can be greatly simplified.

In order to achieve the above object, the high-definition television receiver of the present invention converts the MUSE signal into an analog signal into a digital signal using an AD converter, performs motion correction processing using a vector using a frame memory, and This is a signal that has been subjected to dot interlace interpolation processing such as interpolation processing, and is offset sampled between two fields.
An output terminal is provided that converts the MUSE signal back to an analog signal using an AD converter at a sampling frequency twice (or four times) the original MUSE sampling frequency, and outputs the converted signal. Enables recording by connecting to a VCR. In addition, an AD converter that samples at a sampling frequency twice (or four times) the sampling frequency of the MUSE, and an input terminal for connecting to the MUSE signal or the output of video equipment such as a video camera or VTR are installed. The configuration is such that the signal or the MUSE signal, which is offset sampled between two fields from a video camera or video equipment, can be sampled by the AD converter and converted into a digital signal.

[Embodiments of the invention]

An embodiment of a high-definition television receiver according to the present invention is shown in FIG. In Fig. 2, 1 is the first input terminal, 2 is the first AD converter, 3 is a synchronization and control signal separation circuit, 4 is a noise reducer, and 5
is the first switch, 6 is the frame memory, 7 is the first motion detector, 8 is the LPF, 9 is the second motion detector, 10 is the second switch, 11 is the maximum value selector, and 12 is the second motion detector. 1 is a space-time filter, 13 is a second space-time filter, 14 is a mixer, 15 is a DA converter, 16 is an output terminal, 17 is a second input terminal, 1
8 is the second AD converter, 19 is the third switch,
20 is a first field memory, 21 is a second field memory, 22 is a third motion detector, 23
is the third space-time filter, 24 is YC separation, 25
is a matrix circuit, 26 is a display, 27
is a synchronous separation circuit.

A high-definition television signal transmitted using the MUSE method is input from a first input terminal 1, and sampled and quantized by a first AD converter 2. This sampling frequency can be the same as the MUSE sampling frequency, so
In this case, it is 16.2MHg. From the output of this first AD converter 2, synchronization (vertical, horizontal) information and control signals are transmitted according to the signal pattern.
The control signal separation circuit 3 separates the signals. On the other hand, the input signal is supplied to the first switch 5 after digital noise removal is performed by the noise reducer 4 by adding it to the signal of two frames before. The first switch 5 alternately selects the output of the frame memory 6 and the output of the noise reducer 4 and supplies them to the frame memory 6. Therefore, the frame memory 6 operates at a sampling frequency of 32.4MHg,
That field and the signal two fields before (one frame before) are read alternately. This sampling frequency is stored in the frame memory 6.
Two frames of 32.4MHg signal are stored. In the end, signals for a total of four fields, from the current field to three fields before, are stored in the frame memory 6.

Further, the readout address of the frame memory 6 is controlled by the motion vector included in the output of the synchronization and control signal separation circuit 3, and the readout timing is controlled to change according to the amount of motion, and motion correction is performed. It is done.

In this way, the motion-compensated signals of one frame before and two frames before are read out alternately to the output of the frame memory 6, the signal of two frames before is used by the noise reducer 4, and the signal of one frame before is used. It will be read into the frame memory again.

Furthermore, since the signal from two frames ago is samplewise the same as the signal in the current field, the first motion detector 7 calculates the difference between these two signals, and this is used as motion detection information. On the other hand, the signal from one frame before has high frequency components removed by the LPF 8, is interpolated so that the sampling position coincides with the current field, and is supplied to the second motion detector 9. Here, the difference with the current signal is also taken and used as motion detection information. In the two-frame difference mode, the second switch 10 is off, and the two-frame difference is selected by the maximum value selector 11 and supplied to the mixer 14. In the 1-frame difference mode, the switch 10 is turned on, and the larger difference between 1 frame and 2 frames is selected and supplied to the mixer 14.

On the other hand, in the first spatio-temporal filter 12, interpolation is performed using the signal of the current field from which noise has been removed and the signal of one frame before, as interpolation processing suitable for still images. 13, interpolation using only the signal of the current field is performed as interpolation processing suitable for moving images.

These two signals are mixed in a mixer 14, and the mixing ratio is controlled by the output of the maximum value selector 11. When there is no movement, that is, when the difference is zero, only the output of the first spatiotemporal filter 12 is output, and when there is rapid movement and the screen changes significantly, that is, when the absolute value of the difference is large, the output of the second spatiotemporal filter 12 is output. Only the output of the filter 13 is output, and when it is in between, the output of the first spatio-temporal filter 12 and the output of the second spatio-temporal filter 13 are output depending on the magnitude of the difference.
The output of is mixed.

In this way, motion correction is performed on the output of the mixer 14, dot interlace is interpolated according to the motion of the image, and a signal demodulated to the state of inter-field offset sampling is output.
Normally, this output signal is supplied to the first field memory 20 and the second field memory 21 via the third switch 19. The third motion detector 22 detects the amount of motion from the difference between the output of the second field memory 21 and the third switch 19. In the case of a still image, interpolation of inter-field offset sampling is performed from the output of the first field memory 20 and the output of the third switch 19,
During moving images, the coefficients of the third spatio-temporal filter are switched by the output of the third motion detector 22 so that sample points are interpolated only from signals within the same field.

As can be seen from the above-mentioned literature, MUSE is T.CI
Since it is a time division multiplexed signal called (Time Compressed Integration), the Y and C separation circuit 24
C is time-expanded to match the timing with Y, and the matrix circuit 25 generates R, G, and B signals, which are displayed on the display 26. Here, the second
The first feature of the device shown in the figure is that a DA converter 15 is provided at the output of the mixer 14, and a signal demodulated to the state of inter-field offset sampling can be taken out from the output terminal 16.

The second feature of the device shown in Figure 2 is the second sampling frequency that is twice the sampling frequency of MUSE.
This is because the AD converter 18 is provided.

As a result, the output is output from the above output terminal.
It becomes possible to input the signals recorded and played back on the VTR and the output of the video camera into the receiver from the second input terminal 17. At this time, it would be ideal if a household camera could use offset sampling to widen the bandwidth, but it is not necessary to do so. In other words, even with a 16MHg video signal that has not been offset sampled, the band is only 20% narrower than the 20MHg band of the original high-definition television signal, so you can enjoy sufficiently high-definition images.

A second embodiment of the invention is shown in FIG. In Fig. 3, inter-field offset sampling interpolation is input from the first input terminal 1.
Only for MUSE, and the second input terminal 17
Interpolation is not performed on signals input from the . In this way, the third
Since the switching of the coefficients of the space-time filter can be controlled by the output of the mixer 14, the circuit can be simplified.

Furthermore, it is also possible to further simplify the circuit by using both the first AD converter 2 and the second AD converter 17. An example of this case is shown in FIG.
In this case, MUSE sampling is also performed at a frequency twice the MUSE sampling frequency, and the
It is best to configure the LPE with a transversal type digital filter so that the amplitude characteristics have the desired characteristics while keeping the phase equality flat. Generally, in a highly band-compressed signal like this MUSE, waveform distortion will occur if the sampling phase on the sending side and the sampling phase on the receiving side do not match. Therefore, it is important to flatten the phase characteristics, but this is difficult to achieve with general analog filters and requires expensive filters.

In the above explanation, the sampling frequency of the DA converter 15 has been explained as twice the sampling frequency of MUSE, but the speed of the DA converter is easier to increase than the AD converter. Therefore, by increasing the sampling frequency,
It is also effective to make it easier to remove sampling frequency components. In this case, as shown in FIG. 5, it is preferable to output the demodulated signal into a high-definition television signal after DA conversion. Of course, in this case as well, the sampling frequency of the AD converter may be twice the sampling frequency of MUSE.

Furthermore, although we have proceeded with the discussion on the assumption that the vertical and horizontal synchronization parts are unchanged from those for the MUSE signal, it goes without saying that this is not necessary.
In MUSE, horizontal synchronization is positive polarity synchronization and is within the amplitude range of the video signal, making synchronization detection complicated. Therefore, it is possible to shift the level of only this part and make it negative polarity synchronization to facilitate synchronization detection, or to provide separate output terminals and input terminals for Y and C as shown in Fig. 6, and to adjust the period of each synchronization signal. It is also effective to use a method such as making synchronization detection easier by increasing the length of the synchronization.

〔Effect of the invention〕

As explained above, the high-definition television receiver according to the present invention has an output terminal that demodulates the compressed band signal transmitted by the MUSE method and outputs it as an analog signal using a DA converter, and a sampling frequency that is twice the sampling frequency of the MUSE method. An input terminal is provided for inputting to an AD converter that samples the input signal at a frequency of It became possible to do so.

[Brief explanation of the drawing]

Fig. 1 is a diagram showing a sampling pattern of the MUSE method, Fig. 2 is a diagram showing an embodiment of a high-definition television receiver according to the present invention, and Fig. 3 is a diagram showing a high-definition television receiver according to the present invention with a simplified circuit. FIG. 4 is a diagram showing an embodiment of the high-definition television receiver according to the present invention which also serves as an AD converter, and FIG.
Figure 6 shows an embodiment of a high-definition television receiver according to the present invention in which the operating frequency of the DA converter is increased.
The figure shows an embodiment of a high-definition television receiver according to the present invention in which an AD converter and a DA converter are provided separately for Y and C. 15, 33... AD converter, 16, 32... Output terminal, 17, 34... Input terminal, 18, 35...
...DA converter, 27...Synchronization separation circuit, 28,2
9...Switch, 30...LPF, 31...Resampler, 36...Time expansion, color difference signal separation circuit.

Claims (1)

[Claims] 1. MUSE whose bandwidth is compressed by offset sampling between 4 fields or between 2 fields
(Multipule Sub-Nyquist Sampling
In a high-definition TV receiver that receives high-definition TV signals transmitted using the encoding method, an AD converter that digitizes the received MUSE signal, and an AD converter that digitizes the received MUSE signal,
demodulation means for digitally processing the output of the AD converter and demodulating it into the original wideband high-definition television signal; display means for displaying the demodulated high-definition television signal on a display; It is characterized by comprising means for outputting it as an analog signal, and for sampling the MUSE signal guided as an input signal of the high-definition television receiver at a frequency twice the sampling frequency of the MUSE method using the AD converter. A high-definition television receiver. 2. In claim 1, the demodulation means has a function of demodulating a MUSE signal offset-sampled between four fields into a MUSE signal offset-sampled between two fields, and A high-definition television receiver characterized in that a signal demodulated into a MUSE signal offset-sampled between fields is converted into an analog signal by the DA conversion means and output. 3 In claim 1, the said reception
Two different AD converters are equipped as AD conversion means to digitize the MUSE signal, one AD converter samples at the same frequency as the sampling frequency of the MUSE method, and the other AD converter performs sampling at the same frequency as the MUSE method sampling frequency. A high-definition television receiver characterized by sampling at twice the frequency.