JPS6042672B2 - TV signal digital transmission system - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an efficient digital transmission system for television signals.

Conventionally, in digital transmission of television signals, signals in the entire band from low frequency components to high frequency components are encoded and transmitted on each line on the television screen.

By the way, since the sampling frequency fs needs to be at least twice the highest frequency of the signal, for example, in the case of -4MH2 television, fs≧8MHz. However, if this sampling frequency can be reduced, it will be extremely advantageous in reducing the transmission bit rate. In the conventional method, the lower limit of the sampling frequency is limited by the signal frequency as described above. It is difficult to reduce this.
By the way, the human eye's visual sensitivity is lower than the vertical resolution of high-frequency components, so there is no need to transmit high-frequency components over all lines, and signals containing only low-frequency components and signals containing low-frequency components Even if signals containing both high-frequency components and signals containing both high-frequency components are transmitted alternately every line period, and signals containing only low-frequency components are reproduced using the high-frequency components of signals from neighboring lines, the image quality will not deteriorate significantly. Can be kept small.

Even in the above-mentioned method, image quality deterioration is very small in flat areas of the image, but deterioration in diagonal limbal areas becomes particularly noticeable in areas that contain many high frequency components such as the limbal areas of the image.

In order to eliminate these drawbacks, the present invention provides a signal containing both low frequency components and high frequency components of a television signal;
A signal containing only low-frequency components is transmitted alternately every line period, and the sampling frequency for the signal containing only low-frequency components is set low, and the receiving side transmits high-frequency components according to the direction of the ring of the image. It is characterized by interpolating and reproducing images, thereby reducing the transmission bit rate and reducing image quality deterioration in the limbal region.

The drawings will be explained in detail below. FIG. 1 is a block diagram of the transmitting side in an embodiment of the method of the present invention.
1 is an analog TV signal input terminal, 102 is an A/D converter, 103 is a digital low-pass filter, 104 is a subsampling circuit, 105 is a changeover switch, 106
, 107 is a selection terminal of the selector switch 105, 108 is a transmission buffer memory, 109 is a DPCM encoder, and 110 is a digital signal output terminal.

Thus, an analog television signal is applied to an input terminal 101 and converted into a digital signal by an A/D converter 102 operating at a sampling frequency Fs.

This digital signal is divided into two paths, one of which is a digital low-pass filter 103 where a low frequency component (L) of approximately 1/2 of the original band is extracted, and this low frequency component is then passed through a sampling circuit. At 104, the sample points are read out one by one and sent to the selection terminal 106. Therefore, the sampling rate of the signal at this selection terminal 106 is 112 fs, and the frequency component of the signal is only the low frequency component (L). On the other hand, the selection terminal 107 has A/
The digital signal that is the output of the D converter 102 is sent as is.

Therefore, the sampling rate of the signal at the selection terminal 107 is Fs, and its frequency components include both a low frequency component (L) and a high frequency component (H). Changeover switch 10
5 is a selection terminal 106 for each horizontal period (H) of the television signal;
107.

The signal selected by the changeover switch 105 is written into the transmission buffer memory 108. Therefore, the writing speed of the transmission buffer memory 108 is 112 fs1 when the selector switch 105 is on the selection terminal 106 side.
When 05 is on the selection terminal 107 side, it is Fs. On the other hand, the read speed of the transmission buffer memory 108 is constant and 112
It is set to the average value of fs and Fs, that is, 314 fs. The digital signal obtained in this way consists of a low frequency component (L), a high frequency component (H), and a low frequency component (
L) are arranged alternately, and the sampling frequency averaged over two lines is reduced to 314 fs. In other words, the average number of sent samples is 31.
It has been reduced to 4. This signal is passed to the DPCM encoder 109
The signal is then further band-compressed encoded, subjected to appropriate code conversion, and then sent to a transmission path. Figure 2 shows the line configuration of the sending signal, where the nth line has low frequency components (Ln) and 10 high frequency components (Hn), the n+1st line has only low frequency components (Ln+1), and the n+2nd line has Ln+ .

+Hn+2, the n+3th line is only Ln+3,...
. This shows that L+H and L-only lines are sent out alternately on a line-by-line basis. FIG. 3 is a block diagram of the receiving side, in which 201 is a digital signal input terminal, 202 is a DPCM decoder, 203 is a reception buffer memory, 204 is a changeover switch, 205, 206 is a changeover terminal for the changeover switch 204, 207, 208 is a 1 line time delay circuit, 209 is a 2 line time delay circuit, 210,2
11 is a one-sample time delay circuit, 212 and 213 are digital high-pass filters, 214 is a comparison circuit, 215,
216 is a 2 sample time delay circuit, 217 and 218 are 1
A sample time delay circuit, 219, sets the sampling rate to 1
Interpolation circuit converting from 12fS to Fs, 220, 221
222, 223, 224 are switching terminals of the switching switch 220, 225, 226, 227 are switching terminals of the switching switch 221, 228, 229 are adders, 2
30 is a multiplication circuit for multiplying by the coefficient 112, 231 is a changeover switch, 232 and 233 are changeover terminals of changeover switch 231, 234 is a D/A converter, and 235 is an output terminal for an analog television signal.

The received digital signal input to the digital signal input terminal 201 is decoded by the DPCM decoder 202 and then written to the receive buffer memory 203.

This writing speed is a constant value of 314 fS. Reading from the reception buffer memory 203 is performed alternately at 112 fS and Fs every 1H period. That is, when the read signal is L, the read speed is 112 fS, and the changeover switch 204 is connected to the selection terminal 205, and when the read signal is L + H. In some cases, the readout speed is Fs and the changeover switch 204 is connected to the selection/selection terminal 206.The low frequency component of the n+1th line taken out from the selection terminal 205 and having a sampling rate of 112fS is processed by the delay circuits 207 and 210.
sample interpolator 2 after being delayed by one sample with the line
At step 19, the distance between the sample point and the sample velocity point is interpolated and converted into a sample velocity Fs.

At this time, the high frequency components Hn, Hn of the n-th line and the n+2-th line extracted by the 2-line time delay circuit 209 and the digital high-pass filters 212 and 213
+. The values of three consecutive sample points are taken out to selection terminals 222, 223, 224, 225, 226, 227 by delay circuits 215, 216, 217, 218, respectively. Also, the value Ln of the nth line and the n+1th line
+Hn,Ln+. +Hn+2 is added to the comparator circuit 214, and the optimal interpolation direction is determined to be the vertical direction,
Select from diagonal direction up the left or diagonal direction up the right. The changeover switch 220 and the changeover switch 221 are the comparison circuit 21
According to the signal from 4, one terminal is selected from each of the three terminals and connected, and the high frequency components of appropriate sample points of the n-th line and the n+1-th line are sent to the adder 229. The adder 229 and the multiplication circuit 230 generate an interpolated value H″ of the high frequency component.+1 is created as the average of the high frequency components of the two sample points, and this interpolated value and the low frequency component Ln+1 of the Added by adder 228,
A playback signal Ln+1+H″n+1 is generated and sent to the selection terminal 232.
sent to. This signal is denoted as L+H''. On the other hand, the received low frequency component and ten high frequency components are delayed by one line and one pixel time by delay circuits 208 and 211 and sent to the selection terminal 233. It is written as L+H.The changeover switch 231 is linked to the changeover switch 204 and switches between the selection terminal 232 and the selection terminal 233.
and L+H are selected alternately in one line period. The digital signal thus created is converted into an analog signal by the D/A converter 234 and output from the output terminal 235. FIG. 4 shows the arrangement of six sample points used when interpolating the high frequency component of sample point X on the (n+1)th line.

a-1, a0, a1, b-1, b0, b1 are sample values at each point. Here, first, the comparison circuit 214 outputs A. I: d. B
The difference between O, a-1 and Bl, al, and b-1 is compared, and the line connecting the two sample points of the set with the smallest absolute value of this difference is set as the optimal interpolation direction, and the line between the two points is Let the average of the high frequency components be the reproduction value of the high frequency components at the sample point X. Figure 5 shows the line configuration of the reproduced signal.
Line is Ln+Hnl line n+1 is Ln+1+H''
n+1, n+2 line is Ln+2+Hn+2, n+
The 3rd line is Ln+3+H''n+3...

Here, H'' is a value interpolated from the high frequency components of adjacent lines, indicating that the high frequency components that are not sent for each line are reproduced using the high frequency components of the previous and following lines. In this way, in the method of the present invention, lines on which high frequency components are not transmitted are reproduced using high frequency components of adjacent lines, but in television signals, the correlation between adjacent lines is extremely strong, and the human eye is less sensitive to high frequency components. Since the sensitivity to the image is reduced, the deterioration in image quality due to reproduction processing is extremely small.

Furthermore, since the direction of interpolation of high frequency components is switched according to the direction of the ring of the image, deterioration in oblique ring parts is also reduced. In the description of this embodiment, the ratio of the sampling frequency for the low frequency component to the high frequency component and the sampling frequency for only the low frequency component is set to 112, but other values may be used.

Further, although three types of directions were used as interpolation directions of high frequency components on the receiving side, it is also possible to use five or more types of directions. As explained above, in the method of the present invention, the low frequency components of the signal are transmitted across all lines of the TV screen, and the high frequency components are transmitted every other line, so the number of Sassops to be transmitted is reduced. Therefore, there is an advantage that the transmission bit rate can also be reduced.

Furthermore, when the transmission bit rate is kept constant, the number of encoding bits can be increased by the amount that the number of transmitted samples is reduced, so it is possible to improve the image quality.
Furthermore, on the receiving side of the system of the present invention, the direction of interpolation of high-frequency components is switched according to the direction of the ring of the image for reproduction, so that deterioration in image quality due to reproduction processing can be kept to an extremely small level. Therefore, it would be extremely advantageous to apply the method of the present invention to digital transmission of television signals.

[Brief explanation of drawings]

Fig. 1 is a block diagram of the transmitting side in an embodiment of the method of the present invention, Fig. 2 is a line configuration diagram of the sending signal, Fig. 3 is a block diagram of the receiving side, and Fig. 4 is a sample point used for high frequency component interpolation. The layout diagram, FIG. 5, is a line configuration diagram of the reproduced signal.

Claims (1)

[Claims] 1 On the transmitting side, the sampling frequency for a signal that contains both low frequency components and high frequency components of a television signal, and for a signal that contains only low frequency components and consists of both low frequency components and high frequency components is set. with the signal set low.
It is transmitted alternately line by line, and on the receiving side, for lines containing only low frequency components, the direction of the outline of the image is detected by comparing multiple sample values on adjacent lines with each other, A television signal digital transmission system characterized by performing interpolation reproduction of high frequency components using sample values on adjacent lines in that direction.