JPS5931916B2 - Interframe coding device for color television signals - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an interframe encoding device that interframe encodes a color television signal, transmits it, receives it, and decodes it.

When digitally transmitting television signals, it is known to perform interframe coding as a band compression method to reduce the number of transmission bits as much as possible and improve the efficiency of the transmission path.

This utilizes the frame correlation of television signals, and is a band compression technique that is particularly effective for television signals with little movement. However, conventionally, NTS
When such interframe coding technology is applied to a color television signal in which a luminance signal and a color signal modulated by a subcarrier are frequency multiplexed, such as in the C, PAL, and SECAM systems, the interframe coding for color components is There is a problem that the prediction error increases and effective compression cannot be performed.

In other words, taking an NTSC color television signal as an example, the subcarrier frequency fsc is (455/2) times the horizontal scanning frequency fH, and the horizontal scanning frequency fH is 525 times the frame frequency fF. After all, fs
c is an odd multiple of 1/2 fF, the subcarrier phase differs by 1800 from frame to frame, and the frame difference signal increases approximately twice in terms of color components. For this reason, the above-mentioned effective band compression cannot be performed. Conventionally, in order to solve this subcarrier phase problem, when performing interframe encoding of color television signals, the color signal is first separated and demodulated into baseband signals, then interframe encoded, and then received. A so-called separate coding technique is employed in which the interframe-decoded chrominance signal and luminance signal are converted back to the original NTSC signal. However, in such separate coding, at the transmitting side, NT
A decoder for the SC signal and an A/D (analog/digital) converter corresponding to each of the luminance signal and chrominance signal are required, and on the receiving side, a D/D converter corresponding to each of the luminance signal and chrominance signal is required. This method has the disadvantage that it requires an A (digital/analog) converter and an encoder for the NTSC signal, making the device configuration complicated.

On the other hand, as a method of directly interframe encoding a color television signal without performing separate encoding, the polarity of the carrier color signal of the color television signal is reversed every other frame, and the phase of the carrier color signal is changed between frames. There is a method of matching the frames and then performing interframe encoding.

However, this method has the disadvantage that signal distortion caused by the polarity reversal operation occurs every other frame, causing flickering on the playback screen. SUMMARY OF THE INVENTION An object of the present invention is to provide an interframe encoding apparatus for color television signals which simplifies the apparatus by direct encoding and does not cause flicker.

That is, the interframe encoding device for a color television signal of the present invention matches the phase of the carrier color signal between frames by reversing the positive and negative polarities of the carrier color signal of the color television signal every other line. This color television signal is interframe coded and transmitted, and the receiving side inverts the polarity of the carrier color signal of the interframe decoded color television signal every other line, thereby reproducing the original color television signal. It is characterized by converting into .

In the present invention, the polarity inversion of the carrier color signal is realized by frequency separation and calculation of the carrier color signal, and the demodulation of the color signal to the baseband signal is not performed, so there are disadvantages such as adoption of the above-mentioned separation coding technique 1. That is, there is no need for an NTSC signal decoder, encoder, or extra A/D and D/A converter on the transmitting and receiving sides, and the process of reversing the positive and negative polarities is simple, so there is almost no signal waveform distortion.

Further, since the phase of the carrier color signal is matched between frames by the polarity inversion operation, the color signal can be efficiently compressed during interframe encoding. Furthermore, since the polarity reversal operation is performed every other line, the resulting distortion does not turn into flicker and disturb the screen. Next, the present invention will be explained in detail with reference to the drawings.

FIG. 1 is a block diagram showing one embodiment of the present invention. A color television signal M to be transmitted is applied to the input terminal 1, and the preprocessing circuit 100 performs signal conversion to invert the positive and negative polarities of the carrier color signal every other line. N is interframe encoded by the interframe encoder 200 and sent to the transmission path 3. On the receiving side, the signal from the transmission line 3 is decoded by the interframe decoding device 201 to obtain the signal N, which is converted into the original color television signal M by the post-processing circuit 101. A feature of the present invention is that it includes a pre-processing circuit 100 and a post-processing circuit 101 for inverting the polarity of a carrier color signal, and the interframe encoding device and decoding device will be described later. It has the same functions as the conventional one, except that it is provided with a function of transmitting and receiving a code to distinguish between lines with reversed polarity and lines without polarity inversion.

Furthermore, since the pre-processing and post-processing perform conversion operations that are opposite to each other, they will be collectively described in detail below. FIG. 2 is a diagram showing the preprocessing circuit 100 and postprocessing circuit 101 of FIG. 1 in detail.

First, in the preprocessing circuit 100 , the carrier color signal component of the color television signal M is frequency-separated by a bandpass filter (BPF) 110 , the amplitude is doubled by a coefficient multiplier 113 , and the resultant signal is input to the minus side of a subtracter 114 . be done.

The delay circuit 111 is for compensating for the delay of BPFllO, and the delay-compensated color television signal is supplied to the plus side of the subtracter 114 and also applied to one side of the switch circuit 115. The switch circuit 115 switches these two signals for each line, and the switching signal 1
16 is generated from the original color television signal M via the control signal generator 112, and is sent to the interframe encoding device 200 together with the converted television signal N, and transmitted to the receiving side. The post-processing circuit 101 is similar to the pre-processing circuit 10.
Although it has almost the same configuration as 0, the switching signal 116/ according to the line is supplied from the interframe decoding device 201. The operation of the dojo will be explained below using mathematical formulas.

The input color television signal M(t) is the luminance signal Y(t
) and the color signals C1(t) and C2(t). However, Φ1(t)=Sln2πF,
. t, Φ2(t)=COs2πF8Ot. Also,
Since the signal M(t-TF) after one frame is -Φ2(t).

If the screen is stationary, C2(t-TF)ZC2
(t), so the phase of the carrier color signal differs by 180 cycles for each frame.

A BPF through which the signal M(t) selectively passes the carrier color signal.
Since most of the luminance signal Y(t) is attenuated after passing through llO, the output M2(t) of the subtracter 114 becomes t~ゞ'ゞ▼6~
7'71~1'11~117≦~2'VClゞ'11'
As a result, the positive and negative polarities of the carrier color signal are reversed.

Also, a delay compensation circuit 11 for compensating for the delay of the BPF.
The signal M1(t) passing through 1 is not polarized. A switch circuit 115 switches M,(t) and M2(t) line by line to produce an output signal N(t). Hereinafter, the line whose polarity is reversed will be referred to as the P line, and the line whose polarity will not be reversed will be referred to as the Q line. Now, if the output signal N(t) is on the Q line, one frame after this will be on the P line since one frame has 525 lines.

Therefore, it is possible to obtain a signal in which the phase of the carrier color signal is almost the same for each frame. When the screen is stationary, N(t)ZN(t-TF), the frame difference signal becomes almost zero, and efficient interframe encoding can be performed. On the receiving side, when it is the Q line, N(t) is simply given delay compensation and is passed through as is, and when it is the P line, the positive and negative polarities of the carrier color signal are reversed by the same operation as on the transmitting side. The original color television signal M can be obtained as the output of the post-processing circuit 101.

Although it is of course possible to physically implement this embodiment using an analog circuit, it is preferable to use a digital circuit in order to maintain stable calculation accuracy.

FIG. 3 is a diagram showing another example of the preprocessing circuit 100 of FIG. 1.

The preprocessing circuit 100 shown in FIG. 2 has the advantage of a simple configuration, but has the disadvantage that the amplitude of the frame difference signal increases with respect to the high frequency component of the luminance signal. That is, since the luminance signal Y of the color television signal M includes a high frequency component that shares the band with the carrier color signal, the output that has passed through BPFllO contains the luminance signal Y in addition to the carrier color signal component. A high frequency component YH also appears. Therefore, if we write equation (4) more precisely, we get
The frame difference signal D(t) for a still image appears as a double high-frequency luminance signal.

The configuration of FIG. 3 eliminates this drawback by using a one-line delay line.

In FIG. 3, reference numeral 122 is a one-line delay line, reference numeral 120 is an adder, reference numerals 125 and 126 are 1/2 and -1/2 coefficient multipliers, respectively, and the rest are not used in FIG. It uses the same components.

The switch circuit 115 connects the P line to the b side and the Q line to the a side.
The converted output N for the input color television signal M is as follows.

Here, YL(t)=Y(t)−YH(t), and T
H is one horizontal scanning time. In this conversion, the polarity of the color signal is reversed every other line, and the color signals of two scanning lines are averaged. Also, for the luminance signal, if the vertical correlation of the screen is strong, YH(t)
Since it is H(t-TH), the high frequency component is attenuated by the comb filter. After performing such conversion, the frame difference signal D(t) for the still image becomes the line difference of the high frequency component of the luminance signal.

Since there are few line differences in ordinary images, the frame difference signal becomes almost zero, and efficient interframe coding can be realized. As the post-processing circuit 101 on the receiving side, the one shown in FIG. 2 can be used as is. Therefore, as a third example, a sub-Nyquist encoding method in which encoding and transmission is performed at a sampling frequency lower than the Nyquist frequency (twice the frequency of the signal band) is applied to the preprocessing and postprocessing of the present invention. A sub-Nyquist interframe encoding device that is much more effective than the conventional one will be described. In the sub-Nyquist encoding method, the signal is sampled at every time when the subcarrier phase differs by about 1800, and the sampling phase is shifted by 1800 for each scanning line.
On the receiving side, the intermediate sample points are interpolated from neighboring sample values.

Even with this sub-Nyquist coding method, the positive and negative polarities of the carrier color signal are still inverted every frame, so
If applied directly to interframe coding, the carrier color signal will appear in the frame difference signal as approximately twice as large, and the band compression effect of interframe coding will be small. Therefore, if the preprocessing circuit 100 of the present invention incorporates an operation of reversing the positive/negative polarity of the carrier color signal every other line, the frame difference signal becomes almost zero, and efficient interframe encoding can be realized. Regarding the above-mentioned sub-Nyquist encoding method, the publication ``32 Mb/s Direct Coding (HOD
PCM) color television transmission system (communication system study group materials (
S75-69)" (Reference 1), pages 10 to 11, so further explanation will be omitted. FIG. 4 is a block diagram showing the preprocessing circuit on the transmitting side of the sub-Nyquist interframe encoding device described above. The structure is the same in principle as that shown in FIG. 3, but in order to improve the overall characteristics of transmission and reception. The color signal is averaged with the color signal one scan line later.

Further, FIG. 4 shows a case in which it is constructed using a digital circuit. In FIG. 4, a color television signal M is input to an A/D converter 130 and a pulse generation circuit 136. A pulse generation circuit 136 separates the synchronization signal from the color television signal and generates a sampling pulse 140 for A/D conversion.
, a sub-Nyquist sampling pulse 150, an odd/even line switching signal 141, and a PQ line switching signal 116. The sampling pulse frequency Fp for A/D conversion is selected to be approximately four times the subcarrier frequency Fsc. F,=4f
, 0, the sampling time is a position where the subcarrier phase is shifted by 900. A/D converter 1 as shown in FIG.
If a number is assigned to the sample value A, which is the output of the register 131, the signal delayed by one sample in the register 131 becomes like B. After switching A and B for each odd and even scanning line with the switch circuit 132, the signal is transferred to the register 135. If subsampling is performed at a sub-Nyquist frequency F8 that is 1/2 of the sampling pulse frequency Fp, a sample value of C or D will appear in the output.

The sample values of C and D each have a subcarrier phase of 1.
The sample value is shifted by 800 degrees, and the phase of the subcarrier between C(!:.D is shifted by 90 degrees. Switch circuit 132
switches between A and B according to the odd and even scanning lines, so that the sub-Nyquist encoded sample value is obtained at the output of the register 135. 1 line delay line 122 is NTSC
Fp=4f5 for the signal.

Or F8=2f8. When selected, it consists of a shift register or memory with a delay of 455 samples. The multiplexer 133 is a switch that inverts the positive/negative polarity of the carrier color signal according to the PQ line, and other arithmetic operations are performed by the second switch.
This is the same as the example shown in FIG. The relationship between input M and output N can be easily described using Z transformation. That is, in Z transformation, since z−n represents n sample delay, one line delay is written as Z−H, and BPFllO
The Z transformation of the transfer characteristic of is B(8), 1-B(8)=A(8
), the Z transformation N(8) of the output N becomes, so
The Z transformation D(2) of the frame difference signal is as follows.

However, Z-F represents frame delay. According to the AI formula, the first term represents the frame difference of low frequency components, and the second term represents the frame sum of line differences of high frequency components including the carrier color signal. The [th] term is zero for a still image, the second term is also zero for the carrier color signal, and what remains is only the line difference of the high frequency luminance component, similar to equation (7). A concrete implementation method of BPFllO in a digital circuit is, for example, by configuring B(2) as or (12), so it can be implemented using a register and an adder/subtractor.

Most of the configuration shown in FIG. 4 is a circuit used for sub-Nyquist encoding using a comb filter, and the circuit required only for polarity inversion of the carrier color signal of the present invention is the subtracter 18.
4 and multiplexer 133, the sub-Nyquist encoding method can be easily applied to the present invention.

FIG. 6 is a block diagram showing a post-processing circuit on the receiving side of the sub-Nyquist frame inter-frame coding apparatus.

Sub-Nyquist decoding is performed on the transmitting side by interpolating subsampled pixels. First, the transmitted sample value is
', but one of the outputs of the delay correction circuit 11' directly enters the multiplexer (MPX) 133b, and when it is the Q line, passes through this multiplexer and is output to MPX16O.
The signal is multiplexed with the interpolated pixel by the D/A converter 165, and converted into an analog signal by the D/A converter 165. In this circuit, the signal is at sub-Nyquist sampling frequency F8 up to the input of multiplexer 160.
The sampling frequency becomes 2f5, that is, F, at the multiplexer. For the P line, a signal whose polarity of the carrier color signal is inverted by the subtracter 134/ is selected by the MPXl 33b. The interpolated pixel is created separately into an interpolated value of the low-range luminance component and an interpolated value of the carrier color signal, and these are combined by an adder 162 and a subtracter 163. First, the low-frequency luminance component is obtained by subtracting the high-frequency component containing the carrier color signal that has passed through the BPFll' from the original signal using the subtracter 114/, and this is obtained by passing it through the interpolation circuit 161. Then, the interpolated value is obtained. As the interpolation circuit 161, for example, a circuit that creates an average value of two adjacent sample values may be used. Since the interpolated value of the carrier color signal is obtained by inverting the polarity of the sample value one scanning line before, it is sufficient to use the output of the one-line delay line 122'. Therefore, on the Q line, the interpolated values are combined in a subtracter 163, and on the P line, since the polarity of the carrier color signal is inverted, they are combined in an adder 162. MPX
133a is for switching these interpolation pixels according to the PQ line. Most of the post-processing circuit in FIG. 6 is necessary for sub-Nyquist decoding, and only the subtracter 134', adder 162, and Multiplexer 133a
, 133b only. Next, interframe encoding device 2
An example of a specific configuration for 00 is shown in FIG.

In FIG. 1, the output N of the preprocessing circuit 100 is the frame memory 2.
Subtractor 2 only for the signal of one frame before that appears at the output of 14.
10, noise components are removed by a non-linear circuit 211, and the signal is transmitted to the next-stage intra-frame predictive coding loop.
The nonlinear circuit 211 is a circuit whose gain is less than 1 when the amplitude of the signal is small, and whose gain is approximately 1 when the amplitude is large. Since the difference signal amplitude is sufficiently suppressed, minute noise can be effectively removed. The subtracter 210, the quantizer 212, the adder 215, and the prediction circuit 213 constitute an intra-frame predictive coding rube, and function to remove horizontal redundancy remaining in the frame difference signal. Here, the intra-frame prediction error signal created based on the frame difference signal by the quantizer 212 is quantized, and is unequal-length coded by the unequal-length code conversion circuit 217, so that the quantized code has redundancy. is removed. A signal 116 for switching the PQ line is encoded by a synchronization code generation circuit 216, multiplexed with the unequal length code together with a synchronization code, a coding mode control code, etc. by a multiplexer 218, and written into a puffer memory 219. buffer memory 219
is for temporally equalizing the information generated in the interframe coding loop, and data is read out from the buffer memory 219 at a constant speed and sent to the transmission path.
The above-mentioned interframe encoding device 200 (Fig. 1) is based on a publication published by the Institute of Electronics and Communication Engineers in 1975 [4: Structure of an interframe encoding system for MTV telephones - Communication system study group material (CS75- The structure is almost the same as that shown in detail in FIG. 1 on page 1 and FIG. 6 on page 5 of "68)" (Reference 2), and its operation is also similar, so further detailed description will be omitted. Similarly, the interframe decoding device 201 on the receiving side in FIG. 1 according to the present invention simply performs the opposite operation to the interframe encoding device 200 in FIG. In addition, the configuration shown in FIG.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is a diagram showing an example of the preprocessing circuit 100 and postprocessing circuit 101 shown in FIG. 1, and FIG. 4 is a diagram showing a third example of the preprocessing circuit 100 in FIG. 1, FIG. 5 is a diagram for explaining the sampling phase, and FIG. The diagram showing a second example of the post-processing circuit 101 in the figure and FIG. T are block diagrams showing an example of the interframe encoding device 200. 1 to T, 100... pre-processing circuit, 101... post-processing circuit, 200...
...Interframe encoding device, 201...Interframe decoding device, 110, 110'...Band pass filter, 111, 111'...Delay compensation circuit, 113 , 113'...Coefficient unit, 114,
114'... Subtractor, 115, 115'...
...Switch circuit, 112...Control signal generator, 122, 122ι...1 line delay line, 12
0... Adder, 130... Analog/
Digital converter, 131, 135... Register, 133, 160, 218... Multiplexer, 134, 134'... Subtractor, 61...
...Interpolation circuit (165...Digital/analog converter, 212...Quantization circuit, 213
...Prediction circuit, 214...Frame memory, 21T...Unequal length encoder, 216...
. . . synchronous code generation circuit, 219 . . . buffer memory, 211 . . . non-linear circuit.

Claims (1)

[Claims] 1. A composite color television signal in which a luminance signal and a carrier color signal obtained by modulating a subcarrier with a color signal are frequency multiplexed, and the positive and negative polarities of the carrier color signal are reversed every other line. and a means for interframe-encoding and transmitting the composite color television signal obtained from this means, and a means for interframe-decoding the transmitted interframe-encoded signal; A color television signal which is characterized in that the receiving side is equipped with means for inverting the positive and negative polarities of the carrier color signal of the composite color television signal decoded as described above every other line and converting it into the original composite color television signal. An interframe coding device for television signals.