JPS598989B2 - Color signal interframe coding method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a color signal interframe predictive coding method for efficiently predictively coding a color television signal using interframe correlation.

As is well known, in a composite color television signal such as the NTSC system, a carrier color signal component having a large power is arranged in a high frequency range centered on the color subcarrier frequency, and furthermore, the subcarrier frequency is one part of the frame frequency. Since the value is set to an odd multiple of /2, the carrier color signal components have opposite phases between adjacent frames. In such color signals, the signal correlation between adjacent frames is significantly lower than in monochrome television signals, so it is impossible to directly apply the interframe coding method that is effective for monochrome signals to color signals. It is possible. Therefore, conventionally, this composite color television signal was combined with a luminance signal.
A method has been implemented in which the color signals are divided into two color signals (for example, I and Q signals), the signals are time-division multiplexed to create a TDM signal, and the TDM signal is subjected to interframe coding processing. However, in this method, separation of color signals,
This method has the drawback of increasing the scale of the device because it requires synthesis processing, time-division multiplexing, and separation processing. In addition, by using frame memory for two frames, it is possible to
A method of directly interframe predictive coding of the TSC signal is also considered, but this method requires double the frame memory capacity, which again has the drawback of increasing the size of the device. The present invention uses a frame memory for one frame to
This is a method that directly encodes color signals between frames, in which a composite color signal frame and a luminance signal frame are transmitted alternately every frame, and the content of the transmitting and receiving frame memory is rewritten every two frames. The purpose of the present invention is to provide an efficient and economical color signal interframe coding method.

Hereinafter, embodiments will be described in detail with reference to the drawings. Figure 1 shows
This figure shows one embodiment of the present invention, where 1 is an input terminal, 2 is a comb filter for extracting a luminance signal, 3 is a changeover switch, and 4 is a comb filter for extracting a luminance signal.
, 5 is a selector switch selection terminal, 6 is a subtracter, 7 is a quantizer, 8 is an output terminal, 9 is an adder, 10 is a selector switch, 11 and 12 are selector switch terminals, 13 is one frame memory , 14 is a comb filter for brightness signal extraction, 1
5 is a changeover switch, and 16 and 17 are selection terminals of the changeover switch.

Further, Y+C is a composite color signal consisting of a luminance signal and a carrier color signal, Y is an extracted luminance signal, x is an input Δ input to the interframe coding circuit, and x=Y+C or x=Y, and , x are predicted values for x, and the input signal has already been converted into a PCM signal by an A/D converter.

Next, the operation of this embodiment will be explained.

First, the input signal Y+C applied to input terminal 1 is divided into two paths, one signal is sent to terminal 5, and the other signal is sent to terminal 4 after extracting only the Y signal component by comb filter 2. It will be done. Since the switch 3 alternately selects terminals 4 and 5 every frame, the signal x sent from the switch 3 to the interframe encoding circuit becomes Y+C and Y every frame. FIG. 2 shows the sequence of frames to be processed, where the n-th frame is Yn+Cnl, the n-1-th frame is Yn-1, the n-2-th frame is Yn? 2+
Cn-2,...... and becomes φ. Next, returning to FIG. 1, the above signal x is subtracted from the predicted value x by the subtracter Δ6, and the prediction error is quantized by the quantizer 7, and then sent from the output terminal 8 to the buffer memory. , a code converter, etc., and then sent to the transmission path.

Further, this quantized prediction error is simultaneously added to the predicted value x by an adder 9 to create a decoded signal, which is sent to the switch 10.

The switch 10 is switched every frame in conjunction with the switch 3, and selects the terminal 12 when the frame is x:Y+C, and selects the terminal 11 when the x=Y゛7 frame. The contents are rewritten every two frames. That is, it is designed to be rewritten only when the frame is x=Y+C. This frame memory 1
The output of No. 3 is divided into two paths, one of which is sent directly to the terminal 17, and the other is led to the comb filter 14 and sent to Y.
After the signal is extracted, it is sent to terminal 16. switch 1
5 switches every frame in conjunction with switches 3 and 10, selects terminal 17 when x=Y+C, and selects terminal 17 when x=Y+C.
When the frame is Y, terminal 16 is selected. In this way, the contents of the frame memory 13 are updated every two frames x::
Since it is rewritten only in the Y+C frame, if x=Yn+Cn in the nth frame, the predicted value x that is the output of the switch 15 is x2Y'o-2+C! /n-2.

Also, in the n-1th frame △, X:=YO-1,
At this time, the prediction error xΔ becomes x2Y/n−.

However, Y'K and C'k indicate the k-th frame processed signal (decoded signal) stored in the frame memory. FIG. 3 shows the correspondence between the contents of frames to be processed at the time points of frames n-4, n-3, n-2, n-1, and n and the contents of the frame memory 13 at that time. In the n-4th frame, X:=Yn-4
+Cn-4, and the content of frame memo January 3 at that time is Y'.

-6+C'. -68 ^, therefore, the predicted value X is x=Y'.

-6+C'. ．． 6, and the contents of the frame memory 13 are sequentially rewritten from the processed samples to YA-4+
C'. Changes to -4. Also, in the n-3th frame, x=
Yn-, and the contents of frame memory 13 at that time are Y'.

one. +C'. -4, and the predicted value X8 is obtained by extracting only the luminance component from this, and becomes X=Y'n-4.

However, since the contents of the frame memory 13 are not rewritten in this frame, the contents of the frame memory 13 remain Y' even when one frame processing is completed. -4
+Cn-4 remains. In the n-2nd frame, X=Y
n-, +Cn-, and frame memory 1 at that time
The content of 3 is Y'. -4+Cn-4, so the predicted value x is x-Y7. -4+C'. -4, and the contents of this frame memory 13 are sequentially rewritten from the processed samples to Y'. -2+C'. -Changes to 2. below,
A similar operation is performed to change the contents of the frame memory 13 and the predicted value. FIG. 4 shows a receiving section of an embodiment of the present invention, in which 18 is a receiving input terminal, 19 is an adder, 20 is a changeover switch, 21 and 22 are selection terminals of changeover switch 20, and 23 is a frame. 24 is a comb filter for extracting the 1st carrier color signal; 25 is a subtractor; 26 is a changeover switch; 27 and 28 are selection terminals of the changeover switch;
29 is an output terminal, 30 is a comb filter for extracting a luminance signal, 31 is a changeover switch, 32 and A33 are selection terminals of the changeover switch, X is a predicted 1 value, and E1 is a received quantized prediction error signal. .

Next, the operation of this embodiment will be explained. First, a predicted value x is added to the quantized prediction error signal inputted from the terminal 18 in an adder 19 to produce a decoded signal.

This decoded signal is alternately divided into two luminance signals Y' or a composite color signal (Y'+C●) every frame. This decoded signal is divided into two paths, and one of the signals is sent to the switch 20. This switch 20 is switched every frame, and the decoded signal is Y'+C' (color frame).
The terminal 22 is selected when the decoded signal is Y' (luminance frame), and the terminal 21 is selected and input to the frame memory 23. The contents of this frame 23 are rewritten with new decoded signal values every two frames, that is, only in color frames. The algorithm for creating predicted values in this way is exactly the same as that for the transmitter. The decoded signal is also connected to a subtracter 25 and a terminal 48, and this subtracter 25 is connected to the frame memory 23 which is sent from the comb filter 24.
Subtraction is performed with the transport color component extracted from the contents of . Note that when the decoded signal is Y'k, the contents of the frame memory are Y! Since k-1+Ck-1, the output of the subtracter 25 becomes Y/k-C/k-1. That is, for a frame in which only a luminance signal is transmitted, color components are reproduced using the phase-inverted carrier color signal of the previous frame. The switch 26 is switched every frame, and is connected to a terminal 28 during a color frame to select the decomposed composite color signal, and is connected to a terminal 27 during a luminance frame to select the reproduced color signal described above. The signal (carrier color signal of the previous frame of the decoded luminance signal) is selected and sent out from the terminal 29.

Note that in the explanation of the above embodiment, only the signal values of past frames stored in the frame memory are used when creating predicted values, but it is also possible to use adjacent lines or sample values of the current frame as well. It is. As explained above, according to the present invention, a composite color frame and a luminance signal frame are alternately encoded and decoded frame by frame, and the contents of the frame memory are rewritten only when a composite color signal frame is received. Also, for the composite color signal, the composite color signal stored in the frame memory two frames ago is used as a predicted value, but the carrier color signal two frames ago is in the same phase as that of the current frame. Therefore, the influence of frame-by-frame inversion of the carrier color signal phase, which is the most problematic in interframe coding of NTSC color signals, does not occur at all.

Furthermore, for the luminance signal, the luminance signal extracted from the composite color signal of the previous frame stored in the frame memory is used as a predicted value, but in this case, since there is no carrier color signal component, The effect of phase inversion of the color signal does not occur either. Therefore, prediction accuracy becomes extremely high and efficient interframe coding can be performed. In addition, for a luminance signal frame, color signals are reproduced using the color components of adjacent frames, but since the human eye's characteristics for color signals are dull compared to luminance signals, the effect is It can be considered that it can be ignored. Furthermore, color television signals can be directly interframe encoded by simply using frame memories for one frame each for transmission and reception, which has the advantage of reducing the size of the device.
The present invention can realize an extremely effective color signal interframe coding system in realizing an efficient and economical interframe coding system.

[Brief explanation of drawings]

FIG. 1 is a block diagram of a transmitter in an embodiment of the present invention, FIG. 2 is a diagram showing a series of processing frames, and FIG. 3 is a diagram showing a correspondence relationship between the contents of a frame memory and the contents of a processing frame. FIG. 4 is a block diagram of a receiving section in one embodiment of the present invention. 1. Input terminal, 2... Comb filter for brightness signal extraction, 3... Selector switch, 4, 5...
...Selector switch selection terminal, 6...Subtractor, 7...Quantizer, 8...Output terminal, 9...Adder , 10... changeover switch, 11, 12... changeover switch selection terminal,
13... Frame memory, 1A... Comb filter for brightness signal extraction, 15... Changeover switch, 16, 17... Changeover switch selection terminal, 18... Reception input terminal, 19...
Adder, 20...Switch switch, 21, 22...
...Selector switch selection terminal, 23...
Frame memory, 24... Comb filter for carrier color signal extraction, 25... Subtractor, 26...
...Selector switch, 27,28...Selector switch selection terminal, 29...Output terminal, 30...
...Comb filter for brightness signal extraction, 31...
- Changeover switch, 32, 33...Selection terminal of changeover switch.

Claims (1)

[Claims] 1. For a composite color signal consisting of a luminance signal and a carrier color signal, the transmitting unit transmits a color frame consisting of the composite color signal itself and a luminance frame consisting only of the luminance signal extracted from the composite color signal, frame by frame. and for the input signal of the color frame, create a predicted value using the decoded signal pixel value of the composite color signal two frames before stored in the frame memory,
The difference between the input signal pixel value and the predicted value is quantized and encoded and transmitted, and for the input signal of the luminance frame, a decoded signal of the composite color signal of the previous frame stored in the frame memory is used. Creating a predicted value using the extracted luminance signal pixel value, quantizing and encoding the difference between the input signal pixel value and the predicted value, and transmitting the same; during a frame period in which the input signal of the color frame is encoded; sequentially rewrites the contents of the frame memory pixel by pixel with the new decoded signal pixel value obtained by adding the predicted value and the quantized difference value,
Also, during the frame period in which the input signal of the luminance frame is encoded, rewriting of the contents of the frame memory is stopped, and the receiving section adds the received quantized difference value to the predicted value created by the same algorithm as the transmitting section. and decoding the color frame and luminance frame signals alternately every frame, and during the frame period in which the color frame signal is decoded, the contents of the frame memory are sequentially rewritten with new decoded signal pixel values. , and stopping rewriting of the contents of the frame memory during a frame period in which the signal of the luminance frame is decoded, and
A color signal interframe encoding method, characterized in that, for the decoded signal of the luminance frame, color components are reproduced using a carrier color signal extracted from the decoded signal of an adjacent color frame.