JP2834134B2 - Multiplexing device and multiple decoding device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a multiplex coding apparatus, and more particularly to a high efficiency coding of an image such as a television signal, and more particularly to a high efficiency coding of one television signal. The present invention relates to a multiplex coding apparatus capable of transmitting a plurality of television signals in a time-division manner.

[Conventional technology]

Since a television (hereinafter abbreviated as TV) signal has a wide frequency band, a high transmission rate is required if the signal is converted into a digital signal and transmitted as it is. In order to reduce this speed, TV signal high-efficiency coding and decoding devices (hereinafter abbreviated as TV Codec) have been developed which compress the redundancy of TV signals.

Further, one of the present inventors has realized a method of effectively using an expensive transmission line and a TV codec by transmitting a plurality of TV signals in a time-division manner by one TV codec. 30256. The operation will be briefly described with reference to FIG.

 First, the operation on the transmitting side is as follows.

1) The TV signals read from the three sets of TV cameras 111 are switched by a switch 112, and one of the TV signals is periodically selected. 2) The selected TV signal is digitally converted by an analog / digital conversion circuit 113. 3) The signal is further delayed by three pixels by the delay circuit 114 and then multiplied by a, and is subtracted by the subtraction circuit 115 from the predicted signal value generated and multiplied by a.

4) The prediction error resulting from the subtraction is quantized / encoded by the quantization 116 / encoding circuit 117 and then transmitted to the transmission path.

5) At the same time, the quantization result and the prediction signal are
8 and input to the delay circuit 114 for use as a predicted value three pixels later. The receiving side performs the reverse operation. That is, 1) “the result of the above quantization and encoding” and “delay circuit
And a predicted value generated by multiplying the signal delayed by three pixels by a with respect to “a”.

2) The above addition result is distributed by the switch 123,
The data is displayed on three TV receivers 125 through three digital / analog conversion circuits 124.

In the above-described conventional technique, a case is described in which a so-called D-PCM in which a difference from a signal three pixels before is highly efficient coded, that is, a so-called D-PCM is used. Motion compensation inter-frame prediction, orthogonal transform coding,
It can be applied to any high-efficiency coding method such as vector quantization or a combination thereof.

[Problems to be solved by the invention]

The above configuration has a problem that it cannot be used flexibly as exemplified below. That is, for example, three TV cameras
Depending on whether the TV signal is a still image or a moving image, or a main screen or a sub-screen, the desired spatial resolution and the number of screens transmitted per second cannot be freely set. .

Therefore, an object of the present invention is to determine the properties (movement, importance) of an image.
Accordingly, it is an object of the present invention to realize a multiplex coding apparatus capable of controlling the resolution of an image and the number of screens per second and a transmission system using the same.

[Means for solving the problem]

The above object is achieved by adding means for changing a cycle at which a TV signal is switched and means for notifying a receiving side which TV signal is being transmitted.

[Action]

With the above means, it is possible to transmit a plurality of TV signals in a time-division manner with one TV Codec. At this time, a virtual transmission path is adaptively allocated to each TV signal according to the amount of generated information and the necessity of the user, so that an image quality most suited to the request of the user can be obtained as a whole.

〔Example〕

Hereinafter, an embodiment of the present invention will be described with reference to FIG. In the figure, only the area surrounded by the dotted line is the improved part according to the present invention, and the other parts are the same as the block configuration of the conventionally known apparatus for inter-frame predictive coding of TV signals. First,
The overall configuration of the inter-frame prediction encoding device will be briefly described.

TV camera 1 (can be any of 1A, 1B, 1C. The following 2, 3, 4,
The same applies to 12 and the like), the TV signal is scanned in the scanning line direction, is converted into a digital signal by the analog / digital converter 2, becomes a pixel signal, and is stored in the temporary storage memory 4.

The generated code amount prediction circuit 3 uses the pixel signal to
Predicts the code amount of the entire screen and the result
Transmit to 11.

One embodiment of the generated code amount prediction circuit 3 is shown in FIG.
That is, the screen transmitted immediately before is stored in the memory 51,
The difference between the two screens at the same position is calculated by the difference circuit 52, and squared by the squaring circuit 57. Furthermore, an addition circuit
Using the 54 and the latch 55, the sum of squares of the difference signal between the screen transmitted immediately before, that is, the power is obtained. This power is quantized by ROM56
To generate a generated code amount prediction value.

Note that the generated code amount can be predicted even if an absolute value circuit is used instead of the squaring circuit 57.

When the prediction of the generated code amount is completed, the control unit 11
As shown in FIG. 4, the ROM 62 is searched using the generated code amount prediction value signal 30 in order to keep the code amount actually generated in each screen constant, and the quantization precision is determined.

When the quantization precision is determined, the signals stored in the temporary storage memory 4 are read out collectively in units of several pixels and become blocks. At the same time, a pixel signal at the same position on the previous screen is read from the frame memory 12 to become a prediction signal of the above-described block unit signal, and a prediction error is calculated by the subtraction circuit 6.

This prediction error is quantized by the quantization circuit 7 according to the quantization precision determined above. Furthermore, the quantized signal is
The variable-length encoding circuit 8 assigns a codeword having a length corresponding to the frequency of occurrence. This codeword is transmitted to the transmission line 10 at a constant speed through a buffer 9 for smoothing the transmission speed.

At the same time, the signal quantized by the quantization circuit 7 is stored in the frame memory 12 after being added to the prediction signal. This signal is used for prediction of a subsequent screen.

The overall configuration of the TV signal inter-frame encoding apparatus has been briefly described above. Next, the parts 21 and 22, which are main parts of the present invention, will be described in detail. In the input part 21, TV camera 1, analog /
There are three sets (A, B, C) of the same combination as the digital converter 2, the generated information amount prediction circuit 3, and the temporary storage memory 4, and one of the output signals from the temporary storage memories 4A, 4B, 4C is output. Selection is made by switch 5. The prediction section 22 includes frame memories 12A, 12B, and 12C corresponding to the respective signals of the input section.
3, Select by switch 14. These switches 5,13,1
4 is controlled by the control unit 11.

The switches 5, 13, and 14 are switched on a screen basis. When the control unit 11 detects the end of the encoding of the screen currently being encoded through the signal line 32, it increments the counter 63 by one. The output of the counter 63 is input to the signal source selection ROM 64,
A signal for selecting a signal source to be transmitted next is output to the selection signal line 33. The selection signal line 33 is connected to the switches 5, 13, and 14, and switches a signal source and the like. At the same time, select signal line 33
Is also connected to the switch 65, determines the quantization precision for the selected signal, and outputs it to the quantization precision signal line 31.

When the variable length encoding circuit 8 identifies the signal source to be transmitted next from the selection signal line 33, it encodes and transmits the identifier according to a predetermined method.

Here, the signal source to be transmitted is selected according to a predetermined order (for example, 1 → 2 → 3 → 1 → 2 → 3 →
Or 1 → 2 → 1 → 3 → 1 → 2 → 1 ...), the following selection method is also possible.

1) The user determines or selects the order of selection in advance.

2) The generated code amount prediction values are selected in descending order.

3) Select according to a sequence table corresponding to the combination of the generated code amount prediction values.

For example, assuming that the number of input sources is 3 (A, B, C) and the number of generated code amount prediction values is 3 (generated code amount is small, medium, and large), (A, B, C)
= (Small, small, small), (small, small, medium), (small, small, large),
(Small, medium, small), (small, medium, medium) ... 27 of (large, large, large)
The transmission order is determined according to the combinations.

4) Select a signal source specified by the transmitting or receiving user. After the selection, the signal is continuously selected until the next specification.

5) Analyze the voice input corresponding to each signal source, and preferentially select, for example, a signal source having a voice input.

As means for communicating the selected signal source to the receiving side, other means such as a pair can be considered.

1) Transmit the identifier using another data line.

2) The order of signal selection is determined in advance, or the method of selection is determined in advance, and switching to the next signal source is performed using a code indicating the start or end of each screen. However, at the beginning of transmission, it is necessary to transmit the identifier of the selected signal source.

3) Predetermine some tables that indicate the order of selection in advance,
Which table to use is transmitted at the start of transmission or during transmission, and is usually switched to the next signal source by a code indicating the start or end of each screen.

In some cases, it may be desirable to determine the signal source to be transmitted next and, at the same time, determine the quantization accuracy for the screen. The following method can be considered as a determination method in this case.

1) All have the same accuracy.

2) The quantization accuracy is determined independently for each signal source.

For example, the prediction of the generated code amount is large, the quantization precision is coarse, and the prediction is small.

3) Quantization accuracy is determined by the sum of all generated code amount prediction values. For example, by making the sum of the generated code amounts of all the signal sources always equal to or less than a fixed amount, the number of transmission frames is made uniform.

4) The quantization precision is determined in advance for each combination of the generated code amount prediction values. For example, after equalizing the generated code amount as in the above 3), a large amount of code is allocated to a specific signal source to improve the image quality of the signal source (priority is set for each signal source). wear).

5) Refine the quantization accuracy for a signal source (especially a still image) specified by the user.

Next, the receiving side will be described. FIG. 5 is a block diagram of one embodiment of the receiving side according to the present invention. The operation will be described below.

The signal sent from the transmission line 10 is stored in a buffer 80 and decoded by a variable-length code decoding circuit 81. At the same time, one of the previous screen signals stored in the frame memories 86A, 86B or 86C is selected by the switch 89 to become a predicted value, and is added to the decoded signal by the adder 82. The added signal is stored in one of the frame memories 86A, 86B, and 86C and one of the temporary storage memories 83A, 83B, and 83C selected by the switch 90. The contents of the temporary storage memory 83 are read out as appropriate and
The signal is converted into an analog signal by the A
Is output to

Upon decoding the signal source selection identifier, the variable length code decoding circuit 81 writes the signal source selection signal 100 to the latch 87. The signal 101 stored in the latch 87 is supplied to the switches 88, 89, and 90, and is switched to a path corresponding to the designated signal source, and decoding is performed.

According to this method, multiplexing of a TV signal can be performed only by adding a switching switch of an input unit or the like or a frame memory. Even when the image of one signal source largely moves and many codes occur, if the image of the other signal source does not move much, the occupancy of the moving signal in the transmission path automatically increases, and Deterioration of the image quality of the signal and a decrease in the number of frames can be reduced. When the images of all signal sources generate the same amount of code, the distribution of the transmission paths is automatically divided equally.

In addition to the case illustrated in FIG. 1, the present invention is also applicable to the case where motion compensation inter-frame coding is performed, the case where orthogonal transform (discrete cosine transform or the like) or vector quantization is performed in quantization, or the case where both are used. The invention is applicable. In the above embodiment, the switches 5, 13, and 14 or the switches 88, 89, and 90 are switched at the same time. However, when the encoding and decoding processes are performed in a pipeline manner, the switching is not necessarily performed at the same time. There is no. In the above embodiment, the case where the number of signals to be multiplexed is 3 has been described, but the number can be set to any number of 2 or more.
Also, it is possible to multiplex signals smaller than the set maximum number of multiplexed signals. Further, in this embodiment, the signal source on the transmitting side and the output destination on the receiving side have been described as being fixed on a one-to-one basis, but the transmission A-receiving C, the transmitting B-receiving A, and the transmitting C-receiving B The correspondence may be switched as described above. Also, for example, (N) kb / s
It is also possible to multiplex three TV Codec signals in a time-division manner, receive this on the (3N) kb / s receiving side in FIG. 5, and separate and output it into three screens. In addition, two cameras are used as input sources, each signal is stored in two temporary storage memories on the receiving side, and each signal is given a different color or deflected, and switched to TV frame or TV field units. It is also possible to transmit a stereo screen by outputting to one screen.

〔The invention's effect〕

As described above, according to the present invention, one TV Codec
By simply adding a frame memory, a switching switch, and the like to the system, a plurality of signal sources can be efficiently encoded and transmitted. Considering the recent low cost of LSI memories, the burden for this is small.

Further, since one line is adaptively allocated to a plurality of signal sources, the lines can be used efficiently. For example, (3 × N) kb / s
When three TV signals are multiplexed using the TV codec for each line, when the movement of each signal is small, each (N) kb
/ s is assigned (generally, transmission is possible at Nkb / s or less when stationary). When one signal largely moves, a code amount of (N) kb / s or more is assigned to the signal.
As a result, the image quality is higher than when one TV signal is transmitted through a (N) kb / s line, and the number of transmission frames is increased.

[Brief description of the drawings]

FIG. 1 is a block diagram of an embodiment of an encoding apparatus according to the present invention, FIGS. 2 (a) and 2 (b) are block diagrams of a conventional encoder and decoding block, respectively, and FIG. FIG. 4 is a block diagram of one embodiment of a generated code amount prediction circuit 3, and FIG.
FIG. 5 is a block diagram of an embodiment of the control unit 11, and FIG. 5 is a block diagram of an embodiment of a decoder corresponding to FIG. DESCRIPTION OF SYMBOLS 1 ... TV camera, 2 ... A / D conversion circuit, 3 ... Generation code amount prediction circuit, 4 ... Temporary storage memory, 5 ... Input selection switch, 7 ... Quantization circuit, 8 ... Variable length Encoding circuit,
11 ... Control unit, 12 ... Frame memory, 33 ...
Selection signal line (transmitting side), 57: squaring circuit, 56: Generated code amount predicted value ROM, 62: Quantization parameter ROM, 64:
Signal selection ROM, 81: Variable length code decoding circuit, 84: D / A
Converter 85 TV monitor 101 Selection signal line (reception side) 121 121 Clock delay circuit.

──────────────────────────────────────────────────続 き Continuation of the front page (72) Inventor Hiromi Watanabe 1-280 Higashi Koikebo, Kokubunji-shi, Tokyo Inside the Central Research Laboratory, Hitachi, Ltd. (56) References JP-A-52-71921 (JP, A) JP-A Sho 62-137983 (JP, A) JP-A-58-145284 (JP, A) JP-A-60-143088 (JP, A) JP-A-63-90953 (JP, A) (58) Fields investigated (Int. Cl. ⁶ , DB name) H04N 7/24-7/68 H04N 7/04-7/10

Claims (3)

(57) [Claims] 1. A first method for switching n sets (n ≧ 2) of TV signals before high-efficiency coding in a predetermined order, time-division multiplexing at a predetermined ratio, and outputting one multiplexed signal. Switching means, one high-efficiency encoding means for high-efficiency encoding and transmitting the multiplexed signal, and each multiplexing signal in the first switching means.
Control means for controlling the rate of time-division multiplexing of the TV signal, wherein the high-efficiency coding means performs n sets of prediction signals corresponding to the n sets of TV signals in frame units, respectively. N independent storage means for storing, a second switching means for switching a storage means for reading out a prediction signal among the n storage means, and a prediction signal among the n storage means. Third switching means for switching the storage means for writing, wherein the control means performs each switching control of the second and third switching means in conjunction with the switching control of the first switching means. ,
When the first switching means selects the m-th (m = 1,..., N) TV signal from the n sets of TV signals, the m-th storage signal corresponding to the m-th TV signal is stored. A multiplex coding apparatus characterized by causing a means to read and write a prediction signal. Further, based on the n sets of TV signals before the high-efficiency coding, n generated code amount predictions for predicting the code amounts generated when the respective TV signals are respectively subjected to the high-efficiency coding. The control means controls a rate of time division multiplexing in the first switching means based on a code amount prediction result in each of the generated code amount prediction means. 2. The multiplex encoding device according to 1. 3. Receiving means for receiving a multiplexed coded signal in which n sets (n ≧ 2) of TV signals are time-division multiplexed and highly efficient coded, and multiplexes the multiplexed coded signal by highly efficient decoding. One high-efficiency decoding means for outputting a decoded signal; separating means for separating the multiplexed decoded signal into the n sets of TV signals;
A multiplex decoding apparatus comprising: identification means for identifying which of the n sets of TV signals the received multiplexed coded signal relates to, among the n sets of TV signals, wherein the high-efficiency decoding means comprises the n sets of N storage means for storing n sets of prediction signals corresponding to the TV signals in units of frames, respectively, and a first means for switching the storage means for reading out the prediction signals among the n storage means. Switching means, and a second switching means for switching a storage means for writing a prediction signal among the n storage means, wherein the identification means performs the first and second storage operations based on an identification result.
The respective switching control of the switching means is performed, and the reception means performs the above-mentioned n switching.
When the m-th (m = 1,..., N) TV signal of the set of TV signals is received, the prediction signal is read out from the m-th storage means corresponding to the m-th TV signal and read. A multiplex decoding device for performing writing.