JPS62296633A - Data compressing and transmitting equipment using differential pulse modulation system - Google Patents

Abstract

PURPOSE:To improve the coding efficiency by quantizing only the effective data range of a difference signal while converting it into a characteristic being the object of quantization attended with the fluctuation of a threshold level so as to reduce the quantized error. CONSTITUTION:The change detection circuit of a transmission circuit outputs a difference signal based on the threshold level and a predication difference signal between an input signal and predicate signal calculated by a prediction circuit, a quantization circuit quantizes the difference signal and stores it tentatively in a transmission buffer circuit, and a local decoding circuit decodes the quantization signal locally. Further, a reception circuit is provided with a reception buffer circuit, a local decoding circuit and an adder. Then the quantization circuit consists of an adaptive quantization circuit including a characteristic selection circuit 8 selecting plural quantization characterstics of the circuit depending on the hershold level and a local decoding circuit consists of an adaptive local decoding circuit including a characteristic selection circuit 9 selecting the plural decoding characteristics of the circuit based on the threshold level so as to switch the effective data range of the difference signal into a characteristic being the object of quantization to apply quantization.

Description

[Detailed Description of the Invention] 3. Detailed Description of the Invention [Field of Industrial Application] The present invention is a data compression and transmission device using a differential pulse modulation method, and in particular, a data compression and transmission device for band-compressing and transmitting data such as image signals. This article concerns the improvement of a transmission device suitable for

[Prior Art] A data compression transmission device using a conventional differential pulse modulation method will be described below with reference to the drawings.

(The data compression transmission device using the conventional differential pulse modulation method converts digital input signals into differential pulse modulation (hereinafter referred to as DPC).
a transmitting circuit section that performs data compression transmission using a transmission circuit (abbreviated as M), and a receiving circuit section corresponding to the transmitting circuit;
The figure shows a block diagram of the transmitting circuit section.

In the figure, a prediction circuit (1) predicts a digital input signal as described later and calculates a predicted signal @ (101). This predicted signal (101) is transmitted to the human power signal @
(100), and a prediction error signal (102), which is the difference between the two, is calculated.

Then, in the change detection circuit (3), the prediction error signal (102) is converted into a difference signal (103) and this difference signal (
103) is converted into a change detection signal (113) indicating whether the data is in the valid/invalid data range (whether or not the amount of change is larger than the corresponding threshold value), and in the quantization circuit (4b),
The difference signal (103) is quantized.

The quantization signal (104b) is the output of the quantization circuit (4b).
) is encoded in the variable length encoding circuit (10) and sent to the transmission buffer (11).

Further, the quantized signal (104b) is transmitted to the local decoding circuit (5b).
), thereby calculating a reproduction difference signal (105b) for obtaining a reproduction input signal (106) (for calculating the predicted signal).

The threshold generation circuit (7) is a circuit that generates a threshold value (107) to be supplied to the change detection circuit (3) in accordance with the amount of information in the transmission buff nearer (11).

FIG. 7 is a detailed explanatory diagram of the change detection circuit (3), and the absolute value circuit (12) is the absolute value (
112), and the comparator circuit (13) compares the threshold value (107) with the absolute value (112>).
This is a circuit that determines whether there is a change of more than a predetermined amount, determines whether or not there is a change of data that is less than a predetermined amount, and outputs a change detection signal (113) indicating the validity/invalidity of the data, and a zero allocation circuit (14). is a circuit that assigns ##OH to the prediction difference signal @ (102) when the change detection signal (11'3) is invalid.

Next, the signal flow will be explained. As shown in FIG. 6, digitized input signals such as image signals F1'@(1
00) is s, predicted signal @ (1' 01 ) is [)1, prediction error signal (102) is 01, difference signal (103) is ei
, reproduced differential signal @ (105b) 6 reproduced input signal (1
06) to 94. Then, the following relationship holds between the values of each signal.

eS=8. >-P, 2 ej! = e, 10 d @ 1 = e I +Q where d is the invalid error Q in the change detection circuit is the suspension error A is the prediction count Z4 is the reproduction input signal @@ in the delay prediction circuit (1) at time t, g(106) is delayed for a preset 1 hour, and further (2, coefficient A
The predicted signal Pi (101) is obtained by multiplying by .

Note that the input signal @S, (100) is an image signal and the time t is 1
If set to one frame time, inter-frame DI-) CM transmission equipment, if set to one field time, inter-field D
It becomes a PCM transmission device.

As shown in FIG. 7, in the change detection circuit (3), the absolute value (112) of the prediction error signal eS (102) calculated by the absolute value circuit (12) is set to 10,1. When the threshold (107) is tossed, the comparator circuit (1
The value of the change detection signal (113) V calculated by 3H is determined as follows.

Hl e,l <Th then V=O (invalid) else V=1 (valid) Furthermore, if no change is detected for a predetermined period (in the case of an invalid data range), the zero allocation circuit (14) assigns a prediction error signal to e, (102> is assigned II 011,
The difference signal J (103) is output as 1101#.

Then, the difference molecule f 1 e 1 (103) input to the quantization circuit (4b) is converted into a quantized signal χ, (104b) according to an arbitrary characteristic.

Then, the variable length encoding circuit (10) takes in only the quantized signal χ, (104b) when the change detection signal V (113) is valid (V=, 1), and takes in the change detection signal V (
For the quantized signal χ, (1,04b), run-length encoding is applied to the quantized signal ), and the data accumulated in the transmission buffer circuit (11) is sent to the transmission path as an encoded signal (200>).

Further, the threshold generation circuit (7) monitors the amount of data accumulated in the transmission balance circuit (11), generates an appropriate threshold according to the amount of data accumulated, and smoothes the data code amount.

FIG. 8 shows a block diagram of the receiving circuit section of a data compression transmission device using the conventional differential pulse modulation method. The variable length encoding circuit (22b) is a circuit for restoring the front i encoded signal (200), and the local decoding circuit (23b) outputs the reproduced difference signal (223b). It is a prediction circuit (
24) is a circuit that predicts the reproduced signal (225>).

The reproduced signal @(225> is the predicted signal (224>
The adder (25) by the reproduced difference signal @ (223b)
He was taken out of Korea.

Next, the flow of signals in the receiving circuit section will be explained.

Encoded signal variable-length coded in the transmitter circuit @ (200)
is received by the receiving buffer 7 circuit (21) and sent to the variable length decoding circuit (22b>).

Then, only when the change detection signal @V (113) decoded in the variable length decoding circuit (22b) indicates a valid data range (V=1), the quantized signal (222b) is transferred to the local decoding circuit (23b). ), and when the restored change detection signal V (113) indicates an invalid data range (V=O), 170 fl is output to the local decoding circuit (23b).

Then, the local decoding circuit (23b) decodes the quantized signal @ (222b) into a reproduced difference signal (223b), and the prediction circuit (24) generates a predicted signal (224>
The encoded signal (
200> is played back (225).

[Problems to be Solved by the Invention] Since the conventional data compression data device using DPCM is configured as shown above, the quantization characteristic of the quantization circuit in the transmitting circuit section has a threshold value. Although the dynamic range of effective data becomes narrower as the size increases, there is a problem in that quantization accuracy does not improve.

81 The present invention was made to solve the above problems, and uses a quantization characteristic corresponding to a threshold value to reduce quantization errors and uses a DPCM with high quantization efficiency. The purpose is to obtain a data compression transmission device.

[Means for Solving the Problems] In order to achieve the above object, the quantization circuit in the transmitting circuit section of the present invention has a plurality of quantization characteristics, and the quantization characteristics are determined by the threshold value. The local decoding circuit in the transmitting circuit section and the receiving circuit section in the present invention has a plurality of decoding characteristics corresponding to the quantization circuit, and the local decoding circuit in the transmitting circuit section and the receiving circuit section in the present invention has a plurality of decoding characteristics corresponding to the quantizing circuit, and an adaptive local decoding circuit including a characteristic selection circuit that selects a quantization characteristic based on the threshold value, and the quantization characteristic selected by the quantization circuit is within the invalid data range (within the corresponding threshold value). quantized output of the difference signal (within the specified range) is 'O''
This characteristic is adaptively quantized according to the threshold value.

[Operation] As is clear from the above-described configuration, according to the present invention, as the threshold value changes, a plurality of characteristics are prepared in which only the effective data range of the difference signal is subject to quantization, and the threshold value changes. Since the configuration is configured such that quantization is performed by switching to a corresponding quantization characteristic, it is possible to reduce quantization errors and obtain a data compression transmission device using DPCM with high quantization efficiency.

[Embodiments] Hereinafter, preferred embodiments of the present invention will be described based on the drawings.

A data compression transmission device using a differential pulse modulation method according to the present invention includes a transmission circuit section that performs data compression transmission of a digital input signal using differential pulse modulation, and a reception circuit section corresponding to the transmission circuit, FIG. 1 shows a block diagram of the transmitting circuit section according to this embodiment.

In the present invention, the quantization circuit is an adaptive quantization circuit (4a).
The adaptive quantization circuit (4a) has a plurality of quantization characteristics, selects the characteristic from the threshold value (107), adaptively quantizes the difference signal (103), and adaptively quantizes the adaptive quantization signal (104).
Output a).

The adaptive local decoding circuit (5a) is provided corresponding to the adaptive hanging circuit (4a), and the adaptive local decoding circuit (5a) is provided with the adaptive quantization signal @(1
04a) and outputs a reproduced difference signal (105a>), and the same parts as those in the conventional example are given the same numbers, and the explanation thereof will be omitted.

FIG. 2 is a detailed block diagram of the adaptive quantization circuit (4a), in which the threshold value (107
), the quantization characteristic is selected, and the difference signal (103) is quantized.

FIG. 3 is a detailed block diagram of the adaptive decoding circuit (5a), in which the threshold value (107
), the decoding characteristics are selected by the quantized signal (104a
> is decoded, and further passes through a zero assignment circuit (15) and is assigned a zero by a change detection signal (113).

Next, the signal flow will be explained. As shown in FIG. 1, a digitized input signal (100) such as an image signal
3. ．． 1! As described above, the prediction signal @(101) calculated by the prediction circuit (1) is converted into P2) prediction error signal! (102
> is C5, the differential signal (103) is 61, the reproduced differential signal (105b) is ~, and the reproduced input signal (106) is hl, then the following relationship holds between the values of the respective signals.

C5-8I Pl e p ・-e , 10d 6Jl""e Jl0Q ■ However, d is the invalid error in the change detection circuit Q1 is the quantization error △ is the prediction count Z"1 is the delay prediction circuit at time t ( In 1), the reproduced human power signal @54 (106) is delayed by a preset one hour, and further multiplied by a coefficient A to output the prediction signal PJ! (101).

In the change detection circuit (3), similarly to FIG. 7, the prediction error signal e calculated by the absolute value circuit (12),
If the absolute value (112> of (102) is 10.1 and the threshold value (107) is Th, then the value of the change detection signal @ (113) V calculated by the comparator circuit (13) is the following J
: It is decided by sea urchin.

if Ie, l<Th then V=0 (invalid) else V=1 (valid) Furthermore, if a predetermined amount of change is not detected (in the case of an invalid data range), the zero allocation circuit (14) assigns a prediction error signal to "0" is assigned to e, (102), and the difference signal ei (103) is output as u OIU to the adaptive quantization circuit (4a).

The operation of the adaptive quantization circuit (4a) is as shown in FIG. Quantized signal χ according to the quantization characteristic specified by the characteristic selection circuit (9). (104a) and output to the variable length encoding circuit (10).

Here, the example of the quantization characteristic shown in FIG. 5(a) will be explained.

The dashed-dotted line is a characteristic example when the threshold value is Th=O, and the solid line is a characteristic example when the threshold value (107) Th=3. As shown in this figure, when the data of the difference signal (103) is in the invalid data range (-Th, data < +7h), it is not quantized and is not quantized, but in the valid data range (data <-Th,
Only cases within Or, +Hh<data) are subject to quantization. As is clear from the figure, the quantization characteristic selected by the characteristic selection circuit (8) is such that the quantization output of the difference signal within the invalid data range (within the specified range of the corresponding threshold value) is The quantization accuracy is improved in this way.

Then, in the variable length encoding circuit (10), when the change detection signal V (113) is valid (V=1>), only the quantized signal χ (104a) is taken in, a ) is assigned run-length encoding, and the quantized signal ), and the data accumulated in the transmission buffer circuit (11) is sent to the transmission path as an encoded signal @ (200>).

Also, the quantized signal χ. (104a ><A, as shown in FIG. 1, it is also sent to the adaptive quantization circuit (5a).

On the other hand, in the adaptive section decoding circuit (59), as shown in FIG. 3, the characteristic selection circuit (9) selects a decoding characteristic corresponding to the quantization characteristic selected by the adaptive quantization circuit. Further, the zero allocation circuit (15) allocates invalid data "O'' and outputs it as a reproduced difference signal@(105b). An example of its decoding characteristics is shown in FIG. 5(b).Reproduced difference signal ~(10
5b) is the predicted signal Pj! by the adder (6). (10
1) and Kaki's playback input signal @J! (106) and is input to the prediction circuit (1).

Further, the threshold generation circuit (7) monitors the amount of data accumulated in the transmission buffer circuit (11), generates an appropriate threshold according to the amount of data accumulated, and smoothes the data code amount.

FIG. 4 shows a block diagram of the receiving circuit section of the data compression and transmission apparatus using the differential pulse modulation method according to the present embodiment. The variable length encoding circuit (22a) is a circuit for temporarily storing the encoded signal (200), and the adaptive local decoding circuit (23a) is a circuit for decoding the encoded signal (200).
is a circuit that outputs the reproduced difference signal (223a>), and the prediction circuit (24) is a circuit that predicts the reproduced signal @ (225).

The reproduced signal (225> is the predicted signal (224>
The adder (25) is controlled by the reproduced difference signal (223a).
Calculated from.

Next, the flow of signals in the receiving circuit section will be explained.

Encoded signal (200) variable-length encoded in the transmitter circuit
is received by the reception buffer circuit (21) and sent to the variable length decoding circuit (22a).

Then, only when the change detection signal V (222a3) decoded in the variable length decoding circuit (22a) indicates a valid data range (V=1), the quantized signal (222al) is generated.
and the threshold value (222a2) is the local decoding circuit (23a
> and decoded change detection signal V (113)
indicates an invalid data range (V=O), IT O$
1 is output to the local decoding circuit (23a>).

Furthermore, in the adaptive local decoding circuit (23a), similarly to the adaptive local decoding circuit (5a) in the transmitting circuit section, the threshold value (222a2)G (from) q coding characteristic is selected,
The quantized signal (222a1) is the reproduced difference signal (223a>
The prediction signal (22) is decoded by the prediction circuit (24).
4> and the hl calculator's manipulation, the encoded signal (200) from the transmitting circuit section is reproduced (225>).

In addition, in this embodiment, the same effect can be obtained even if the threshold value to be transmitted is reduced by generating (updating) the threshold value once at an arbitrary time t (for example, one frame time). Obtainable.

[Effects of the Invention] As described above, the quantization circuit of the data compression and transmission device using the differential pulse modulation method according to the present invention has a plurality of quantization characteristics, and the quantization characteristics are determined based on the threshold value. The local decoding circuit in the transmission circuit and the reception circuit of the transmission device has a plurality of decoding characteristics corresponding to the quantization circuit, and the local decoding circuit in the transmitting circuit and the receiving circuit of the transmission device has a plurality of decoding characteristics corresponding to the quantization circuit, and Since it is composed of an adaptive local decoding circuit including a characteristic selection circuit that selects a characteristic based on the threshold value, the characteristic can be adaptively switched to a characteristic in which only the valid data range of the difference signal is subject to quantization as the threshold value changes. This has the effect of providing a transmission device that can reduce quantization errors and has high encoding efficiency.

[Brief explanation of the drawing]

FIG. 1 is a block configuration diagram showing a transmitting circuit section of a preferred embodiment of a transmission device using a differential pulse modulation method according to the present invention, and FIG. 2 is a detailed block configuration diagram of an adaptive quantization circuit in FIG. 1. FIG. 3 is a detailed block diagram of the adaptive local decoding circuit in FIG. 1, FIG. 4 is a block diagram showing the receiving side of the embodiment of FIG. 1, and FIGS. A diagram showing an example of adaptive quantization/reversion characteristics in
FIG. 6 is a block configuration diagram showing the transmitting circuit section of a transmission device using the conventional differential pulse modulation method, FIG. 7 is a detailed block configuration diagram of the change detection circuit in FIG. 6, and FIG.
FIG. 2 is a block configuration diagram showing a receiving side of a conventional example. In the figure, (1) is a prediction circuit, (2) is a subtracter, and (3
) is a change detection circuit, (4a) is an adaptive quantization circuit, (5a
) is an adaptive local decoding circuit, (6) is an adder, (7) is a threshold generation circuit, and (8). 9) is a characteristic selection circuit, (11) is a transmission buffer circuit, (
12) is an absolute value circuit, (13) is a comparison circuit, (14),
(15) is a zero assignment circuit, (21) is an absolute value circuit, (22
a> is a variable length decoding circuit, (23a) is an adaptive local decoding circuit, (24) is a prediction circuit, (25) is a trunk unit, (10
0) is the digital input signal, (101) is the predicted signal, (1
02> is the prediction error signal, (103) is the difference signal, (10
4a) is the quantized signal (105a) is the decoded signal, (10
6) is the reproduction input signal, (107) is the threshold value, (113
) is a change detection signal, (200> is a coded signal, (222
al> is the quantized signal, (222a2) is the threshold value, (2
22a3) is a change detection signal (223a) is a reproduced difference signal, (224> is a predicted signal, and (225) is a reproduced signal. In the drawings, the same members are given the same reference numerals and explanations will be omitted.

Claims (2)

[Claims] (1) In a data compression transmission device using a differential pulse modulation method, which includes a transmitting circuit that compresses and transmits digital input signals using differential pulse modulation, and a receiving circuit corresponding to the transmitting circuit, the transmitting circuit outputs a difference signal from a prediction circuit that calculates a prediction signal that predicts a digital input signal, a prediction error signal that is the difference between the input signal and the prediction signal, and a threshold that smoothes the amount of transmitted data. a change detection circuit, a quantization circuit that quantizes the difference signal, a transmission buffer circuit that temporarily stores the quantized transmission signal, and a threshold generation circuit that generates and controls the threshold; a local decoding circuit that locally decodes the quantized signal in order to calculate the early prediction signal; the receiving circuit includes a receiving buffer circuit that temporarily stores the received signal; and a receiving buffer circuit that decodes the received signal. The quantization circuit includes a local decoding circuit and an adder that calculates a reproduced signal from the decoded signal and the predicted signal, and the quantization circuit has a plurality of quantization characteristics, and the quantization characteristics are set to the threshold value. The local decoding circuit in the transmitting circuit and the receiving circuit has a plurality of decoding characteristics corresponding to the quantization circuit, and the local decoding circuit in the transmitting circuit and the receiving circuit has a plurality of decoding characteristics corresponding to the quantization circuit, and A data compression and transmission device using a differential pulse modulation method, comprising an adaptive local decoding circuit including a characteristic selection circuit that selects based on the threshold value. (2) In the data compression transmission device using the differential pulse modulation method according to claim (1), the quantization characteristic selected by the quantization circuit is within the invalid data range (corresponding threshold value). 1. A data compression transmission device using a differential pulse modulation method, characterized in that the quantized output of a differential signal (within a specified range) is "0".