JPS6252984B2 - - Google Patents

Description

[Detailed Description of the Invention] The present invention relates to an encoding device that converts multilevel digital signals, such as video signals and audio signals, in which temporally close signals are likely to have a strong correlation, into an efficient code sequence. .

In a system that digitally records a video signal obtained by scanning a picture containing halftones, it is necessary to efficiently transmit a multi-value digital signal. Therefore, a case where such a multilevel digital signal is double encoded will be explained as an example.

Conventionally, when binary-encoding a multilevel digital signal, the most commonly used method is level-based run-length encoding, and examples of this encoding are shown in Figures 1a to 1.
It is shown in Figure 1e. Figure 1a shows the input signal when the number of levels is 6 (level 1 to level 6),
Fig. 1b shows a predictive conversion signal obtained by predictively converting this input signal using its left neighbor signal as a predictive signal.
“0” indicates a prediction match, and “1” indicates a prediction mismatch. In Figure 1c, the predicted conversion signal in Figure 1B has many "0s" that are prediction matches, so it is further converted to a code sequence with a shorter length using the code example for the predictive conversion signal shown in Figure 2. This is what I did. Then, on the receiving side, it is possible to decode the signal of the code sequence shown in FIG. 1c (predictive conversion signal after encoding) into the predictive conversion signal shown in FIG. In this case, it is only known that the predictions do not match, so it is necessary to send a transition code word indicating a transition from the prediction conversion signal. Therefore, it is necessary to send the transition code word shown in FIG. 1d every time the predicted conversion signal shown in FIG. 1b is "1". and,
The predictive conversion signal shown in FIG. 1b and the transition code word shown in FIG. 1d are combined, and the transition code shown in FIG. 1e is transmitted. Therefore, on the receiving side, this
The signal shown in FIG. 1a is decoded by receiving the transition code shown in FIG.

Note that FIG. 3 shows an example of measuring the generation transition destination for each level of predicted signals, and FIG. 4 shows an example of a conventional transition code.

However, in conventional encoding, it is necessary to send a predictive transform signal after code sequencing and also to send a transition code word when the predictive transform signal is "1", so the code length of the transition code must be sufficiently shortened. I can't. As a result of a detailed study of conventional coding, it was found that this is because conventional transition codes do not sufficiently utilize information on the direction of change of the preceding change point. That is, FIG. 5 is an example of measuring the transition destination that occurs when predictions do not match when the direction of change in the most recent past is focused and classification is performed in positive and negative directions. From FIG. 5, it can be seen that level changes tend to occur in the same direction as the change in the most recent past.

Accordingly, an object of the present invention is to provide an encoding device that reduces the code length of a transition code based on the above-mentioned principle.

In order to achieve this purpose, this invention
A device for level-based run-length encoding of an input multilevel digital signal includes a predictor that stores the closest past input signal to the current input signal as a predicted signal, and a predicted signal from this predictor. a memory that stores the direction in which the value indicated by the current input signal changes with respect to the value; and a memory that stores the direction in which the value indicated by the current input signal changes; and a transition code generator that generates a transition code indicating the value of the difference and the direction of change, and will be described in detail below using examples.

FIG. 6 is a block diagram showing an embodiment of the encoding device according to the present invention. In the figure, (1) is a predictor that stores the left adjacent signal, that is, the past input signal closest to the current input signal, as a predicted signal, and (2) is a predictor that stores the value indicated by the current input signal and the predicted signal. (3) is a memory that stores the direction in which the value indicated by the current input signal changes with respect to the value indicated by the predicted signal; (4) is a predicted signal encoder; (4) is a predicted signal encoder; 5) is a transition signal encoder configured with a read-only memory (ROM), and (6) is the output code of the predictive signal encoder 4 and the transition signal encoder 5.
This is a code controller that switches and outputs the output code.

Next, the operation of the encoding method according to the above configuration will be explained with reference to FIG. 2.

The input signals are input to a predictor 1, a comparator 2 and a transition signal encoder 5, respectively. For this reason, the predictor 1 outputs a predicted signal to the comparator 2. Therefore, the comparator 2 compares each value of the current input signal and the predicted signal, and if they are equal, outputs a "0" symbol to the predicted signal encoder 4. For example, when the number of "0" symbols reaches a certain number based on FIG. 2, the predictive signal encoder 4 creates (two in the example of FIG. 2) "0" codes and sends them out. On the other hand, if the input signal and the predicted signal are not equal, a "1" symbol is output to the predicted signal encoder 4. Therefore, when using the code shown in FIG. 2, the predictive signal encoder 4 always creates a code of "10" or "11". At this time, the transition signal encoder 5 indicates the value and change direction of the difference between the current input signal and the predicted signal based on the current input signal, the predicted signal from the predictor 1, and the output signal from the memory 3. Generates a transition code, code controller 6
Output to.

Therefore, the code controller 6 is the predictive signal encoder 4
After outputting the output signal from , this transition code is output as an output code to a line (not shown). Thereafter, the contents of the memory 3 are rewritten to a new change direction signal.

On the other hand, the receiving side is similarly equipped with a predictor, change direction memory, predicted signal decoder, and transition signal decoder, and it goes without saying that the received signal can be easily demodulated by performing the opposite operation to the transmitting side. be.

Note that FIG. 7 shows an example of transition encoding by the transition signal encoder 5 of FIG. 6, and the method of creating transition codes from transition probabilities is based on a known highly efficient method. Examples of encoding according to the present invention are shown in FIGS. 8a to 8h. In this case, Figures 8a to 8
FIG. 1e corresponds to FIGS. 1a to 1e, respectively, and the transition code shown in FIG. 8e is encoded as shown in FIG. 6 to further shorten the code length. FIG. 8f is a diagram showing the direction of change in the most recent past for the input signal shown in FIG. 8a, and FIG. 8g is a diagram showing the direction of change in the input signal shown in FIG. In the predicted transformed signal,
8 is a diagram showing a transition code word that needs to be sent every time a symbol of “1” appears, and as shown in FIG.
The transition code signal shown in Fig. g is combined and transmitted. Therefore, the receiving side receives the binary code shown in FIG. 8h and reproduces the signal shown in FIG. 8a.

The binary code sequence shown in Fig. 8h is shown in Fig. 1e.
It is clear that the transition code sequence is shorter than the transition code sequence shown in .

As described above in detail, according to the encoding device according to the present invention, by storing the direction of change in the most recent past, the code length of the transition code can be shortened, and the writing/picture signal including halftones can be shortened. , which has a great effect on reducing the transmission time in digital transmission of audio signals.

[Brief explanation of the drawing]

Figures 1a to 1e are diagrams for explaining conventional encoding, Figure 2 is a diagram showing an example of a predictive conversion signal code, Figure 3 is a diagram showing an example of measurement of a transition destination, and Figure 4 5 is a diagram showing an example of a conventional transition code, FIG. 5 is a diagram showing an example of measurement that also takes into account the direction of the immediately preceding change, FIG. 6 is a block diagram showing an embodiment of the encoding device according to the present invention, and FIG. This figure is a diagram showing an example of the transition code used in FIG. 6, and FIGS. 8a to 8h are diagrams showing encoding according to FIG. 6. 1...Predictor, 2...Comparator, 3...Memory device,
4... Prediction signal encoder, 5... Transition signal encoder,
6...Sign controller. Note that the same reference numerals in the figures indicate the same or corresponding parts.

Claims (1)

[Claims] 1. A device for level-based run-length encoding of an input signal of a multilevel digital signal, comprising: a predictor that stores a past input signal closest to a current input signal as a prediction signal; a memory for storing the direction in which the value indicated by the current input signal has changed with respect to the value indicated by the predicted signal from the predictor; and the current input signal, a predicted signal from the predictor, and an output signal from the memory. and a transition code generator that generates a transition code indicating the value and direction of change of the difference between the current input signal and the predicted signal based on the current input signal and the predicted signal. 2. Claim 1, characterized in that the transition signal encoder is constituted by a read-only memory.
Encoding device described in Section 2.