JPH0366854B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an apparatus for encoding information such as television signals and audio signals.

[Conventional technology]

Television signals and audio signals themselves contain many redundant components, and redundancy suppression is performed by predictive coding such as delta modulation and DPCM in order to reduce the amount of transmitted information.

This predictive encoding device, as shown in FIG.
Difference circuit 22 that takes the difference between the input signal and the prediction circuit output
, a quantizer 25 that quantizes the difference circuit output, and a prediction circuit 2 that predicts the input signal from the quantizer output.
It consists of 6 and 27. When encoding television signals, the sampling frequency is extremely high at 10 MHz, so loop processing for differential encoding must be performed at nearly 100 nsec, and each processing circuit is required to operate at high speed.

One important step is to suppress the amount of transmitted information through predictive coding.
In other words, the effect of redundancy suppression becomes greater as the characteristics of the prediction circuit are closer to the spectrum of the input signal. However, since input signals constantly change, it is difficult to maximize the redundancy suppression effect for all input signals using one prediction characteristic. Therefore, an adaptive predictive coding method has been proposed in which a prediction circuit having a plurality of prediction characteristics is provided, the prediction characteristic that gives the most accurate prediction is always selected, and redundancy suppression is increased. In this method, the process of finding and switching to the predictor output with the best prediction accuracy is increased in the predictive coding loop process.

[Problem that the invention seeks to solve]

When applying such an adaptive predictive coding method to a television signal, there is a problem in that it requires a high-speed element and is difficult to implement.

An object of the present invention is to provide a plurality of predictive characteristics in a predictive encoding device, and when applying adaptive prediction that can be switched adaptively, to minimize the increase in processing time of the predictive encoding loop, and to improve operation efficiency. The objective is to realize a stable predictive coding device.

[Means for solving problems]

In order to solve the above-mentioned drawbacks, the present invention includes a plurality of difference circuits that take the difference between an input signal and a plurality of prediction circuit outputs, and a switch A that selects one of the difference circuit outputs by a selection signal. , a quantizer that quantizes the output of the switch A, a switch B that selects one of the plurality of prediction circuit outputs using a selection signal, and an adder circuit that adds the output of the quantizer output. , comprising a plurality of prediction circuits that predict input signals from their outputs, and an analyzer that analyzes the outputs of the plurality of prediction circuits using a given algorithm, determines an optimal prediction circuit output, and outputs the selected signal, The processing time of the analyzer and a plurality of difference circuits are parallelized to minimize the increase in the processing time of the analyzer.

〔Example〕

Next, the present invention will be explained in detail with reference to the drawings.

FIG. 1 shows the configuration of a predictive encoding device for color television signals according to an embodiment of the present invention, and FIG. 2 shows the configuration of predictive encoding.

In FIG. 1, a color television signal from a terminal 1 is sampled and quantized at a sampling frequency of 10.7MHz by an A/D converter 2, converted to a PCM signal, predictively encoded by a predictive encoder 3, and then sent to a terminal. 4 to the transmission path. In FIG. 2, the difference between the input color television signal from the terminal 21 and the outputs of the prediction circuits 26 and 27 is taken by the difference circuits 22 and 23, and the difference between the input color television signal from the terminal 21 and the outputs of the prediction circuits 26 and 27 is taken by the difference circuits 22 and 23.
Either of the outputs 2 or 23 is selected, quantized by the quantizer 25, and sent out from the terminal 9 to the transmission line. The switch 30 is the prediction circuit 26, 27
One of the outputs is selected by the selection signal and added to the output of the quantizer 25 in the adder circuit 31. The output (locally decoded signal) of the adder circuit 31 is sent to the prediction circuit 2
6 and 27, each input signal is predicted and input to difference circuits 22 and 23. The analyzer 28 calculates the difference between the input signal and each prediction circuit, and selects the prediction circuit output whose absolute value of the difference is small.

Equations (1) and (2) show the prediction characteristics of the prediction circuits 26 and 27.

H ₁ (z)=0.5・z ^-1 +z ^-3 −0.5・z ^-4 ...(1) H ₂ (z)=z ^-2H ...(2) In the above equation, z ^-1 is one sampling period shows the delay of
z ^-2H indicates a delay of two scan lines.

Equation (1) is a prediction based on sample points on the same scanning line, and Equation (2) is a prediction based on sample points two lines before, and the prediction is accurate for images with high correlation in the vertical direction. Figure 3 shows the addition, prediction, and
The processing times for analysis, switching, difference processing, and quantization are shown. In FIG. 3, (a) shows the processing time of the differential encoding loop in the conventional prediction circuit shown in FIG. 4, and (b) shows the processing time of the differential encoding loop in the prediction circuit according to the present invention. . As is clear from FIG. 3(b), in the configuration of the present invention, analysis processing and difference processing are performed redundantly, so compared to the conventional configuration, the overlapped time (in this embodiment, the difference processing time is approximately
It can be seen that the processing time is reduced by 24nsec).

〔Effect of the invention〕

As described above, according to the present invention, since the processing time of the differential encoding loop can be shortened, a timing margin can be given to the loop processing time, and the processing can be stabilized. Additionally, since conventional low-speed elements can be used, low power can be achieved. Furthermore, the system with a high sampling frequency has advantages such as the possibility of realizing a device using existing elements. Although the present invention has been described with reference to the case where there are two prediction circuits and two difference circuits, the present invention is not limited to these numbers or characteristics. Further, the analysis method is not limited to the method shown in the embodiment, and the prediction signal output from the prediction circuit whose absolute value of the difference between each prediction circuit output and the local decoded signal is small at the time before one sampling. Various methods can be considered, such as selecting the current prediction signal as the current prediction signal, and selecting the one with the smaller absolute value of the difference between the selected prediction circuit output before sampling and the current output of each prediction circuit.
Alternatively, the quantizer may be omitted and the quantizer may be passed through.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of a predictive encoding device for color television signals which is an embodiment of the present invention, FIG. 2 is a block diagram showing the configuration of a predictive encoding circuit, and FIG. 3 is a coding loop. FIG. 4 is a block diagram showing the configuration of a conventional predictive encoding circuit. In FIG. 1 and FIG. 2, 1...input terminal;
2... A/D converter, 3... Predictive encoding circuit, 4
...Output terminal, 21...Input terminal, 22, 23...
... difference circuit, 24 ... switch, 25 ... quantizer,
26, 27... Prediction circuit, 28... Analyzer, 29
. . . output terminal, 30 . . . switch B, 31 . . . adder circuit.

Claims (1)

[Claims] 1. A plurality of difference circuits that take differences between an input signal and a plurality of prediction circuit outputs, a first switch that selects one of the difference circuit outputs by a selection signal, and this first switch. a quantizer that quantizes the output of the prediction circuit; a second switch that selects one of the plurality of prediction circuit outputs according to the selection signal; and an output of the second switch that quantizes the output of the quantizer; an adding circuit for adding, a plurality of prediction circuits for predicting the input signal from the output of the addition circuit, and an analysis for analyzing the outputs of the plurality of prediction circuits to determine an optimal prediction circuit output and outputting it as the selection signal. What is claimed is: 1. A predictive encoding device comprising: a predictive coding device, wherein the processing times of the analyzer and the plurality of difference circuits are made parallel to each other.