JP3443629B2 - Predictive coding / decoding device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a predictive coding / decoding device for predictively coding a digitized television signal and then transmitting the coded signal. The present invention relates to an encoding / decoding device.

[0002]

2. Description of the Related Art Usually, a prediction error signal of a difference between a digitized television signal at a certain time and a predicted signal of the television signal is calculated and predictively coded, and the predictively coded quantized output signal is coded. As one of the predictive coding / decoding apparatuses for performing the adaptive transmission, the quantization of the prediction error signal is performed by the PCM (P
There is a differential PCM device (DPCM) performed by the pulse code modulation. In this DPCM quantizer, when the image input signal is "n" bits, the difference signal (prediction error signal) between this input signal and the prediction signal is
More than "n + 1" bits.

Therefore, a quantizer which quantizes and outputs a prediction error signal needs to have the same dynamic range as the prediction error signal, and quantizes and outputs a prediction error signal having a dynamic range of at least "n + 1" bits. It is necessary to have a quantization property that can be. At this time, it is difficult to set, as the quantization characteristic, a characteristic that can sufficiently satisfy both the high S / N characteristic and the excellent overload characteristic, with a limited number of levels (quantization bit number). In order to make both the quantization noise and the deterioration due to overload inconspicuous, a large number of quantization bits were required.

Conventionally, in this type of predictive coding / decoding apparatus, as a predictive coding circuit, a method by folding quantization is adopted in order to solve the above-mentioned drawbacks. For details of this technology, see, for example, Bostelmann.
Is disclosed in German Patent No. 2405534.

Generally, a predictive coding circuit inputs an image input signal and a predictive signal, performs a subtraction operation to obtain a predictive error signal, quantizes the obtained predictive error signal, and outputs a quantized output signal. ing. The prediction signal is obtained by calculating the next prediction value according to the prediction characteristic from the locally decoded signal obtained by adding and calculating the quantized output signal and the prediction signal.

[0006] The dynamic range of the prediction error signal is calculated by subjecting the subtraction operation for obtaining the prediction error signal and the addition operation for obtaining the locally decoded signal to the modulo operation ignoring the overflow and underflow bits.
The same "n" bits as the input signal are output. That is,
This method has half the dynamic range of "n"
The overload characteristic is improved by defining the quantization characteristic so that only bits are needed.

In this configuration, when DPCM coding is performed, it is necessary to prevent the locally decoded signal (= input signal + quantization noise) from exceeding the dynamic range of “n” bits due to the influence of quantization noise. . Therefore, a limiter circuit is provided on the input side of the predictive coding circuit, and the limiter circuit limits the upper and lower limits of the input signal by the magnitude of the maximum quantization noise in advance to solve the above problem. There is.

[0008]

In the conventional predictive coding / decoding apparatus described above, a limiter circuit is provided on the input side of the predictive coding circuit, and the upper and lower limits of the input image signal are set to the maximum quantization noise magnitude. It is necessary to limit the amplitude only. As a result, the actual dynamic range of the image signal is reduced, and when the analog signal is converted into a PCM signal with a large signal range, the limiter circuit clips the tip of the horizontal synchronization signal during the horizontal synchronization period. is there.

More specifically, for example, CCIR Recommendation 601
As for the video signal level according to, the black level value is set to "16" and the white level value is set to "235" in the 8-bit level (0 to 255) of 256 steps. Therefore, in the range of "-128 to 127", the value of the black level is "-112" (= -128 +
16), and the white level value is “107” (= −128)
+235).

However, when the maximum quantization noise of the quantizer is, for example, "18", the black level value is "-1".
At 10 ″ (= −128 + 18), it will be clipped by the limiter circuit. When the maximum quantization noise is large and the clamp level of the limiter circuit is set further below, the clipped part is It gets even bigger.

An object of the present invention is to add an offset signal to an input signal and then perform predictive coding via a limiter, so that a signal with a wide dynamic range can be obtained even if the limiter applies a large amplitude limitation due to maximum quantization noise. On the other hand, it is an object of the present invention to provide a predictive coding / decoding device capable of transmitting a high-quality image without the leading edge of the horizontal synchronization signal being clipped and without causing an overload due to quantization.

[0012]

According to the first aspect of the present invention, an analog / digital converter for converting an input analog television signal into a digital television signal at a predetermined sampling timing.
A digital converter, a limiter circuit for limiting the upper and lower limits of the amplitude value of the digital television signal, a predictor for generating a predicted signal of the digital television signal according to a locally decoded signal and a prediction characteristic, and the digital TV after the amplitude limitation. A subtraction calculator that calculates a difference between a signal and the prediction signal to generate a prediction error signal, a quantizer that quantizes the prediction error signal to generate a quantized output signal, and the quantized output signal as a transmission code. In a predictive coding apparatus including a code converter for converting and an adder for adding the quantized output signal and the predicted signal to generate the local decoded signal, the input analog television signal is input, A synchronization detection circuit for detecting a horizontal synchronization signal included in the signal, and the horizontal detection signal in the synchronization section when the horizontal synchronization signal is detected by the synchronization detection circuit. It is possible to obtain a predictive coding apparatus characterized by comprising an offset circuit for an offset value of positive value is added to Ijitaruterebi signal output to the limiter circuit.

According to a second aspect of the present invention, the predictive coding apparatus further detects the amplitude level of the digital television signal and determines to generate an offset control signal when the amplitude value is equal to or larger than a desired threshold value. A circuit, the offset circuit, when detecting the offset control signal,
It is possible to obtain a predictive coding device characterized by adding a negative offset value to the digital television signal and performing level shift.

According to the third aspect of the present invention, the amplitude level of the locally decoded signal is detected instead of the determination circuit,
A second determination circuit that generates an offset control signal when the amplitude value is equal to or larger than a desired threshold value is provided, and the offset circuit, when detecting the offset control signal, has a negative offset value in the digital television signal. It is possible to obtain a predictive coding device that is characterized by performing addition and level shifting.

[0015]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, the present invention will be described with reference to the drawings.

FIG. 1 is a functional block diagram showing a first embodiment of the present invention. In the predictive coding / decoding apparatus shown in FIG. 1, on the transmission side, an offset circuit 2 adds an offset signal to a television (hereinafter TV) signal in a horizontal synchronization section digitized by an A / D converter 1. , The level limiter circuit 4 outputs the signal after level shifting in advance.
Inputs this output and performs a predetermined amplitude limitation, then outputs it to the predictive coding circuit 10, and the predictive coding circuit 10 uses the TV.
The horizontal sync signal portion of the signal is prevented from being clipped. On the receiving side, the inverse offset circuit 14 subtracts the offset signal from the decoded signal to restore the original TV signal.

This will be described for each of the constituent elements with reference to FIG. 1. The predictive coding / decoding apparatus sends the A / D signal to the transmitting side.
Converter 1, offset circuit 2, synchronization detection circuit 3, limiter circuit 4, code converter 9, and predictive coding circuit 10.
And a code inverse converter 11 and an addition calculator 1 on the receiving side.
2, predictor 13, inverse offset circuit 14, and D / A
The converter 15 is provided. The predictive coding circuit 10 includes a subtraction calculator 5, a quantizer 6, an addition calculator 7, and a predictor 8.
Have.

The A / D converter 1 samples the input NTSC color TV signal at a sampling frequency three times as high as the color subcarrier frequency Fsc, and outputs "n = 8" bits of PCM.
The signal is converted into a signal (level of -128 to 127) and output to the offset circuit 2.

The offset circuit 2 receives a signal indicating the horizontal synchronization section from the synchronization detection circuit 3, and adds a positive offset signal to the TV signal only during this horizontal synchronization section. As the offset signal, a constant value (for example, K), or a trapezoidal waveform signal in which the beginning and the end of the horizontal synchronization section gently increase and decrease is predetermined.

The sync detection circuit 3 inputs a TV signal, detects a horizontal sync signal from the TV signal, outputs the horizontal sync signal to the code converter 9, and outputs a signal indicating a horizontal sync section to the offset circuit 2.

The limiter circuit 4 limits the amplitude of the upper and lower parts of the TV signal to which the offset signal is added. This amplitude limitation is only for the value K of the maximum quantization noise generated in the quantizer 6, and sets the level range of the TV signal to "-128 + K to 127-K".

The subtraction calculator 5 inputs the TV signal whose amplitude is limited and the prediction signal received from the predictor 8, is subtracted by an 8-bit modulo operation, and outputs an 8-bit prediction error signal to the quantizer 6. To do.

The quantizer 6 quantizes the 8-bit prediction error signal according to the quantization characteristic, and outputs it to the addition operator 7 and the code converter 9 as a quantized output signal.

As a concrete example of the quantization characteristic, the companding rule of the quantization step is defined as "0, 1, 3, 5, 1 with positive and negative symmetry.
1, 21, 33, 36 ", 4 bits (15 levels)
The quantization characteristic of is used. In this case, the maximum quantization noise magnitude K of the quantization characteristic is "18". Therefore, the level range of the TV signal output from the limiter circuit 4 is "-110 to 109".

The addition calculator 7 adds the quantized output signal and the prediction signal output from the predictor 8 to be described later by an 8-bit modulo operation to output a locally decoded signal, which is supplied to the predictor 8. The predictor 8 calculates the next predicted value according to the prediction characteristic and outputs an 8-bit prediction signal to the subtraction calculator 5 and the addition calculator 7.

When sampling is performed at a frequency three times the color subcarrier frequency Fsc, a color television signal is directly and efficiently predicted, and the Z function is shown in the following prediction function "P (Z ^-1 )". There is an example.

The P (Z ⁻¹ ) = 0.5Z ⁻¹ + Z ⁻³ −0.5Z ⁻⁴ code converter 9 converts the quantized output signal received from the quantizer 6 into a transmission code representing each quantization level. It is converted, multiplexed with the signal representing the horizontal sync signal received from the sync detection circuit 3, and output to the receiving side. Even if the horizontal sync pulse signal section to which the offset is applied varies, the code converter 9 accurately limits the horizontal sync pulse signal section to the horizontal sync pulse signal section to process the horizontal sync pulse signal section. The transmission code to be transmitted is encoded by run-length encoding or the like and transmitted so that the receiving side can determine.

On the receiving side, the code inverse converter 11 inversely converts the received transmission code to reproduce the quantized output signal and supplies it to the addition calculator 12, and separates the information of the synchronizing signal to form an inverse offset circuit. It is sent to 14. When a coded horizontal synchronization pulse signal section is sent in addition to the information of the sync signal, the code inverse converter 11 decodes the received coded signal and applies a reverse offset as a pulse signal section. It is sent to the offset circuit 14.

The addition calculator 12 adds the quantized output signal received from the code inverse converter 11 and the prediction signal received from the predictor 13 by an 8-bit modulo operation and outputs it as a decoded signal, which is the inverse of the predictor 13. It is supplied to the offset circuit 14. The predictor 13 has the same prediction characteristic as that of the predictor 8 on the transmitting side, calculates the next predicted value from the decoded signal according to this prediction characteristic, and outputs an 8-bit predicted signal.

The inverse offset circuit 14 includes a code inverse converter 1
In the horizontal synchronization section obtained from the horizontal synchronization section information received from 1, the same offset signal as that added on the transmission side is subtracted from the decoded signal received from the addition calculator 12 and the decoded TV signal is output to the D / A converter 15. To do. The D / A converter 15 is
The output signal of the reverse offset circuit 14 is converted into an analog signal and output.

Next, a second embodiment will be described with reference to FIG.

In the second embodiment, a block determination circuit 21 is added on the transmission side shown in FIG. 1, and along with this addition, an offset circuit 22, a code converter 23, a code inverse converter 24, and an inverse offset are provided. The function is added in the circuit 25.

With this additional function, the TV signal can be prevented from being clipped by the limiter circuit even when the TV signal has the maximum amplitude in the image signal section in addition to the horizontal synchronization section.

That is, the block determination circuit 21 divides the received TV signal into predetermined blocks, and the TV signal in each block is divided into a predetermined limit level (127-
The presence / absence of a pixel having a value exceeding K) is determined, and if there is a pixel having a value exceeding this, this information is sent to the offset circuit 22. The offset circuit 22 is the TV in the block.
A negative offset signal (for example, "-K") is added to the signal to lower the level of the TV signal to a level at which a value exceeding the limit level is not clipped and is sent to the limiter circuit 4.

The information of the area to which the offset signal is added is encoded and transmitted by a method such as run length coding so that the offset signal added on the transmitting side can be subtracted from the decoded signal on the receiving side. If the size of the divided blocks is too small, the number of blocks increases. As a result, when the runs of the blocks are not continuous, the number of bits forming the block information for each block does not decrease significantly, so that a large number of bits are required to transmit the block information. For this reason, the blocks to be divided need to have an appropriate size.

The A / D converter 1 digitizes the input TV signal and supplies it to the block determination circuit 21 and the offset circuit 22. The synchronization detection circuit 3 detects horizontal synchronization and sends a signal indicating a synchronization section to the offset circuit 22 and the code converter 23.

The block determination circuit 21 is composed of the A / D converter 1
From the block, the offset circuit 22 detects the region where the amplitude of the TV signal exceeds the value of "127-K" and is close to the upper limit for each block in the image signal section. To the code converter 23.

The offset circuit 22 adds each offset signal to the synchronization section received from the synchronization detection circuit 3 and the block area exceeding the limit level received from the block determination circuit 21, and outputs it to the limiter circuit 4.

The limiter circuit 4, the subtraction calculator 5, the quantizer 6, the addition calculator 7, and the predictor 8 have the same functions as those in the first embodiment. That is, in the limiter circuit 4, the TV signal is amplitude-limited by the magnitude of “K” between the upper part and the lower part, and the subtraction calculator 5 obtains the prediction error signal. The prediction error signal is added as a quantized output signal by the quantizer 6 to the addition calculator 7 and the code converter 23.
Is output to. The predictor 8 outputs a prediction signal according to the prediction characteristic.

The code converter 23 multiplexes the quantized output signal, the signal obtained by coding the block information, and the signal indicating horizontal synchronization, and outputs the multiplexed signal to the receiving side.

On the other hand, on the receiving side, the code inverse converter 24
The received transmission code is inversely converted, the quantized output signal is reproduced and supplied to the addition arithmetic unit 12, the information of the synchronization signal is separated, and the block information exceeding the limit level is decoded to the inverse offset circuit 25. Send out.

The addition calculator 12 adds the quantized output signal and the prediction signal and outputs the decoded signal to the inverse offset circuit 25.

The inverse offset circuit 25 responds to the horizontal synchronization section and the decoded signal of the block exceeding the limit level,
Each offset signal added on the transmission side is subtracted and a decoded TV signal is output from the quantized output signal to the D / A converter 15. The D / A converter 15 converts the received decoded TV signal into an analog signal and outputs it.

Next, a third embodiment will be described with reference to FIG.

In the third embodiment, a decision circuit 31 is added on the transmitting side and a decision circuit 34 is added on the receiving side shown in FIG. 1, and the offset circuit 32 and the reverse offset circuit 33 of the related circuits function. Has been added.

With this additional function, instead of controlling the offset signal using the information of the block exceeding the predetermined limit level in the second embodiment, the locally decoded signal generated by the predictive coding circuit is replaced with the local decoded signal. It is used to control the offset signal for the next pixel.

The A / D converter 1 digitizes the input TV signal and supplies it to the offset circuit 32. The synchronization detection circuit 3 detects horizontal synchronization and sends a signal indicating a synchronization section to the offset circuit 32 and the code converter 9.

The decision circuit 31 generates an offset signal based on the locally decoded signal output from the addition calculator 7 of the predictive coding circuit 10 and supplies it to the offset circuit 32.

Explaining a specific example of this offset signal generation, the level of the locally decoded signal in the image section is "1".
When it becomes 27-2K "or more or" -128 + 2K "or less, a value obtained by accumulating constant values of" -K / 4 "or" K / 4 "for each pixel is calculated, and the offset signal of the next pixel is obtained. The maximum value is limited by "-K" or "K." The level of the locally decoded signal is "-K".
In the range of 128 + 4K to 127-4K ", a constant value is subtracted from this cumulative value for each pixel until the cumulative value becomes" 0 ".

The offset circuit 32 sets the offset signal to T based on the signal from the horizontal synchronizing section and the determination circuit 31.
Add to V signal. The limiter circuit 4 limits the amplitude of the upper part and the lower part of the TV signal by the magnitude of "K".

The remaining predictive coding circuit 10 has the same function as that of the first embodiment. That is, the subtraction calculator 5 outputs the prediction error signal, the quantizer 6 outputs the quantized output signal to the addition calculator 7 and the code converter 9, and the code converter 9 sends the transmission code to the receiving side. On the other hand, the predictor 8
A prediction signal is output according to the prediction characteristic.

The code converter 9 converts the quantized output signal into a transmission code representing each quantization level, multiplexes it with the signal representing the horizontal synchronizing signal, and outputs the multiplexed code to the receiving side.

On the receiving side, the code inverse converter 11 inversely converts the encoded quantized output, reproduces the quantized output signal and supplies the quantized output signal to the addition calculator 12, and separates the information of the synchronization signal. And supplies it to the reverse offset circuit 33. The addition calculator 12 adds the quantized output signal and the prediction signal from the predictor 8 and outputs a decoded signal, which is output to the inverse offset circuit 33.

The reverse offset circuit 33 eliminates the offset signal by subtracting each offset signal added on the transmitting side based on the horizontal synchronization section and the signal from the determination circuit 34, and outputs the offset signal. The judging circuit 34 has the same function as the judging circuit 31 on the transmitting side,
An offset signal is generated based on the decoded signal from.

The D / A converter 15 converts the decoded signal from which the offset signal has been removed into an analog signal and outputs it.

In the embodiment, the sync detection circuit 3 is configured to input an analog TV signal, but a digital TV signal output from the A / D converter 1 can be used as an input.

Although the functional blocks are illustrated in the above description, they are merely examples, and the functions can be separated and merged freely depending on the distribution of the functions. Therefore, the above description does not limit the present invention.

[0058]

As described above, according to the present invention, an offset means for adding an offset signal to an input television signal and outputting the input signal, and an output of the offset means are applied to the transmitting side to limit the amplitude. Limiter means for calculating the difference between the prediction signal and the prediction error signal and outputting the result as a television signal. The receiver side subtracts the offset signal from the decoded signal and outputs the decoded television signal. Offset means are provided.

With this configuration, even if the limiter performs a large amplitude limitation due to the maximum quantization noise, the tip of the horizontal synchronizing signal is not clipped with respect to a signal having a wide dynamic range, and overload due to quantization occurs. It is possible to obtain a predictive coding / decoding device capable of transmitting a high-quality image without using the predictive coding / decoding device.

[Brief description of drawings]

FIG. 1 is a functional block diagram showing a first embodiment of the present invention.

FIG. 2 is a functional block diagram showing a second embodiment of the present invention.

FIG. 3 is a functional block diagram showing a third embodiment of the present invention.

[Explanation of symbols]

1 A / D converter 2,22,32 offset circuit 3 Sync detection circuit 4 limiter circuit 5 Subtraction calculator 6 quantizer 7,12 Addition calculator 8,13 Predictor 9,23 Code converter 10 predictive coding circuit 11, 24 code inverse converter 14, 25, 33 Reverse offset circuit 15 D / A converter 21 Block judgment circuit 31, 34 Judgment circuit

Continuation of front page (56) Reference JP-A-4-79690 (JP, A) JP-A-2-216917 (JP, A) JP-A-63-318818 (JP, A) JP-A-63-209236 (JP , A) JP 63-172527 (JP, A) JP 2-135886 (JP, A) JP 2-214380 (JP, A) JP 6-303459 (JP, A) JP 5-83692 (JP, A) JP-A-5-64102 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) H04N ⁷ /24-7/68 H04N 1/41-1 / 419

Claims (3)

(57) [Claims] 1. An input analog television signal is sampled into a predetermined sample.
A conversion to digital TV signal at ring timing
Analog / digital converter and the digital television signal
Limiter circuit for limiting the upper and lower limits of the amplitude value of the signal
According to the locally decoded signal and the prediction characteristics.
A predictor for generating a prediction signal of a bi-signal, and after the amplitude limitation
Calculates the difference between the digital television signal and the predicted signal
Subtraction calculator that generates a prediction error signal and the prediction error
A quantizer that quantizes the signal to produce a quantized output signal;
Code converter for converting this quantized output signal into a transmission code
And adding the quantized output signal and the prediction signal
Predictive coding with adder for generating locally decoded signal
In the device, the input analog television signal is input and included in the signal.
A synchronous detection circuit for detecting the Murrell horizontal synchronization signal, when the horizontal synchronizing signal detected by the synchronization detecting circuit
A positive value is added to the digital television signal during the synchronization period.
The offset value is added and output to the limiter circuit.
A predictive coding device having a fusing circuit
Place 2. The predictive coding apparatus according to claim 1, wherein:
Detects the amplitude level of the digital TV signal,
Generate an offset control signal when it is above a desired threshold
2. The predictive code according to claim 1 , further comprising a determination circuit , wherein when the offset control signal is detected, the offset circuit adds a negative offset value to the digital television signal to perform level shift. Device. 3. The predictive coding apparatus according to claim 1, wherein:
The amplitude level of the locally decoded signal is detected and the amplitude value is
Judgment time for generating an offset control signal when the threshold value is exceeded
A path, the offset circuit detects the offset control signal.
The digital TV signal, the negative offset
Contracts characterized by adding a set value and performing a level shift.
The predictive coding apparatus according to claim 1.