JPH02253782A - Encoder - Google Patents

Abstract

PURPOSE:To obtain the encoder of high picture quality with simple constitution by arranging sampling points in the shape of a lattice in a field when field off-set sub sample is executed and using the sampling value of a picture element adjacent to a picture element corresponding to the sampling value in a vertical direction as the forecasting value of the input sampling value. CONSTITUTION:A sub sampler 3 arranges the picture elements to be transmitted in the shape of the lattice concerning the respective fields, inputs a local decoding value Vi, which is outputted from a limiter 22, to a 1H delay line 28, executes delay for one horizontal scanning and outputs the value to a subtracter 12 as a forecasting value Pi by the field off-set sub sampling. This forecasting value Pi goes to be the sampling point in the upper direction of one horizontal scanning line to be adjacent in the vertical direction on the picture. Thus, since a forecasting distance is obtained to shorten the force of this execution example rather than a conventional example, forecasting accuracy is improved, distortion caused by quantization is suppressed and the high picture quality can be obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an encoding device that encodes a difference value between a sample value and its predicted value.

[Prior Art] When image information, audio information, etc. is digitized and transmitted, encoding is performed to reduce the amount of information to be transmitted.

One of the encoding methods is differential encoding (Differential encoding).
There is a Pulse Code Modulation (hereinafter referred to as DPCM), which compresses the amount of information by utilizing the correlation between adjacent sample values. in particular,
The coded sample value is once decoded, the decoded value is used to find the predicted value for the next sample value, and the difference between this predicted value and the input sample value is quantized and encoded. .

FIG. 6 is a block diagram showing an example of the configuration of an encoding device based on the above principle. Also, Figure 7 shows the sub-sampler 3 in Figure 6.
An explanatory diagram showing a sampling pattern according to FIG.

In FIG. 6, 1 is an input terminal to which input data Xi (for example, a digital television signal), which is a video signal, is applied. 2 is a prefilter that attenuates the two-dimensional spatial frequency domain in input data Xi; 3 is a prefilter 2
4 is a DPCM encoder that performs differential encoding on the output signal of the sub-sampler 3. 5 is a transmission path, 6 is a DPCM decoder that decodes the encoded signal from the 3P CM encoder 4 and performs band expansion processing, and 7 is a transmission path for the signal expanded by the DPCM decoder 4. An interpolation filter performs interpolation, and 8 is an output terminal that outputs output data.

In the above configuration, a digital television signal of, for example, 8 bits) is applied to the input terminal -fx, and the pre-filter 2 attenuates a predetermined one-dimensional spatial frequency region by 6°C. Offset subsampling is performed by Jibzambla 3. This offset . . . subsampling is processed as in the pattern of FIG.

In FIG. 7, the actual picture represents the scanning line of the first field, and the broken line represents the scanning heat of the second field. It is assumed here that the input signal is a 2:1 interlaced television signal. Then, the Derehi's signal of 2 is sampled in a grid pattern at the points indicated by O and is the sampling frequency). In this way, by performing the well-known field 1-off trt subsampling in a 1,000-aI pattern and interpolating the sample points that were not transmitted by the subsequent interpolation filter 7 on the receiving side, horizontal and vertical It is possible to transmit the same resolution without any loss of resolution. However, the resolution in the diagonal direction (deteriorates to 5), but human visual characteristics are such that the resolution in the diagonal direction is smaller than that in the horizontal and vertical directions, so the deterioration will not be as large as that in the pre-filter 2. The characteristics are designed to change the spatial frequency in the diagonal direction, and prevent the aliasing distortion caused by the aliasing described in F.

The signal subsampled by subsample 3 is D
The PCM encoder 4 compresses, for example, a signal of 8 bits per sample point (pixel) into 4 bits, and sends it to the transmission path 5. The 4-bit encoded signal from the transmission line 5 is decoded by the DPCMPCM decoder and converted to the original 8-bit I
- Stretched into a signal. Then, the J-sample points indicated by the X marks in FIG.
The interpolated output data 9 is output from the output terminal 8.

Next, details of the configuration of the DPCM encoder 4 shown in FIG. 6 will be explained. FIG. 8 is a block diagram showing an example of the configuration of a general PCM encoder 4. As shown in FIG. The encoding here is similar to No. 1 above, where the encoded sample values are first decoded (
A method is used in which the decoded value is used to obtain a predicted value for the next tree value, and the error between this predicted value and the actual value is quantized and encoded.

10 is an input terminal into which the sample value x1 is input; 12 is a subtractor that performs M calculation between the sample value [5] and the predicted value P (predetermined decoded value); and 14 is the difference value output from the subtracter 12. Quantized D
This is a quantizer which outputs a PCM coat YA to an output terminal 1G. 18 is an inverse quantizer that decodes the DPCM boolean into a difference value; 20 is an adder that adds the predicted value PI and the output signal of the inverse quantizer 18 to restore the difference m to a sample value; 22 is a limiter that limits the amplitude of the output signal @ of the adder 20 to a predetermined range and obtains the local decoded value ■.
24 is the local decoded value 8 output from the limiter 22 as the next sample value x, and the predicted value P of 41. This is a D-type flip-flop 24 that applies the signal to the subtracter 12 as 1.

Next, the operation of the DPCM encoder 4 having the configuration shown in FIG. 8 will be explained.

Input to input terminal 10 in7'! ': The sample value X is subtracted from the predicted value P by the subtracter 12. The difference value resulting from this subtraction is quantized by the quantizer 14, and the DPCM
:1lff-do Y, and then output to the output terminal 16. Further, the output of the quantizer 14 is applied to the inverse quantizer 18, and 1
, here, after the DPCM 7' Y i is decoded into a difference value and output to the adder 20, the adder 20 adds the predicted value PI output from the D-type flip-flop 24 to the difference value. As a result, the difference value (quantized representative value) is restored to the sample value (locally decoded value). , the restored sample value is
After its amplitude is limited to a predetermined range by the limiter z2, it is applied to the Dfi flip-flop 24. , D-type flip-flop z4 outputs the locally decoded value output from the limiter 22 in synchronization with the clock.

It is applied to the subtracter 12 as the predicted value PI+1 for the next sampled value Xil.

Generally, the distribution of difference values of predicted values is biased toward small values, and compressed transmission of information becomes possible by encoding and transmitting the difference values using nonlinear quantization.

[Problems to be Solved by the Invention] In the conventional encoding device as described above, the previous value, that is, the value two sampling points before, is used as the predicted value, so when subsampling is performed, the difference between the sampling points As the spatial distance increases, the prediction accuracy decreases. For this reason, distortion due to quantization appears in the image, resulting in edge business (deterioration of image quality at the edges) and granular noise (roughness of the reproduced image in flat areas).
This has made it difficult to transmit high-quality images.

The present invention was made to solve these problems, and provides an encoding device that improves prediction accuracy while performing subsample processing on video signals, and prevents distortion caused by quantization from appearing in images. The purpose is to

[Means for Solving the Problems] In order to achieve the above object, the present invention provides a sampler that performs field offset subsampling of a bidet signal so that the sample pixels of each field are arranged in a grid pattern, sampled values by the subsampler, and the like. An encoder that encodes a difference value from a predicted value, and a prediction that generates the predicted value using only sample values of pixels adjacent in the vertical direction of the screen to the pixel to be encoded by the encoder. The structure is equipped with a container.

[Operation] According to the encoding device configured as described above, the sample pixels of each field are arranged in a grid pattern, so that the two pixels between adjacent horizontal scanning lines are
The two transmission pixels will be placed very close together.

Since only sample values adjacent to each other in the horizontal direction of the screen are used to generate the predicted value, the distance on the screen between the pixel to be encoded and the pixel used to generate the predicted value can be kept extremely short. It is possible to reduce distortion due to accurate L deviation and quantization. Furthermore, since the predicted value is generated using only vertically adjacent pixels, the configuration is not complicated.

[Embodiment] FIG. 1 is a block diagram showing the configuration of the main part of an embodiment of the encoding device of the present invention. Note that the device of this embodiment is the first one.
The encoder shown in the figure is incorporated as the DPCM encoder 4 in FIG. In addition, in FIG. 1, the same reference numerals are used for the same parts as in FIG. 7, so the redundant explanation will be omitted below.

In this embodiment, by field offset subsampling by the subsampler 3, the pixels to be transmitted for each field are arranged in a lattice pattern.On the other hand, in the encoder of FIG. 1, the D-type flip-flop 24 of FIG. 1) The structure is characterized by using one (one horizontal scanning period) delay line 28.

Next, the operation of the configuration shown in FIG. 1 will be explained. here,
The process of generating the DPCM code yi and decoding using it is the same as in the case of FIG. 7, so the explanation will be omitted.

The local decoded value V, output from the limiter 22 is input to 1) 1 delay line 28; IH delay line 28 receives locally decoded value V
, is delayed by l horizontal scans, and this is calculated as the predicted value P,
It is output to the subtracter 12 as . This predicted value PL is not a previous value, but is a sample point above l horizontal scanning lines adjacent to each other in the vertical direction on the screen. FIG. 142 is a diagram for explaining the generation of this predicted value.

In the conventional encoding device shown in FIGS. 6 to 8,
The distance on the image between the sample value . The sampling frequency f is set to be slightly more than twice the transmission bandwidth of the television signal. , the sampling frequency f is approximately 44M
Usually, the transmission band of a television signal is determined by multiplying the vertical conversion by a coefficient called the Kel factor. In the case of high-definition, the Kel factor is approximately 0.56
When the sampling frequency f1 is twice the transmission band, the distance between sampling points in the horizontal direction, 1/f, is the distance between scanning lines, 1/f.
0, which is 110.56=1.78 times farther away from ν,
In the past, the sampling frequency f was set at 2.2 of the transmission band, considering the feasibility of a pre-filter when A/D converting a television signal.
Take at least twice as much. Taking this into account, if the sampling frequency f for a high-definition signal is set to -41 MHz, l/f will be about 1.45 times the distance of 1/v. This ratio corresponds to the ratio of the predicted distance between the conventional example and the present example, and since the predicted distance is shorter in the present example, the prediction accuracy is increased, distortion due to quantization is suppressed,
High image quality becomes possible.

Next, FIG. 3 is a block diagram showing a configuration example of a DPCM decoder corresponding to the encoder of FIG. 1, and is used in addition to the DPCMPCM decoder shown in FIG.

In FIG. 3, 3.0 is an input terminal into which the DPCM code Yi is input, 18 is an inverse quantizer that decodes the DPCM code yi into a difference value, and 20 is a predicted value P for the decoded difference value.
An adder that adds i to restore an image, 22 is an adder 20
a limiter that limits the output amplitude of the output signal to a predetermined range and outputs it as a decoded image signal Vi;
8 is an output terminal for outputting the decoded image signal V, and 40 is an IH delay line for obtaining the predicted value P.

In the DPCM decoder with this configuration, the 4-bit DPCM code Yi input from the input terminal 30 is input to the inverse quantizer 1.
8 to decode the difference value (quantized representative value). about,
The adder 20 adds this difference value and the predicted value Pi to restore the original image signal.The output amplitude of the adder 20 is limited to a predetermined range by the limiter 22, and the decoded image signal V , and is applied to the iH delay line 40. The 18 delay line 40 delays the decoded image signal Vl by IH to generate a predicted value P1.

FIG. 4 is a block diagram showing another example of the encoder according to the present invention. In this example, the same reference numerals are used for the same parts as those shown in FIG. 1, so redundant explanations will be omitted.
Whereas each limiter process was performed separately, the ROM contains a table that describes the relationship between the sample value X and predicted value P, and the DPCM code Yi and local decoded value V.
42 is used. A locally decoded value Vi output from the ROM 42 is applied to the IH delay line 28 . in this case,
The range that (x, -pm) can take is limited by the value of the local decoded value V. By utilizing this and changing the quantization characteristics depending on the predicted value Pi, it is possible to configure a sliding quantizer that reduces the quantization step width. Therefore, it is possible to achieve even higher image quality than the configuration using the encoder shown in FIG.

Note that the principle operation of the embodiment using the encoder shown in FIG. 4 is the same as that of the previous embodiment, so a description thereof will be omitted. Furthermore, since the encoder of this example can be configured without using individual circuits, the configuration can be simplified and the number of parts can be reduced, and the encoder can be made smaller and lighter.

FIG. 5 is a block diagram showing another example of the DPCM decoder. The difference between this example and the configuration shown in Figure 3 is that the inverse quantization,
Whereas addition and limiter processing were performed separately,
Local decoded value V for DPCM code Yi and predicted value P1
, using the ROM 44 containing a table that describes the relationships between . The IH delay line 40 is applied with a locally decoded value ① output from the ROM 44 . In this case too,
Similar to the encoder shown in FIG. 4, a sliding quantizer corresponding to this encoder can be constructed. According to this configuration, since it can be configured without using individual circuits, it is possible to simplify the configuration and reduce the number of parts, and when combined with the encoder shown in FIG. 4 described above, even higher image quality can be achieved. It becomes possible.

[Effects of the Invention] As is clear from the above, according to the present invention, when applying field offset subsampling, the sample points are arranged in a grid pattern in each field and correspond to the sample value as a predicted value of the input sample value. Since the configuration uses sample values of pixels adjacent to each pixel in the vertical direction, a high-quality encoding device can be obtained with a simple configuration.

[Brief explanation of drawings]

FIG. 1 is a block diagram without showing the main configuration of an embodiment of the encoding device of the present invention, 2IN is a diagram for explaining the generation of predicted values in this embodiment, and FIG. 3 is a code for the 11th leap. FIG. 4 is a block diagram showing another example of the structure of the encoder, and FIG. 5 is a block diagram showing another example of the DPCM decoder. FIG. 6 is a block diagram showing a general encoding device, and FIG.
The figure is an explanatory diagram showing the sampling pattern by the sub-sampler in Figure 6, and Figure 8 is! 86 is a block diagram showing a conventional configuration example of the DPCM encoder in FIG. In the figure. 12: Subtractor 14/Quantizer Fig. 1 ×---~→D- --C)---
−−×−−−()−Fig. 2 18・Inverse ya generator 20 Adder 22: Limiter 28.40:1) 1 delay line 42・ROM P,:′T−measured value ■,:decoded value (local recovery value) x,: sample value

Claims (1)

[Claims] a subsampler that performs field offset subsampling of a video signal so that the sample pixels of each field form a grid; an encoder that encodes a difference value between a sampled value by the subsampler and its predicted value; An encoding device comprising: a predictor that generates the predicted value using only sample values of pixels horizontally adjacent to a pixel to be encoded.