JPH07121121B2 - Decryption device - Google Patents

Description

The present invention relates to a decoding device for decoding a coded code in which image information is predictively coded.

[Conventional technology]

Conventionally, when a video signal of, for example, the NTSC system is digitalized and transmitted, it is necessary to reduce the amount of information in order to suppress the expansion of the signal band required for transmission of the digitalized signal.

There is a predictive coding method as one method of the band compression described above, and this predictive coding method will be described below.

FIG. 2 is a block diagram of the predictive coding system.

In FIG. 2, 1 is a video signal input terminal, 2 is a low-pass filter (referred to as "LPF"), 3 is an analog digital converter (hereinafter referred to as "A / D"), 4 is a subtractor, 5 is a quantizer, 6 is an inverse quantizer, 7 is an adder, 9 is a predictor, 10 is a transmission line, 11 is an inverse quantizer, 12 is an adder, 13 is a predictor, 14 is a digital analog A converter (hereinafter referred to as "D / A"), 15 is an LPF, and 16 is a video signal output terminal.

The video signal is supplied to the LPF 2 via the video signal input terminal 1. LPF2 is used to remove unwanted components. The output signal of LPF2 is converted into digital information by A / D3. The subtracter 4 calculates the difference between the output of the A / D converter 3 and the output of the predictor 9, that is, the prediction error. The calculated prediction error is quantized by the quantizer 5 and supplied to the transmission line 10. On the other hand, the output of the quantizer 5 is added to the output of the predictor 9 by the adder 7 via the inverse quantizer 6 and input to the predictor 9.

The signal output from the transmission line 10 is applied to the adder 12 via the inverse quantizer 11 and added to the output of the predictor 13. The output of the adder 12 is input to the D / A converter 14 and the predictor 13, respectively, and is restored to an analog signal by the D / A converter 14. Further, the LPF 15 removes unnecessary components and the video signal is taken out via the video signal output terminal 16.

As is well known, the amount of information of the prediction error signal generated by the subtractor 4 is smaller than the amount of information transmitted from the A / D converter 3, and the A / D converter 3 allocates 8 bits per pixel. Therefore, in order to transmit the above-mentioned prediction error, it is possible to reduce it to, for example, about 4 bits per pixel. That is, the calculated prediction error can be transmitted with 16 (24) representative values. The operation of converting to any one of 16 patterns is executed by the quantizer 5 and converted into 4-bit data. The inverse quantizers 6 and 11 convert the 4-bit data into a predetermined representative value (8
Bit configuration) will be converted back.

As can be seen from the above description, in the predictive coding device,
The prediction error largely depends on the prediction method. Therefore, the degree of band compression depends on the prediction method.

As a prediction method, the previous value prediction is a typical one, and the previous value prediction is the value of the pixel immediately before one pixel as the predicted value of the pixel to be transmitted now.

Due to the nature of the video signal, the correlation between adjacent pixels is very high, and thus such a prior-value prediction method is simple and allows relatively good prediction.

[Problems to be solved by the invention]

However, the band compression method using the pre-prediction may deteriorate the image information transmitted by the quantization noise such as edge business.

The edge business is, as shown in FIG. 3, particularly in the vertical edge portion E of the image, since the quantization error changes between the screens at a portion where the change in the luminance signal level is sharp, the edge business is added to the edge portion. There is a disturbance that looks like a jerk called a twitch. In FIG. 3, the circles represent sampling points with a high luminance signal level, and the black circles represent sampling points with a low luminance signal level.

The above phenomenon is caused by the quantizer 5 shown in FIG. 2 performing non-linear quantization in which the quantization step becomes coarser as the input level becomes larger, as shown in FIG. This is because the difference .epsilon. With .epsilon. Is finely quantized when .epsilon. Is small, and roughly quantized when .epsilon. Is large. When such a quantizer is used, the difference value ε is usually concentrated in the vicinity of 0 level due to the nature of the image, so that a large difference value ε rarely occurs, but in the outline portion of the image, Since the difference value ε becomes large, it is roughly quantized, and if the input level is in the vicinity of FIG. 4A, the representative value B or C is taken for each screen. Such interference is noticeable.

The present invention has been made in view of the above problems, and an object thereof is to provide a decoding device capable of detecting a large change portion of an image at the time of restoration with a simple configuration and making deterioration in the portion inconspicuous. To do.

[Means for solving problems]

For the above-mentioned object, in the present invention, in the device for decoding the coded code obtained by coding the difference value between the image information and the predicted value of the image information, the device transmitted through the transmission line is A large change in the image information is detected by using input means for inputting an encoded code, decoding means for decoding the encoded code input from the input means, and the encoded code input from the input means. And a switching means for switching the filtering characteristic of the image information decoded by the decoding means according to the output of the detecting means.

[Operation] With the above configuration, a large change in image information can be detected with an extremely simple configuration by using the value of the predictive-coded encoded code, and when this change occurs, that portion of the image information can be detected. It is possible to reduce the deterioration of the signal, and it is possible to obtain a reproduced image without deterioration even if there is a large change in the image.

〔Example〕

 Hereinafter, the present invention will be described using examples.

FIG. 1 shows a schematic configuration of a predictive coding system decoder for decoding a signal encoded by the predictive coding system encoder shown in part A of FIG. 2 as an embodiment of the present invention. It is a figure.

In FIG. 1, a 4-bit digital difference image information signal encoded and transmitted by the encoder shown in FIG. 2 is input from an input terminal 17, a representative value setting circuit 18, and one horizontal scanning period (H ) Is supplied to the delay circuit 19.

The representative value setting circuit 18 converts the 4-bit digital difference image signal into an 8-bit digital difference image signal. The representative value setting circuit 18 is equivalent to the inverse quantizer 6 shown in FIG. 2 and its explanation is omitted.

As described above, the 8-bit digital difference image signal output from the representative value setting circuit 18 is added to the prediction value signal output from the predictor 27 in the adder 20 to restore the original 8-bit digital image signal. Is output. Since the predictor 27 is the same as the predictor 9 shown in FIG. 2, its explanation is omitted.

The 8-bit digital image signal output from the adder 20 is branched into four directions, one of which is input to the predictor 27, which is used to generate the next prediction value signal, and the other three directions.
The 2H delay circuit 21 delays for two horizontal scanning periods and a coefficient unit a ₂ 22
Is delayed by one horizontal scanning period by the 1H delay circuit 23
It is supplied to a ₁ 24 and to the coefficient unit a ₀ 25 without delay.

The signals input to the coefficient units (a ₀ , a ₁ , a ₂ ) 22, 24, 25 are multiplied by the coefficients a ₀ , a ₁ , a _{2 respectively} , and added by the adder 26 to the E side of the switch 28. Supplied.

That is, the restored digital image information signal forms a vertical direction filter by the 2H delay circuit 21, the 1H delay circuit 23, and the coefficient units (a ₀ , a ₁ , a ₂ ) 22, 24, 25, and in the vertical direction filter, , The coefficients of the coefficient unit (a ₀ , a ₁ , a ₂ ) 22, 24, 25 a ₀ , a ₁ ,
Image information is averaged in the vertical direction by setting a _{2 so} that high-frequency components in the vertical direction are suppressed. For example, when the coefficients are set to a ₂ = 0.25, a ₁ = 0.5, a ₀ = 0.25, the vertical direction filter has a low-pass characteristic in the vertical direction with respect to the image information signal.

An output signal from the 1H delay circuit 23 is supplied to the N side of the switch 28, and the signal supplied to the N side is delayed by one horizontal scanning period without being processed by the vertical direction filter. An image information signal is provided.

The switching operation of the switch 28 is controlled by the judging device 29. The switching operation will be described below.

The decision unit 29 is supplied with the 4-bit digital difference image information signal delayed by one horizontal scanning period by the 1H delay circuit 19 as described above, and the digital difference image information signal is supplied in accordance with the level of the digital difference image information signal. A switch control signal is output to the switch 28 by determining whether the switch 28 is connected to the E side or the N side.

That is, when the digital difference image information signal is 4 bits, there are 15 levels (−7 to 0 to +7) that can occur, and the image information signal having the difference as the level is farther from “0”. The level difference between the comrades is large.
That is, in the image information, since the image information signal sharply changes at the contour portion of the image and the like, such a large level difference is generated. Therefore, by monitoring the digital difference image signal as described above, the signal is changed. It is possible to determine whether is the contour portion of the image.

Therefore, when the digital difference information signal of a predetermined level (for example, the signal of level -7 or +7) is input, the decision unit 29 detects it and determines that it is the signal of the contour portion of the image, and connects the switch 28 to the E side. The switch control signal for
, And otherwise outputs a switching control signal for connecting the switch 28 to the N side.

Incidentally, it is also possible to determine the contour portion by providing a threshold TH in advance in the determiner 29, and by comparing this threshold TH with the digital difference information signal, and further making the threshold TH variable, The switching operation of the switch 28 can be controlled according to the state of the image information output by performing variable control manually or by evaluating the image signal and automatically performing the evaluation according to the result.

By controlling the switch 28 as described above, the signal processed by the vertical direction filter is output to the output terminal 30 in the case of the contour portion of the image, and in other cases, the processing is performed by the vertical direction filter. By outputting a non-existing signal, the image information is averaged in the vertical direction in the contour portion of the image, and the image information signal can be restored so that interference such as edge business becomes inconspicuous.

By the way, in the present embodiment, when the contour portion of the image is detected, this portion is processed by the vertical direction filter. However, a horizontal filter or a diagonal filter is constructed by using a delay circuit for one sample period, and further, the digital filter is used. By monitoring the level of the differential image information signal for several horizontal scanning periods, it is determined which of the vertical direction, the horizontal direction or the diagonal direction the contour portion is in, and the filter is adaptively switched and processed. By doing so, it is possible to further suppress the image quality deterioration due to the interference of the edge business and the like.

〔The invention's effect〕

As described above, according to the present invention, by detecting a large change in the image from the encoded code, the deterioration of the image information signal in the contour portion of the image at the time of restoration can be made inconspicuous with a very simple structure. It is possible to provide a decoding device capable of doing so.

[Brief description of drawings]

FIG. 1 is a diagram showing a schematic configuration of a predictive coding system decoder as an embodiment of the present invention. FIG. 2 is a block diagram of a conventional predictive coding apparatus and decoding apparatus. FIG. 3 is a diagram showing an edge portion of an image on the screen. FIG. 4 is a diagram showing the quantization characteristic of the non-linear quantizer. 19,23 …… 2 horizontal scanning period delay circuit, 21 …… 1 horizontal scanning period delay circuit, 22,24,25 …… coefficient unit, 26 …… adder, 28 …… switch, 29 …… judder.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Nao Ishikawa 770 Shimonoge, Takatsu-ku, Kawasaki-shi, Kanagawa Canon Inc., Tamagawa Plant (72) Inventor Katsuji Yoshimura 770 Shimonoge, Takatsu-ku, Kawasaki-shi, Kanagawa Canon Inc. Company Tamagawa Plant (56) Reference JP-A-61-199346 (JP, A) JP-A-48-71174 (JP, A)

Claims (1)

[Claims] 1. An apparatus for decoding an encoded code obtained by encoding a difference value between image information and a predicted value of the image information, and inputting means for inputting the encoded code transmitted through a transmission line. And decoding means for decoding the coded code input from the input means, and using the coded code input from the input means,
A decoding device comprising: a detection unit that detects a large change in the image information; and a switching unit that switches a filtering characteristic of the image information decoded by the decoding unit according to an output of the detection unit.