JP2611555B2 - Video encoding 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 moving picture coding apparatus for separating a moving picture into a DC component and a residual component and coding the moving picture with high efficiency.

[0002]

2. Description of the Related Art For example, a document (Ibaraki: A study of a video coding method for storage systems, 1989 Shingaku Spring University, D-111)
Of the conventional storage system moving image encoding apparatus (CD-ROM) shown in FIG.
As shown in FIG. 6, the frame memory 1 temporarily stores a plurality of frames of an image frame sequence 11 obtained by AD-converting a moving image signal from an information source. The subtractor 9 calculates the predictive encoding means 10 from the input signal 12 to be encoded read out from the frame memory 1 in a predetermined encoding order.
Is subtracted from the prediction signal 25. Predictive encoding means 10
Encodes the prediction error signal 23 from the subtractor 9 for each pixel block, and generates output encoded data 24. On the other hand, the output coded data 24 is decoded, and a predetermined coding mode (unidirectional interframe prediction / bidirectional interframe prediction /
The prediction signal 25 is generated according to (in the frame).

[0003] The above-mentioned conventional storage system moving picture coding apparatus comprises:
A method of collectively encoding the signal components of a moving image (signal component collective method) is employed.

As shown in FIG. 7, a predictive encoding means 10 first performs a discrete cosine transform (DCT) and a scalar quantization on a prediction error signal 23 from a subtractor 9 for each pixel block by an encoder, and outputs encoded data. 24 is generated. Next, the output coded data 24 from the encoder is input to the decoder, and the generated decoded prediction error signal 26 is input to the adder. The prediction signal 2 from the encoding mode switching predictor is added to the decoded prediction error signal 26.
5, and the generated decoded signal 27 is temporarily stored in the frame memory. Further, according to the unidirectional inter-frame prediction, bidirectional inter-frame prediction, and each encoding mode in the frame, which are set in advance by the encoding mode switching predictor as shown in FIG. Using one or both decoded signals 27, the prediction signal 25
Generate In the unidirectional inter-frame prediction encoding mode, a decoded signal 27 of the preceding image frame is generated as a prediction signal 25 in order to perform forward prediction encoding. In the bidirectional inter-frame prediction encoding mode, in order to encode one or both of forward prediction and backward prediction, a decoded signal 27 of any of a preceding image frame, a succeeding image frame, and a preceding and succeeding arithmetic average image frame is generated. Generated as the prediction signal 25. In the intra-frame coding mode, the prediction signal 25 is set to zero in order to perform direct coding without performing predictive coding.

[0005]

The above-described conventional moving picture coding apparatus employs a method of collectively coding signal components of a moving picture (signal component collective method), so that the brightness of the entire screen is reduced. If there is a large change over time, a large change (prediction error) is encoded in most of the pixel blocks of the moving image, and there is a problem that the encoding efficiency is reduced.

An object of the present invention is to separate a DC component and a residual component of a moving image by a moving image encoding apparatus.
It is an object of the present invention to provide a system (signal component separation system) capable of encoding each image with high efficiency even when the brightness of the entire screen changes greatly with time.

[0007]

According to the present invention, there is provided a moving image encoding apparatus comprising: a plurality of frames of an image frame sequence of a moving image signal from an information source;
Frame memory for temporarily storing frames, and this frame
Image from the input signal to be encoded read from the memory
Calculates and separates the DC component of a predetermined area in the frame,
Signal component generator that generates each input signal of component and residual component
Stage and the residual component generated by the signal component separating means.
Output encoded data of the residual component based on the input signal of
And a signal component separating means,
Subtract the DC component prediction signal from the generated DC component
Subtractor that outputs a prediction error signal, and a signal from the subtractor.
The prediction error signal of the DC component is encoded and the output code of the DC component is encoded.
Encoded data and output encoded data
To decode a predetermined region of the preceding or succeeding image frame
Generated as a decoded signal of the DC component of
DC component predictive encoding means for outputting as a measurement signal,
For a separate image frame generated by the signal component separation means
Compare the DC component input signals in the
Scene change determining means for determining the presence or absence of
The DC component predictive encoding means is adapted to perform the scene change determination.
If it is determined that there is a scene change by the
Is the image frame not including the changed image frame?
Generates a prediction signal for the current image frame from
Output as a prediction signal of the flow component.
You.

[0008]

[0009]

[0010]

[0011]

[0012]

[0013]

[0014]

The moving picture coding apparatus according to the present invention converts a DC component input signal calculated / separated into an image frame unit or a predetermined area unit within an image frame from an input signal to be coded from an information source by the above means. Is adaptively predictively coded based on the determination result of. On the other hand, the residual component input signal is predictively coded or directly coded. At the time of each predictive encoding of the DC component and the residual component, each predictive signal is adaptively generated based on a difference evaluation value between each input signal of the DC component and the residual component and each predictive signal.

[0015]

FIG. 1 is a block diagram of a moving picture coding apparatus according to an embodiment of the present invention, and FIG. The signal component separation means 2 separates the DC component input signal 13 by arithmetically averaging the input signal 12 to be encoded read from the frame memory 1 for each image frame by a DC component calculator. On the other hand, a DC component input signal 13 is subtracted from the input signal 12 by a subtracter to separate a residual component input signal 14. The DC component and residual component subtractors 3 and 4 respectively convert the DC component and residual component input signals 13 and 14 from the DC component and residual component predictive coding means 5 and 6 from the DC component and residual component. The respective prediction signals 19 and 20 of the difference component are respectively subtracted. The DC component and residual component predictive encoding means 5 and 6 convert the DC component and residual component prediction error signals 15 and 16 from the DC component and residual component subtracters 3 and 4 into DC components. For each image frame, the residual component is encoded for each pixel block, and the output encoded data 17 and 1 of the DC component and the residual component are respectively encoded.
8 is generated. On the other hand, the output coded data 17 and 18 are respectively decoded, and the prediction signals 19 and 20 of the DC component and the residual component are converted according to a predetermined coding mode (unidirectional inter-frame prediction / bidirectional inter-frame prediction / in-frame). Generate.

The subtracter 4 and the residual component predictive coding means 6
A difference component encoding unit is configured. Moving picture coding apparatus of the above embodiment, each separate a DC component and residual component of the moving image you coding.

Predictive encoding means 5 for DC component and residual component
And 6 generate output coded data 17 and 18 of DC component and residual component and prediction signals 19 and 20, respectively, as in FIG. However, the DC component predictive encoding means 5 outputs the prediction error signal 1 from the DC component subtractor 3.
5 is scalar-quantized and encoded by an encoder. In the intra-frame encoding mode, a predetermined fixed value is generated as the DC component prediction signal 19 in order to perform the fixed value prediction encoding.

In the above embodiment, the signal component separating means 2 may calculate and separate the DC component from the prediction error signal of the input signal 12 to be encoded as shown in FIG.

In the above embodiment, the DC and residual component predictive encoding means 5 and 6 have been described as setting each encoding mode in advance. However, each of the DC and residual component input signals and Each encoding mode may be selected such that the difference evaluation value (the sum of the absolute difference or the sum of the squared differences) of the image signal and the prediction signal is minimized.

[0020] As in the above embodiments the DC component prediction coding means 5 in Example 3, the scene change determination unit 7 is provided in order to not decrease the coding performance at scene change, the scene change from the scene change determination unit 7 Using the detection signal 21, an encoding mode in which prediction is performed only from an image having no scene change is selected . The scene change determination means 7 temporarily stores the DC component input signal 13 from the signal component separation means 2 in a frame memory. On the other hand, the scene change detector compares each DC component input signal 13 between the current image frame to be encoded from the signal component separation means 2 and the preceding or succeeding encoded image frame from the frame memory, and calculates the absolute value of the difference. The presence or absence of a scene change is determined based on the magnitude of D and the predetermined threshold value TH. The DC component predictive encoding means 5 receives the scene change detection signal 2 from the scene change determining means.
1 When there is a scene change of (D> TH) is, scene
Each encoding mode is selected so as not to use the DC component decoded signal 27 of the preceding or succeeding image frame having a change, and the DC component input signal 13 of the current image frame is predictively encoded.

In the above embodiment, the residual component encoding means is
As shown in FIG. 4, a residual component direct encoding unit 8 is provided instead of the residual component subtractor 4 and the predictive encoding unit 6, and the residual component input signal 14 is directly encoded, The output encoded data 22 may be generated. Since the power of the residual component obtained by separating the DC component of the moving image is generally small, it is possible to efficiently encode even direct encoding.

In the above embodiment, the signal component separating means 2
Although the DC component in the image frame unit has been described as being calculated and separated, a DC component in a predetermined area unit in the image frame as shown in FIG. 5 may be used. Therefore, each of the predictive encoding means 5 and 6 for the DC component and the residual component has been described as using the decoded signal 27 of one or both of the preceding image frame and the succeeding image frame. The decoded signal 27 may be one or both of the predetermined area in the image frame, or one or both of the preceding predetermined area and the subsequent predetermined area in the current image frame. In addition, although it has been described that each encoding mode is adaptively selected based on a difference evaluation value in image frame units, a difference evaluation value in a predetermined area unit in an image frame may be used. Similarly, the scene change determination means 7 has been described as using the DC component input signal 13 of a separate image frame for the scene change determination, but may be the DC component input signal 13 of a predetermined area in a separate image frame.

[0023]

In the moving picture encoding apparatus of the above-described present invention, according to the present invention, the predetermined territory in the signal component or et picture frame of the moving image
Since the DC component of the area is calculated and separated, and the generated DC component and residual component are encoded respectively, even if the brightness of the entire screen changes significantly over time, many of the changes are directly
We table as change in flow component, the change of the residual component is reduced, when it encoded with high efficiency as a whole together
In addition, since each encoding operation at the time of prediction is appropriately selected according to a scene change, a prediction error when a scene change occurs can be reduced, and there is an effect that optimal encoding can be performed.

[Brief description of the drawings]

FIG. 1 is a functional block diagram of a moving picture coding apparatus according to an embodiment of the present invention.

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

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

FIG. 4 is a functional block diagram of another embodiment showing the present invention.

FIG. 5 is a diagram showing a predetermined area unit in an image frame.

FIG. 6 is a functional block diagram of a conventional moving picture encoding device.

FIG. 7 is a functional block diagram of a predictive encoding unit shown in FIG. 6;

FIG. 8 is a diagram showing an encoding mode set in advance.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Frame memory 2 Signal component separation means 3 DC component subtractor 4 Residual component subtractor 5 DC component prediction coding means 6 Residual component prediction coding means 7 Scene change judgment means 8 Residual component direct coding means 11 Image frame Column 12 Input signal to be encoded 13 DC component input signal 14 Residual component input signal 15 DC component prediction error signal 16 Residual component prediction error signal 17 DC component output encoded data 18 Residual component output predicted encoded data 19 DC Component prediction signal 20 Residual component prediction signal 21 Scene change detection signal 22 Residual component output directly encoded data In the drawings, the same reference numerals indicate the same or corresponding parts.

 ──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-2-188079 (JP, A) IEEE GLOBECOM'87 (November 1987) 1392-1396

Claims (2)

(57) [Claims] 1. A frame memory for temporarily storing a plurality of frames of image frame sequence of a moving image signal from the information source, the DC component of a predetermined area in the image frame from the coded input signal read from the Furemumemori was calculated and separated, and a signal component separating means for generating a respective input signal of the DC component and residual component, the input of the residual component generated by the signal component separating means
Generate output encoded data of a residual component based on the signal
The residual component encoding means and the DC component generated by the signal component
A subtractor that subtracts the prediction signal of the component and outputs a prediction error signal; and encodes the prediction error signal of the DC component from the subtractor.
Output encoded data of the DC component is generated, and
Decodes the output encoded data and decodes the preceding or succeeding image file.
Generated as a decoded signal of a DC component in a predetermined area of the frame,
DC component prediction code output as the DC component prediction signal
Encoding means and a separate image frame generated by the signal component separating means.
Comparing the DC component input signals in the predetermined area of
Scene change determination means for determining the presence or absence of
In addition, the DC component predictive encoding means performs the scene change determination.
If it is determined that there is a scene change by the
Is the image frame not including the changed image frame?
Generates a prediction signal for the current image frame from
A moving picture encoding apparatus for outputting a stream component as a prediction signal . 2. The residual component encoding means according to claim 1, wherein
Directly encodes the output signal to generate residual component output encoded data.
The moving picture coding apparatus according to claim 2, wherein the moving picture coding is performed.