JPH031688A - High efficiency picture coder - Google Patents

Abstract

PURPOSE:To improve the coding efficiency and the reduction picture quality in the case of reproduction of a reduced picture by providing a macro block processing means, and inter-field block processing means, an in-field block processing means and a block selection means. CONSTITUTION:A picture signal subject to 2:1 interlace scanning is inputted from an input terminal 1 and a macro block processing circuit 2 splits the signal into plural macro blocks each comprising 8-picture element X 16-line. Then a moving still deciding circuit 3 decides whether or not a video image in each macro block is moved or at standstill and when it is decided that the picture is at standstill, switches 4, 7 are connected to an inter-field block processing circuit 5. When it is decided that the picture is moving, the switches 4, 7 are connected to in-field block processing circuit 6. Thus, the forming method of the block is switched adoptively in the unit of macro blocks to improve the coding efficiency and when a reduced picture is reproduced, the quality of the reduced picture is improved.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a high-efficiency image encoding device for efficiently encoding a 2:1 interlaced scanned image signal such as a television signal. It is something. [Prior Art] When a 2:1 interlaced scanned image signal is encoded in units of blocks, in the prior art, a block is formed for each field. There is also a conventional technology that reproduces an image that is reduced in size from the original image by canting the high frequency components included in one image block by block and reducing the number of samples.An example of this conventional technology is "Compatible HDTV".
Coding for Broadband ISDN
'(proceeding of 19881EEE
GLOBECOM, VOI.

2, 24.1), which means that on the transmitting side, the entire image is divided into blocks of N pixels x M lines, and orthogonal transformation is performed for each block, resulting in N x M orthogonal transformation coefficients per block. is obtained, the transform coefficients are quantized and transmitted, and the receiving side performs orthogonal inverse transform on all N×M transform coefficients per block, thereby obtaining a reproduced image of the same size as the original image.

By orthogonally inversely transforming only nxm (n < N, m < M) transform coefficients corresponding to low frequency components among NxM transform coefficients per l block, n/N A reproduced image reduced to m/M is obtained.

[Problem to be solved by the invention]

When the input image signal to be encoded has been subjected to 2:l interlaced scanning, one frame is composed of two fields, and the fields are replaced every l line. therefore.

For example, in an area where the video is moving, blocks formed within the field (hereinafter referred to as intra-field blocks)
Generally, the correlation between pixels is stronger in blocks formed within a frame across fields (hereinafter referred to as inter-field blocks), so it is more efficient to encode each intra-field block. . On the other hand, in an area where the video is still, inter-field blocks generally have stronger correlation between pixels than intra-field blocks, so it is more efficient to encode each inter-field block. However, in the conventional technology, no consideration has been given to these points.

In addition, in the conventional technology for playing back reduced images, if coding is performed for each inter-field block in an area where the video is moving, the movement of the reduced playback image will become unnatural, or conversely, the video will become static. If coding is performed for each intra-field block in a region where the image is displayed, the spatial resolution of the reduced reproduced image will be lowered more than necessary.The conventional technology does not solve these problems at all.

The present invention has been made in view of the problems of the prior art, and when encoding an image signal subjected to 2:1 interlaced scanning for each block, it is possible to adaptively switch the block formation method. The present invention aims to provide a high-efficiency image encoding device that not only improves encoding efficiency but also improves the quality of a reduced image when reproducing the reduced image.

[Means to solve the problem]

In order to achieve the above object, the present invention has a quaternary configuration.

By macroblocking means, adjacent N
Let the pixels of the pixel X2M line be a macroblock.

The interfield blocking means divides a macroblock of N pixels by 2M lines into a block from the 1st line to the Mth line and a 7' block from the (M+1)th line to the 2Mth line. Therefore, 2
Each block has one.

Adjacent N within a frame across two fields
It is formed from pixels x M lines of pixels.

The intra-field blocking means divides a macroblock of N pixels by 2M lines into a block consisting of pixels only on odd-numbered lines and a block consisting of pixels only on even-numbered lines. Therefore, each of the two blocks is formed from adjacent N pixels×M lines of pixels within the field.

The block selecting means adaptively selects one of the blocks obtained by the inter-field blocking means and the blocks obtained by the intra-field blocking means in units of macroblocks.

[Effect]

According to the present invention, either one of the blocks obtained by the interfield blocking means and the blocks obtained by the intrafield blocking means is selected for each macroblock by the block selection means. Therefore, in a high-efficiency image encoding device that encodes a 2=1 interlace scanned image signal block by block, the block formation method is adaptively switched for each macroblock, so that the encoding efficiency is lower than that of the conventional technology. The encoding efficiency is improved, and when the reduced image is reproduced, the quality of the reduced image is improved compared to the quality of the reduced image according to the prior art.

〔Example〕

FIG. 1 is a block diagram of an embodiment of the present invention.

The image signal subjected to the 21 interlaced scanning is inputted from the input terminal 1, and is divided into a plurality of macroblocks each consisting of 8 pixels x 16 lines in the macroblock forming circuit 2 as shown in FIG. 2 (A>). The black circle in the figure is the first
The pixels of the field are represented, and the white circles represent the pixels of the second field.

The motion/static determination circuit 3 determines whether the video in each macroblock is moving or stationary. If it is determined that it is stationary, switch 4 and switch 7 are connected to interfield blocking circuit 5. In a region where the video is still, the correlation between adjacent lines within a frame across fields is considered to be stronger than the correlation between adjacent lines within a field. Therefore, the inter-field blocking circuit 5 divides the macroblocks into blocks from line 51 to line 8 and line 9 as shown in FIG. 2(B).
It is divided into blocks from line 1 to line 16. The size of both blocks is 8 pixels x 8 lines.

On the other hand, if it is determined that the video within the macroblock is moving, the switches 4 and 7 are connected to the intra-field blocking circuit 6. In areas where the video is moving, the correlation between adjacent lines within a field is considered to be stronger than the correlation between adjacent lines within a frame across fields. Therefore, the intra-field blocking circuit 6 divides the macroblock into a block consisting of pixels only on odd-numbered lines and a block consisting of pixels only on even-numbered lines, as shown in FIG. 2(C). The size of both blocks is 8 pixels x 8 lines. Switching information indicating how the switches 4 and 7 are switched is outputted from the output terminal 16 as additional information.

In the discrete cosine transform circuit 8, the image signal is discrete cosine transformed for each block formed by the inter-field blocking circuit 5 or the intra-field blocking circuit 6.
X8 cosine transform coefficients are obtained. Discrete cosine transform is a type of orthogonal transform, and efficient encoding can be achieved by performing H-losing cosine transform.

Of the obtained 8×8 cosine transform coefficients, 4×4 cosine transform coefficients corresponding to the low frequency components shown in FIG. It is output from Remaining (8X8)-
The (4×4) cosine transform coefficients are quantized by a quantization circuit 10 and encoded by an encoding circuit 12. Encoding circuit 1
The code from the encoder 2 is added to the code from the encoding circuit 11 in the adder 13 and output from the output terminal 15.

Note that by decoding the code output from the output terminal 15 and performing inverse discrete cosine transformation with a block size of 8×8, a reproduced image of the same size as the original image can be obtained.

Furthermore, by decoding the code output from the output terminal 14 and performing inverse discrete cosine transformation with a block size of 4×4, a reduced reproduced image having a size 1/2 vertically and horizontally of the original image can be obtained.

〔Effect of the invention〕

As described above, according to the third aspect of the present invention, in a high-efficiency image encoding device, the way blocks are formed is switched in units of macroblocks so that lines that are considered to have a stronger correlation become the same block. , the encoding efficiency is improved compared to the conventional technique, and when the reduced image is reproduced, the quality of the reduced image is improved compared to the conventional technique.

circuit, ■ 2... Encoding circuit.

...Output terminal.

15...output terminal

Claims (1)

[Scope of Claims] A high-efficiency image in which an image signal in which one frame is composed of two fields by 2:1 interlaced scanning is encoded for each block consisting of pixels of N pixels x M lines. In the encoding device, N pixels adjacent in the frame x 2
A macroblock forming means for forming pixels of M lines into macroblocks, and dividing the macroblock into blocks from the 1st line to the Mth line and blocks from the (M+1)th line to the 2Mth line. , inter-field blocking means for forming two adjacent blocks of N pixels x M lines in a frame, and forming the macroblock into a block consisting of pixels only on odd-numbered lines and a block consisting of pixels only on even-numbered lines. By dividing into N adjacent in the field
Intra-field blocking means for forming two blocks of pixels x M lines, one of the blocks formed by the inter-field blocking means, and the block formed by the intra-field blocking means, is processed using a macro. A high-efficiency image encoding device, comprising: block selection means for selecting each block, and encoding the image signal for each block selected by the block selection means.