KR100424681B1 - Apparatus for encoding shape information - Google Patents

Abstract

PURPOSE: An apparatus for encoding shape information is provided to reduce the quantity of transmission data by rounding shape information of an object image, moving a grid according to shape information so that shape information exists in the minimum macro block, and encoding shape information. CONSTITUTION: A binary rounding unit(31) performs the binary rounding of an original object image. A shape adaptive region dividing unit(32) moves a grid according to shape information of the object image rounded in the binary rounding unit(31), and outputs an offset value which exists in the minimum macro block. A binary quadtree shape information coding unit(33) performs the binary quadtree shape information coding of an image of an object which exists in the minimum macro block according to the movement of the grid in the shape adaptive region dividing unit(32).

Description

Shape information encoding device

The present invention relates to a shape information encoding apparatus for encoding shape information (shape) of an object image in which a Video Object Plane (VOP) is formed in Moving Picture Experts Group (MPEG-4).

MPEG-4, which is currently being standardized, is based on the concept of VOP (Video Object Plane).

In this case, when a plurality of target images such as a predetermined object and a region exist in a single video screen, the VOP is divided into VOPs, and the encoding of the separated target images is performed as a basic skeleton. have.

When the meditation of the encoded object is transmitted, the horizontal value and the vertical value for the minimum size of the image and the information of the object are transmitted together.

The VOP has the advantage of freely synthesizing or decomposing a natural image and an artificial image as a unit of an object image, and is fundamental to processing an object image in the field of computer graphics and multimedia.

FIG. 1 is a block diagram showing the configuration of a VM (Verification Model) encoder determined primarily by the International Standards Organization (ISO / IEC JTC1 / SC29 / WG11 MPEG96 / N1172 January).

Here, the VOP Formation unit 11 forms a VOP when an image sequence to be transmitted or stored is input.

The VOP is formed of a background image and a different VOP for each object image when multiple object images exist in one screen, and is divided into a background image and each object image, and the separated background image. The smallest rectangle that contains the object image is defined as VOP.

2 shows an example of only one VOP being formed by setting an image of "cat" as an image of an object.

Here, the horizontal size of the VOP is defined as the VOP width, and the vertical size is defined as the VOP height.

The formed VOP is divided into M × N macroblocks having M and N pixels on the X-axis and the Y-axis, respectively, with the upper left as the grid starting point. For example, it is divided into 16x16 macroblocks each having 16 pixels on the X-axis and the Y-axis.

In this case, when the number of X and Y axes of the macro block formed on the right and bottom of the VOP is not M and N, respectively, the size of the VOP is expanded to have M pixels of both the X and Y axes of each macro block. And N.

Each VOP formed by the VOP forming unit 11 is input to the VOP encoders 12A, 12B, 12C, ..., respectively, encoded for each VOP, multiplexed by the multiplexer 13, and transmitted as a bit stream.

FIG. 3 is a block diagram showing the configuration of the VOP encoders 12A, 12B, 12C, ... of the VM encoder primarily determined by the international standard subdivision.

The VOP for each object image formed by the VOP forming unit 11 is input to a motion estimation unit 21 to estimate the motion in units of macro blocks.

The motion information estimated by the motion estimator 21 is input to a motion compensation unit 22 to compensate for the motion.

The VOP whose motion is compensated by the motion compensator 22 is input to the adder 23 together with the VOP formed by the VOP forming unit 11 to detect a difference value, and the difference value detected by the adder 23 is The internal information of the object is encoded by the object internal encoding unit 24 in units of sub blocks of the macro block. For example, the object internal encoding unit 24 subdivides the X-axis and the Y-axis of the macroblock into 8x8 subblocks each having eight pixels, each of M / 2 x N / 2, and then stores the internal information of the object. Encode

The VOP whose motion is compensated by the motion compensator 22 and the internal information of the object encoded by the object internal encoder 24 are input to the adder 25, and the output signal of the adder 25 is added to the previous VOP. Input to the detection unit (Previous Reconstructed VOP) 26 to detect the VOP of the previous screen. The VOP of the previous screen detected by the previous VOP detector 26 is input to the motion estimator 21 and the motion compensator 22 to be used for motion estimation and motion compensation.

The VOP formed by the VOP forming unit 11 is input to a shape coding unit 27 to encode shape information.

Here, the output signal of the shape information encoder 27 is determined according to the field to which the VOP encoders 12, 12A, 12B, ... are applied. As indicated by the dotted line, the motion estimation unit 21, the motion compensator 22, and the object internal encoder 24 may be input to the motion estimation unit, the motion compensation unit, and the internal information of the object.

The motion information estimated by the motion estimation unit 21, the internal information of the object encoded by the object internal encoder 34, and the shape information encoded by the shape information encoder 27 are multiplexed by the multiplexer 27, The buffer 29 is output to the multiplexer 13 of FIG. 1 and transmitted as a bit stream.

In the VOP encoders 12A, 12B, 12C, ..., when the shape information of the object is encoded by the shape information encoder 27 without adjusting the position of the grid according to the position where the image of the object exists. The number of macro areas in which the image of the object exists is very large and the amount of encoded information is very large.

Therefore, in 1996 Patent Application No. 11367, which was filed by the present applicant, a shape adaptation region segmentation unit is provided, and a macro to be encoded by moving a grid according to the position of the object image so that the image of the object exists in the smallest number of macroblocks. It is possible to improve the coding efficiency by reducing the number of blocks.

However, the 1996 patent application No. 11367 describes the amount of data transmitted from the shape information encoder by moving the grid according to the original shape information of the object image so that the image of the object exists in the smallest number of macro blocks. In the case of performing lossy coding with rounding, shape information of an object image to be transmitted and reproduced is different from the original shape information.

Therefore, when the motion estimation and the object encoding of the object image are performed, there is a problem in that the position of the macro block must be adjusted by moving the grid according to the shape information of the object image.

Therefore, an object of the present invention is to round the shape information of the object image, move the grid in accordance with the shape information to exist in the minimum macro block, shape shape coding to reduce the amount of transmission data by performing shape information encoding To provide a device.

1 is a block diagram showing the configuration of a VM encoder primarily confirmed by the International Standards Organization.

2 is a diagram illustrating a state in which a VOP having shape information is divided into macroblocks by a grid as an example.

3 is a detailed view showing a VOP encoder of a VM encoder primarily determined by an international standard subdivision.

4 is a hearth diagram showing an embodiment of the shape information encoding apparatus of the present invention.

5 is a circuit diagram showing another embodiment of the shape information encoding apparatus of the present invention.

* Explanation of symbols for the main parts of the drawings

11: VOP forming unit, 31: binary rounding unit, 31A: gray scale rounding unit, 32: shape adaptation region splitting unit, 33: binary quadtree shape information encoding unit, 33A: grayscale quadtree shape information encoding unit

According to the shape information encoding apparatus of the present invention, the rounding unit performs rounding on the shape information of the object image input from the VOP forming unit, and the shape adaptation region partitioner performs the rounding on the shape information of the rounded object image. After the grid is moved to exist in the minimum macro block, quadtree shape information encoding is performed.

Hereinafter, the shape information encoding apparatus of the present invention will be described in detail with reference to the accompanying drawings of FIGS. 4 and 5, where the same reference numerals are given to the same portions as in the prior art.

4 is a circuit diagram showing an embodiment of the shape information encoding apparatus of the present invention.

As shown in the drawing, the exemplary embodiment of the present invention moves the grid according to the binary rounding unit 31 for binary rounding the original object image and the shape information of the binary image rounded by the binary rounding unit 312. Shape-adaptation region dividing unit 32 for outputting an offset value present in the minimum macroblock, and an image of an object present in the minimum macroblock according to the movement of the grid in the shape-adapting region dividing unit 32 It consists of a binary quadtree shape information encoding unit 33 for encoding quadtree shape information.

According to the exemplary embodiment of the present invention configured as described above, when the original object image of the VOP formed by the VOP forming unit 11 is input, the binary rounding unit 31 performs binary rounding.

That is, the binary rounding unit 31 displays shape information by distinguishing a region where an image of the object exists from a region where the image of the object does not exist with two distinct values, for example, '0' and '255'. do.

When the binary rounding of the object image is completed in the binary rounding unit 31, the shape adaptation area divider 32 moves the grid according to the shape information of the object image to offset the image of the object in the minimum macro block. The binary quadtree shape information encoder 33 performs binary quadtree shape information encoding and outputs the binary quadtree shape information.

That is, the binary quadtree shape information encoder 321 sums the shape information of the target image in units of 4 × 4 blocks, and adds the shape information based on a threshold value preset for adjusting the bit rate. If the value is greater than or equal to the threshold value, the entire 4x4 block is filled with '255'. Here, the threshold value is set high if the total shape information amount (bit amount) is large (e.g., the sum of shape information is 3), and set low if the shape information amount is small (e.g., The sum is 1).

Blocks that do not meet the threshold value are further subdivided into 2 × 2 blocks to sum all the values of the shape information, and if the sum of the shape information is less than the threshold value, all of the '0' If the sum of the shape information or the sum of the shape information is more than the threshold value, it is filled with '255'. Here, another threshold value is set to a value equal to or smaller than the threshold value applied to the 4x4 block (for example, the sum of the shape information is 1).

5 is a circuit diagram showing another embodiment of the shape information encoding apparatus of the present invention. As shown in the drawing, a gray scale rounding unit 31A for binary rounding an original object image and a minimum macroblock are moved by moving a grid according to shape information of the grayscale rounded object image in the gray scale rounding unit 31A. The shape adaptation region segmentation unit 32 that outputs an offset value present in the image, and the image of the object present in the minimum macro block according to the movement of the grid in the shape adaptation region segmentation unit 32 are gray scale quadtree shape information. It consists of the gray scale quadtree shape information coding part 33A to encode.

According to another embodiment of the present invention configured as described above, when the original object image of the VOP formed by the VOP forming unit 11 is input, the gray scale rounding unit 31A performs binary rounding.

That is, the gray scale rounding unit 31A is widely used for a blue screen technique for distinguishing a background image from an image of an object, and has a value other than '0' and '255', that is, between '0' and '255'. Encodes shape information that exists as a value.

When the gray scale rounding of the object image is completed in the gray scale rounding unit 31S, the shape adaptation region dividing unit 32 moves the grid according to the shape information of the object image so that the image of the object exists in the minimum macro block. The offset value is output, and the gray scale quadtree shape information encoder 33A performs grayscale quadtree shape information encoding to output the offset value.

That is, the gray scale quadtree shape information encoder 322 sums the shape information of the target image in 4 × 4 block units or 2 × 2 block units, and sets a threshold value previously set to adjust the bit rate. If the sum of the shape information is less than the threshold value, all are filled with '0' or if the sum of the shape information is more than the threshold value, the entire 4x4 block is filled with '255'. Here, the threshold value is set high if the total shape information amount (bit amount) is large (e.g., the sum of shape information is 3), and set low if the shape information amount is small (e.g., The sum is 1).

At this time, when dividing the image of the object into 4 × 4 blocks or 2 × 2 blocks, shape adaptation region division is performed to move the grid in sequence, and the number of macroblocks in which the image of the object exists is minimized. By determining the position of the grid, the effect of rounding is enhanced.

In the above description, the shape information encoding apparatus of the present invention is applied to MPEG-4 as an example. However, in implementing the present invention, various methods such as video telephony, video conferencing, PCS PDA, mobile phone, mobile multimedia, etc. It can be applied to an image processor.

As described above, according to the present invention, when the shape information is encoded, the rounding is performed after the shape adaptation region division to perform rounding for each position of the changed grid so that the object image is positioned in the smallest number of macro blocks. The amount of data to be transmitted can be effectively reduced.

Claims (3)

A rounding unit for rounding the VOP output from the VOP forming unit 11 and a shape adaptation region dividing unit for shifting the output signal of the rounding unit according to the shape information of the object image so as to exist in the minimum macro block. And a quadtree-shape information encoder for quadtree-shape information coding on the output signal of the shape adaptation region dividing unit (32). 2. The shape information encoding apparatus according to claim 1, wherein the rounding portion is a binary rounding portion (31), and the quadtree shape information encoding portion is a binary quadtree shape information encoding portion (33). 2. The shape information encoding apparatus according to claim 1, wherein the rounding portion is a gray scale rounding portion (31A), and the quadtree shape information encoding portion is a gray scale quadtree shape information encoding portion (33A).