KR100424686B1 - Method and apparatus for encoding object image - Google Patents

Abstract

PURPOSE: A method and an apparatus for encoding an object image are provided to improve shape information coding efficiency by moving a macro block of a VOP(Video Object Plane) to an information quantity reduction position according to a contour of the object image and encoding shape information. CONSTITUTION: A contour shape adaptive block dividing unit(40) moves a macro block of a VOP with respect to an object image formed in a VOP forming unit(11), and searches an information quantity reduction position. A shape coding unit(37) encodes shape information of an object which exists in a macro block of the information quantity reduction position searched in the contour shape adaptive block dividing unit(40). A shape adaptive block dividing unit(41) moves the macro block encoded in the shape coding unit(37), and searches an information quantity reduction position according to a contour of the object image. A motion estimation unit(31) estimates the motion of the object using output signals of the VOP forming unit(11) and the shape adaptive block dividing unit(41). A motion compensation unit(32) compensates the motion of the object using motion information estimated in the motion estimation unit(31) and the output signal of the shape adaptive block dividing unit(41). An adder(33) detects a difference value between the output signal of the motion compensation unit(32) and the output signal of the VOP forming unit(11). An object internal coding unit(34) encodes internal information of the object according to the output signal of the adder(33) and the output signal of the shape adaptive block dividing unit(41). An adder(35) adds output signals of the motion compensation unit(32) and the object internal coding unit(34). A previous VOP detecting unit(36) detects a VOP of a previous picture using the output signal of the adder(35) to use the detected VOP in the motion estimation and compensation of the motion estimation unit(31) and the motion compensation unit(32). A multiplexer(38) selectively outputs motion information estimated in the motion estimation unit(31), internal information of the object encoded in the object internal coding unit(34) and shape information encoded in the shape coding unit(37).

Description

Encoding Method and Encoding Device of Object Image

The present invention relates to an encoding method and an encoding apparatus of an object image for improving motion estimation and encoding efficiency by moving a macroblock partitioning a video object plane (VOP) in a moving picture expert group (MPEG-4) to an information amount reduction position. .

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

In the VOP, when a target image such as a predetermined object and a region exists in a video screen, the target image is separated from each other, and encoding the separated target image is performed as a basic frame.

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 / M1172 January).

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

Formation of the VOP is to form a background screen and a different VOP for each object image when there are several object images on one screen, and separates the background screen and each object image, and the separated background image The smallest rectangle that contains the object image is defined as VOP.

2 illustrates an example in which one VOP is formed by setting an image of a 'cat' as an object image.

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 MXM macro blocks having M and M pixels on the X and Y axes, respectively, with the upper left as the grid starting point. For example, it is divided into 16 X 16 macroblocks each having 16 pixels on the X and Y axes.

In this case, when there are not M and N pixels of the X and Y axes of the macroblocks formed on the right and the bottom of the VOP, the size of the VOP is expanded so that both the X and Y axes of the macroblock are M. And N.

In addition, M and N are set to an even number in order to perform encoding in units of sub-blocks in an object encoding unit (Texture Coding) to be described later.

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.

3 is a block diagram showing the configuration of a VM decoder primarily determined by an international standard subdivision.

The encoded signal of the VOP, which is encoded through the VM encoder and transmitted in the bit stream, is separated for each VOP in the demultiplexer 21 of the VM decoder.

Each of the separated VOP coded signals is decoded by the VOP decoders 22A, 22B, 22C, ..., respectively, and the output signals of the VOP decoders 22A, 22B, 22C, ... are synthesized by the combiner 23 to produce the original signal. It is output as an image on the screen.

4 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 estimator 31 to estimate the motion in units of macro blocks. The motion information estimated by the motion estimation unit 31 is input to a motion compensation unit 32 to compensate for the motion.

The VOP whose motion is compensated by the motion compensator 32 is input to the adder 33 together with the VOP formed by the VOP forming unit 11 to detect a difference value, and the difference value detected by the adder 33 is The internal information of the target is encoded by the target internal encoder 34 to be encoded in sub-block units of the macroblock.

For example, the object internal encoding unit 34 subdivides the X-axis and the Y-axis of the macroblock into M / 2X N / 2 into 8X8 subblocks each having eight pixels, and then encodes the internal information of the object.

The VOP whose motion is compensated by the motion compensator 32 and the internal information of the object encoded by the object internal encoder 34 are input to the adder 35, and the output signal of the adder 35 is added to the previous VOP. It is input to the detection unit (Previous ReconstructedVOP) 36 to detect the VOP of the previous screen. The VOP of the previous screen detected by the previous VOP detector 36 is input to the motion estimator 31 and the motion compensator 32 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 37 to encode shape information.

Here, the output signal of the shape encoder 37 is determined according to the field to which the VOP encoders 12, 12A, 12B, ... are applied, and the output signal of the shape encoder 37 is indicated by a dotted line. The motion estimation unit 31, the motion compensator 32, and the object internal encoder 34 may be input to the motion estimation unit 31, the motion compensator 32, and the object internal encoder 34 to encode the motion information, the motion compensation, and the internal information of the object.

The motion information estimated by the motion estimation unit 31, the internal information of the object encoded by the object internal encoder 34, and the shape information encoded by the shape encoder 37 are multiplexed by the multiplexer 38, and a buffer ( 39 is output to the multiplexer 13 of FIG. 1 and transmitted as a bit stream.

In the MPEG-4, when the VOP is formed, it is conventionally formed as a macroblock of MXM having M and M pixels in the X and Y axes from the VOP grid start point on the upper left.

In this case, when the object image, that is, the pixel having the shape information of the object, exists in a large number of macro blocks, the number of macro blocks for encoding the image of the object increases, thereby lowering the coding efficiency.

Therefore, according to the conventional technology of forming the object image at the upper left of the grid as the grid starting point and forming the macroblock unit of the MXN, the number of macroblocks to encode the object image is large, and thus there is a limit in improving the encoding efficiency of the VOP.

For example, in the drawing of FIG. 2 in which "cat" is displayed as an object, the number of macro blocks in which the pixels of "cat" are located is 20, and the 7 to 8 macro blocks include only small size pixels. The technique had to encode all 20 macroblocks in which "cat" pixels exist, resulting in low coding efficiency.

In addition, even when the shape encoder 37 performs motion estimation, motion compensation, and intra-object encoding on a signal in which shape information is encoded, since the amount of information output from the shape encoder 37 is large, motion estimation, motion compensation, and intra-object encoding are performed. There was a problem with a large amount of information.

Accordingly, an object of the present invention is to provide an encoding method and an encoding apparatus of an object image which improves encoding efficiency of shape information by moving shape macros of a VOP to a position of information reduction according to the contour of the object image and then encoding shape information. .

Another object of the present invention is to move the macroblock of the shape-information-encoded VOP coded signal to the information reduction position according to the object image, and then perform motion estimation, motion compensation, motion compensation, and intra-object encoding. The present invention provides a method and an encoding apparatus for encoding an object image that can improve efficiency and reduce an amount of information.

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 VOP having shape information divided into macroblocks according to a conventional method.

3 is a block diagram showing the configuration of a VM decoder primarily determined by the International Standards Organization.

4 is a block diagram showing the configuration of a VOP encoder determined primarily by an international standard subdivision.

FIG. 5 is a block diagram showing an embodiment in which a contour shape adaptation block divider and a shape adaptation block divider by an encoding apparatus of the present invention are configured in a VOP encoding unit determined primarily by an international standard subdivision.

FIG. 6 is a circuit diagram showing an embodiment of the contour shape adaptation block divider and the shape adaptation block divider of FIG.

7 is a circuit diagram showing an embodiment of the address generator of FIG.

8 is a diagram illustrating an example of outputting images stored in a memory of FIG. 6 in macroblocks according to the X-axis and Y-axis addresses output by the address generator.

9 is a circuit diagram showing the configuration of the block number counter of FIG.

10 is a signal flowchart showing an encoding method of the present invention.

FIG. 11 is a diagram illustrating a state in which a macroblock of a VOP is moved to an information amount decreasing position according to an outline of an object image according to the encoding method of the present invention as an example.

12 is a signal flow diagram illustrating an embodiment of finding a location of a reduced amount of information according to an outline of an object image of a macroblock according to the encoding method of the present invention.

FIG. 13 is a signal flowchart illustrating another embodiment of finding a location of a reduced amount of information according to an outline of an object image of a macroblock according to an encoding method of the present invention. FIG.

FIG. 14 is a signal flowchart illustrating another embodiment of finding a location of a reduced amount of information according to an outline of an object image of a macroblock according to an encoding method of the present invention. FIG.

(A) and (b) of FIG. 15 show zigzag movement of the grid starting point of the macro block according to another embodiment of FIG.

FIG. 16 is a signal flowchart illustrating another embodiment of finding a location of a reduced amount of information according to an outline of an object image of a macroblock according to an encoding method of the present invention. FIG.

FIG. 17 is a signal flow diagram illustrating another embodiment of finding a location of a reduced amount of information according to an outline of an object image of a macroblock according to an encoding method of the present invention. FIG.

(A) and (b) of FIG. 18 show a state in which the macro block is moved to the information amount reduction position according to another embodiment of FIGS. 16 and 17. FIG.

FIG. 19 is a signal flowchart illustrating an embodiment of finding a location of a reduced amount of information according to an object image of a macroblock according to an encoding method of the present invention. FIG.

FIG. 20 is a signal flow diagram illustrating another embodiment of finding a reduced amount of information according to an object image of a macroblock according to an encoding method of the present invention.

21 is a signal flow diagram illustrating another embodiment of finding a location of a reduced amount of information according to an object image of a macroblock according to an encoding method of the present invention.

FIG. 22 is a signal flowchart illustrating another embodiment of finding a location of reduced information amount according to an object image of a macroblock according to an encoding method of the present invention. FIG.

FIG. 23 is a signal flowchart illustrating another embodiment of finding a location of reduced information amount according to an object image of a macroblock according to an encoding method of the present invention. FIG.

Explanation of symbols for the main parts of the drawings

11: VOP forming portion, 12A, 12B, 12C,... : VOP encoder, 13, 38 multiplexer, 21, 38: demultiplexer, 22A, 22B, 22C,... ; VOP, 23: synthesis unit, 31: motion estimation unit, 32: motion compensation unit, 33, 35: adder, 34: object internal encoding unit, 36; Previous VOP detector, 37: shape encoder, 39: buffer, 40: contour adaptive block divider, 41: shape adaptive block divider, 51: address generation controller, 52: address generator, 53: memory, 54: number of blocks 55, minimum macroblock grid selector, 521: X-axis size determining unit, 522: Y-axis size determining unit, 523: block address generator, 541: macro convex counter, 542: judging unit, 543: adder

In order to achieve the above object, the present invention defines a VOP formed according to an object image, and finds an information reduction position while moving the X and Y axis grid start points of the macro block of the defined VOP.

The information amount reduction position is, for example, a position where the number of macro blocks in which the outline of the object image is present is minimized, and the X-axis and Y-axis grid starting point where the number of macro blocks in which the outline of the object image is present is minimized is optimal. And the Y-axis grid starting point, and reshaping the macro block according to the determined optimal X-axis and Y-axis grid starting point, and then the shape encoder performs encoding of shape information.

In addition, while the shape coding unit moves the X and Y axis grid starting points of the shape coded macroblock, the position where the number of macroblocks in which the object image exists is minimized is determined as the optimal X and Y axis grid starting point. Reconstruct the macroblocks according to the optimal X- and Y-axis grid starting points to perform motion estimation, motion compensation, and intra-object encoding.

Hereinafter, a method and an encoding apparatus for encoding an object image according to an exemplary embodiment of the present invention will be described in detail with reference to the accompanying drawings of FIGS. 5 to 9, where the same reference numerals are given to the same portions as in the prior art.

FIG. 5 shows the VOP encoders 12A, 12B, 12C, ... of the VM encoder determined primarily by the International Standards Organization, which constitute the contour shape block dividing unit and the shape adaptation region dividing unit by the encoding apparatus of the present invention. A block diagram showing an embodiment.

As shown in the drawing, the present invention includes an outline shape adaptation block divider 40 between the VOP forming unit 11 and the shape encoder 37, while moving the X and Y axis grid starting points of the macro block. The first information amount decreasing position is determined according to the contour position of the object image of the VOP.

As the first information amount reduction position, the position at which the number of macro blocks in which the contour of the object image is present is minimized is determined as the optimal X-axis and Y-axis grid starting points, and the determined optimal X-axis and Y-axis grid starting points are output. The shape encoding unit 37 is inputted.

The shape encoder 37 reconstructs the macro block according to the optimum X-axis and Y-axis grid start points output from the outline shape adaptation block dividing unit 40, and encodes shape information of the object.

The shape adaptation block dividing unit 41 is provided between the shape encoder 37, the motion estimation unit 31, the motion compensator 32, and the object internal encoder 34.

The shape adaptation block dividing unit 41 finds and outputs the second information amount decreasing position again while moving the X-axis and Y-axis grid starting points of the macroblock in which the shape encoding unit 37 encodes the shape information.

The second information amount reduction position determines an optimal X-axis and Y-axis grid start point at which the number of macroblocks in which the object image exists is minimized, and outputs the determined optimal X-axis and Y-axis grid start points.

The motion estimator 31, the motion compensator 32, and the object internal encoder 34 reconstruct the macroblock according to the optimum X-axis and Y-axis grid starting points determined by the shape adaptation block divider 41. Encode motion estimation, motion compensation, and object internal information.

FIG. 6 is a circuit diagram showing an embodiment of the outline shape adaptation block divider 40 and the shape adaptation block divider 41 of FIG. Reference numeral 51 is an address generation controller for moving the address start position at which the address is to be generated at regular intervals within the X-axis and Y-axis sizes of one macro block.

Reference numeral 52 denotes an address generator for generating X-axis and Y-axis addresses so that the target image is divided into macroblocks according to the address start position output by the address generation controller 51 and sequentially output. Reference numeral 53 is a memory that sequentially stores an object image having shape information and / or shape information inputted therein, and divides the image into macroblocks according to an address generated by the address generator 52.

In the address generator 52, as illustrated in FIG. 7, the X-axis size determiner 521 determines the X-axis size of the macro block according to the size information of the input target image, and the Y-axis size determiner 522. ) Determines the Y-axis size of the macro block.

Here, when the size of the X and Y axes of the macro block is the same, the size of the macro block may be determined using only one size determining unit.

In the address generator 52, the macro block address generator 523 determines the X-axis size and the Y-axis size of the macro block determined by the X-axis size determiner 521 and the Y-axis size determiner 522. On the basis of the address start position output from the address generation controller 51, the X-axis and Y-axis addresses for the macroblock of the object image stored in the memory 53 are distinguished, and the X-axis and Y-axis addresses are generated sequentially.

According to the X-axis and Y-axis addresses in which the macro block address generator 523 is sequentially generated, the memory 53 sequentially divides the stored object image into macro blocks. For example, as shown in FIG. 8, the object images of one macro block are sequentially output, and the object images of the next macro block are sequentially output.

Reference numeral 54 is a block number counter that counts the number of macro blocks in which the contour of the object image exists among the output signals of the memory 53 in the case of the contour shape adaptation block divider 40.

As shown in FIG. 9, the block number counter 54 distinguishes macro blocks by counting clock signals by the macro block counter 541. FIG. The determination unit 542 of the block number counter 54 divides the macroblock of the object image output from the memory 53 according to the output signal of the macroblock counter 541, and whether or not the outline of the object image exists. Judge.

Here, in the macro block in which the contour of the object image exists, the determination unit 542 determines that the macro block satisfying the two conditions as the object image exists in one or more pixels and does not exist in all the pixels of the object. It is determined as a macro block in which the outline of the object image exists.

The adder 543 adds the determination signal of the determination unit 542 to output the number of macro blocks in which the outline of the object image exists.

In the case of the shape adaptation block dividing unit 41, the block number counter 54 counts the number of macro blocks in which the shape information of the target image exists among the output signals of the memory 53.

The determination unit 542 of the block number counter 54 classifies the macroblock of the object image output from the memory 53 according to the output signal of the macroblock counter 541, and the shape information of the object image exists. Judge whether

The adder 543 adds the determination signal of the determination unit 542 to output the number of macro blocks in which the shape information of the object image exists.

Reference numeral 55 is a minimum macroblock grid selector that selects the X-axis and Y-axis grid starting points of the smallest number of macroblocks among the number of macroblocks counted by the block number counter 54 as information reduction positions.

The minimum macroblock grid selector 55 stores the count value when the count of the block number counter 54 is completed, and controls the address generation controller 51 to determine the start positions of the X-axis and Y-axis addresses. It is controlled to occur by moving at regular intervals within the X and Y axis sizes of the macro block. That is, the minimum macroblock grid selector 55 moves the start positions of the X-axis and Y-axis addresses, and then completes the operation of counting the number of macroblocks in which the contour of the object image exists. The control is repeated to move the start position of the X-axis and Y-axis addresses.

And while moving the start position of the X-axis and Y-axis address within the X-axis and Y-axis size of one macro block, the operation of counting the number of macro blocks in which the contour of the object image or the shape information of the object image exists is completed. In this case, the minimum macroblock grid selector 55 determines the start position of the X-axis and Y-axis address counting the smallest number of macroblocks among the counted macroblocks, and determines the determined start position of the X-axis and Y-axis addresses. Is determined as the optimum X-axis and Y-axis grid start position, which is the information reduction position, and is output.

10 is a signal flowchart showing an encoding method of the present invention.

First, in step S11, the VOP formed in the VOP forming unit 11 is defined. For example, the VOP is defined for the image of the "cat" shown in FIG. 11, and the VOP width and the VOP height are determined.

In the next step (S12), as shown by a dotted line in FIG. 11 on the basis of the grid starting point ratio, which is the upper left corner of the defined VOP, each block is divided into macro blocks having MXM pixels.

When the macro block is partitioned, the information amount reduction position is determined while moving the grid starting point of the macro block in step S13. For example, as shown by the solid line in FIG. 11, the optimal grid start point at which the contour of the object image exists in the smallest number of macroblocks is found to determine the information amount reduction position.

When the information amount reduction position is determined in step S13, the shape encoding unit 37 encodes shape information according to the information amount reduction position in the next step S14.

In the next step S15, the shape information of the object image is present in the smallest number of macro blocks while moving the grid starting point with respect to the macro block in which the shape encoder 37 encodes the shape information, as in the step S13. Find and print the information reduction location.

In step S16, the motion estimation unit 31 estimates the motion in units of macroblocks based on the information amount reduction position output in step S15, and the motion compensation unit 32 compensates for the motion. The encoder 34 encodes the object internal information in units of subblocks.

In the next step S17, the shape information, motion information, and object internal information encoded above are output through the multiplexer 38 and the buffer 39.

FIG. 12 is a signal flow diagram illustrating an embodiment of finding an information amount reduction position of a macro block according to the contour of an object image in step S13 according to the encoding method of the present invention.

First, in step S21, the initial X-axis and Y-axis are used to find the optimal X-axis and Y-axis grid starting points, which are information reduction positions where the number of macro blocks in which the contour of the object image having shape information and / or shape information exists is minimized. Initialize the axis grid starting point (XM) (YN).

The initial values of the X- and Y-axis grid starting points (XM) (YN) are given by XM = 0 and YN = 0.

In the next step S22, the number of macro blocks in which the contour of the object image exists is counted based on the X-axis and Y-axis grid starting points initialized in the step S21.

When the operation of counting the number of macroblocks is completed in step S22, the counted number of macroblocks is stored in step S23.

In a next step S24, the X-axis grid starting point XM is moved from the first macro block to the X-axis by a predetermined distance K to reconstruct the macro block, and in step S25, the X-axis grid starting point XM is X Determine if the axis has been moved more than M times. That is, it is determined whether the X-axis grid start point XM is moved beyond the X-axis size of one macro block.

When the X-axis grid starting point XM has not moved more than M times in the X-axis at step S25, the steps S22 to S25 are performed to set the X-axis grid starting point XM at a predetermined interval (X-axis). The method moves repeatedly by K) and counts and stores the number of macro blocks in which the contour of the object image exists.

When the X-axis grid start point XM is moved more than M times in the step S25 in the step S25, the X-axis grid start point XM in which the smallest number of macroblocks are counted in step S26 is optimized for the X-axis. Determined by the grid starting point (XOM).

In step S27, the macroblock is reconstructed with the determined optimal X-axis grid starting point XOM and Y-axis grid starting point YN = 0, and the number of macro blocks in which the contour of the object image exists is counted. When the operation of counting the number of macroblocks is completed in step S27, the counted number of macroblocks is stored in step S28.

In the next step S29, the Y-axis grid starting point YT is moved by the predetermined interval L on the Y-axis, and in step S30, it is determined whether the Y-axis grid starting point YN has been moved N times or more on the Y-axis. .

That is, it is determined whether the Y-axis grid start point YN has moved beyond the Y-axis size of one macro block.

If the Y-axis grid start point YN is not moved more than N times by a predetermined distance L in the Y-axis at step S30, the steps S27 to S30 are repeatedly performed to obtain an optimal X-axis grid starting point ( The X-axis grid starting point YM is moved by a predetermined interval L based on XOM), and the number of macroblocks in which the contour of the object image exists is counted and stored.

If the Y-axis grid start point YN is moved N or more times by a predetermined interval L in step S30, the Y-axis grid start point XM that counted the smallest number of macroblocks in step S31 is optimal. The Y-axis grid start point YON is determined, and the optimal X-axis grid start point XOM and Y-axis grid start point YON determined in the step S26 and S31 are output to the information amount reduction position.

That is, in the exemplary embodiment of the present invention shown in FIG. 12, the X-axis grid counting the smallest number of macro blocks in which the contour of the object image exists while moving the entire grid M times by a predetermined interval K on the X-axis. Determine the starting point (XM) as the optimal X-axis grid starting point (XOM). Y-axis grid counting the smallest number of macroblocks in which the contour of the object image exists while moving the entire grid N times by a predetermined interval (L) on the Y-axis based on the determined optimal X-axis grid start point (XOM). The starting point YN is determined as the optimal Y axis grid starting point YON, and the determined optimal X axis grid starting point XOM and the optimal Y axis grid starting point YON are output to the information amount reduction position.

Therefore, an embodiment of the present invention is to move the grid starting point M times on the X axis and N times on the Y axis. The X and Y axis grids are optimally moved by moving the grid starting points (XM, YM) M + N times. Find the starting point (XOM, YON) and print it.

13 is a signal flowchart showing another embodiment of finding the information amount reduction position according to the encoding method of the present invention.

In order to find the optimal X-axis and Y-axis grid starting points (XOM, YON) in which the number of macro blocks in which the contour of the object image having the VOP is formed in step Smini is minimum, the X-axis and Y-axis grid starting points (XM) Initialize (YN) to XM = 0, YN = 0.

In step S42, the number of macroblocks in which the contour of the object image exists is counted based on the X-axis and Y-axis grid starting points XM = 0 (YN = 0) initialized in step S41, and In S43, the counted number of macroblocks is stored.

In the next step S44, the X-axis grid starting point XM is moved by the predetermined distance K on the X-axis, and in step S45, the X-axis grid starting point XM is moved by the X-axis more than M times to the X-axis. It is determined whether the macro block has moved beyond the size of one macro block.

When the X-axis grid starting point XM has not moved more than M times in the X-axis at step S45, the steps S42 to S45 are performed to move the X-axis grid starting point XM to the X-axis at a predetermined interval K. ) And then counting and storing the number of macro blocks in which the contour of the object image exists.

When the X-axis grid starting point XM is moved more than M times in the X-axis at step S45, the Y-axis grid starting point YM is moved by a predetermined distance L in the Y-axis direction at step S46.

In the next step S47, it is determined whether the Y-axis grid start point YM has been moved more than M times by a predetermined interval L on the Y-axis.

When the Y-axis grid start point YN has not moved more than N times on the Y-axis in step S47, the steps S42 to S47 are performed to move the Y-axis grid by a predetermined distance L on the Y-axis. After that, the X-axis grid is moved to the X-axis by a predetermined interval K within the size of one macro block, and the number of macro blocks in which the contour of the object image exists is counted and stored.

When the Y axis grid start point YM is moved more than M times in the Y axis in step S47, the X axis grid start point XM in which the smallest number of macro blocks in which the contour of the object image exists is counted in step S48. ) And the Y-axis grid start point (YN) as the optimal X-axis grid start point (XOM) and Y-axis grid start point (YON), and the determined optimal X-axis and Y-axis grid start point (XOM) (YON) is reduced. Output to position

That is, in another embodiment of the present invention shown in FIG. 13, the entire grid starting point is moved M times by a predetermined interval K on the X axis, and then the entire grid starting point is moved by the Y axis. After repeating the movement of M intervals by the predetermined interval (K) on the X axis and the predetermined interval (L) on the Y axis, the number of macro blocks in which the contour of the object image exists is counted, and the smallest number of macro blocks The X-axis and Y-axis grid starting points (XM) YN, each of which is counted, are output to the information amount decreasing position.

Therefore, another embodiment of the present invention shown in FIG. 13 counts the smallest number of macroblocks by moving the X- and Y-axis grid starting points (XM) (YN) by MXN times at regular intervals (K) (L). Outputs the X- and Y-axis grid start points (XM) (YN).

In the embodiments of FIGS. 12 and 13, the grid starting point is moved by a predetermined interval (K) on the X axis, and then by the predetermined interval (L) on the Y axis, so that the optimal X and Y axis grid starting points (XOM) are moved. YON) is described as an example.

However, in the practice of the present invention, as shown in the parenthesis in FIGS. 12 and 13, the contour of the object image is shifted by a certain distance (L) along the Y axis and by a predetermined distance (K) by the X axis. It is also possible to find and output the optimal X- and Y-axis grid starting points (XOM, YON) that exist in the smallest number of macro blocks.

14 is a signal flow diagram illustrating another embodiment of finding the information amount reduction position according to the encoding method of the present invention.

In step S51, both the X-axis and Y-axis grid starting points (XM, YN) are found in order to find the optimal X-axis and Y-axis grid starting points (XOM, YON) in which the number of macro blocks in which the contour of the object image exists is minimized. Initialize to '0'.

In the next step S52, the number of macroblocks in which the contour of the object image exists at the initializing X-axis and Y-axis grid start points XM and YN is initialized, and in step S53, the number of macroblocks counted is counted. Save it.

In the next step S54, it is determined whether the X-axis grid start point XN has been moved M times in the X-axis, and the Y-axis grid start point YN has been moved N times in the Y-axis.

In the step S54, when the X-axis grid starting point XM is not moved more than M times in the X axis or the Y-axis grid starting point YM is not moved more than M times in the Y axis, in step S55 The axis and Y-axis grid starting point (XM) (YN) is moved by a predetermined distance (K) (L) in the zigzag direction within the size of one macro block, and the steps S52 to S55 are performed so that the outline of the object image is The number of macro blocks present is counted and stored, and the X and Y axis grid starting points (XM) YN are repeated in the zigzag direction.

Here, there are two methods for moving the X- and Y-axis grid starting points (XM) (YN) in the zigzag direction by a predetermined interval (K) (L). For example, it can move along the direction of the arrow shown to (a) of FIG. 15, or can move along the direction of the arrow shown to (b) of FIG.

When the X-axis grid start point XM is moved M times in the X-axis and the Y-axis grid start point YN is moved N times in the Y-axis in step S54, the smallest number of macroblocks are moved in step S56. The counted X-axis grid start point (XM) and Y-axis grid start point (YN) are determined as the optimal X-axis grid start point (XOM) and Y-axis grid start point (YON), and the determined optimal X-axis grid start point (XOM) is determined. And the Y-axis grid start point YON to the information amount reduction position.

That is, another embodiment of the present invention shown in FIG. 14 counts the number of macro blocks in which the outline of the object image exists while moving the grid starting point in a zigzag within the X and Y axis sizes of one macro block. .

Therefore, another embodiment of the present invention shown in FIG. 14 moves the smallest number of macroblocks by moving the X- and Y-axis grid starting points (XM) (YN) by MXN times in total by a predetermined distance (K) (L). Outputs the counted optimal X- and Y-axis grid starting points (XM) (YN).

An example of reforming the macro block according to the optimal grid start point XM (YN) determined according to the embodiment of FIGS. 12 to 14 described above is shown in solid line in FIG.

FIG. 16 is a signal flow diagram illustrating another embodiment of finding the information amount reduction position of a macroblock according to the encoding method of the present invention.

In step S61, the X- and Y-axis grid starting points XM and YN are initially initialized to '0' in order to find the information amount decreasing position where the number of macro blocks in which the contour of the object image exists is minimized.

In step S62, the number of macroblocks in which the contour of the object image exists at the X-axis and Y-axis grid starting points (XM = 0, YN = 0) initialized in step S61 is counted, and the next step ( In step S63, the number of macroblocks counted in step S62 is stored.

In the next step S64, it is determined whether the Y-axis grid start point YN has been moved by the predetermined distance L on the Y-axis and the Y-axis grid start point YN has been moved N times or more by the Y-axis in step S65. .

If the Y-axis grid start point YN has not moved more than N times on the Y-axis, the steps S62 to S65 may be performed to move the Y-axis grid start point YM by a predetermined interval L on the Y-axis. Counting and storing the number of macro blocks in which the contour of the image exists is repeated.

If the Y-axis grid starting point YN is moved more than N times by a predetermined interval L on the Y-axis in step S65, the Y-axis grid starting point YN in which the smallest number of macro blocks are counted in the next step S66. ) Is determined as the optimal Y-axis grid start point (YON).

When the optimal Y-axis grid start point YON is determined in which the smallest number of macroblocks are counted in step S66, the grid is determined based on the determined optimal Y-axis grid start point YON in step S67. In step S68, the macroblock of the current X-axis row in which the contour of the object image exists is counted.

That is, as shown in (a) of FIG. 18, the macroblock in which the contour of the object image exists is counted among the macroblocks in the X (1) row, which is the first row on the X axis.

In the next step S69, the counted number of macro blocks is stored.

In the next step S70, the grid starting point XM of row X (1) is moved by a predetermined distance K on the X axis, and in step S71, the grid starting point XM of the row X (1) is X-axis. It is determined whether the M has been moved more than M times by a predetermined interval (K).

When the X axis grid start point XM of row X (1) is not moved more than M times to the X axis in step S71, the steps S68 to S71 are performed to perform the X axis grid of row X (1). The starting point XM is moved by a predetermined interval K on the X-axis, and the macroblocks in which the contour of the object image is present are counted and stored among the macroblocks in the X (1) row.

If the grid starting point XM of the X (1) row is moved more than M times by a predetermined interval K in the step S71, the smallest number of macroblocks among the number of macroblocks counted so far in the step S72. The grid starting point XM of the X (1) row which counted is determined as the optimal grid starting point X1M of the X (1) row.

In the next step S73, it is determined whether it is the last row on the X axis, and if it is not the last row on the X axis, the process moves to the next row on the X axis in step S74, and the steps S68 to S74 are performed.

This operation is repeated to sequentially move the rows of the X axis to the rows X (1), X (2), X (3), X (4), and X (5), with the smallest contours of the object image. The grid starting point (XM) in rows X (1), X (2), X (3), X (4), and X (5), which counted the number of macroblocks, is the optimal X (1), X (2) , X (3), X (4) and X (5) row grid start points (X1M, X2M, ...).

In the case of the last row on the X-axis in step S74, the optimal Y-axis grid start point YON and X (1), X (2), X (3), X (4) and The optimal grid starting point (X1M, X2M, ...) in row X (5) is output to the information amount reduction position.

FIG. 17 is a signal flow diagram illustrating another embodiment of finding the information amount reduction position of a macroblock according to the encoding method of the present invention.

In step S81, the X-axis and Y-axis grid start points XM and YN are initialized to '0' in order to find a position where the number of macro blocks in which the contour of the object image exists is reduced.

In a next step S82, the macroblock in which the contour of the object image exists in the macroblock of the current X-axis row is counted. That is, the macroblock in which the contour of the object image exists among the macroblocks in row X (1) is detected and counted.

When the counting of the macroblocks is completed in step S82, the number of macroblocks counted in step S83 is stored.

In the next step S84, the grid starting point XM of row X (1) is moved by a predetermined interval K on the X axis, and in step S85, the grid starting point XM of the row X (1) is fixed by a certain interval ( It is determined whether K) has moved more than M times.

When the grid starting point XM of row X (1) has not been moved more than M times by a predetermined distance K in step S85, the steps S82 to S85 are performed to perform the X-axis of row X (1). The grid start point XM is moved by a predetermined interval K, and the number of macroblocks in which the contour of the object image exists is counted and stored.

When the X-axis grid start point XM of the X (1) row is moved more than M times by a predetermined distance K on the X axis in the step S85, the X (1) row counted to the present time in the step S86 The grid starting point XM of the X (1) row that counted the smallest number of macroblocks among the macroblocks is determined as the optimal grid starting point X1M of the X (1) row.

In the next step S87, it is determined whether it is the last row on the X axis, and if it is not the last row on the X axis, the next row on the X axis, that is, X (2), X (3), X (4) and After sequentially moving to row X (5), the steps S82 to S88 are performed to count the smallest number of macro blocks in which the contour of the object image exists, and X (2), X (3), and X (4). ) And the grid starting point (XM) of row X (5) is sequentially determined as the optimal grid starting point (M1M, X2M, ...) of X (2), X (3), X (4) and X (5) rows. Repeat that.

In the case of the last row on the X-axis in step S87, it is determined in step S89 whether the Y-axis grid start point YM has moved M times or more by a predetermined interval L in the Y-axis.

If the Y-axis grid starting point (YM) in the step (S89) has not moved more than N times by a predetermined interval (L) to the Y-axis in step (S90) Y-grid grid starting point (YM) in the Y-axis at a predetermined interval ( Move by L) and repeat steps S82 to S90.

That is, the Y-axis grid starting point YN is moved to the Y-axis by a predetermined distance L, and the X-axis grid starting point YN is moved to X (1), X (2), X (3), X The number of macro blocks in which the contour of the object image exists is counted while moving the starting point (X) of the row grid (4) and the X (5) row by a predetermined interval (K), and the smallest number of macro blocks is counted. (1), X (2), X (3), X (4), and X (5) row grid starting points (XM) with optimal X (1), X (2), X (3), X (4) ) And X (5) row grid starting points (X1M, X2M, ...).

If the Y-axis grid starting point (YM) is moved N or more times by a predetermined interval (L) in the step (S90), in the next step (S51), the optimal X determined at the position where the Y-axis grid starting point (YM) is moved 1), X (2), X (3), X (4), and X (5) all of the number of macroblocks counted at the grid start points X1M, X2M, ... are added together.

In the next step S92, the Y-axis grid start point YN in which the smallest number of macro blocks are counted is determined, and the determined Y-axis grid start point YN is determined as the optimal Y-axis grid start point YON. do. The X (1), X (2), X (3), X (4) and X (5) row grid starting points (X1M) counting the smallest number of macro blocks at the optimal Y-axis grid starting point (YON) , X2M,…) is determined as the optimal X (1), X (2), X (3), X (4) and X (5) row grid starting points (X1M, X2M,…) and the determined optimal A Y-axis grid start point YON and the X (1), X (2), X (3), X (4), and X (5) row grid start points (X1M, X2M, ...) are output as information reduction positions. .

Looking at the result of reconstructing the macro block according to the optimal grid starting point found in the embodiment of FIG. 16 and FIG. 17, it is as shown in (a) of FIG.

Here, in the embodiments shown in FIGS. 16 and 17, the Y-axis grid start point YN is moved and the X-axis grid start point XM is moved to find the information amount reduction position as an example.

In the present invention, as described in Figs. 16 and 17, the X-axis grid starting point (XM) and Y-axis grid starting point (YM) are interchanged so that the contour of the object image exists in the minimum macro block. You can also find the location of the information reduction.

Looking at the result of reconstructing the macro block to the optimal grid starting point found by swapping the X-axis grid starting point (XM) and the Y-axis grid starting point (YM), as shown in (b) of FIG.

FIG. 19 is a signal-flow diagram showing an embodiment of finding an information amount reduction position of a macroblock according to shape information of an object image in step S15 according to the encoding method of the present invention.

First, in step S101, the initial X-axis and Y-axis are searched to find the optimal X-axis and Y-axis grid starting points, which are information reduction positions where the number of macroblocks in which the shape information of the object image having shape information and / or shape information exists is minimized. Initialize the Y-axis grid start point (XM) (YN). The initial values of the X- and Y-axis grid starting points (XM) (YN) are given by XM = 0 and YN = 0.

In the next step S102, the number of macro blocks in which the shape information of the object image exists is counted based on the X-axis and Y-axis grid starting points initialized in the step S101.

In the example of the macroblock of the VOP shown by the dotted line in FIG. 11, the number of macroblocks in which the image of the "cat" exists from the first row on the X-axis is 3, 4, 4, 4, and 5, respectively. The images of "Cat" all exist in 20 macro blocks.

When the operation of counting the number of macroblocks is completed in step S102, the counted number of macroblocks is stored in step S103.

In a next step S104, the X-axis grid start point XM is moved from the first macro block to the X-axis by a predetermined distance K to reconstruct the macro block, and in step S105, the X-axis grid start point XM is X Determine if the axis has been moved more than M times.

That is, it is determined whether the X-axis grid start point XM is moved beyond the X-axis size of one macro block.

When the X-axis grid starting point XM has not moved more than M times in the step S105 in the step S105, the steps S102 to S105 are performed to move the X-axis grid starting point XM to the X axis by a predetermined interval ( The method moves repeatedly by K) and counts and stores the number of macro blocks in which the object image exists.

When the X-axis grid start point XM is moved more than M times in the step S25 in the step S25, the X-axis grid start point XM in which the smallest number of macroblocks are counted in step S26 is optimized for the X-axis. Determined by the grid starting point (XOM).

In step S27, the macroblock is reconstructed using the determined optimal X-axis grid starting point XOM and Y-axis grid starting point YN = 0, and the number of macroblocks in which an object image exists is counted.

When the operation of counting the number of macroblocks is completed in step S27, the counted number of macroblocks is stored in step S28.

In the next step S29, the Y-axis grid starting point YN is moved by the predetermined interval L on the Y-axis, and in step S30, it is determined whether the Y-axis grid starting point YM has been moved N times or more on the Y-axis. .

That is, it is determined whether the Y-axis grid start point YN is moved beyond the Y-axis size of one macro block.

When the Y axis grid start point YN is not moved more than N times by a predetermined distance L in the Y axis in step S30, the steps S27 to S30 are repeatedly performed to obtain an optimal X axis grid start point ( The X-axis grid starting point YM is moved by a predetermined interval L, and the number of macroblocks in which the object image exists is counted and stored.

If the Y-axis grid starting point YM is moved N or more times by a predetermined interval L in step S30, the Y-axis grid starting point XM that counted the smallest number of macroblocks in step S31 is optimal. The Y-axis grid start point YON is determined, and the optimal X-axis grid start point XOM and Y-axis grid start point YON determined in the step S26 and S31 are output to the information amount reduction position.

That is, according to the exemplary embodiment of the present invention shown in FIG. 12, the X-axis grid starting point (counting the smallest number of macroblocks in which the object image exists) while moving the entire grid by a predetermined interval (K) on the X-axis ( XM) is determined as the optimal X-axis grid starting point (XOM). Y-axis grid starting point counting the smallest number of macroblocks in which an object image exists while moving the entire grid M times by a predetermined distance (L) on the Y-axis based on the determined optimal X-axis grid starting point (XOM) ( YN) is determined as the optimal Y-axis grid starting point YON, and the determined optimal X-axis grid starting point XOM and optimal Y-axis grid starting point YON are output to the information amount reduction position.

Therefore, an embodiment of the present invention is to move the grid starting point M times on the X axis and N times on the Y axis. The X and Y axis grids are optimally moved by moving the grid starting points (XM, YM) M + N times. Find the starting point (XOM, YON) and print it.

13 is a signal flowchart showing another embodiment of finding the information amount reduction position according to the encoding method of the present invention.

In order to find the optimal X-axis and Y-axis grid starting points (XOM, YON) in which the number of macro blocks in which the object image having the VOP-formed object is present in step S41 is minimized, the X-axis and Y-axis grid starting points (XM) (YN ) Is initialized to XM-0, YN = 0.

In step S42, the number of macroblocks in which the object image exists is counted based on the X-axis and Y-axis grid starting points XM = 0 (YN = 0) initialized in step S41, and in step S43 ) Stores the counted number of macroblocks.

In the next step S44, the X-axis grid starting point XM is moved by the predetermined distance K on the X-axis, and in step S45, the X-axis grid starting point XM is moved by the X-axis more than M times to the X-axis. It is determined whether the macro block has moved beyond the size of one macro block.

When the X-axis grid starting point XM has not moved more than M times in the X-axis at step S45, the steps S42 to S45 are performed to move the X-axis grid starting point XM to the X-axis at a predetermined interval K. ), And counting and storing the number of macro blocks in which the object image exists.

When the X-axis grid starting point XM is moved more than M times in the X-axis at step 545, the Y-axis grid starting point YN is moved by a predetermined interval L in the Y-axis direction in step S46.

In the next step S47, it is determined whether the Y-axis grid start point YN has been moved N times or more by a predetermined interval L on the Y-axis.

When the Y-axis grid start point YN has not moved more than N times on the Y-axis in step S47, the steps S42 to S47 are performed to move the Y-axis grid by a predetermined distance L on the Y-axis. After that, the X-axis grid is moved on the X-axis by a predetermined interval K within the size of one macro block, and the number of macro blocks in which the target image exists is counted and stored.

If the Y-axis grid start point YN is moved N times or more in the step S47 in the step S47, the X-axis grid start point XM counting the smallest number of macro blocks in which the object image exists in step S47 and Determine the Y-axis grid starting point (YN) as the optimal X-axis grid starting point (XOM) and Y-axis grid starting point (YON), and determine the optimal X-axis and Y-axis grid starting point (XOM) (YON) as the information reduction position. Output

That is, in another embodiment of the present invention shown in FIG. 13, the entire grid starting point is moved M times by a predetermined interval K on the X axis, and then the entire grid starting point is moved by the Y axis. Repeat the operation of moving M intervals by a certain interval (K) on the X axis and a certain interval (L) on the Y axis, counting the number of macroblocks in which the object image exists, and counting the smallest number of macroblocks. One X-axis and Y-axis grid start point (XM) YN is output to the information amount reduction position.

Therefore, another embodiment of the present invention shown in FIG. 13 counts the smallest number of macroblocks by moving the X- and Y-axis grid starting points (XM) (YN) by MXN times at regular intervals (K) (L). Outputs the X- and Y-axis grid start points (XM) (YN).

In the embodiments of FIGS. 12 and 13, the grid starting point is moved by a predetermined interval (K) on the X axis, and then by the predetermined interval (L) on the Y axis, so that the optimal X and Y axis grid starting points (XOM) are moved. YON) is described as an example.

However, in the practice of the present invention, as shown in parentheses in the drawings of FIGS. 12 and 13, the optimum X-axis and the predetermined X-axis are moved along the X-axis by a certain distance (K). The Y-axis grid start point (XOM, YON) can also be found and output.

14 is a signal flowchart showing another embodiment of finding the information amount reduction position according to the encoding method of the present invention.

In step S51, both the X and Y axis grid starting points (XM, YN) are set to '0' in order to find the optimal X and Y axis grid starting points (XOM, YON) in which the number of macro blocks in which the object image is located is minimized. Initialize to '.

In the next step S52, the number of macroblocks in which an object image exists at the initializing X-axis and Y-axis grid start points XM and YN is counted, and the counted number of macroblocks is stored in step S53. .

In the next step S54, it is determined whether the X-axis grid start point XM has been moved M times in the X-axis, and the Y-axis grid start point YN has been moved N times in the Y-axis.

In the step S54, when the X-axis grid start point XM is not moved more than M times in the X axis or the Y-axis grid start point YN is not moved more than N times in the Y axis, in step S55 The axis and Y-axis grid starting point (XM) (YN) is moved by a predetermined distance (K) (L) in the zigzag direction within the size of one macro block, and steps S52 to S55 are performed to present the object image. The number of macro blocks is counted and stored, and the X and Y axis grid starting points (XM) YN are repeated in the zigzag direction.

Here, there are two methods for moving the X- and Y-axis grid starting points XM and YM by a predetermined interval K and L in the zigzag direction. For example, it can move along the direction of the arrow shown to (a) of FIG. 15, or can move along the direction of the arrow shown to (b) of FIG.

When the X-axis grid start point XM is moved M times in the X-axis and the Y-axis grid start point YN is moved N times in the Y-axis in step S54, the smallest number of macroblocks are moved in step S56. The counted X-axis grid start point (RLFM) and Y-axis grid start point (YN) are determined as the optimal X-axis grid start point (XOM) and Y-axis grid start point (YON), and the determined optimal X-axis grid start point (XOM) is determined. And the Y-axis grid start point YON to the information amount reduction position.

That is, another embodiment of the present invention shown in FIG. 14 counts the number of macro blocks in which an object image exists while moving the grid starting point in a zigzag within the X and Y axis sizes of one macro block.

Therefore, another embodiment of the present invention shown in FIG. 14 moves the smallest number of macroblocks by moving the X- and Y-axis grid starting points (XM) (YN) by MXN times in total by a predetermined distance (K) (L). Outputs the counted optimal X- and Y-axis grid starting points (XM) (YN).

An example of reforming the macro block according to the optimal grid start point XM (YN) determined according to the embodiment of FIGS. 12 to 14 described above is shown in solid line in FIG.

Here, as a result of reconstructing the macroblock by moving the optimal X-axis and Y-axis grid starting point (XOM) (YON) according to the position of the object image forming the VOP, the number of macroblocks in which the image of "cat" is present is X From the first row on the axis we can see that 2, 3, 4, 3 and 5 are reduced to 17 in 20 macro blocks.

FIG. 16 is a signal flow diagram illustrating another embodiment of finding the information amount reduction position of a macroblock according to the encoding method of the present invention.

In step S61, the X- and Y-axis grid starting points XM and YN are initially initialized to '0' in order to find the information amount decreasing position where the number of macro blocks in which the object image exists is minimized.

In step S62, the number of macroblocks in which the object image exists at the X and Y-axis grid starting points XM = 0 and YN = 0 initialized in the step S61 is counted, and in the next step S63 In step S62, the number of macro blocks counted in step S62 is stored.

In the next step S64, it is determined whether the Y-axis grid start point YN is moved to the Y axis by a predetermined distance L, and in step S65, the Y-axis grid start point YN is moved N times or more to the Y-axis. do.

If the Y-axis grid start point YN has not moved more than N times on the Y-axis, the steps S62 to S65 are performed to move the Y-axis grid start point YN to the Y-axis by a predetermined interval L. Counting and storing the number of macro blocks in which an image exists is repeated.

If the Y-axis grid start point YN is moved N or more times by a predetermined interval L in the step S65 in the step S65, the Y-axis grid start point YN in which the smallest number of macro blocks are counted in the next step S66. ) Is determined as the optimal Y-axis grid start point (YON).

When the optimal Y-axis grid start point YON is determined in which the smallest number of macroblocks are counted in step S66, the grid is determined based on the determined optimal Y-axis grid start point YON in step S67. In step S68, the macroblock of the current X-axis row in which the object image exists is counted.

That is, as shown in (a) of FIG. 18, the macroblock in which the object image exists is counted among the macroblocks in row X (1) which is the first row on the X axis.

In the next step S59, the counted number of macro blocks is stored.

In the next step S70, the grid starting point XM of row X (1) is moved by a predetermined distance K on the X axis, and in step S71, the grid starting point XM of the row X (1) is X-axis. It is determined whether the M has been moved more than M times by a predetermined interval (K).

When the X axis grid start point XM of the X (1) row is not moved more than M times to the X axis in the step S71, the steps S68 to S71 are performed to perform the X axis grid of the X (1) row. The starting point XM is moved by a predetermined interval K on the X axis, and the macro blocks in which the object image exists in the macro blocks in the X (1) row are counted and stored.

If the grid starting point XM of the X (1) row is moved more than M times by a predetermined interval K in the step S71, the smallest number of macroblocks among the number of macroblocks counted so far in the step S72. The grid starting point XM of the X (1) row which counted is determined as the optimal grid starting point X1M of the X (1) row.

In the next step S73, it is determined whether it is the last row by the slaughter, and when it is not the last row by the X axis, the process moves to the next row of the X axis in step S74, and the steps S68 to S74 are performed.

This operation is repeated to sequentially move rows on the X-axis to rows X (1), X (2), X (3), X (4), and X (5), with the smallest number of target images present. The grid starting point (XM) in rows X (1), X (2), X (3), X (4), and X (5) where the macroblock is counted is set to the optimal X (1), X (2), X (3), X (4) and X (5) row grid start points (X1M, X2M, ...) are determined.

In the case of the last row on the X-axis in step S74, the optimal Y-axis grid start point YON and X (1), X (2), X (3), X (4) and The optimal grid starting point (X1M, X2M, ...) in row X (5) is output as the information amount / decrease position.

FIG. 17 is a signal flow diagram illustrating another embodiment of finding the information amount reduction position of a macroblock according to the encoding method of the present invention.

In step S81, the X-axis and Y-axis grid start points XM and YN are initialized to '0' in order to find a position where the number of macro blocks in which the object image exists is reduced in the amount of information.

In a next step S82, the macroblock in which the object image exists among the macroblocks in the current X-axis row is counted. That is, the macroblock in which the object image exists among the macroblocks in row X (1) is detected and counted.

When the counting of the macroblocks is completed in step S82, the number of macroblocks counted in step S83 is stored.

In the next step S84, the grid starting point XM of row X (1) is moved by a predetermined interval K on the X axis, and in step S85, the grid starting point XM of the row X (1) is fixed by a certain interval ( It is determined whether K) has moved more than M times.

When the grid starting point XM of row X (1) has not been moved more than M times by a predetermined distance K in step S85, the steps S82 to S85 are performed to perform the X-axis of row X (1). The grid starting point XM is moved by a predetermined interval K, and the number of macroblocks in which the object image exists is counted and stored.

When the X-axis grid start point XM of the X (1) row is moved more than M times by a predetermined distance K on the X axis in the step S85, the X (1) row counted to the present time in the step S86 The grid starting point XM of the X (1) row that counted the smallest number of macroblocks among the macroblocks is determined as the optimal grid starting point of the X (1) row.

In the next step S87, it is determined whether it is the last row on the X axis, and if it is not the last row on the X axis, the next row on the X axis, that is, X (2), X (3), X (4) and After sequentially moving to row X (5), steps S82 to S88 are performed to count the smallest number of macroblocks in which the object image exists, and X (2), X (3), X (4) and Repeatedly determine the grid starting point (XM) in row X (5) as the optimal grid starting point (X1M, X2M, ...) in rows X (2), X (3), X (4) and X (5). do.

In the case of the last row on the X-axis in step S87, it is determined in step S89 whether the Y-axis grid start point YN is moved N times or more by a predetermined interval L in the Y-axis.

When the Y-axis grid start point YN has not moved more than N times by a predetermined distance L in the Y axis in step S89, the Y-axis grid start point YN is fixed in Y-axis in step S90. Move by L) and repeat steps S82 to S90.

That is, the Y-axis grid starting point YN is moved to the Y-axis by a predetermined distance L, and the X-axis grid starting point YN is moved to X (1), X (2), X (3), X The number of macroblocks in which an object image exists is counted while moving the grid starting point (XM) of (4) and X (5) rows sequentially by a predetermined interval (K), and X (1) of the smallest number of macroblocks is counted. ), X (2), X (3), X (4), and X (5) row grid starting points (XM) are set to optimal X (1), X (2), X (3), X (4) and X (5) row Grid start points (X1M, X2M, ...) are determined sequentially.

If the Y-axis grid start point YN is moved N or more times by a predetermined interval L in step S90, in the next step S91, the optimal X determined at the position where the Y-axis grid start point YN is moved, respectively ( 1), X (2), X (3), X (4), and X (5) all of the number of macroblocks counted at the grid start points X1M, X2M, ... are added together.

In the next step S92, the Y-axis grid start point YN in which the smallest number of macro blocks are counted is determined, and the determined Y-axis grid start point YN is determined as the optimal Y-axis grid start point YON. do. The X (1), X (2), X (3), X (4) and X (5) row grid starting points (X1M) counting the smallest number of macro blocks at the optimal Y-axis grid starting point (YON) , X2M,…) is determined as the optimal X (1), X (2), X (3), X (4) and X (5) row grid starting points (X1M, X2M,…) and the determined optimal A Y-axis grid start point YON and the X (1), X (2), X (3), X (4), and X (5) row grid start points (X1M, X2M, ...) are output as information reduction positions. .

Looking at the result of reconstructing the map and the row block according to the optimal grid starting point found in the embodiment of FIGS. 16 and 17, it is as shown in (a) of FIG.

Here, in the embodiments shown in FIGS. 16 and 17, the Y-axis grid start point YN is moved and the X-axis grid start point XM is moved to find the information amount reduction position as an example.

In the present invention, as shown in the parenthesis in FIGS. 16 and 17, the X-axis grid starting point (XM) and the Y-axis grid starting point (YN) are interchanged with each other to reduce the amount of information in which the object image exists in the minimum macro block. You can also find a location.

Looking at the result of reconstructing the macro block to the optimal grid starting point found by swapping the X-axis grid starting point (XM) and the Y-axis grid starting point (YN), as shown in (b) of FIG.

In the above-described embodiment of the present invention, the predetermined interval K (L) for moving the X-axis and Y-axis grid start points XM (YN) is based on the number of pixels existing in the macro block.

For example, the X-axis and Y-axis grid starting points (XM) (YN) within the X-axis and Y-axis sizes of the macroblock may be moved at intervals of unit pixels.

However, since the information on the color signal in the video signal is 1/2 of the information on the low light signal, when the information on the color signal and the brittleness signal is taken into consideration, the moving interval K of the X- and Y-axis grid start points (XM) (YN) ( L) is preferably at two unit pixel intervals.

In addition, in the above-described embodiment, only one optimal grid starting point exists in the smallest number of macroblocks in which the object image having the VOP is formed as an example. However, one or more optimal grid starting points may occur.

Therefore, in the present invention, when multiple optimal grid starting points occur, the macro block is subdivided into sub-blocks having the size of M / 2 XN / 2, and the grid starting point (XM) within the X- and Y-axis sizes of the sub-blocks. (YN) is moved by a predetermined interval (K) (L), it is possible to find and output the information reduction position where the object image is present in the smallest number of macro blocks.

In other words, if it is assumed that the size of the macro block is 16 × 16 pixels, the grid starting point XM (as shown in the above embodiment) is subdivided into subblocks having 8 × 8 pixels and subpixels subdivided into 8 × 8 pixels. YN) is moved by a predetermined interval K (L) to find the X and Y axis grid starting points having the smallest number of macro blocks in which the object image exists, and outputs them to the information amount reduction position.

Further, even when the macro block is subdivided into sub-blocks, a plurality of information amount reduction positions that count the smallest number of macro blocks may occur.

Therefore, in the present invention, even when subdivided into sub-blocks, one should be selected when a plurality of optimal X- and Y-axis grid starting points (XOM, YON) occur.

At this time, the one information amount reduction position to be selected is smaller as the motion vector becomes smaller as the initial X-axis and Y-axis grid starting points (XM = 0, YN = 0) decrease, and the prediction error when the motion is estimated Since the incidence of is low, the X and Y-axis grid starting points with the closest straight line distance are determined based on the initial grid starting point (XM = 0, YN = 0).

In the above-described embodiment of the present invention, the predetermined interval K (L) for moving the X-axis and Y-axis grid start points XM (YN) is based on the number of pixels existing in the macro block.

For example, the X-axis and Y-axis grid starting points (XM) (YN) within the X-axis and Y-axis sizes of the macroblock may be moved at intervals of unit pixels.

However, since the information about the color signal in the image signal is 1/2 of the information about the luminance signal, when the information about the color signal and the luminance signal is taken into account, the moving interval K (X) of the X- and Y-axis grid start points (XM) (YN) ( L) is preferably at two unit pixel intervals.

In addition, in the above-described embodiment, only one optimal grid starting point exists in which the contour of the object image on which the VOP is formed is present in the smallest number of macro blocks. However, one or more optimal grid starting points may occur.

Therefore, in the present invention, when a plurality of optimal grid starting points occur, the macro block is divided into sub-blocks having the size of M / 2 XN / 2, and the grid starting point (XM) within the X-axis and Y-axis sizes of the sub-blocks. ) YN may be moved at predetermined intervals K and L, and the position of the amount of information reduction in which the contour of the object image exists in the smallest number of macroblocks may be found and output.

In other words, if it is assumed that the size of the macroblock is 16 × 16 pixels, the grid starting point XM (YN) is subdivided into subblocks having 8 × 8 pixels and subpixels subdivided into 8 × 8 pixels as described above. ) By a predetermined interval (K) (L) to find the starting point of the X-axis and Y-axis grid with the smallest number of meckle blocks in which the contour of the object image exists, and outputs them to the information amount reduction position.

Further, even when the macro block is subdivided into sub-blocks, a plurality of information amount reduction positions that count the smallest number of macro blocks may occur.

Therefore, in the present invention, even when subdivided into sub-blocks, one should be selected when a plurality of optimal X- and Y-axis grid starting points (XOM, YON) occur.

At this time, the one information amount reduction position to be selected is smaller as the motion vector becomes smaller as the initial X-axis and Y-axis grid starting points (XM = 0 and YN = 0) decrease, and the prediction error when the motion is estimated Since the incidence of is low, the X and Y-axis grid starting points with the closest straight line distance are determined based on the initial grid starting points (XM = 0 and YN = 0).

As described above, the present invention encodes the shape information by adjusting the grid starting point so that the contour of the object image is located at the minimum macro block, and also moves after adjusting the grid starting point so that the shape information encoding signal is located at the minimum macro block. By performing estimation and encoding of the object internal information, the encoding efficiency is improved, and the amount of information to be transmitted and stored is reduced.

Claims (24)

A partitioning step of dividing the VOP formed according to the object image into a macroblock having pixels of MXN, and a first search step of finding a position of information reduction according to the contour of the object image while moving the grid starting point of the partitioned macroblock in the partitioning step And a shape information encoding step of encoding shape information of the object image existing in the macro block at the information reduction position found in the first search step, and a reduction of the amount of information according to the object image contour in the macro block encoded in the shape information encoding step. A second search step for finding a position, a motion estimation / encoding step for estimating and compensating for motion in units of macro blocks at a reduced amount of information found in the second search step, and encoding object internal information in units of sub blocks; Shape information and images encoded in the information encoding step Motion estimation / motion information and the coding method of the target image for encoding the output of the object inside the information characterized by the output stage made of an estimation by the encoding step. The method of claim 1, wherein the first and second search steps include: a moving step of moving the entire macroblock on the X-axis and / or a Y-axis; and an outline of the object image at the position where the entire macroblock is moved in the moving step. And an output step of finding and outputting a position where the information amount decreases in the determination step. The method of claim 2, further comprising: a first shift count step of counting the number of macroblocks in which the contour of the object image exists while moving the entire macroblock on the X axis, and the smallest number of macroblocks in the first shift count step; The first setting step of setting the counted X-axis position, and the macro block is reconstructed to the X-axis position set in the first setting step, and the contour of the object image exists while moving the entire reconstructed macro block to the Y axis A second shift count step of counting the number of macroblocks, a second shift step of setting a Y-axis position counting the smallest number of macroblocks in the second shift count step, and the first and second set steps And a determination step of determining the X-axis and Y-axis positions set as the information amount reduction positions. The method of claim 2, further comprising: a first shift count step of counting the number of macroblocks in which the contour of the object image exists while moving the entire macroblock on the Y-axis, and the smallest number of macroblocks in the first shift count step; The first setting step of setting the counted Y-axis position, and the contour of the object image exists while reconstructing the macro block to the Y-axis position set in the first setting step and moving the entire reconstructed macro block to the X-axis A second movement counting step of counting the number of macroblocks, a second setting step of setting an X-axis position counting the smallest number of macroblocks in the second movement counting step, and the first and second setting steps And a determination step of determining the Y-axis and the X-axis positions set as the information amount reduction positions. The method of claim 2, further comprising: a first shift count step of counting the number of macroblocks in which the contour of the object image exists while the entire macroblock is moved on the X-axis, and the entire macro when the X-axis shift is completed in the count step; A second move count step of repeating the count step after moving the block to the Y axis, and X having the smallest number of macroblocks counted when the Y axis move is completed in the first and second move count steps; And a determination step of determining the position of the axis and the Y-axis as the information amount reduction position. The method of claim 2, further comprising: a first shift count step of counting the number of macroblocks in which the contour of the object image exists while moving the entire macroblock on the Y-axis; and when the Y-axis shift is completed in the count step A second move count step of repeating the count step after moving the block to the X axis, and X having the smallest number of macroblocks counted when the X axis move is completed in the first and second move count steps And a determination step of determining the position of the axis and the Y-axis as the information amount reduction position. The method according to claim 2, further comprising: a movement counting step of counting macroblocks in which the contour of the object image exists while moving all macroblocks in a zigzag, and a position where the smallest number of macroblocks are counted in the counting step as an information amount reduction position; And a determination step of determining the object image. The method of claim 1, wherein the first and second search steps include: a moving step of moving the entire macroblock on the X-axis or the Y-axis, and each of the contours of the object image at the position where the macroblock is moved in the moving step. And a determination step of determining the amount of information of the X-axis or the Y-axis macroblock of the row or column, and an outputting step of finding and outputting the information amount reduction position in the determination step. The method of claim 8, wherein the first count step of counting the macro block in which the contour of the object image exists while moving the entire macro block on the Y axis, and the position where the smallest number of macro blocks are counted in the first count step. A first setting step of setting the Y-axis position and a row of each X-axis where the contour of the object image is positioned while moving the macroblock of each row of the X-axis to the X-axis at the Y-axis position set in the first setting step A second counting step of counting the macroblocks of the second counting step, a second setting step of setting the position at which the smallest number of macroblocks are counted in the second counting step to a position of each row of the X-axis, and the first setting step And a determining step of determining the position of each of the Y-axis position and the X-axis row set in the second setting step as the information amount reduction position. . The method of claim 8, wherein the first count step of counting the macro block in which the contour of the object image exists while moving the entire macro block on the X axis, and the position where the smallest number of macro blocks are counted in the first count step. The first setting step of setting the X-axis position and the macro of each column of the Y-axis in which the contour of the object image is positioned while moving the macroblock of each column of the Y-axis at the X-axis position set in the first setting step to the Y-axis. A second counting step of counting blocks, a second setting step of setting a position of counting the smallest number of macroblocks in the second counting step as a position of each column of the Y axis, the first setting step and the second setting step And a determination step of determining the position of each of the X-axis position and the Y-axis column set in the setting step as the information amount reduction position. . The method according to claim 8, wherein the number of macroblocks in which the contour of the object image exists and the smallest number of macroblocks are counted while moving the macroblocks of each row of the X-axis in the X-axis. An X-axis positioning step of determining the position of the row; a moving step of moving the entire macroblock to the Y-axis and repeating the X-axis positioning step when the count is completed in the X-axis positioning step; A Y-axis positioning step of summing the count values at the position of each row of the X-axis determined in the X-axis positioning step and determining the Y-axis position with the smallest sum when the Y-axis movement is completed in the step; The position of each row of the X-axis in which the Y-axis position determined in the Y-axis positioning step and the smallest macroblock in the determined Y-axis position is counted as the information amount reduction position. Coding method of the object image, characterized by determining made of an crystal phase. The method of claim 8, wherein the number of macroblocks in which the contour of the object image exists and the smallest number of macroblocks are counted while moving the macroblocks of each column of the Y-axis to the Y-axis. Y-axis position setting step of setting to position, a moving step of moving the entire macroblock to X-axis and repeating the Y-axis positioning step when counting is completed in the Y-axis positioning step, and in the moving step An X-axis position setting step of adding up count values at the position of each column of the Y-axis set in the Y-axis position setting step when the X-axis movement is completed, and setting an X-axis position having the smallest sum value; In the X axis position setting step, set the position of each column of the Y axis that counted the least macro block in the X axis position set in the X axis position as the information reduction position. Coding method of the object image, characterized by determining made of an crystal phase. The encoding method according to any one of claims 1 to 12, wherein the information amount decreasing position is a position where the information amount is minimum. The encoding method according to any one of claims 1 to 12, wherein the information amount reduction position is a position where the number of macro blocks in which the contour of the object image exists is minimized. The method according to any one of claims 1 to 12, wherein the moving range of the macro block is within the X and Y axes of one macro block. The method according to any one of claims 1 to 12, wherein the movement range of the macro block is within the X-axis and Y-axis sizes of one macro block, and moves at unit pixel intervals. The encoding method of any one of claims 1 to 12, wherein the movement range of the macroblock is within the X-axis and Y-axis sizes of one macroblock, and is moved at intervals of two unit pixels. Way. The X and Y axes of the macro block according to any one of claims 3 to 7, and 9 to 12, when there are a plurality of X and Y axis positions in which the smallest number of macro blocks are counted. And dividing the size into half, and counting the number of macroblocks in which the contour of the target image exists and determining the position where the smallest number of macroblocks is counted as the information amount reduction position. . 19. The method of claim 18, wherein the X- and Y-axis sizes of the macro block are divided into 1/2, and then the closest to the X- and Y-axis grid starting points when there are a plurality of positions where the smallest number of macro blocks are counted. And a position of the information amount decreasing position. A first outline shape adaptation block dividing unit 40 for moving the macro block of the VOP with respect to the object image formed by the VOP forming unit 11 to find a position of information reduction according to the outline of the object image, and the outline shape adaptation block dividing unit A shape encoder 37 for encoding the shape information of the object present in the macroblock at the frequently reduced amount of information at 40 and the shape block 37 moves the macro block encoded with the shape information to contour the object image. A motion of estimating the motion of the object by the output signal of the second outline shape adaptation block dividing unit 41 and the VOP forming unit 11 and the second outline shape adaptation block dividing unit 41 to find the location of reduced information amount The motion estimator 31 compensates for the motion of the object using the motion signal estimated by the motion estimator 31 and the output signal of the second contour shape adaptation block divider 41. An adder 33 for detecting a difference value between the unit 32, an output signal of the motion compensator 32, and an output signal of the VOP forming unit 11, an output signal of the adder 33, and the first signal; 2 The object internal encoder 34 for encoding the internal information of the object according to the output signal of the contour shape block dividing unit 41, and the output signals of the motion compensation unit 32 and the object internal encoder 34 An adder 35 to add and an output signal of the output signal of the adder 35 detect the VOP of the previous screen and use the motion estimator 31 and the motion compensator 32 for motion estimation and motion compensation. Outputs the previous VOP detector 36, the motion information estimated by the motion estimator 31, the internal information of the object encoded by the object internal encoder 34, and the shape information encoded by the shape encoder 37 Characterized in that consisting of a multiplexer 38 to The encoder in the object image. 21. The apparatus of claim 20, wherein the first and second contour shape adaptation block dividing units (40) (41) comprise: address generation control means for shifting and outputting a start position at which an address is to be generated at regular intervals within the size of the macro block; An address generating means for generating an address to divide and output an object image into macroblocks according to an address start position output by the address generation control means, and storing an object image having input shape information and generating the address image generating means. A memory means for outputting according to an address, a block number counting means for counting the number of macroblocks in which the contour of the object image exists among the signals output from the memory means, and a number of macroblocks counted by the block number counting means Draw the X- and Y-axes with the smallest number of macro blocks Of the object to the image encoding apparatus, characterized by at least consists of a macroblock grid selecting means for selecting the starting point. 22. The apparatus according to claim 21, wherein the address generating means comprises: X-axis size determining means and Y-axis size determining means for respectively determining the X-axis size of the macro block and the Y-axis size of the macro block according to the size information of the input target image; From the address start position output from the address generation control means, the size of the X-axis and the Y-axis of the macroblock determined by the X-axis size determining means and the Y-axis size determining means are divided, And an area address generating means for sequentially generating the corresponding addresses. 23. The apparatus of claim 22, wherein the size of the X and Y axes of the macro block is determined by one size determining means when the sizes of the X and Y axes of the macro block are the same. 22. The apparatus of claim 21, wherein the block order counting means comprises: region counting means for counting clock signals to classify macroblocks, and macroblocks of an image output from a memory means according to output signals of the region counting means, and Contour existence judging means for judging whether or not the shape information exists, and a block number adding means for outputting the number of macro blocks in which the contour of the object image exists by counting the determination signal of the contour existence judging means. An apparatus for encoding an object image to be played.