JPH1169352A - Shape encoding method and shape decoding method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a shape encoding method which does not increase the number of the bits of picture element encoding while the visual distortion of object shape encoding is reduced. SOLUTION: This method is provided with a shape encoder 211 executing an encoding processing on an object shape signal Sk, generating a shape encoding bit string Dk, decoding a signal in the middle of the encoding processing and generating a local decoding shape signal Pek and a shape post-processor 212 executing a shape post-processing on the local decoding shape signal Pek. A selection switch 13 selects the output Pek of the shape encoder 11 or the output PFk of the shape post-processor 212. The selection output of the switch 13 is outputted as an encoding processing reference signal Rc used for the encoding processing of a picture element signal.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a shape coding method and a shape coding device, a shape decoding method and a shape decoding device, and a program for realizing processing by these methods or devices by software. In particular, the present invention relates to a process for encoding and decoding an object shape signal (a signal indicating the shape of an object) necessary for encoding and decoding an image display signal on an object basis. .

[0002]

2. Description of the Related Art In recent years, a multimedia era has been entered in which voice, image, and other data are handled in an integrated manner. Has become a feature of multimedia. Generally, multimedia means not only characters but also graphics, sounds, and especially images, etc. are simultaneously associated with each other. To make the above-mentioned conventional information media the target of multimedia, the information must be expressed in digital form. Is an essential condition.

However, when the amount of information of each of the above information media is estimated as a digital information amount, the amount of information per character is 1 to 2 bytes in the case of a character, while the amount of information per character is 64 kbits per second in the case of voice. (Telephone quality), and more than 100 Mbits per second (current television reception quality) for moving images, it is not practical to handle the vast amount of information in digital form, especially in information media related to image information. . For example, a videophone is an integrated service digital network (ISDN) having a transmission speed of 64 kbps to 1.5 Mbps.
Services Digital Network), but it is not possible to send the video of a TV camera directly through ISDN.

Therefore, what is required is an information compression technique. For example, in the case of a videophone, ITU-T
(International Telecommunication Union Telecommunication Standardization Sector) The H.261 standard moving image compression technology is used. Further, according to the information compression technology of the MPEG1 standard, it is possible to store image information together with audio information in a normal music CD (compact disc).

Here, MPEG (Moving Picture Exper)
(ts Group) is an international standard for data compression of moving images, and MPEG1 is a standard for compressing moving image data to 1.5 Mbps, that is, TV signal information to about 1/100. Also, since the transmission speed for the MPEG1 standard is mainly limited to about 1.5 Mbps, the MPE standardized to meet the demand for higher image quality
In G2, the moving image data is compressed to 2 to 15 Mbps.

Further, at present, MPEG1, MPEG2
And the working group that has promoted standardization (ISO / IEC JTC1 / SC2
9 / WG11) MPE that enables image information to be encoded for each object on the display screen and performs necessary signal processing to realize new functions required in the multimedia age
G4 is being standardized.

In MPEG4, first, an object shape signal indicating the shape of an object included in an image display signal corresponding to each object on a display screen is encoded, and the encoded object shape signal (hereinafter referred to as shape encoding) is encoded. With reference to a bit string, a pixel value signal for displaying an object in color, which is included in the image display signal, is encoded.

Here, pixel values corresponding to pixels inside the object need to be coded in order to display the object.
In MPEG4, since a signal for image display is processed for each object on the display screen, information on pixels outside the object is unnecessary, and therefore, it is not necessary to encode pixel values of pixels outside the object. Therefore, in MPEG4, the encoding process is performed only on the pixel values inside the object with reference to the object shape, and the encoding process is omitted on the pixel values outside the object, thereby saving the number of encoded bits.

FIG. 15 is a diagram for explaining the encoding process of a pixel value signal in MPEG4. Here, a fish-shaped object (object) O displayed on the display screen R is displayed.
Assume that pixel value encoding processing is performed on b. FIG.
3A shows the pixel values of the pixels constituting the object Ob for each macroblock Mb, and the dot display area in the figure shows the inside of the object.

The encoding process of pixel values in MPEG4 divides the display screen R in the same manner as MPEG1 and MPEG2.
A unit area (macroblock) M composed of 16 × 16 pixels
b, the pixel value does not need to be coded for a pixel in a macroblock located completely outside the object, but the pixel value is not coded for a pixel in a macroblock that also includes part of the object shape. Need to be encoded. FIG.
In the process of encoding the pixel value of the object Ob shown in FIG.
The macroblock to be encoded is shown in FIG.
Is a macroblock Mbx indicated by oblique lines.

FIG. 16 is a diagram for more specifically explaining the encoding process for pixel values in MPEG4.
In FIG. 16A, the pixel values of the luminance signal corresponding to the pixels constituting the object Ob are shown for each unit area (sub-block) Sb composed of 8 × 8 pixels. The display area shows the inside of the object.

In MPEG, motion compensation of pixel values and coding in a coding mode indicating whether to perform intra-coding or inter-coding are performed in units of 16 × 16 pixel macroblocks. As shown in FIG. 16A, the luminance signal included in the pixel value signal for performing color display is obtained by DCT transform and quantization in units of 8 × 8 pixel sub-block Sb, which divides a macroblock into four. Processing is performed and encoded.

Here, it is not necessary to encode the pixel values of the luminance signal for the pixels in the sub-block Sb located completely outside the object. Needs to encode the pixel value of the luminance signal. In the process of encoding the pixel value of the luminance signal of the object Ob shown in FIG. 16B, the block to be encoded is a sub-block Sbx indicated by oblique lines in FIG. 16C.

As described above, since the number of macroblocks Mb and sub-blocks Sb to be subjected to the pixel value encoding process depends on the object shape, the encoding process of the object shape signal is performed by the encoding process. It is important that the number of macroblocks and the number of sub-blocks be small.

Hereinafter, the conventional encoding and decoding processes in MPEG4 will be specifically described. FIG. 12 shows a conventional image display signal encoding method based on MPEG4.
FIG. 12A is a diagram for explaining a decoding process, and FIG. 12A shows a configuration of an encoding circuit that performs an encoding process of the image display signal.

In FIG. 12A, an encoding circuit 200a performs an encoding process on an image display signal for displaying an object on a display screen. This encoding circuit 200
a is a shape encoding device 210a that encodes an object shape signal Sk representing the shape of the object, which is included in the image display signal.
And a pixel value encoding device 2 for encoding a pixel value signal Sg for displaying an object in color included in the image display signal
20a.

The shape coding device 210a performs a shape coding process including arithmetic coding on the object shape signal Sk included in the image display signal to form a shape coded signal Ck.
And a shape encoding unit 211a that locally decodes (information source decodes) the shape-encoded signal Ck to generate a local shape-decoded signal PEk. A shape post-processing unit 212 that performs a shape post-processing to make the shape when displaying the object shape smoother based on the local shape decoded signal PEk and outputs a shape post-processing output PFk; Signal Ck
And a variable-length coding unit 211b for performing a variable-length coding process to generate a shape-coded bit string Dk.

Here, the shape encoder 211a and the variable-length encoder 211b constitute a shape encoder 211. The shape post-processor 212 includes a noise removal filter, For example, the configuration corresponds to a block distortion removal filter, a mosquito distortion removal filter, or the like.

The encoding circuit 200a performs a discrete cosine transform (DCT) and a quantum cosine transform (DCT) on the pixel value signal Sg included in the image display signal based on the shape post-processing output PFk. Pixel value encoding section 221a that performs an encoding process to generate a pixel value encoded signal Cg, and performs a variable length encoding process on the pixel value encoded signal Cg to generate a pixel value encoded bit sequence Dg And a variable length encoding unit 221b.

FIG. 12 (b) shows the configuration of a decoding circuit for decoding the image display signal.
, 200b is a decoding circuit that performs a decoding process on the image display signal encoded by the encoding circuit 200a, and the decoding circuit 200b decodes and decodes the shape encoded bit sequence Dk. A shape decoding device 210b for generating an encoded object shape signal Fk, and a decoded pixel value signal E for decoding the pixel value encoded bit string Dg based on the decoded object shape signal Fk to display an object in color.
and a pixel value decoding device 220b for generating g.

The shape decoding device 210b performs a variable length decoding process on the shape encoded bit sequence Dk corresponding to the object shape signal to generate a variable length decoded shape signal C'k. A long decoding unit 213a, a shape decoding unit 213b that performs a shape decoding process including arithmetic decoding on the variable-length decoded shape signal C ′ k to generate a decoded shape signal Ek, For the generalized shape signal Ek,
A shape which is subjected to shape post-processing to make the shape when displaying the object shape smoother based on the shape decoded signal Ek, and outputs the shape post-processed output as the decoded object shape signal Fk And a post-processing unit 214. Here, the variable length decoding unit 213a and the shape decoding unit 213b constitute a shape decoder 213.

The decoding circuit 200b performs a variable length decoding process on the coded bit sequence Dg to generate a variable length decoded pixel value signal C'g. 222a and the variable-length decoded pixel value signal C ′ g
Are inverse quantized and inverse discrete cosine transformed (inverse D
CT) to generate a decoded pixel value signal Eg.

Next, the operation will be described. FIG. 13 is a flowchart showing an encoding process of an object shape signal by a conventional encoding circuit. In the encoding circuit 200a, the shape encoding device 210a performs an encoding process on the object shape signal Sk for each macroblock.

That is, when the object shape signal Sk is externally input to the shape encoding device 210a (step S
1) A shape coding process including arithmetic coding is performed on the object shape signal Sk by a shape coding unit 211a constituting the shape coding unit 211 in the device 210a, and the shape coding unit 211a Outputs a shape-encoded signal Ck (step S2). Subsequently, the shape encoded signal Ck is
Variable length encoding section 211 constituting the shape encoder 211
b, a variable-length encoding process is performed.
11b outputs the shape-encoding bit string Dk (step S3). At this time, the shape encoding section 211a outputs a locally decoded shape signal PEk obtained by locally decoding (information source decoding) the shape encoded signal Ck.

Thereafter, the shape post-processing unit 212 performs the above-described shape post-processing on the locally decoded shape signal PEk (step S4), and the shape post-processing unit 212 outputs a post-shape processing output Fk. (Step S5). By this shape post-processing, distortion of the object shape due to encoding of the object shape signal Sk becomes visually inconspicuous on the display screen.

In the encoding circuit 200a, the pixel value encoding unit 221a performs the DCT process and the quantization process on the pixel value signal Sg based on the post-shape processing output PFk. The obtained encoded pixel value signal Cg is subjected to a variable length encoding process by a variable length encoding unit 221b, and the variable length encoding unit 221b outputs a pixel value encoded bit sequence Dg.

FIG. 14 is a flowchart showing the decoding processing of the object shape signal by the conventional decoding circuit.
First, in the decoding circuit 200b, the shape decoding bit string Dk is decoded by the shape decoding device 210b.

That is, when the shape encoded bit string Dk is input to the shape decoding device 210b (step S1).
1), the shape-encoded bit sequence Dk is
The variable-length decoding unit 213a included in 10b performs variable-length decoding, and outputs a variable-length decoded shape signal C′k from the variable-length decoding unit 213a. Further, the variable-length decoded shape signal C ′ k is
b is subjected to shape decoding processing by the shape decoding unit 213b, and a decoded shape signal Ek is output from the shape decoding unit 213b (step S12). Thereafter, the decoded shape signal Ek is subjected to shape post-processing by the shape post-processor 214 (step S13), and the shape post-processor 2
14 outputs a shape post-processing output Fk (step S
14). The shape post-processing in step S13 and the shape output operation in step S14 are performed in steps S4 and S4 in the conventional object shape encoding process shown in FIG.
This is the same as the processing in S5.

In the decoding circuit 200b, when the pixel value encoded bit sequence Dg is input to the pixel value decoding device 220b, the pixel value encoded bit sequence Dg constitutes the pixel value decoding device 220b. Variable length decoding section 222a
Performs a variable-length decoding process, and the variable-length decoding unit 2
The variable length decoded pixel value signal C'g is output from 22a. Further, the variable length decoded pixel value signal C ′ g is output to the pixel value decoding unit 222 constituting the pixel value decoding device 220b.
At b, a decoding process is performed based on the shape post-processing output (decoded object shape signal) Fk from the shape post-processing unit 214, and a decoded pixel value signal Eg is output from the pixel value decoding unit 222b. Is output.

Here, the shape post-processing output PEk output from the shape post-processing unit 212 of the encoding circuit 200a,
Since the shape post-processing output Ek output from the shape post-processing unit 214 of the decoding circuit 200b matches, the encoding process and the decoding process of the pixel value signal can be correctly performed based on the object shape signal. .

[0031]

Generally, when a pixel value signal is encoded, the locally decoded shape signal PEk of the object shape signal is referred to in encoding and decoding of the pixel value signal. This is because the decoding process of the pixel value signal is performed based on the decoded object shape signal (decoded shape signal) Ek depending on the object shape and in the decoding process.

However, in the above-described conventional encoding processing of an image display signal, there is a problem that the number of encoded bits of a pixel value signal increases. That is, FIG. 17 is an explanatory diagram of the conventional shape post-processing. As in FIGS. 15 and 16, the inside of the object is represented by a dot-to-dot display area, and blocks to be subjected to pixel value signal encoding processing are indicated by hatching.

In the following description of the problem, 16 × 16
The same can be said for the macroblock Mb composed of pixels and the sub-block Sb composed of 8 × 8 pixels.
I will call it.

FIG. 17A shows a portion of the object shape obtained from the object shape signal, which is represented on a display area composed of four blocks B1 to B4. FIG.
(b) is a locally decoded shape signal PE for the object shape signal.
k, and the object shape obtained from the decoded shape signal Ek,
FIG. 17C shows a portion represented on a display area including four blocks B1 to B4. FIG. 17C shows a shape obtained by performing post-shape processing on the local decoded shape signal PEk and the decoded shape signal Ek. It shows a portion of an object shape obtained from post-processing outputs PFk and Fk, which is displayed on a display area including four blocks B1 to B4.

As can be seen from FIG. 17, shape post-processing is an important process for reducing visual coding distortion of an object shape signal, and the shape post-processing includes an object shape. The number of blocks may increase. That is, in FIG. 17 (a) and FIG.
Reference numeral 2 denotes a block located outside the object, but as shown in FIG.
2 will be located inside the object. In this case, a new encoding process of the pixel value signal corresponding to the upper right block B2 is required. This increases the number of bits in the encoding process of the pixel value signal, and deteriorates the encoding efficiency as a whole encoding method.

Referring to FIG. 18, the shape post-processing and the blocks located outside the object area occupied by the object on the display screen by the shape post-processing (hereinafter also referred to as “out-of-object blocks”). Will be briefly described as a mechanism that changes into an intra-object block (a block having at least an overlap with a part of the object region).

In the shape post-processing, it is determined whether the post-processing pixel Gx to be subjected to the shape post-processing is a pixel inside the object or a pixel outside the object, based on the post-processing pixel Gx and its surroundings. This is performed using the located peripheral pixels Ga to Gh. Here, the positional relationship between the peripheral pixels Ga to Gh and the post-processed pixel Gx is, as shown in FIG.
Gb and Gc are the upper left, right above, and
The peripheral pixels Gd, Ge are located on the upper right, right and left of the post-processed pixel Gx, respectively, and the peripheral pixels Gf, Gg,
Gh is located at the lower left, right below, and lower right of the post-processing pixel Gx, respectively. Note that X (0,0) to X
(2, 2) indicates the position coordinates of the post-processing pixel Gx as X (1, 2).
The position coordinates of each pixel in the case of 1) are shown.

In the shape post-processing, whether the post-processing pixel Gx is a pixel inside the object or a pixel outside the object is determined by the post-processing pixel Gx and the peripheral pixels Ga to For the nine pixels of Gh, the determination is performed based on the result of the majority decision which determines which of the pixels inside the object and the pixels outside the object is larger.

For example, as shown in FIG. 18B, peripheral pixels Ga, Gd, G located around the post-processing pixel Gx.
f, Gg, Gh are located inside the object, and the post-processing pixel Gx
When the peripheral pixels Gb, Gc, and Ge around the object are located outside the object, the post-processing pixel Gx and nine peripheral pixels Ga to Gh located around the pixel Gx have more pixels inside the object. . Therefore, in this case, the post-processed pixel Gx is obtained from the result of the majority decision based on the number of pixels inside the object (5) and the number of pixels outside the object (4), as shown in FIG.
As shown in (c), it is a pixel inside the object. FIG.
Medium and post-processed pixels Gx are indicated by dot display, and pixels inside the object are indicated by oblique lines.

Therefore, as shown in FIG. 18 (b), a block is formed between the post-processing pixel Gx and the peripheral pixel Gd on the left side thereof, and between the post-processing pixel Gx and the peripheral pixel Gg below the block. When the boundary is located, even if the block including the post-processed pixel Gx is an outer block, the block including the post-processed pixel Gx becomes an inner block by the shape post-processing on the object shape signal.

Further, when information indicating whether each block is a transparent block or an opaque block changes between the object shape signal and the locally decoded shape signal due to the shape signal encoding process, an arbitrary There is a problem that encoding processing of a digital image signal corresponding to an object having a shape becomes complicated.

That is, in the encoding system corresponding to the MPEG4 standard, when encoding a pixel value signal, a motion detecting process for detecting a motion vector and determining whether each block is an intra-coded block or an inter-coded block. The encoding mode is determined and the processing such as variable-length encoding is switched depending on whether the target block is transparent or opaque, and thus the encoding of the shape signal of the target block to be encoded is completed. Unless this is done, it cannot be determined whether the target block is transparent or opaque, and the encoding process cannot be performed on the pixel value signal of the corresponding target block.

Thus, there is a problem that, in parallel with the shape encoding process for each block, the pixel value encoding process for the corresponding block, especially the time-consuming motion vector detection process, cannot be performed. More specifically, in the encoding process of the pixel value signal, the arithmetic process in the encoding process is simplified by omitting processes such as detection of a motion vector for a block outside the object and compression of the pixel value.

For this reason, when the encoding process of the shape signal and the encoding process of the pixel value signal are performed in parallel, a block (image space) formed by the object shape signal due to encoding distortion caused by the shape encoding process. ) Is a block outside the object, but when a block (image space) formed by the locally decoded shape signal falls within the object, processing such as motion vector detection and pixel value compression of this block is performed. Since it is omitted, the image quality in this block is greatly deteriorated.

Further, in the encoding system corresponding to the MPEG4 standard, an advanced motion compensation (overlap motion compensation) in which a calculation method for motion compensation differs depending on whether an adjacent block is a transparent block or an opaque block. ) Has also been adopted. For this reason, when encoding processing of a shape signal and encoding processing of a pixel value signal are performed in parallel,
Due to the encoding distortion caused by the shape encoding process, the block (image space) formed by the locally decoded shape signal is out of the object even though the block (image space) formed by the object shape signal is an intra-object block. In such a case, the calculation in the overlap motion compensation is not correctly performed on the decoding side, and the image quality of the reproduced image is deteriorated.

The present invention has been made to solve the above-described problem, and it is possible to reduce an increase in the number of bits in encoding a pixel value signal while reducing visual distortion due to encoding processing of an object shape signal. It is an object to obtain a shape encoding method and a shape encoding device, a shape decoding method and a shape decoding device, and a data storage medium that can be suppressed.

Further, according to the present invention, a shape encoding process for a shape signal constituting an image signal corresponding to an object having an arbitrary shape and a pixel value encoding process for a pixel value signal constituting the image signal are performed in parallel. A shape encoding method and a shape encoding device, a shape decoding method and a shape decoding device capable of preventing image quality deterioration due to encoding distortion accompanying the shape encoding process, The purpose is to obtain a storage medium.

[0048]

Means for Solving the Problems According to the present invention (claim 1)
Is a shape encoding method for encoding an object shape signal corresponding to a predetermined object shape in an image display signal, wherein the object shape signal has a predetermined size on a display screen. Encoding processing is performed for each unit area to generate a shape-encoded bit string, and a signal in the middle of the encoding processing is decoded to generate a locally decoded shape signal. In accordance with the analysis result, the type of shape post-processing for smoothing the object shape obtained from the locally decoded shape signal with respect to the locally decoded shape signal, and the local decoding Determine at least one of whether or not to perform the shape post-processing on the shape signal, and apply the predetermined shape post-processing to the locally decoded shape signal, or without performing the shape post-processing. And outputs it as coding processing shape signal used for the encoding of signals other than the object shape signal in an image display signal.

According to a second aspect of the present invention, in the shape encoding method according to the first aspect, the determination regarding the shape post-processing according to the analysis result is performed by the local decoding shape signal corresponding to a certain unit area. When the unit area indicates that the unit area is located outside the object, the locally decoded shape signal subjected to the shape post-processing is also performed so as to be a signal indicating that the certain unit area is located outside the object. is there.

According to a third aspect of the present invention, in the shape encoding method according to the second aspect, the shape post-processing is performed such that an object shape obtained from the locally decoded shape signal becomes smoother. This is a process for converting the locally decoded shape signal using a filter having predetermined characteristics.

A shape encoding method according to the present invention (claim 4) is a shape encoding method for encoding an object shape signal corresponding to a predetermined object shape in an image display signal. Is downsampled to obtain a downsampled shape signal in which the number of coded pixels is reduced, and the downsampled shape signal is subjected to an encoding process for each unit area of a predetermined size on the display screen to form a shape encoded bit string. Is generated, a signal in the middle of the encoding process is decoded to generate a locally decoded shape signal, the locally decoded shape signal is upsampled, and an upsampled shape signal having an increased number of pixels is generated. Analyzing the upsampled shape signal for each of the unit regions, and obtaining the upsampled shape signal with respect to the upsampled shape signal according to the analysis result. Determine the type of shape post-processing to make the object shape smoother and at least one of whether or not to perform the shape post-processing on the up-sample shape signal, determine the up-sample shape signal, Is subjected to the above-mentioned predetermined shape post-processing, or without performing the above-mentioned shape post-processing, and is output as an encoding processing shape signal used for encoding a signal other than the object shape signal in the image display signal. is there.

A shape decoding method according to the present invention (claim 5) decodes a shape-encoding bit sequence corresponding to a predetermined object shape in an image display signal, and decodes the decoded object shape corresponding to the object shape. A shape decoding method for obtaining a signal, wherein the shape coding bit string is decoded for each unit area of a predetermined size on a display screen to generate a decoded shape signal,
Analyzing the decoded shape signal for each unit area, and performing shape post-processing on the decoded shape signal to make the object shape obtained from the decoded shape signal smoother according to the analysis result And at least one of whether or not to perform the shape post-processing on the decoded shape signal, and subject the decoded shape signal to the predetermined shape post-processing, The signal is output as a decoded object shape signal used for decoding a signal other than the object shape signal in the image display signal without performing post-processing.

According to a sixth aspect of the present invention, in the shape decoding method according to the fifth aspect, the determination regarding the shape post-processing in accordance with the analysis result is performed by the decoding shape signal corresponding to a certain unit area. When the region indicates that it is located outside the object, the decoded shape signal that has been subjected to the shape post-processing is also performed so as to be a signal that indicates that the certain unit region is located outside the object.

According to a seventh aspect of the present invention, in the shape decoding method according to the sixth aspect, the shape post-processing is performed such that an object shape obtained from the decoded shape signal becomes smoother. This is a process for converting the converted shape signal using a filter having a predetermined characteristic.

A shape decoding method according to the present invention (claim 8) decodes a shape coded bit string corresponding to a predetermined object shape in an image display signal, and decodes the decoded object shape corresponding to the object shape. A shape decoding method for obtaining a signal, wherein the shape coding bit sequence is decoded for each unit area of a predetermined size on a display screen to generate a decoded shape signal,
The decoded shape signal is up-sampled to obtain an up-sampled shape signal having an increased number of pixels, the up-sampled shape signal is analyzed for each unit area, and the up-sampled shape signal is analyzed in accordance with the analysis result. The type of shape post-processing for making the object shape obtained from the up-sample shape signal smoother, and at least one of performing or not performing the shape post-processing on the up-sample shape signal Then, the up-sampled shape signal is subjected to the predetermined shape post-processing or not subjected to the shape post-processing, and is used for decoding of a signal other than the object shape signal in the image display signal. It is output as a simulated object shape signal.

A shape coding apparatus according to the present invention (claim 9) is a shape coding apparatus for coding an object shape signal corresponding to a predetermined object shape in an image display signal. A shape code for performing a coding process for each unit area of a predetermined size on the display screen to generate a shape coded bit string, and decoding a signal in the middle of the coding process to generate a locally decoded shape signal Means,
The local decoded shape signal is analyzed for each of the unit regions, and according to the analysis result, for the locally decoded shape signal,
Determine at least one of a type of shape post-processing for making the object shape obtained from the locally decoded shape signal smoother, and whether or not to perform the shape post-processing on the locally decoded shape signal Shape analyzing means for performing, a shape post-processing means for performing a predetermined shape post-processing on the locally decoded shape signal based on an output of the shape analyzing means, and the shape encoding based on an output of the shape analyzing means. Means for selecting any of a locally decoded shape signal from the means and a locally decoded shape signal subjected to shape post-processing from the shape post-processing means, and an output of the signal selecting means for image display. The signal is output as an encoding shape signal used for encoding a signal other than the object shape signal in the application signal.

According to a tenth aspect of the present invention, there is provided a shape encoding apparatus for encoding an object shape signal corresponding to a predetermined object shape in an image display signal. Down-sampling means for obtaining a down-sampled shape signal in which the number of encoded pixels is reduced, and performing an encoding process on the down-sampled shape signal for each unit area of a predetermined size on the display screen Shape encoding means for generating a shape-encoded bit string, decoding the signal in the middle of the encoding process to generate a locally decoded shape signal, and up-sampling the locally decoded shape signal to reduce the number of pixels Up-sampling means for generating an increased up-sample shape signal; analyzing the up-sample shape signal for each unit area; The type of shape post-processing for making the object shape obtained from the up-sample shape signal smoother with respect to the up-sample shape signal, and whether or not to perform the shape post-processing on the up-sample shape signal Shape analysis means for determining at least one of the above; shape post-processing means for performing a predetermined shape post-processing on the up-sample shape signal based on the analysis result; and Signal selection means for selecting any of a sample shape signal and an upsampled shape signal subjected to shape post-processing from the shape post-processing means, and outputting the signal selection means as an object shape signal in an image display signal. The signal is output as a shape signal for encoding used for encoding other signals.

A shape decoding apparatus according to the present invention (claim 11) decodes a shape coded bit string corresponding to a predetermined object shape in an image display signal, and decodes the decoded object shape corresponding to the object shape. A shape decoding device for obtaining a signal, a shape decoding means for decoding the shape-encoded bit sequence for each unit area of a predetermined size on a display screen to generate a decoded shape signal, and Analysis is performed for each unit area, and a type of shape post-processing for smoothing an object shape obtained from the decoded shape signal with respect to the decoded shape signal according to the analysis result; Analysis means for determining at least one of whether or not to perform the shape post-processing on the decoded shape signal, and shape post-processing means for performing a predetermined shape post-processing on the decoded shape signal according to the analysis result. In accordance with the result of the analysis, a signal selecting means for selecting any one of a decoded shape signal from the shape decoding means and a decoded shape signal subjected to shape post-processing from the shape post-processing means, The output of the signal selection means
The signal is output as a decoded object shape signal used for decoding a signal other than the object shape signal in the image display signal.

A shape decoding apparatus according to the present invention (claim 12) decodes a shape-encoding bit string corresponding to a predetermined object shape in an image display signal, and decodes the decoded object shape corresponding to the object shape. A shape decoding device for obtaining a signal, a shape decoding means for decoding the shape-encoded bit sequence for each unit area of a predetermined size on a display screen to generate a decoded shape signal, and Up-sampling means for up-sampling and generating an up-sample shape signal having an increased number of pixels, analyzing the up-sample shape signal for each unit area, and responding to the analysis result for the up-sample shape signal. A type of shape post-processing for making the object shape obtained from the up-sample shape signal smoother; Shape analysis means for determining at least one of whether or not to perform the shape post-processing, shape post-processing means for performing a predetermined shape post-processing on the upsampled shape signal according to the analysis result, In response to,
Signal selecting means for selecting any of an up-sample shape signal from the up-sampling means and an up-sample shape signal subjected to shape post-processing from the shape post-processing means, and outputting the signal of the signal selecting means to an image. The signal is output as a decoded object shape signal used for decoding a signal other than the object shape signal in the display signal.

A data storage medium according to the present invention (claim 13) is a data storage medium storing a program for performing the shape encoding method according to claim 1 by a computer. The object shape signal is subjected to an encoding process for each unit area of a predetermined size on a display screen to generate a shape-encoded bit string, and a signal in the middle of the encoding process is decoded to generate a locally decoded shape signal. Generating, analyzing the locally decoded shape signal for each unit region, and, according to the analysis result, a smoother object shape obtained from the locally decoded shape signal with respect to the locally decoded shape signal A type of shape post-processing for determining, and a process of determining at least one of whether or not to perform the shape post-processing on the locally decoded shape signal, and The decoded shape signal is subjected to the above-mentioned predetermined shape post-processing or to the above-mentioned shape post-processing without being subjected to the above-mentioned shape post-processing. Output as a use shape signal.

A data storage medium according to the present invention (claim 14) is a data storage medium storing a program for performing the shape decoding method according to claim 5 by a computer. A process of decoding the shape-encoded bit string for each unit area of a predetermined size on the display screen to generate a decoded shape signal, analyzing the decoded shape signal for each unit area, and according to the analysis result, Type of shape post-processing for making the object shape obtained from the decoded shape signal smoother with respect to the decoded shape signal, and whether or not to perform the shape post-processing on the decoded shape signal The process of determining at least one of the above, and the decoded shape signal, subjected to the predetermined shape post-processing, or without performing the shape post-processing,
And outputting as a decoded object shape signal used for decoding a signal other than the object shape signal in the image display signal.

In the shape encoding method according to the present invention (claim 15), an object shape signal indicating the shape of an object, which is included in an image display signal forming an image space including an object having an arbitrary shape, is input. A shape encoding method for encoding the input object shape signal for each unit area of a predetermined size that divides the image space, wherein the object shape signal corresponding to the target unit area to be encoded is To perform a compression process at a predetermined compression rate to generate a compressed shape signal, and at this time, as a compression / decompression process for performing a decompression process corresponding to the compression process on the compressed shape signal to generate a locally decoded shape signal, By performing a plurality of compression / decompression processes having different compression ratios, and analyzing the locally decoded shape signals obtained by the respective compression / decompression processes having different compression ratios, the target unit included in the input object shape signal The information indicating that the area is located outside the object is retained in the locally decoded shape signal, and the compression / expansion processing with the highest compression rate is selected from the plurality of compression / expansion processing. , And encodes the compressed shape signal obtained by the selected compression / decompression process and outputs the encoded signal as a shape-encoded signal.

According to a sixteenth aspect of the present invention, in the shape encoding method according to the fifteenth aspect, the compression processing in the compression / decompression processing with different compression ratios for the object shape signal corresponding to the target unit area is performed by the object encoding method. For the shape signal, it is assumed that a down-sampling process having a predetermined down-sampling rate is included.
An up-sampling process corresponding to the down-sampling process is included for the compressed shape signal. Regardless of the generated coding distortion, the information indicating that the position of the target unit area is outside the object, which is included in the input object shape signal, is assumed to be held in the locally decoded shape signal. Things.

In the shape encoding method according to the present invention (claim 17), an object shape signal indicating the shape of an object, which is included in an image display signal forming an image space including an object having an arbitrary shape, is input. A shape encoding method for encoding the input object shape signal for each unit area of a predetermined size that divides the image space, wherein the object shape signal corresponding to the target unit area to be encoded is To generate a compressed shape signal by performing a compression process at a predetermined compression rate, and at this time, as a compression / decompression process for performing a decompression process corresponding to the compression process on the compressed shape signal to generate a locally decoded shape signal, The target unit included in the input object shape signal is obtained by performing a plurality of compression / decompression processes having different compression ratios and analyzing the locally decoded shape signals obtained by the respective compression / decompression processes having different compression ratios. Information indicating that the area is not located outside the object is retained in the local decoded shape signal, the compression / expansion processing with the highest compression rate is selected from among the plurality of compression / expansion processings, The compression shape signal obtained by the selected compression / decompression processing is encoded and output as a shape encoded signal.

According to the present invention (claim 18), in the shape encoding method according to claim 17, the compression processing in the compression / decompression processing with different compression ratios for the object shape signal corresponding to the target unit area is performed by the object For the shape signal, it is assumed that a down-sampling process having a predetermined down-sampling rate is included.
An up-sampling process corresponding to the down-sampling process is included for the compressed shape signal, and one compression / expansion process selected from the plurality of compression / expansion processes having different compression rates is performed by the down-sampling process and the up-sampling process. Regardless of the encoding distortion that occurs, information indicating that the target unit area is not located outside the object, which is included in the input object shape signal, is held in the locally decoded shape signal. It is what it was.

In the shape encoding apparatus according to the present invention (claim 19), an object shape signal indicating the shape of an object, which is included in an image display signal forming an image space including an object having an arbitrary shape, is input. A shape encoding device that encodes the input object shape signal for each unit area of a predetermined size that divides the image space, wherein the object shape signal corresponding to the target unit area to be encoded is Performs compression processing at a predetermined compression ratio to generate a compressed shape signal, and performs expansion processing corresponding to the compression processing on the compressed shape signal to generate a locally decoded shape signal. A plurality of compression / expansion units having different compression rates are provided. By analyzing the locally decoded shape signals output from the compression / expansion devices having different compression ratios, the target unit shape signal included in the input object shape signal is analyzed. Analysis means for specifying a compression / expansion unit for performing compression / expansion processing with the highest compression ratio, wherein information indicating that the region is located outside the object is held in the locally decoded shape signal; A signal selecting unit that receives an output of the analyzing unit and selects a compression shape signal as an output of the compressing / expanding unit designated by the analyzing unit from the outputs of the plurality of compressing / expanding units; And a coder for coding the compressed shape signal from the compression / decompression device and outputting the coded signal as a shape coded signal.

According to a twentieth aspect of the present invention, in the shape encoding apparatus according to the nineteenth aspect, each of the compression / decompression units includes a predetermined down-sampling rate for the object shape signal as part of the compression processing. Performing down-sampling processing, and performing up-sampling processing corresponding to the down-sampling processing on the compressed shape signal as a part of the decompression processing, wherein the analyzing unit performs the down-sampling processing and the up-sampling processing. Regardless of the encoding distortion caused by the above, information indicating that the target unit area is located outside the object, which is included in the input object shape signal, is held in the locally decoded shape signal. It is configured to specify a compression / expansion unit that performs compression / expansion processing.

In the shape encoding apparatus according to the present invention (claim 21), an object shape signal indicating the shape of an object, which is included in an image display signal forming an image space including an object having an arbitrary shape, is input. A shape encoding device that encodes the input object shape signal for each unit area of a predetermined size that divides the image space, wherein the object shape signal corresponding to the target unit area to be encoded is Performs compression processing at a predetermined compression ratio to generate a compressed shape signal, and performs expansion processing corresponding to the compression processing on the compressed shape signal to generate a locally decoded shape signal. A plurality of compression / expansion units having different compression rates are provided. By analyzing the locally decoded shape signals output from the compression / expansion devices having different compression ratios, the target unit shape signal included in the input object shape signal is analyzed. Analysis means for specifying a compression / expansion unit for performing compression / expansion processing with the highest compression rate, wherein information indicating that the region is not located outside the object is retained in the locally decoded shape signal; Means for selecting a compression shape signal as an output of the compression / decompression device specified by the analysis means from among the outputs of the plurality of compression / decompression processes, and a signal selected by the signal selection means. And a coder for coding the compressed shape signal from the compression / decompression device and outputting the coded signal as a shape coded signal.

According to a twenty-second aspect of the present invention, in the shape encoding apparatus according to the twenty-first aspect, each of the compression / decompression units includes a predetermined down-sampling rate for the object shape signal as part of the compression processing. Performing down-sampling processing, and performing up-sampling processing corresponding to the down-sampling processing on the compressed shape signal as a part of the decompression processing, wherein the analyzing unit performs the down-sampling processing and the up-sampling processing. Irrespective of the encoding distortion caused by the compression, the information included in the input object shape signal and indicating that the target unit area is not located outside the object is held in the locally decoded shape signal. It is configured to specify a compression / decompression device that performs decompression processing.

In the data storage medium according to the present invention (claim 23), an object shape signal indicating the shape of an object, which is included in an image display signal forming an image space including an object having an arbitrary shape, is input, A data storage medium storing a program for causing a computer to perform a shape encoding process of encoding the input object shape signal for each unit area of a predetermined size that divides the image space. A program is executed by a computer to generate a compressed shape signal by performing a compression process at a predetermined compression rate on an object shape signal corresponding to the target unit area to be encoded. As compression / decompression processing for performing decompression processing corresponding to the processing to generate a locally decoded shape signal,
By performing a first process of performing a plurality of compression / decompression processes having different compression ratios, and analyzing a locally decoded shape signal obtained by each of the compression / decompression processes having different compression ratios, an object included in the input object shape signal is obtained. The information indicating that the unit area is located outside the object is held in the locally decoded shape signal, and the compression / expansion processing with the highest compression rate is selected from the plurality of compression / expansion processing. And a third process of encoding a compressed shape signal obtained by the selected compression / decompression process and outputting the encoded shape signal as a shape-encoded signal.

In the data storage medium according to the present invention (claim 24), an object shape signal indicating the shape of an object, which is included in an image display signal forming an image space including an object having an arbitrary shape, is input, A data storage medium storing a program for causing a computer to perform a shape encoding process of encoding the input object shape signal for each unit area of a predetermined size that divides the image space. A program is applied to a computer to generate a compressed shape signal by performing a compression process at a predetermined compression rate on an object shape signal corresponding to the target unit region to be encoded, and the compression process is performed on the compressed shape signal. A first process of performing a plurality of compression / decompression processes having different compression ratios as a compression / decompression process of generating a locally decoded shape signal by performing a decompression process corresponding to the compression ratio; By analyzing the locally decoded shape signal obtained by each of the compression / decompression processes, information indicating that the target unit area is not located outside the object, which is included in the input object shape signal, The second processing for selecting the compression / expansion processing having the highest compression ratio, which is held by the shape signal, from the plurality of compression / expansion processing, and the compression / shape signal obtained by the selected compression / expansion processing. It is configured to perform a third process of encoding and outputting as a shape-encoded signal.

[0072]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described below with reference to FIGS. Embodiment 1 FIG. FIG. 1 is a block diagram illustrating a configuration of a shape encoding device according to Embodiment 1 of the present invention, and FIG. 2 is a diagram illustrating a flowchart of an encoding process performed by the shape encoding device. The first embodiment is described in claims 1 to 3, 9
It corresponds to. In the figure, 110a is MPE
This is a shape encoding device of the first embodiment based on G4, which corresponds to the shape encoding device 210a in the conventional encoding circuit 200a shown in FIG.

[0073] Shape coding apparatus 110a of the present embodiment
Represents a predetermined unit area (the macroblock or subblock described above) on the display screen with respect to the object shape signal Sk.
The encoding process is performed every time, and similarly to the conventional shape encoding device 210a, the object shape signal Sk included in the image display signal of the object is received, and the shape encoding process and the variable length encoding process are performed on the object shape signal Sk. To generate a shape-encoded bit string Dk, and to decode the signal in the middle of the above-mentioned encoding process (the signal obtained by subjecting the object shape signal Sk to the information source encoding process once) to generate a locally decoded shape signal PEk. And a shape post-processor 212 that performs post-shape processing on the locally decoded shape signal PEk to generate a post-shape output PFk.

Then, the shape encoding device 1 of the present embodiment
10a is an open / close switch 12 provided between the output of the shape encoder 211 and the input of the shape post-processor 212.
And output of the shape post-processing 212 (shape post-processing output) PF
and a selection switch 13 for selecting one of k and the output (locally decoded shape signal) PEk of the shape encoder 211. Further, the shape encoding device 110a analyzes the locally decoded shape signal PEk for each of the unit regions, and determines whether or not to perform the shape post-processing on the locally decoded shape signal PEk. It has a shape analyzer 11 for controlling each of the switches 12 and 13 by a switch control signal SWc corresponding to the result. The selection switch 13 outputs a reference signal for encoding (shape signal for encoding) Rc as a selection output. Is output.

Next, the operation will be described. When the object shape signal Sk is input to the present shape encoding device 110a (step S1), the shape encoder 211 outputs the object shape signal Sk.
Is shape-coded (step S2), and the shape-coded signal is subjected to variable-length coding to form a shape-coded bit sequence D
k is output from the shape encoder 211 to the outside (step S3). At this time, in the shape encoder 211,
A process of shape-decoding the shape-encoded signal is performed, and the shape encoder 211 also outputs a locally decoded shape signal PEk obtained by shape-decoding the shape-encoded signal.

When the locally decoded shape signal PEk is input to the shape analyzer 11, the shape analyzer 11 analyzes the locally decoded shape signal PEk, and analyzes the locally decoded shape signal PEk. Then, it is determined whether or not to perform the shape post-processing (step S4a). At this time, the shape analyzer 11
From the switch control signal SWc corresponding to the above determination result.
Is output, and the open / close switch 12 outputs the switch control signal S
The selector switch 13 opens and closes according to the switch control signal SWc.
One of k and the output PFk of the shape post-processor 212 is selected.

That is, as a result of the analysis processing by the shape analyzer 11, if the block to be coded includes at least a part of the object (if the block is located within the object), the switch The open / close switch 12 is turned on by the control signal SWc, and the selection switch 13 selects the output of the shape post-processor 212. As a result, the shape encoder 21
1 is subjected to shape post-processing on the locally decoded shape signal PEk (step S4b).
The shape post-processing output PFk from No. 12 is output as the above-mentioned reference signal for encoding process Rc.

On the other hand, when the block to be encoded does not include an object (when the block is located outside the object), the open / close switch 12 is turned off by the switch control signal SWc. The selection switch 13 selects the output of the shape encoder 211. Thus, the locally decoded shape signal PEk output from the shape encoder 211 is output as it is as the encoding reference signal Rc.

Hereinafter, the shape post-processing will be described in detail. FIG. 3 shows that blocks located outside the object area occupied by the object on the display screen (hereinafter, also referred to as “outside object blocks”) are overlapped with blocks inside the object (at least overlap with at least a part of the object area). This is an example in which the block changes to 3, the same reference numerals as those in FIGS. 17 and 18 denote the same components as those in these drawings.

In this shape post-processing, it is determined whether the post-processing pixel Gx to be subjected to the shape post-processing is a pixel inside the object or a pixel outside the object, based on the post-processing pixel Gx and its surroundings. Using the located peripheral pixels Ga to Gh, as in the case of the conventional shape post-processing described with reference to FIG. 18, nine pixels of the post-processed pixel Gx and the peripheral pixels Ga to Gh located around the pixel Gx are located inside the object. This is performed based on the result of a majority decision that determines which of the pixels outside the object is larger.

Therefore, as shown in FIG. 3 (a), the block between the post-processed pixel Gx and the peripheral pixel Gd on the left side thereof and between the post-processed pixel Gx and the peripheral pixel Gg below the block Gx. When the boundary is located, even if the block B2 including the post-processed pixel Gx is an out-of-object block, as shown in FIG. The block B2 including Gx is an intra-object block. In FIG. 3, the blocks indicated by hatching are:
This is a block to be subjected to the pixel value encoding process, and a dotted portion indicates the inside of the object.

On the other hand, in Embodiment 1 of the present invention, the locally decoded shape signal PEk is analyzed by shape analyzer 11, and the block to be coded (encoded block) is an object. It is determined whether or not the block is an outer block, and only when the block to be coded is not an outer block, shape post-processing is performed on the locally decoded shape signal PEk.

Specifically, in the first embodiment, FIG.
As shown in (b), the locally decoded shape signal PEk corresponding to the upper right block B2 indicates that the upper right block B2 is an extra-object block, and the upper left, lower left, and lower right blocks B1, B3, When the locally decoded shape signal PEk corresponding to B4 indicates that these blocks are intra-object blocks, the post-shape processing is not performed on the locally decoded object shape of the upper right block B2, and other blocks are not processed. B1,
Shape post-processing is performed only on the locally decoded shape signals PEk of B3 and B4. For this reason, the locally decoded shape signal PEk of the other blocks is subjected to shape post-processing, and becomes a shape post-processed output PFk indicating a smooth object shape, and this becomes the encoding processing reference signal Rc. Also, block B on the upper right
The 2 locally decoded shape signal PEk is not subjected to shape post-processing, and is directly used as the encoding process reference signal Rc.
Therefore, the encoding reference signal Rc corresponding to the upper right block does not indicate an intra-object block.

Here, an object shape (FIG. 3 (b)) obtained from a signal obtained by applying a shape encoding process to the locally decoded shape signal PEk corresponding to the above four blocks, and a locally decoded shape of a necessary block When comparing the signal PEk with the object shape (FIG. 4 (c)) obtained from the post-shape-processed signal, the object shape shown in FIG. 4 (c) is discontinuous at the upper right block boundary. Such distortion is visually acceptable. FIG. 4 (a) shows the same one as FIG. 17 (a).

As described above, the conventional shape encoding device 21
In the case of 0a, the block outside the object may become the block inside the object due to the shape post-processing, but in the present invention,
The block outside the object does not become the block inside the object by the shape post-processing. In other words, the blocks necessary for encoding the pixel value signal never increase due to the shape post-processing.

As a result, an increase in the number of coded bits required for coding the pixel value signal can be prevented while removing visual distortion of the object shape in the shape post-processing. In addition, as for the locally decoded shape signal PEk of the intra-object block, the shape post-processing is performed as in the related art, so that a visually favorable object shape can be obtained.

Embodiment 2 FIG. 5 is a block diagram showing a configuration of a shape encoding device according to Embodiment 2 of the present invention. The second embodiment corresponds to claims 4 and 10. In the figure, reference numeral 120a denotes a shape encoding device according to the second embodiment, and the same reference numerals as those in FIG. 1 denote the same components as those in the shape encoding device 110a according to the first embodiment.

That is, the shape encoding device 120a according to the second embodiment is provided before the shape encoder 211, downsamples the object shape signal Sk input from the outside, and calculates the number of encoded pixels. A down-sampler 14a that generates a down-sampled shape signal Skd in which
An up-sampler 14b provided at a stage subsequent to the shape encoder 211 to up-sample the locally decoded shape signal PEkd from the shape encoder 211 to generate an up-sampled shape signal PEku having an increased number of pixels. And Here, the shape encoder 211 outputs the downsampled shape signal S from the downsampler 14a.
kd is subjected to a shape encoding process and a variable-length encoding process to generate a shape-encoded bit sequence Dkd, and a signal in the middle of the encoding process (that is, a signal obtained by performing a shape encoding process on the down-sampled shape signal Skd) Is subjected to shape decoding to generate a locally decoded shape signal PEkd.

The shape encoding device 120a has a shape analyzer 11 and a shape post-processing device 212, like the shape encoding device 110a of the first embodiment. The up-sample shape signal PEku from the up-sampler 14b is analyzed for each unit area to determine whether or not to perform the shape post-processing on the up-sample shape signal PEku. The post-processor 212 outputs the up-sampled shape signal PE
Ku is subjected to the same shape post-processing as in the first embodiment. Also, the shape encoding device 120
a is the up-sampler 14b and the shape post-processor 21
2, an open / close switch 15 provided between the control unit 2 and a selection switch 16 for selecting any one of the output of the shape post-processing unit 212 and the output of the up-sampler 14b.
These switches are controlled by a switch control signal SWc from the shape analyzer 11 based on the determination result. And the selection switch 1
6, one of the output of the shape post-processor 212 and the output of the up-sampler 14 b outputs the reference signal R for encoding processing.
This is output as cu.

The shape encoding device 120 according to the second embodiment
The operation a is performed by the shape encoding device 110 according to the first embodiment.
The operation is basically the same as that of a. The input object shape signal Sk is down-sampled and supplied to the shape encoder 211, and the output of the shape encoder 211 is up-sampled to obtain a shape analyzer. Embodiment 11 is different from Embodiment 1 only in that it is supplied to the shape post-processor 212 and the selection switch 16.

In the second embodiment having such a configuration, in addition to the effects of the first embodiment, the down-sampler 14a preceding the shape encoder 211 outputs the object shape signal Sk.
Is downsampled, and the shape encoder 211 encodes the object shape signal whose pixel value has been reduced by the downsampling, so that the number of bits of the shape encoded bit sequence Dkd can be reduced.

Also, in the second embodiment, the locally decoded shape signal PEkd from the shape encoder 211 is up-sampled by the subsequent up-sampler 14b. The possibility that the block outside the object becomes the block inside the object is higher. Therefore, in the second embodiment, the shape analyzer 11 analyzes the upsampled locally decoded object shape (upsampled shape signal) PEku, and the shape post-processor (filter) 212 based on the analysis result. The configuration for switching the presence / absence of processing is more significant than in the first embodiment.

In the first and second embodiments, the shape encoders 110a and 120a determine the shape of the locally decoded shape signal PEk or PEkd according to the analysis result of the locally decoded shape signal PEk or PEkd. Although the switching of whether or not to perform the processing is shown, the switching of the post-processing is not limited to the switching of whether or not to perform the processing.

For example, a configuration is provided in which a plurality of shape post-processors having different processing contents are provided, and a shape analyzer is obtained as an analysis result by a number corresponding to the number of the plurality of shape post-processors. And, according to the analysis result, select an output of a shape post-processing unit suitable for this, and when a specific analysis result is obtained, do not perform the shape post-processing on the locally decoded shape signal. Is also good.

In the second embodiment, as the shape analyzer 11, based on the output PEku of the up-sampler 14b, the output of the up-sampled shape signal PEk
Although the case where it is determined whether or not to perform the shape post-processing on u is described above, the shape analyzer performs the local decoding shape signal PE from the shape encoder 211 in the preceding stage of the up-sampler 14b.
kd, the up-sampled shape signal PEku
It may be configured to determine whether or not to perform shape post-processing on. In this case, the amount of data processing in the shape analyzer 11 can be reduced.

Embodiment 3 FIG. 6 is a block diagram showing a configuration of a shape encoding device according to Embodiment 3 of the present invention. Note that the third embodiment is similar to the first embodiment.
This corresponds to claims 1 to 3 and 9. In the figure,
Reference numeral 130a denotes a shape encoding device according to the third embodiment, and the same reference numerals as those in FIG.
The same thing as a is shown. The shape encoding device 130a according to the third embodiment includes a specific shape signal BG indicating that the block to be encoded is a block outside the object.
In addition to having the specific shape signal generator 17 for generating k, the shape post-processing unit 212 is used instead of the selection switch 13 in the shape encoding device 110a of the first embodiment.
And an output BGk from the specific shape signal generator 17 are provided.

Here, the shape analyzer 11 is
When it is determined based on the locally decoded shape signal PEk of the block to be coded that there are almost no pixels constituting the object in this block, for example, three or more pixels, the open / close switch 12 is turned on and off. It turns off and generates a switch control signal SWc that causes the selection switch 18 to select the output of the specific shape signal generator 17. Otherwise, the open / close switch 12 is turned on and the selection switch 18 is turned on.
To generate a switch control signal SWc that causes the output of the shape post-processor 212 to be selected.

That is, in the third embodiment, the selection switch 18 allows the object shape signal after shape post-processing (shape post-processing output) PFk and the pixels in the block to be coded to be all outside the object. This embodiment differs from the first embodiment only in that any one of the specific shape signals BGk indicating a pixel is selected, and the other configuration is the same as that of the shape encoding device 110a of the first embodiment.

In the third embodiment having such a configuration, when the shape analyzer 11 determines that there is almost no pixel in the object in the block to be encoded, the selection switch The output of the specific signal generator 17 is selected by 18 and a specific shape signal BGk indicating that the block to be subjected to the encoding process is a non-object block is output as the encoding process reference signal Rc. Therefore, when encoding the pixel value signal, it is possible to further reduce the number of blocks that are determined to be intra-object blocks that need to be encoded as compared with the one based on the encoding reference signal Rc in the first embodiment. Thus, there is an effect that the number of bits of pixel value encoding can be further reduced by reducing the number of blocks of pixel values requiring encoding.

In the third embodiment, the shape encoding device 130a switches whether or not to perform post-shape processing on the locally decoded shape signal PEk according to the analysis result of the locally decoded shape signal PEk. However, as described in the first and second embodiments, the switching of the shape post-processing is not limited to the switching of whether or not to perform the processing.

For example, a configuration is provided in which a plurality of shape post-processors having different processing contents are provided, and the shape analyzer is used as an analysis result to obtain different results corresponding to the number of the plurality of shape post-processors. According to the analysis result, an output of the shape post-processor suitable for this is selected, and when a specific analysis result is obtained, the selection switch 1
8, the output of the specific shape signal generator 17 may be selected so that the shape post-processing is not performed.

Embodiment 4 FIG. 7 is a block diagram showing a configuration of a shape decoding apparatus according to Embodiment 4 of the present invention.
FIG. 4 is a flowchart showing a decoding process by the shape decoding device. The fourth embodiment is described in claim 5
7 and 11, respectively. 12, the same reference numerals as those in FIG. 12 denote the same parts as in the conventional shape decoding apparatus 210b, and 110b denotes a shape decoding apparatus according to the fourth embodiment based on MPEG4, which is shown in FIG. This corresponds to the shape decoding device 210b in the conventional decoding circuit 200b.

[0103] Shape coding apparatus 110 of the fourth embodiment
b performs a decoding process on the variable-length coded bit string Dk for each predetermined unit area (the above-described macroblock or sub-block) on the display screen. As in the case of the shape encoding device 210b, a shape decoder 213 that receives a variable-length coded bit string Dk and performs a variable-length decoding process and a shape decoding process to generate a decoded shape signal Ek; Shape post-processing is performed on the shape signal Ek to output a shape post-processing output Fk.
And a shape post-processing unit 214 for generating

Then, the shape decoding apparatus 1 of the present embodiment
10b is an on / off switch 20 provided between the output of the shape decoder 213 and the input of the shape post-processor 214.
And the output of the shape post-processing 214 (shape post-processing output) Fk
And output (decoded shape signal) Ek of shape decoder 213
And a selection switch 21 for selecting one of the two. Further, the shape decoding device 110b analyzes the decoded shape signal Ek for each block, which is an encoding processing unit, and determines whether or not to perform the shape post-processing on the decoded shape signal Ek. , A switch control signal S corresponding to the determination result.
It has a shape analyzer 19 for controlling each of the switches 20 and 21 according to Wc, and a reference signal for decoding processing (decoded object shape signal)
Is output.

Next, the operation will be described. When the variable length coded bit sequence Dk is input to the present shape decoding device 110b (step S11), the shape decoder 213 performs variable length decoding and shape decoding on the variable length coded bit sequence Dk. (Step S12), and outputs a decoded shape signal Ek. The variable length decoded bit string Dk
However, the shape encoding device 110a according to the first embodiment shown in FIG.
, The decoded shape signal Ek output from the shape decoder 213 is converted to the locally decoded shape signal PEk in the shape coding device 110a of the first embodiment. It is equivalent.
Therefore, the subsequent processing is exactly the same as the processing in the shape encoding device 110a according to the first embodiment.

In brief, the decoded shape signal E
When k is input to the shape analyzer 19, the shape analyzer 19 analyzes the decoded shape signal Ek and determines whether or not to perform shape post-processing on the decoded shape signal Ek. Is performed (step S13a). At this time, the shape analyzer 19
From the switch control signal SWc corresponding to the above determination result.
Is output, and the open / close switch 20 outputs the switch control signal SW.
The selection switch 21 opens and closes according to the switch control signal SWc, and selects one of the output of the shape decoder 213 and the output of the shape post-processor 214 in accordance with the switch control signal SWc.

That is, as a result of the analysis processing by the shape analyzer 19, if the block to be coded includes at least a part of the object (when the block is located in the object), the switch The open / close switch 20 is turned on by the control signal SWc, and the selection switch 21 selects the output of the shape post-processing unit 214. Thereby, the shape decoder 21
3 is subjected to shape post-processing on the decoded shape signal Ek (step S13b), and the shape post-processor 214
Is output as the decoding processing reference signal Re (step S14).

On the other hand, when the block to be coded does not include an object (when the block is located outside the object), the open / close switch 20 is turned off by the switch control signal SWc. The selection switch 21 selects the output of the shape decoder 213. As a result, the decoded shape signal Ek output from the shape decoder 213 is output as it is as the decoding processing reference signal Re (step S14).

Here, the decoded shape signal E obtained by decoding the shape encoded bit string Dk in step S12 is obtained.
k is the same as the locally decoded shape signal PEk obtained by locally decoding the shape-encoded signal of the object shape signal Sk in step S2 of the first embodiment. The output decoding reference signal Re
It is evident that the signal coincides with the encoding process reference signal Rc output in step S5 of the first embodiment. Therefore, the decoding process of the pixel value signal can be correctly performed by the correct decoding process reference signal Re that matches the coding process reference signal Rc in the coding circuit.

Embodiment 5 FIG. FIG. 9 is a block diagram showing a configuration of a shape decoding apparatus according to Embodiment 5 of the present invention. The fifth embodiment corresponds to claims 8 and 12. In the figure, reference numeral 120b denotes a shape decoding device for decoding the shape-encoded bit sequence Dkd encoded by the shape encoding device 120a of the second embodiment, and the same reference numerals as those in FIG. This shows the same one as the encoding device 110b.

That is, the shape encoding device 120b according to the fourth embodiment is provided at the subsequent stage of the shape decoder 213, and the decoded shape signal Ekd from the shape decoder 213 is output.
Up-sampling device 22 which up-samples the signal to generate an up-sampled shape signal Eku having an increased number of pixels
have. Here, the shape decoder 213 receives the variable-length encoded bit sequence Dk from the shape encoding device 120a.
d is subjected to variable-length decoding and shape decoding to generate a decoded shape signal Eku.

The shape decoding device 120b has a shape analyzer 19 and a shape post-processor 214, like the shape encoding device 110b of the fourth embodiment. The up-sampled decoded shape signal Eku from the up-sampler 14b is analyzed for each block to determine whether or not to perform the shape post-processing on the up-sampled decoded shape signal Eku.
Further, the shape post-processor 214 is configured to perform the same post-processing as in the fourth embodiment on the up-sampled decoded shape signal Eku. Further, the shape encoding device 120b includes an open / close switch 23 provided between the output of the up-sampler 22 and the input of the shape post-processor 214, and the output of the shape post-processor 214 and the up-sampler 22. Switch 24 for selecting one of the outputs
These switches are controlled by a switch control signal SWc from the shape analyzer 19 based on the determination result.

The shape decoding apparatus 120 according to the fifth embodiment
The operation b is performed by the shape decoding apparatus 110 according to the fourth embodiment.
b is basically the same as the operation of FIG.
The third embodiment differs from the fourth embodiment only in that the output of sample No. 3 is up-sampled and supplied to the shape analyzer 19, shape post-processor 214 and selection switch 24.

In the configuration decoding apparatus 120b of the fifth embodiment having such a configuration, the variable length coded bit string Dk
d is the shape encoding device 120 according to the second embodiment shown in FIG.
a, the decoded shape signal Ekd output from the shape decoder 213 is the shape coding device 120a according to the second embodiment.
And the local decoded shape signal PEkd, which is the output of the shape encoder 211 in FIG. Therefore, shape analysis processing and shape post-processing for the decoded shape signal Ekd are performed in exactly the same manner as in the second embodiment.

As a result, the shape decoding device 120b according to the fifth embodiment has the shape encoding device 1 according to the second embodiment.
It is possible to obtain a correct decoding processing reference signal Re that matches the encoding processing reference signal Rc obtained in 20a,
This makes it possible to correctly perform the decoding processing of the pixel value signal.

In the fourth and fifth embodiments, the shape decoding devices 110b and 120b operate as the decoded shape signals E
k, Ekd, the decoded shape signal Ek
Alternatively, the switching of whether or not to perform the shape post-processing on Ekd is described, but the switching of the shape post-processing is not limited to the switching of whether or not to perform the processing.

For example, a configuration is provided in which a plurality of shape post-processors having different processing contents are provided, and the shape analyzer is obtained as an analysis result by a number corresponding to the number of the plurality of shape post-processors. According to the analysis result, an output of the shape post-processing device suitable for this may be selected, and the shape post-processing may not be performed on a specific analysis result.

In the fifth embodiment, based on the output Eku of the up-sampler 22, the shape analyzer 19 determines whether or not to perform post-shape processing on the output up-sample shape signal Eku. However, the shape analyzer determines whether or not to perform shape post-processing on the up-sampled shape signal Eku based on the decoded shape signal Ekd from the shape encoding 214 in the preceding stage of the up-sampler 22. It is good also as a structure which performs. In this case, the data processing amount in the shape analyzer 19 can be reduced.

Embodiment 6 FIG. FIG. 10 is a block diagram showing a configuration of a shape decoding device according to Embodiment 6 of the present invention. The sixth embodiment corresponds to claims 5 to 7 and 11. In the drawing, reference numeral 130b denotes a shape encoding device according to the third embodiment, and the same reference numerals as those in FIG. 7 denote the same components as those in the shape encoding device 110b according to the fourth embodiment. Shape decoding device 130b according to the sixth embodiment
Has a specific shape signal generator 25 that generates a specific shape signal BGk indicating that a block to be decoded is a non-object block, and has the fourth embodiment.
Is provided with a selection switch 26 for selecting either the output Fk of the shape post-processor 214 or the output BGk of the specific shape signal generator 25 in place of the selection switch 21 in the shape decoding device 110b.

Here, the shape analyzer 19 is
Decoded shape signal Ek of the block to be encoded
, When it is determined that there are almost no pixels constituting the object in this block, for example, three or more pixels, the open / close switch 20 is turned off, and the output of the specific shape signal generator 25 is output to the selection switch 26. Is generated. Otherwise, the open / close switch 20 is turned on, and the switch control signal Sc causes the selection switch 26 to select the output of the shape post-processing unit 214.
It is configured to generate Wc.

That is, in the sixth embodiment, the object shape signal (shape post-process output) Fk after shape post-processing and the pixels in the block to be decoded are all outside the object by the selection switch 26. The fourth embodiment is different from the fourth embodiment only in that any one of the specific shape signals BGk indicating a pixel is selected, and the other configurations are different from the fourth embodiment.
Is the same as the shape encoding device 110b.

Shape decoding apparatus 130 according to the sixth embodiment
In b, if the shape encoding bit sequence Dk input to the shape decoder 213 is the shape encoding bit sequence Dk created by the shape encoding device 130a of the third embodiment shown in FIG. The decoded shape signal Ek output from the encoder 213 matches the locally decoded object shape signal PEk output from the shape encoder 211 in the shape encoder 130a according to the third embodiment.

Therefore, the shape analysis processing and the shape post-processing for the decoded shape signal Ek are performed in exactly the same manner as in the third embodiment. As a result, the shape decoding device 130b according to the sixth embodiment generates a correct decoding process reference signal Re that matches the encoding process reference signal Rc obtained by the shape encoding device 130a according to the third embodiment. Thus, the decoding process of the pixel value signal can be performed correctly.

In the sixth embodiment, the shape encoder 130b switches whether or not to perform shape post-processing on the decoded shape signal Ek according to the analysis result of the decoded shape signal Ek. As described above, as described in the fourth and fifth embodiments, the switching of the shape post-processing is not limited to the switching of whether or not to perform the processing.

For example, a configuration is provided in which a plurality of shape post-processors having different processing contents are provided, and the shape analyzer is provided with different results corresponding to the number of the plurality of shape post-processors as analysis results. According to the analysis result, the output of the shape post-processor suitable for this is selected, and for the specific analysis result, the output BGk of the specific shape signal generator 25 is selected by the selection switch 26. The shape post-processing may not be performed.

Embodiment 7 FIG. FIG. 19 is a diagram for describing a shape encoding device according to Embodiment 7 of the present invention.
The seventh embodiment corresponds to claims 15 to 24. Here, an image processing apparatus 1000 that encodes a digital image signal forming an image space including an object having an arbitrary shape for each unit area (block) that divides the image space will be described as an example. The digital image signal includes a shape signal indicating the shape of the object by binary information, and a pixel value signal for displaying the object in a gray scale. It comprises a shape encoding device 1100 that performs the encoding process, and a pixel value encoding device 1200 that performs the encoding process of the pixel value signal.

Shape coding apparatus 110 according to the seventh embodiment
0 indicates that the shape signal Sk corresponding to each frame supplied to the input terminal 1101a is represented by 16 as a unit of encoding processing.
A blocker 1101 that divides the block to correspond to a block (image space) composed of × 16 pixels and outputs shape data corresponding to each block, and a target block (a block to be encoded) from the blocker 1101 ) Includes first to third shape encoding units (compression / expansion means) 1100a to 1100c for performing compression / expansion processing on the shape data Bk.

For example, the second shape encoding unit 1100
b is 1/1 with respect to the shape data Bk of the target block.
A second down-sampler (D) that performs down-sampling processing at a ratio of
S2) 1111b, and a second shape encoder (ENC2) 1112b that performs shape encoding on the shape sample data Skd2 to generate compressed shape data Ck2. The second shape encoder 1112b constitutes a data compressor. Further, the second shape encoding unit 1100b performs a shape decoding process on the shape compressed data Ck2 to generate restored sample data PEkd2, and a second shape decoder (DEC2) 1113b; PE
a second upsampler (US2) 1114b for performing upsampling processing at a ratio of 1/2 to kd2 to generate shape-extended data PEku2; A data decompressor is constituted by two up-samplers (US2) 1114b. Further, the second shape encoding unit 1100b
Is a second shape post-processing device 11 that performs the same processing as the shape post-processing in the first embodiment on the shape extension data PEku2 to generate the shape post-processing data PFku2.
15b.

Note that the first and third shape encoding units 11
00a and 1100c correspond to the second shape encoding unit 110
0b is substantially the same as that of the first
Third downsampler (DS1, 3) 1111a, 1
111c and the first and third shape encoders (ENC1,
3) 1112a, 1112c, first and third shape decoders (DEC1, 3) 1113a, 1113c, and first and third upsamplers (US1, 3) 1114
a, 1114c and first and third shape post-processors 1115
a, 1115c. Here, the first down-sampler 1111a performs down-sampling processing at a ratio of 1/1 to the shape data Bk of the target block, and the first up-sampler 1114a performs down-sampling on the restored shape data PEkd1. On the other hand, the up-sampling process is performed at a ratio of 1/1. However, the first down-sampler 1111a and the second up-sampler 1114a, as an actual circuit configuration, directly output signals from the preceding stage. Further, the third down-sampler 1111c performs down-sampling processing on the shape data Bk of the target block at a ratio of 1/4, and the third up-sampler 1114c includes
The up-sampling process is performed on the restored shape data PEkd3 at a ratio of 1/4. The above first and third
The other devices in the shape encoding units 1100a and 1100c have exactly the same configuration as that in the second shape encoding unit 1100b.

Further, the shape encoding apparatus 1100 according to the seventh embodiment includes first to third shape encoding sections 1100a to 1100a.
A data selection switch 1130 for selecting one of the compressed shape data Ck1 to Ck3 output from the shape encoders 1112a to 1112c of the block 1100c based on the selection control signal SWc2, and the shape data Bk of the target block
And the corresponding first to third shape encoding units 110
Output data (shape post-processing data) P of 0a to 1100c
By comparing Fku1 to PFku3, the contents of the block position information included in the shape data Bk of the target block and indicating whether the target block is located inside the object or outside the object is unchanged. The shape post-processing data held by the above-mentioned shape post-processing data PFku
1 to PFku3, and a signal corresponding to the result of the determination is used as the selection control signal SWc2.
And an analyzer 1120 for supplying 130. Here, the data selection switch 1130 is connected to the analyzer 11.
20 based on the selection control signal SWc2 indicating the result of the discrimination, the compression / decompression processing with the highest compression ratio among the shape encoding units that output the shape post-processing data holding the contents of the block position information is performed. The configuration is such that the compressed shape data output from what is performed is selected. The block located outside the object is a block in which the pixels constituting the block do not include the pixels constituting the object (transmission block), and the target block located inside the object is the pixel constituting the block. Are those that partially or completely include the pixels that constitute the object (opaque blocks).

Then, the shape encoding device 1100
A variable length coder (VLC) 1140 that performs variable length coding on the shape compression data SCk selected by the data selection switch 1130 and outputs it as a shape coded signal (shape bit stream) Dk.

The shape encoding apparatus 1100 receives the output Bk of the blocking unit 1101 and forms the target block together with block position information Hb indicating whether the target block is located inside or outside the object. A pixel determiner 11 that outputs pixel position information Hg indicating whether each pixel is located inside the object or outside the object.
It also has 50.

On the other hand, the pixel value encoding apparatus 1200 according to the seventh embodiment converts the pixel value signal Sg corresponding to each frame supplied to the input terminal 1201a into 16 × 16 pixels as a unit of encoding processing. A blocker 1201 that divides a block to correspond to a block (image space) and outputs pixel data corresponding to each block;
And a prediction data generation unit 1200c that generates prediction data (pixel data of the prediction block) Prg corresponding to the pixel data Bg of the target block (block to be encoded) from 01.

Further, the pixel value encoding device 1200 includes:
A subtractor 1202 for outputting difference data between the pixel data Bg of the target block and the pixel data Prg of the prediction block as pixel data (prediction error signal) DEg of the residual block; An information compressor 1200a that performs processing and outputs pixel data Qg of a compressed residual block;
00 block position information Hb from the pixel determiner 1150
, A variable length encoding process is performed on the output Qg of the information compressor 1200a, and only the pixel value encoded signal Dg corresponding to the block located in the object is output to the output terminal 12g.
And a variable length coder (VLC) 1211 for outputting the data to the output unit 01b.

Here, the information compressor 1200a converts the pixel data DEg of the residual block into insignificant pixels (positions outside the object) among the pixels constituting the image space formed by the pixel data DEg. The pixel determining unit 1100 of the shape encoding apparatus 1100 performs the interpolation process of replacing the pixel value of the pixel to be replaced with a predetermined interpolation value (for example, “0”).
A first compensator (PAD1) 1203 based on the pixel value position information Hg from 150 and the e1 compensator 12
A DCT unit 1204 for performing a discrete cosine transform process on the output (pixel data of the interpolation residual block) Pg of No. 03;
A quantizer 1205 for quantizing the output Tg of the DCT unit 1204 and outputting a quantized coefficient as pixel data Qg of the compressed difference block. The discrete cosine transform process is performed in units of small blocks of 8 × 8 pixels obtained by dividing the block of 16 × 16 pixels into four.

Further, the pixel value encoding device 1200 includes:
An information decompressor 1200b that performs information decompression processing on the output Qg of the information compressor 1200a and outputs data ITg of a decompression difference block, and adds pixel data ITg of the decompression difference block to pixel data Prg of the prediction block. Adder 12 that outputs pixel data Rg of the reproduction block
10 is provided. Here, the information decompressor 1200
b denotes an inverse quantizer 1208 for inversely quantizing the output Qg of the information compressor 1200a, and an inverse DCT for transforming frequency domain data into spatial domain data for the output IQg of the inverse quantizer 1208. IDCT unit 1209 that performs processing and outputs pixel data ITg of the expanded difference block
It is composed of

Further, the prediction data generation unit 1200c
The frame memory (FM) 1213 for storing the pixel data Rg of the reproduction block as reference image data for the next processing frame, and the reference image data Mg for the current processing frame stored in the frame memory 1213, A second compensator (PAD2) 1212 for performing a compensation process based on the pixel position information Hg;
A motion detector (ME) 1214 for obtaining and outputting a motion vector MV corresponding to the target block of the current processing frame based on the output PMg of the compensator 1212 and the pixel data Bg of the target block. Of the frame memory 1213 based on the motion vector MV and the block position information Hb, and from the supplementary data PMg of the reference image data for the current processing frame, data corresponding to the memory area specified by the address is obtained. And a motion compensator (MC) 1215 for extracting and outputting as the pixel data Prg of the prediction block.

FIG. 20 shows a configuration of an image processing apparatus 2000 for decoding a digital image signal encoded by the image processing apparatus shown in FIG. 19, and will be briefly described below. The image processing apparatus 2000 receives the shape encoded signal Dk from the image processing apparatus 1000 at an input terminal 2
101a, and performs a data analysis process and an arithmetic decoding process on each frame to output a shape reproduction signal Rk to an output terminal 2101b.
, The pixel value coded signal Dg from the image processing apparatus 1000 is received at the input terminal 2201a. And an output pixel value decoding device 2000b.

The shape decoding device 2000a receives the shape coded signal Dk corresponding to the target area (target block) to be decoded, analyzes the shape coded signal Dk, and performs variable length decoding. A data analyzer 2102, a shape decoder (DEC) 2103 that performs shape decoding processing on the output Ck of the data analyzer 2102 to generate shape restoration data Ekd, The up-sampling process is performed at the ratio of the down-sampling process corresponding to this data, and the shape extension data E
an upsampler (US) 2104 for generating ku;
A shape post-processor 2105 for performing shape post-processing on the shape extension data Eku and outputting a shape reproduction signal Rk to an output terminal 2101b. Here, the shape decoder 2103, the upsampler 2104, and the shape post-processor 2105 constitute a shape decoding unit 2100a.

Further, the shape decoding device 2000a
Upon receiving the reproduced shape signal Rk output from the shape decoding unit 2100a, block position information Hb indicating whether the target block of the decoding process is located inside the object or outside the object.
In addition, it also has a pixel determiner 2150 that outputs pixel position information Hg indicating whether each pixel constituting the target block is located inside or outside the object.

On the other hand, the pixel value decoding device 2000b
A data analyzer 2202 that receives a pixel value coded signal Dg corresponding to a target unit area (target block) to be decoded, analyzes the pixel value coded signal Dg, and performs variable length decoding; An information decompressor 2200a that performs decompression processing on the output Qg of the data analyzer 2202 and outputs pixel data ITg of a decompression block of the current processing frame;
An adder 2206 for adding the pixel data ITg of the decompressed residual block from the information decompressor 2200a to the pixel data Prg of the prediction block corresponding to the decompressed residual block
And the output Rg of the adder is output as a pixel value reproduction signal.

Here, the information decompressor 2200a includes an inverse quantizer 2204 for performing an inverse quantization process on the output Qg of the data analyzer 2202, and an inverse frequency transform for the output IQg of the inverse quantizer 2204. IDCT, a type of processing
(Inverse Discrete Cosine Transform) An IDCT converter 2205 that performs processing and outputs pixel data ITg of the decompressed block.

Further, the pixel value decoding unit 2000b
Upon receiving the pixel data Rg of the reproduced block output from the adder 2206, the motion vector MV from the data analyzer 2202 and the pixel decision unit 2 of the shape decoding device 2000a are received.
150 from the pixel position information Hg and the block position information H
and a prediction data generation unit 2200b that generates pixel data Prg of the prediction block based on b.

Here, the prediction data generation unit 2200b
Is a frame memory (FM) 2209 for storing the output Rg of the adder 2206 as reference data, and a pixel determination unit 2150 for the data Mg stored in the frame memory.
A compensator (PAD) 2208 that performs the same compensation processing as that on the encoding side based on the pixel position information Hg from the encoder and an output PMg of the compensator 2208, and the motion vector M
And a motion compensator (MC) 2210 that generates the prediction data based on the V and the block position information Hb.

Next, the operation will be described. When a digital image signal is input to the image processing apparatus 1000, a shape signal Sk constituting the digital image signal is sequentially encoded for each frame by a shape encoding process in a shape encoding apparatus 1000a. The pixel value signal Sg constituting the signal is sequentially encoded for each frame by the information compression process and the variable length encoding process in the pixel value encoding unit 1000b.

That is, when the shape signal Sk is supplied to the input terminal 1101a of the shape encoding device 1000a, the shape signal Sk is converted by the block generator 1101 into each of a plurality of blocks for dividing the object region including the object. The data is divided so as to correspond to each other and output as shape data Bk corresponding to each block.

The shape data Bk of the target block from the block generator 1101 is subjected to compression / expansion processing at different compression ratios by first to third shape encoding units (compression / expansion means) 1100a to 1100c. Will be applied.

For example, the second shape encoding unit 1100
In b, the shape data Bk of the target block is converted into shape sample data Skd2 by a downsampling process at a ratio of 1/2 in a second downsampler 1111b.
Is subjected to shape encoding processing by the shape encoder 1112b, and is converted into shape compressed data Ck2. Further, the shape compression data Ck2 is subjected to shape decoding processing by the second shape decoder 1113b, and the restored sample data PEkd
2 and the restored sample data PEkd2
Is processed by the second up-sampler 1114b by the up-sampling process at a ratio of 1/2 to obtain the shape extension data PEk.
Converted to u2. Then, the shape extension data PEk
u2 is subjected to the same processing as the shape post-processing in the first embodiment in the second shape post-processor 1115b, and is output as shape post-processing data PFku2.

The shape data Bk of the target block
The above-described compression / decompression processing for
In the shape encoding units 1100a and 1100c, the processing is performed in substantially the same manner as in the second shape encoding unit 1100b.

However, in the first down-sampler 1111a of the first shape encoder 1100a, the shape data Bk of the target block is subjected to down-sampling processing at a ratio of 1/1, that is, substantially, The shape data Bk is output as it is without down-sampling. In the first upsampler 1114a, upsample processing is performed at a ratio of 1/1 to the restored shape data PEkd1, that is, substantially upsampled is performed on the restored shape data PEkd1. It is output as it is without being. Further, the third down-sampler 1111c of the third shape encoding unit 1100c outputs a quarter of the shape data Bk of the target block.
Is performed at a ratio of the third shape encoder 1114c.
Up-sample processing is performed on the restored shape data PEkd3, which is the output of 12c, at a ratio of 1/4. Then, the first and third shape encoding units 1100a and 1100c
Are performed in the same manner as in the second shape encoding unit 1100b.

The analyzer 1120 of the shape encoding apparatus 1100 according to the seventh embodiment outputs the shape data Bk of the target block and the corresponding output of the first to third shape encoding units 1100a to 1100c. By comparing with data (shape post-processing data) PFku1 to PFku3, block position information included in the shape data Bk of the target block and indicating whether the target block is located inside the object or outside the object. Is stored as it is, the shape post-processing data P
A selection control signal SWc2 determined from Fku1 to PFku3 and corresponding to the determination result is supplied to the selection switch 1130.

Then, in the data selection switch 1130, the first to third shape encoding units 1100a to 1100c
Of the compressed shape data Ck1 to Ck3 output from the shape encoders 1112a to 1112c are selected based on the selection control signal SWc2 for each block. Specifically, a local reproduction block located outside the object is formed when the target block is located outside the object, and a local reproduction block located inside the object is formed when the target block is located inside the object. Shape compression data output from the shape encoding unit that generates such shape post-processing data and that has the highest compression rate is selected by the selection switch 1130.

The data selection switch 1130
Is subjected to variable-length encoding processing by a variable-length encoder 1140, and is output as a shape-encoded signal (shape bit stream) Dk to an output terminal 1
It is output to 101b. The downsampling rate for the shape data of each block is sent to the decoding side as a flag (side information) of each block together with the shape encoded signal Dk corresponding to each block.

At this time, the shape data Bk of the target block is supplied to the pixel determiner 1150. The pixel determiner 1150 sends block position information Hb indicating whether the target block is located outside the object. Pixel position information Hg indicating whether or not each pixel constituting the target block is located outside the object is output.

On the other hand, at this time, the pixel value encoding device 100
0b, the pixel value signal S supplied to the input terminal 101a
g is divided by the block generator 1101 so as to correspond to each of a plurality of blocks that divide the object region including the object, and is output as pixel data Bg corresponding to each block.

Then, the pixel data Bg of the target block to be encoded is input to the motion detector 1214, and the pixel data Mg of the past reproduced image (reference image) stored in the frame memory 1213 is converted to the pixel data Mg. 2 is read out to the second compensator 1212, compensated based on the pixel position information Hg, and further compensated for pixel data P of the reference image.
When Mg is input to the motion detector 1214, the motion detector 1214 uses the block data matching or the like to convert the current processing frame into pixel data Bg of the target block based on the pixel data PMg of the supplemented reference image. On the other hand, motion displacement information giving a prediction block having pixel data with the smallest error is output as a motion vector MV.

This motion vector MV is used as the motion compensator 121.
5, the motion compensator 1215 converts the pixel data Prg of the prediction block corresponding to the target block from the pixel data PMg of the supplemented reference image corresponding to the current processing frame based on the block position information Hb. Generated. At this time, the motion vector MV is sent to a variable length encoder (VLC) 1211, converted into a variable length code, and output to an output terminal 1201b.

The pixel data Bg of the target block and the pixel data Pg of the prediction block are subjected to arithmetic processing in the first adder 1202 to obtain difference data of these image data as pixel data DEg of the residual block.

Next, the pixel data DE of the residual block is obtained.
An information compression process is performed on g. That is, in the present embodiment, the pixel data DEg is stored in the first complementer 12.
At 03, a predetermined compensation process is performed based on the pixel position information Hg. The supplementary pixel data P obtained by this
g is converted to a frequency component Tg by a DCT unit 1204 by a discrete cosine transform process.
g is converted into a quantization coefficient Qg by quantization in a quantizer 1205, and is output as pixel data of a compressed block. The pixel data Qg of the compressed block is converted into a variable length code by the variable length encoder 1211 and output from the output terminal 1201b together with the encoded data of the other side information including the motion vector as an image encoded signal Dg. Is done.

On the other hand, the pixel data Qg of the compressed block
Is subjected to information decompression processing by the information decompression device 1200b. In the present embodiment, the pixel data Q
g is converted into a frequency component IQg by inverse quantization by an inverse quantizer 1208, and this frequency component IQg is
The IDCT unit 1209 restores the data ITg in the spatial area. The data ITg in this space area is a restoration signal (pixel data of the expanded block) for the pixel data Dg of the above-described interpolation residual block. The pixel data ITg of the decompressed block is added to the pixel data Prg of the prediction block by the second adder 1210 and supplied to the frame memory 113 as pixel data Rg of the reproduction block.

The coded shape signal Dk and coded pixel value signal Dg output from the image processing apparatus 1000 are supplied to an image processing apparatus 2000 shown in FIG.
Then, the shape-encoded signal Dk is applied to the shape decoding device 200.
0a is sequentially decoded for each frame by data analysis processing and shape decoding processing, and the pixel value encoded signal Dg is sequentially decoded for each frame by data analysis processing and information decompression processing by the pixel value decoding device 2000b. Is done.

The operation will be briefly described below. The shape-encoded signal Dk corresponding to the target area (target block) to be decoded is generated by the shape decoding unit 2000a.
Is input to the input terminal 2101a, the shape-encoded signal Dk is subjected to variable-length decoding by the data analysis 2102 by the data analysis and restored to the compressed shape data Ck.
The compressed shape data Ck is subjected to a shape decoding process in a shape decoder 2103 to generate shape restoration data Ekd. Further, the upsampler 2104 applies the shape restoration data Ekd to this data. Up-sample processing is performed at the ratio of down-sample processing. That is, the compression ratio of the shape data for each block is sent from the shape encoding device 1000a for each block as side information for each block.
At 104, based on the side information, the upsampling process is performed on the shape restoration data Ekd for each block by appropriately changing the compression ratio. The shape extension data Eku output from the upsampler 2104 is
Shape post-processing is performed by the shape post-processing unit 2105, and the result is output to the output terminal 2101b as a shape reproduction signal Ek.

On the other hand, the pixel value coded signal Dg input to the input terminal 2201a of the pixel value decoding device 2000b is
The data analyzer 2202 performs the analysis. In the data analyzer 2202, the pixel value encoding unit 100 shown in FIG.
The pixel data Qg and the motion vector MV of the compressed residual block at 0b are generated. This data analyzer 22
02 is output to the information decompressor 2200a, and the data analyzer 220
2 is output to the motion compensator 2210.

In the information decompressor 2200a, the information decompression process is performed on the pixel data Qg of the compression residual block.
The pixel data Pg of the interpolation residual block in the pixel value encoding unit 1000b is the pixel data I of the expanded residual block.
It is restored as Tg. In the present embodiment, the pixel data Qg of the compressed residual block is converted into frequency-domain pixel data IQg by inverse quantization at an inverse quantizer 2204, and the pixel data IQg is subjected to inverse discrete cosine transform (IDCT). ) Unit 2205 converts the pixel data into pixel data in the spatial area and outputs the pixel data ITg of the expanded residual block.

In the motion compensator 2210, based on the motion vector MV from the data analyzer 2202,
An address for accessing the frame memory 2209 is generated. Based on this address, storage data Mg, which is reference image data for the current processing frame in the frame memory 2209, is read out to the compensator 2208, and the compensator 2208 reads the stored data Mg. Compensated supplementary storage data P
Mg is supplied to the motion compensator 2210. Then
In the motion compensator 2210, pixel data Prg of the prediction block is generated from the supplementary storage data PMg. The adder 2206 performs an addition process of the pixel data Prg of the prediction block and the pixel data ITg of the decompression residual block, and the pixel data Eg of the reproduction block is output as a pixel value reproduction signal by the addition process.
Is output to

As described above, in the seventh embodiment, the shape signal of the target block to be coded is subjected to compression processing at a predetermined compression ratio to generate compressed shape data, and the compressed shape data is compressed. First to third shape encoding units (compression / expansion units) 1 having different compression ratios as compression / expansion means for performing expansion processing corresponding to the processing to generate expanded shape data
100a to 1100c, and the locally decoded shape signals PFku1 output from the respective compression / decompression units having different compression ratios.
According to the analysis of .about.PFku3, information indicating that the position of the target block is located inside or outside the object, which is included in the shape data Bk of the target block, is held in the locally decoded shape signal. A shape encoder (compressor / decompressor) that performs compression / expansion processing at a high rate is specified, and the compressed shape data S output from the specified shape encoder is specified.
Since Ck is coded and output as shape coded signal Dk, if the target block to be coded is located within the object, the corresponding target block to be decoded must be Will be located inside the object,
Also, when the target block to be encoded is located outside the object, the corresponding target block to be decoded is always located outside the object. As a result, the decoding side can perform the motion compensation process using the motion vector on the coded data of all the target blocks to perform the decoding process satisfactorily. As a result, the coding of the shape signal can be performed. Even when performing post-shape processing to reduce visual distortion due to the processing, it is possible to avoid significant image quality degradation of the reproduced image.

For example, even when the encoding process of the shape signal and the encoding process of the pixel value signal are performed in parallel, a block (image space) formed by the object shape signal due to encoding distortion caused by the shape encoding process. It is possible to prevent a block (image space) formed by the locally decoded shape signal from being inside the object even though is a block outside the object, thereby detecting a motion vector or compressing a pixel value for the block. Large image quality deterioration due to the lack of the above processing can be prevented.

[0168] Also, MPEG4 encoding that employs advanced motion compensation (overlap motion compensation), in which the operation method for motion compensation differs depending on whether the adjacent block is a transparent block or an opaque block, is used. According to the method, even when the encoding process of the shape signal and the encoding process of the pixel value signal are performed in parallel, a block (image space) formed by the object shape signal may be an object due to encoding distortion caused by the shape encoding process. Despite being an in-vivo block, it is possible to avoid that a block (image space) formed by the locally decoded shape signal is out of the object, so that the operation in the overlap motion compensation is correctly performed on the decoding side. It is possible to eliminate such a problem that the reproduction is not performed and the image quality of the reproduced image is deteriorated.

In the seventh embodiment, the process of selecting a predetermined shape encoding unit from a plurality of shape encoding units having different compression ratios is performed. When the block is outside and the target block is in the object, the case where the local decoding block is performed in the object has been described. The processing may be performed only so that the local decoding block is outside the object when the target block is outside the object, or such that the local decoding block is inside the object when the target block is inside the object. You may go only.

Furthermore, an encoding or decoding program for realizing the configuration of the shape encoding device or the shape decoding device shown in each of the above embodiments is recorded on a data storage medium such as a floppy disk. Thus, the encoding process or the decoding process in each of the above-described embodiments can be easily performed by an independent computer system.

FIG. 11 shows a case where the encoding or decoding processing of the first to seventh embodiments is performed by a computer system using a floppy disk storing the above encoding or decoding program. FIG. FIG. 11B shows an external appearance, a sectional structure, and a floppy disk of the floppy disk as viewed from the front, and FIG. 11A shows an example of a physical format of the floppy disk as a recording medium body. The floppy disk FD is built in the case F, and a plurality of tracks Tr are formed concentrically from the outer circumference toward the inner circumference on the surface of the disk, and each track is 16
Sector Se. Therefore, in the floppy disk storing the program, data as the program is recorded in an area allocated on the floppy disk FD.

FIG. 11C shows a configuration for recording and reproducing the program on the floppy disk FD. When recording the program on the floppy disk FD, data as the program is written from the computer system Cs via the floppy disk drive FDD. When the encoding or decoding device is constructed in a computer system by a program in the floppy disk FD, the program is read from the floppy disk FD by the floppy disk drive FDD and transferred to the computer system Cs.

In the above description, a case where a floppy disk is used as a data recording medium has been described, but an optical disk may be used as a data storage medium. Also,
The data recording medium is not limited to this, and any of the above-described encoding processes or decoding processes can be similarly executed as long as a program can be recorded thereon, such as an IC card or a ROM cassette.

[0174]

As described above, according to the shape encoding method according to the present invention (claim 1), the object shape signal is subjected to the encoding process for each unit area on the display screen, and the shape encoding is performed. A bit string is generated, and a signal in the middle of the encoding process is decoded to generate a locally decoded shape signal. In accordance with the analysis result of the locally decoded shape signal, the local decoded shape signal is set to a predetermined value. Since the post-shape processing is performed or the post-shape processing is not performed on the locally decoded shape signal, it is output as an encoding processing shape signal used for encoding processing of an image display signal other than the object shape signal. It is possible to suppress the processing of encoding pixel value signals for a unit area located outside the object on the display screen, thereby saving the number of bits of pixel value encoding without substantially impairing the image quality of the object shape. Can .

According to the present invention (claim 2), claim 1
In the shape encoding method described above, when the locally decoded shape signal corresponding to a certain unit region indicates that the unit region is located outside the object, the locally decoded shape signal subjected to the shape post-processing is Also, since the shape post-processing is performed so as to be a signal indicating that the certain unit area is located outside the object, the encoding process of the pixel value signal for the unit area located outside the object on the display screen is performed. Can be reliably avoided.

According to a third aspect of the present invention, in the shape encoding method according to the second aspect, the shape post-processing is performed so that an object shape obtained from the locally decoded shape signal becomes smoother. Since the locally decoded shape signal is converted by using a filter having a predetermined characteristic, shape post-processing can be performed on the locally decoded shape signal with a simple configuration.

According to the shape encoding method of the present invention (claim 4), an object shape signal is down-sampled, and a down-sampled shape signal in which the number of encoded pixels is reduced is obtained.
The down-sampled shape signal is subjected to an encoding process for each unit area of a predetermined size on the display screen to generate a shape-encoded bit string, and a signal in the middle of the encoding process is decoded to generate a locally decoded shape signal. Is generated, and the locally decoded shape signal is up-sampled to generate an up-sample shape signal having an increased number of pixels, and according to the analysis result of the up-sample shape signal, the up-sample shape signal is Is subjected to a predetermined shape post-processing, or is not subjected to the shape post-processing to the upsampled shape signal, and is output as an encoding shape signal used for the encoding process of an image display signal other than the object shape signal. Therefore, it is possible to reduce the number of bits in the shape-encoding bit sequence without substantially deteriorating the image quality of the object shape, and to further reduce the number of bits in pixel value encoding. It is possible.

According to the shape decoding method of the present invention (claim 5), a shape-encoded bit string is decoded for each unit area of a predetermined size on a display screen to generate a decoded shape signal.
Analyzing the decoded shape signal for each of the unit areas, according to the analysis result of the decoded shape signal, the decoded shape signal,
A decoded object shape signal (decoded) used for decoding an image display signal other than the object shape signal without subjecting the decoded shape signal to a predetermined shape post-processing or performing a shape post-processing on the decoded shape signal. Since this is output as the processing reference signal, the decoding process of the pixel value signal using the decoding reference signal can be performed accurately while the number of bits of the pixel value encoding is kept low.

According to the present invention (claim 6), claim 5
In the shape decoding method described, when the decoded shape signal corresponding to a certain unit area indicates that the unit area is located outside the object, the decoded shape signal subjected to the shape post-processing is also Since the shape post-processing is performed so as to be a signal indicating that the certain unit area is located outside the object, the decoding processing of the pixel value signal for the unit area located outside the object on the display screen is performed. Can be avoided reliably.

According to a seventh aspect of the present invention, in the shape decoding method according to the sixth aspect, in the shape post-processing, the decoding is performed such that the object shape obtained from the decoded shape signal becomes smoother. Since the transformed shape signal is converted by using a filter having a predetermined characteristic, the shape post-processing can be performed on the decoded shape signal with a simple configuration.

According to the shape decoding method according to the present invention (claim 8), a shape-encoded bit string corresponding to a predetermined object shape is decoded for each unit area of a predetermined size on a display screen to obtain a decoded shape signal. Is generated, the decoded shape signal is up-sampled, an up-sampled shape signal having an increased number of pixels is obtained, the up-sampled shape signal is analyzed for each unit area, and the analysis result of the up-sampled shape signal is obtained. Depending on the above, the up-sample shape signal is subjected to a predetermined shape post-processing, or the up-sample shape signal is not subjected to the shape post-processing, and the encoding process of the image display signal other than the object shape signal is performed. Is output as a decoded object shape signal (reference signal for decoding processing) used in Savings can be realized in several.

According to the shape encoding apparatus of the present invention (claim 9), an object shape signal indicating the shape of an object is subjected to an encoding process for each unit area of a predetermined size on the display screen, and the shape encoding is performed. Shape encoding means for generating a localized bit string, decoding the signal in the middle of the encoding process to generate a locally decoded shape signal, and shape analyzing means for analyzing the locally decoded shape signal for each unit area. In accordance with the analysis result of the locally decoded shape signal, the locally decoded shape signal is subjected to a predetermined shape post-processing, or without performing the shape post-processing to the locally decoded shape signal. The pixel value signal is encoded for a unit area located outside the object on the display screen, since the signal is output as an encoding shape signal used for encoding an image display signal other than the object shape signal. Processing Dividing the can be suppressed, thereby without impairing most the quality of object shape, it is possible to reduce the number of bits of the coded pixel value.

According to the shape encoding apparatus of the present invention (claim 10), down-sampling means for down-sampling an object shape signal indicating an object shape and obtaining a down-sampled shape signal having a reduced number of encoded pixels. And performing an encoding process on the downsampled shape signal for each unit area of a predetermined size on the display screen to generate a shape-encoded bit string, and decoding the signal in the middle of the encoding process to perform local decoding. Shape encoding means for generating a shape signal; up-sampling means for up-sampling the locally decoded shape signal to generate an up-sampled shape signal having an increased number of pixels; And a shape analyzing means for analyzing each of the up-sample shape signals according to the analysis result of the up-sample shape signal. The post-shape processing is performed, or the up-sample shape signal is not subjected to the shape post-processing, and is output as an encoding processing shape signal used for encoding an image display signal other than the object shape signal. Therefore, it is possible to reduce the number of bits in the shape-encoded bit sequence without substantially deteriorating the image quality of the object shape, and to further reduce the number of bits in pixel value encoding.

According to the shape decoding device of the present invention (claim 11), a shape-encoded bit string corresponding to a predetermined object shape is decoded for each unit area of a predetermined size on the display screen, and the decoded shape signal is decoded. And a shape analyzing means for analyzing the decoded shape signal for each unit area, and, in accordance with an analysis result of the decoded shape signal, the decoded shape signal, A decoded object shape signal (for decoding processing) used for decoding an image display signal other than the object shape signal without performing predetermined shape post-processing or without performing shape post-processing on the decoded shape signal. Since the pixel value signal is output as the reference signal, the decoding process of the pixel value signal using the decoding process reference signal can be performed accurately while the number of bits of the pixel value encoding is kept low.

According to the shape decoding apparatus of the present invention (claim 12), a shape-encoded bit string corresponding to a predetermined object shape is decoded for each unit area of a predetermined size on the display screen, and the decoded shape signal is decoded. , And up-sampling means for up-sampling the decoded shape signal to generate an up-sampled shape signal having an increased number of pixels, and analyzing the up-sampled shape signal for each of the unit regions. A shape analyzing means for performing a predetermined shape post-processing on the up-sample shape signal according to an analysis result of the up-sample shape signal, or subjecting the up-sample shape signal to a shape post-processing. Instead, it is output as a decoded object shape signal (decoding process reference signal) used for decoding an image display signal other than the object shape signal. It is possible to reduce the number of bits of the bit string, so it is possible to realize the savings in the number of bits of the pixel value decoding.

According to the data storage medium of the present invention (claim 13), the object shape signal is subjected to an encoding process for each unit area on the display screen to generate a shape-encoded bit string, and A locally decoded shape signal is generated by decoding the signal in the middle of the encoding process, and according to the analysis result of the locally decoded shape signal, the local decoded shape signal is subjected to a predetermined shape post-processing. Or, without subjecting the locally decoded shape signal to post-shape processing, the computer outputs, as a shape signal for encoding, a signal used for encoding of an image display signal other than the object shape signal. The encoding method and the encoding method of saving the number of bits of the pixel value encoding by loading the program into the computer without substantially deteriorating the image quality of the object shape. It is possible to realize a location.

According to the data storage medium of the present invention (claim 14), a shape-encoded bit string is decoded for each unit area of a predetermined size on a display screen to generate a decoded shape signal, and the decoded shape signal is generated. The signal is analyzed for each unit area, and the decoded shape signal is subjected to a predetermined shape post-processing according to the analysis result of the decoded shape signal, or the decoded shape signal is subjected to the shape post-processing. A program for causing a computer to perform a process of outputting as a decoded object shape signal (decoding process reference signal) used for decoding an image display signal other than the object shape signal without performing the process is stored. Therefore, by loading the program into the computer, the decoding process of the pixel value signal using the decoded object shape signal can be performed accurately while the number of bits of the pixel value encoding is kept low. It is possible to realize a decoding method and a decoding apparatus.

According to the shape encoding method of the present invention (claim 15), the object shape signal corresponding to the target unit area to be encoded is subjected to the compression processing, and the object shape signal subjected to the compression processing is simultaneously processed. A compression / decompression process that converts a signal into a locally decoded signal by decompression, from among a plurality of compression / decompression processes with different compression ratios, where the target unit area included in the input object shape signal is located outside the object Is selected as the compression / decompression process with the highest compression ratio, which is held in the local decoded shape signal, and the compression / decompression signal obtained by the selected compression / decompression process is encoded to form a shape. Since this is output as a coded signal, while improving coding efficiency, these blocks are positioned outside the object between the target block to be coded and the target block to be decoded. It is possible to avoid the inconsistency that occurs in the information indicating that to those.

As a result, it is possible to perform the shape encoding process for the shape signal constituting the image signal corresponding to the object having an arbitrary shape and the pixel value encoding process for the pixel value signal constituting the image signal in parallel. In this case, it is possible to prevent the image quality from being deteriorated due to the encoding distortion caused by the shape encoding process.

For example, it is possible to improve the coding efficiency while avoiding the visual distortion of the reproduced image due to the encoding processing of the object shape signal by the shape post-processing and the like, and to realize the large image quality of the reproduced image caused by the shape post-processing and the like. Deterioration can be prevented. According to this invention (claim 16), claim 1
In the shape encoding method according to the fifth aspect, since each of the compression / expansion processing having a different compression rate includes a down-sampling processing and an up-sampling processing, the compression / expansion processing of the object shape signal can be realized with a simple configuration. it can.

According to the shape coding method of the present invention (claim 17), the object shape signal corresponding to the target unit area to be coded is subjected to the compression processing and the object shape signal subjected to the compression processing is simultaneously processed. A compression / decompression process that converts a signal into a locally decoded signal by decompression, from among a plurality of compression / decompression processes with different compression ratios, where the target unit area included in the input object shape signal is located outside the object Is stored in the locally decoded shape signal, the compression / decompression process having the highest compression rate is selected, and the compression shape signal obtained by the selected compression / decompression process is encoded to form a shape code. Output as a coded signal, while improving coding efficiency, these blocks are located outside the object between the target block to be coded and the target block to be decoded. It is possible to avoid the resulting mismatch information indicating that it is not intended to location, as described above,
The shape encoding process and the pixel value encoding process can be performed in parallel, and at this time, it is possible to prevent image quality deterioration due to encoding distortion accompanying the shape encoding process.

According to the present invention (claim 18), in the shape encoding method according to claim 17, each of the compression / decompression processes having different compression rates includes a down-sample process and an up-sample process. The compression / decompression processing of the object shape signal can be realized with a simple configuration.

According to the shape encoding apparatus of the present invention (claim 19), the object shape signal corresponding to the target unit area to be encoded is subjected to the compression processing, and the object shape signal subjected to the compression processing is simultaneously processed. As compression / decompression means for converting a signal into a locally decoded signal by decompression processing, a plurality of compression / decompression units having different compression ratios are provided, and the locally decoded shape signal output from each of the compression / decompression units with different compression ratios is provided. According to the analysis, information indicating that the target unit area is located outside the object, which is included in the input object shape signal, is held in the locally decoded shape signal. Analyzing means for designating a compression / expansion unit for performing an expansion process, selecting a predetermined compression / expansion device based on the output of the analysis means, and generating a compression shape signal output from the selected compression / expansion device; Since the output as a coded shape signal and Goka, while improving the coding efficiency,
Between the target block to be encoded and the target block to be decoded, it is possible to avoid a mismatch in information indicating that these blocks are located outside the object, It is possible to prevent large deterioration in image quality of a reproduced image due to shape post-processing or the like.

According to a twentieth aspect of the present invention, in the shape encoding apparatus according to the nineteenth aspect, each of the compression / decompression units is provided with a predetermined down-sampling rate for the object shape signal as part of the compression processing. The down-sampling process is performed, and as a part of the decompression process, the up-sampling process corresponding to the down-sampling process is performed on the compressed shape signal. It can be realized by the configuration.

According to the shape coding apparatus of the present invention (claim 21), the object shape signal corresponding to the target unit area to be coded is subjected to the compression processing, and the object shape signal subjected to the compression processing is simultaneously processed. As compression / decompression means for converting a signal into a locally decoded signal by decompression processing, a plurality of compression / decompression units having different compression ratios are provided, and the locally decoded shape signal output from each of the compression / decompression units with different compression ratios is provided. According to the analysis, information indicating that the target unit area is not located outside the object, which is included in the input object shape signal, is held in the locally decoded shape signal. Analyzing means for designating a compression / expansion unit for performing processing; selecting a predetermined compression / expansion device based on the output of the analysis means; Since the output as a coded shape signal and Goka, while improving the coding efficiency,
It is possible to avoid a mismatch between the target block to be coded and the target block to be decoded, which indicates that these blocks are not located outside the object. It is possible to prevent a large deterioration in image quality of a reproduced image due to post-processing or the like.

According to the present invention (invention 22), in the shape encoding apparatus according to claim 21, the compression / expansion device is configured to perform a predetermined down conversion on the object shape signal as part of the compression processing. A down-sampling process having a sample rate is performed, and as a part of the decompression process, an up-sampling process corresponding to the down-sampling process is performed on the compressed shape signal. It can be realized with a simple configuration.

According to the data storage medium of the present invention (claim 23), the program for encoding the object shape signal is compressed into the object shape signal corresponding to the target unit area to be encoded. At the same time, the compression / decompression processing for converting the compressed object shape signal into a locally decoded signal by decompression processing is included in the input object shape signal from among a plurality of compression / decompression processings having different compression rates. Information indicating that the target unit area is located outside the object is held in the locally decoded shape signal,
Since a series of processing of selecting the compression / expansion processing with the highest compression rate, encoding the compressed shape signal obtained by the selected compression / expansion processing, and outputting the encoded signal as a shape-encoded signal is configured to be performed by a computer. It is possible to avoid the occurrence of a mismatch between the target block to be encoded and the target block to be decoded, which indicates that these blocks are located outside the object. It is possible to realize, by a computer, an image encoding process in which a shape encoding process and a pixel value encoding process can be performed in parallel without causing image quality deterioration due to encoding distortion accompanying the encoding process.

A data storage medium according to the present invention (claim 24) performs a compression process on an object shape signal corresponding to a target unit area to be encoded with a program for encoding the object shape signal. At the same time, as a compression / decompression process for converting the compressed object shape signal into a locally decoded signal by a decompression process, a plurality of compression / decompression processes having different compression rates are included in the input object shape signal. The information indicating that the target unit area is not located outside the object is retained in the locally decoded shape signal, the compression / expansion processing with the highest compression rate is selected, and obtained by the selected compression / expansion processing. The computer is configured to perform a series of processes of encoding a compressed shape signal to be output as a shape-encoded signal, so that the encoding efficiency is improved. It is possible to avoid the occurrence of a mismatch between the target block to be decoded and the target block to be decoded, indicating that these blocks are not located outside the object. An image encoding process in which the shape encoding process and the pixel value encoding process can be performed in parallel without causing image quality degradation due to encoding distortion accompanying the process can be realized by a computer.

[Brief description of the drawings]

FIG. 1 is a block diagram for describing a configuration of a shape encoding device according to a first embodiment of the present invention.

FIG. 2 is a flowchart showing an encoding process performed by the shape encoding device according to the first embodiment;

FIG. 3 is a diagram for explaining a case where a block outside an object becomes inside an object by post-processing in the shape encoding process.

FIG. 4 is a diagram for explaining post-shape processing by the shape encoding device according to the first embodiment;

FIG. 5 is a block diagram illustrating a configuration of a shape encoding device according to a second embodiment of the present invention.

FIG. 6 is a block diagram for describing a configuration of a shape encoding device according to a third embodiment of the present invention.

FIG. 7 is a block diagram illustrating a configuration of a shape decoding device according to a fourth embodiment of the present invention.

FIG. 8 is a flowchart showing a decoding process by the shape decoding apparatus according to the fourth embodiment.

FIG. 9 is a block diagram illustrating a configuration of a shape decoding device according to a fifth embodiment of the present invention.

FIG. 10 is a block diagram illustrating a configuration of a shape decoding device according to a sixth embodiment of the present invention.

FIG. 11 is a diagram for describing a data storage medium for storing a program for realizing the object shape encoding or object shape decoding method according to each of the above-described embodiments using a computer system.

12A and 12B are diagrams showing a conventional image signal processing system based on MPEG4, wherein FIG. 12A shows a configuration of an encoding circuit for performing an encoding process, and FIG. 12B shows a decoding circuit for performing a decoding process. Is shown.

FIG. 13 is a diagram illustrating, by a flowchart, an encoding process in a conventional image signal processing system.

FIG. 14 is a flowchart showing a decoding process in a conventional image signal processing system.

FIG. 15 is a diagram illustrating a pixel value encoding process performed in a macroblock unit in MPEG4.

FIG. 16 is a diagram for describing a pixel value encoding process performed on a subblock basis in MPEG4.

FIG. 17 is a diagram for explaining a case where an outer block becomes an inner block by the shape post-processing in the shape encoding processing.

FIG. 18 is a diagram for explaining a conventional post-shape process.

FIG. 19 is a block diagram for describing a configuration of a shape encoding device according to a seventh embodiment of the present invention.

FIG. 20 is a block diagram illustrating a configuration of a shape decoding device that decodes a shape-encoded signal encoded by the shape encoding device according to the seventh embodiment.

[Explanation of symbols]

11, 19 Shape analyzer 12, 15, 20, 23 On / off switch 13, 16, 18, 21, 24, 26 Select switch 14a Down sampler 14b, 22, 2104 Up sampler 17, 25 Specific shape signal generator 110a, 120a, 130a Shape encoder 110b, 120b, 130b Shape decoder 210a, 1000a Shape encoder 211 Shape encoder 211a Shape encoder 211b, 221b Variable length encoder 212, 214, 2105 Shape post-processor 213 shape decoders 213a, 222a variable length decoding units 213b, 2100a shape decoding units 220a, 1000b pixel value coding devices 220b, 2000b pixel value decoding devices 221a pixel value coding units 222b pixel value decoding units 1000, 2000 Image processing equipment 1100a to 1100c First to third shape encoders 1111a to 1111c First to third downsamplers 1112a to 1112c First to third shape encoders 1113a to 1113c First to third shape decoding Units 1114a to 1114c First to third up-sampling units 1115a to 1115c First to third shape post-processors 1120 Analyzers 1130 Data selection switches 1140 Variable-length encoders 1150 Pixel determiners 1200a Information compressors 1200b Information decompression 1200c, 2200b Prediction data generator 2102, 2202 Data analyzer 2200a Information expander Sk Object shape signal Skd Down sample shape signal Dk, Dkd Shape coding bit string PEku Up sample shape signal PEk, PEkd Locally decoded shape signal Fk, F u, PFk, PFku Shape post-processing output Ek, Ekd Decoded shape signal BGk Specific shape signal Hb Block position information Hg Pixel position information SWc1, SWc2 Switch control signal Rc, Rcu Selection output (reference signal for encoding process) Re, Reu Selection output (reference signal for decoding) Cs Computer system FD Floppy disk FDD Floppy disk drive

Claims (24)

[Claims] 1. A shape encoding method for encoding an object shape signal corresponding to a predetermined object shape in an image display signal, wherein the object shape signal corresponds to a unit area of a predetermined size on a display screen. To generate a shape-encoded bit sequence by decoding the signal, generate a locally decoded shape signal by decoding the signal in the middle of the encoding process, and analyze the locally decoded shape signal for each unit area. , According to the analysis result, for the locally decoded shape signal,
Determine at least one of a type of shape post-processing for making the object shape obtained from the locally decoded shape signal smoother, and whether or not to perform the shape post-processing on the locally decoded shape signal The locally decoded shape signal is subjected to the predetermined shape post-processing or to the above-mentioned shape post-processing without being subjected to the shape post-processing. A shape encoding method characterized by outputting as a shape signal for encoding processing. 2. The shape encoding method according to claim 1, wherein the determination regarding the shape post-processing according to the analysis result includes: determining a locally decoded shape signal corresponding to a certain unit area by positioning the unit area outside the object; The local decoding shape signal that has been subjected to the shape post-processing is also a signal indicating that the certain unit area is located outside the object. Method. 3. The shape encoding method according to claim 2, wherein the shape post-processing is performed so that the locally decoded shape signal is determined so that an object shape obtained from the locally decoded shape signal becomes smoother. Characterized in that the conversion is performed using a filter having the following characteristics: 4. A shape encoding method for encoding an object shape signal corresponding to a predetermined object shape in an image display signal, wherein the object shape signal is down-sampled to reduce the number of encoded pixels. The obtained down-sampled shape signal is obtained, an encoding process is performed on the down-sampled shape signal for each unit area of a predetermined size on the display screen to generate a shape-encoded bit string, and the signal in the middle of the encoding process is obtained. Decoding to generate a locally decoded shape signal, upsample the locally decoded shape signal, generate an upsampled shape signal having an increased number of pixels, and analyze the upsampled shape signal for each unit area And
According to the analysis result, a type of shape post-processing for making the object shape obtained from the up-sample shape signal smoother with respect to the up-sample shape signal,
And determining whether or not to perform the shape post-processing on the up-sample shape signal, and performing the predetermined shape post-processing on the up-sample shape signal, or performing the shape post-processing on the up-sample shape signal. Without giving
A shape encoding method characterized by outputting as an encoding shape signal used for encoding a signal other than an object shape signal in an image display signal. 5. A shape decoding method for decoding a shape-encoding bit sequence corresponding to a predetermined object shape in an image display signal to obtain a decoded object shape signal corresponding to the object shape, the shape decoding method comprising: The coded bit string is decoded for each unit area of a predetermined size on the display screen to generate a decoded shape signal, and the decoded shape signal is analyzed for each unit area. For the shape signal, at least one of a type of shape post-processing for making the object shape obtained from the decoded shape signal smoother, and whether or not to perform the shape post-processing on the decoded shape signal Is determined, the decoded shape signal is subjected to the predetermined shape post-processing, or without performing the shape post-processing, and is used for decoding of signals other than the object shape signal in the image display signal. And outputting the decoded object shape signal. 6. The shape decoding method according to claim 5, wherein the determination on the shape post-processing according to the analysis result is such that a decoded shape signal corresponding to a certain unit area is located outside the object. When it indicates
A shape decoding method, wherein the decoded shape signal subjected to the shape post-processing is also performed so as to be a signal indicating that the certain unit area is located outside the object. 7. The shape decoding method according to claim 6, wherein the post-shape processing is performed by applying a predetermined characteristic to the decoded shape signal so that an object shape obtained from the decoded shape signal becomes smoother. A shape decoding method characterized by performing conversion using the filter of (1). 8. A shape decoding method for decoding a shape-encoding bit sequence corresponding to a predetermined object shape in an image display signal to obtain a decoded object shape signal corresponding to the object shape. Decoding the encoded bit string for each unit area of a predetermined size on the display screen to generate a decoded shape signal, up-sampling the decoded shape signal, and obtaining an up-sampled shape signal having an increased number of pixels, The upsampled shape signal is analyzed for each unit area,
According to the analysis result, the type of shape post-processing for making the object shape obtained from the up-sample shape signal smoother with respect to the up-sample shape signal, and the shape with respect to the up-sample shape signal Determine at least one of whether to perform post-processing, the up-sample shape signal, subjected to the predetermined shape post-processing, or without performing the shape post-processing,
A shape decoding method characterized by outputting as a decoded object shape signal used for decoding a signal other than the object shape signal in the image display signal. 9. A shape encoding apparatus for encoding an object shape signal corresponding to a predetermined object shape in an image display signal, wherein the object shape signal is encoded for each unit area of a predetermined size on a display screen. To generate a shape-encoded bit string by performing an encoding process on the data, and to decode a signal in the middle of the encoding process to generate a locally decoded shape signal. Analyzed for each region, and according to the analysis result, for the locally decoded shape signal,
Determine at least one of a type of shape post-processing for making the object shape obtained from the locally decoded shape signal smoother, and whether or not to perform the shape post-processing on the locally decoded shape signal Shape analysis means for performing, based on an output of the shape analysis means, a shape post-processing means for performing a predetermined shape post-processing on the locally decoded shape signal, and the shape encoding based on an output of the shape analysis means Signal selecting means for selecting one of a locally decoded shape signal from the means and a locally decoded shape signal subjected to shape post-processing from the shape post-processing means. The output of the signal selecting means is displayed as an image. A shape encoding device for outputting as a shape signal for encoding used for encoding a signal other than an object shape signal in an application signal. 10. A shape encoding apparatus for encoding an object shape signal corresponding to a predetermined object shape in an image display signal, wherein the object shape signal is down-sampled to reduce the number of encoded pixels. Down-sampling means for obtaining a down-sampled shape signal obtained by performing the above-described down-sampled shape signal, and performing a coding process for each unit area of a predetermined size on a display screen to generate a shape-coded bit string; Shape encoding means for decoding an intermediate signal to generate a locally decoded shape signal; and up-sampling means for up-sampling the locally decoded shape signal to generate an up-sampled shape signal having an increased number of pixels. Analyzing the upsampled shape signal for each unit area;
According to the analysis result, a type of shape post-processing for making the object shape obtained from the up-sample shape signal smoother with respect to the up-sample shape signal,
And a shape analyzing means for determining at least one of whether or not to perform the shape post-processing on the up-sample shape signal, and a shape for performing a predetermined shape post-processing on the up-sample shape signal based on the analysis result. Post-processing means, based on the analysis result, a signal selecting means for selecting any of an up-sample shape signal from the up-sampling means and an up-sample shape signal subjected to shape post-processing from the shape post-processing means; A shape encoding apparatus comprising: an output unit configured to output an output of the signal selection unit as an encoding shape signal used for encoding a signal other than an object shape signal in an image display signal. 11. A shape decoding apparatus for decoding a shape-encoded bit sequence corresponding to a predetermined object shape in an image display signal to obtain a decoded object shape signal corresponding to the object shape, Shape decoding means for decoding the coded bit sequence for each unit area of a predetermined size on the display screen to generate a decoded shape signal; analyzing the decoded shape signal for each unit area; The type of shape post-processing for making the object shape obtained from the decoded shape signal smoother with respect to the decoded shape signal, and whether to perform the shape post-processing on the decoded shape signal Shape analysis means for determining at least one of the following; shape post-processing means for performing a predetermined shape post-processing on the decoded shape signal in accordance with the analysis result; and shape decoding in accordance with the analysis result. Signal selecting means for selecting one of a decoded shape signal from the stage and a decoded shape signal subjected to shape post-processing from the shape post-processing means, and outputs the signal from the signal selecting means to an image display signal. A shape decoding device for outputting as a decoded object shape signal used for decoding a signal other than the object shape signal in. 12. A shape decoding apparatus for decoding a shape-encoded bit sequence corresponding to a predetermined object shape in an image display signal to obtain a decoded object shape signal corresponding to the object shape, Shape decoding means for decoding a coded bit string for each unit area of a predetermined size on a display screen to generate a decoded shape signal; and an upsampler for upsampling the decoded shape signal and increasing the number of pixels thereof Up-sampling means for generating a shape signal; analyzing the up-sampled shape signal for each of the unit regions;
According to the analysis result, a type of shape post-processing for making the object shape obtained from the up-sample shape signal smoother with respect to the up-sample shape signal,
And a shape analyzing means for determining at least one of whether or not to perform the shape post-processing on the up-sample shape signal, and a shape for performing a predetermined shape post-processing on the up-sample shape signal according to the analysis result Post-processing means, and signal selecting means for selecting any of an up-sample shape signal from the up-sampling means and an up-sample shape signal subjected to shape post-processing from the shape post-processing means in accordance with the analysis result. A shape decoding device for outputting an output of the signal selection means as a decoded object shape signal used for decoding a signal other than the object shape signal in the image display signal. 13. A data storage medium storing, by a computer, a program for performing the shape encoding method according to claim 1, wherein the program stores the program on a display screen with respect to the object shape signal. A process of generating a shape-encoded bit string by performing an encoding process for each unit area of the size, and a process of generating a locally decoded shape signal by decoding a signal in the middle of the encoding process; Analyzed for each unit area, and according to the analysis result, for the locally decoded shape signal,
Determine at least one of a type of shape post-processing for making the object shape obtained from the locally decoded shape signal smoother, and whether or not to perform the shape post-processing on the locally decoded shape signal Processing, and subjecting the locally decoded shape signal to the predetermined shape post-processing, or without performing the shape post-processing, to encode a signal other than the object shape signal in the image display signal. Outputting a shape signal for encoding processing to be used. 14. A data storage medium storing a program for performing the shape decoding method according to claim 5 by a computer, wherein the program stores the shape encoded bit string in a predetermined size on a display screen. Processing for generating a decoded shape signal by decoding for each unit area, analyzing the decoded shape signal for each unit area, and, according to the analysis result, the decoded shape signal for the decoded shape signal A process of determining at least one of a type of shape post-processing for making the object shape obtained from the smoother, and whether or not to perform the shape post-processing on the decoded shape signal; and The shape signal is subjected to the above-mentioned predetermined shape post-processing or to the above-mentioned shape post-processing without being subjected to the above-mentioned shape post-processing, for decoding of signals other than the object shape signal in the image display signal. Outputting a decoded object shape signal to be used. 15. An object shape signal indicating the shape of an object included in an image display signal forming an image space including an object having an arbitrary shape is input, and the input object shape signal is converted to the image space. A shape encoding method for encoding for each unit area of a predetermined size by dividing the object shape signal, the object shape signal corresponding to the target unit area to be encoded is subjected to a compression process at a predetermined compression rate to perform a compression shape. A signal is generated, and at this time, a plurality of compression / decompression processes having different compression ratios are performed as a compression / decompression process of performing a decompression process corresponding to the compression process on the compressed shape signal to generate a locally decoded shape signal. Analysis of the locally decoded shape signal obtained by each compression / decompression process having a different compression ratio indicates that the target unit area included in the input object shape signal is located outside the object. The compression / expansion processing with the highest compression ratio, in which the information is held in the local decoded shape signal, is selected from the plurality of compression / expansion processings, and the compression shape obtained by the selected compression / expansion processing is selected. A shape encoding method comprising encoding a signal and outputting the encoded signal as a shape encoded signal. 16. The shape encoding method according to claim 15, wherein the compression processing in each of the compression / decompression processings having different compression ratios for the object shape signal corresponding to the target unit area includes a predetermined down-conversion to the object shape signal. The decompression process in each of the compression and decompression processes includes an upsampling process corresponding to the downsampling process on the compressed shape signal, and a plurality of compressions having different compression rates. One of the compression / decompression processes selected from the decompression process is such that the position of the target unit area included in the input object shape signal is outside the object regardless of the encoding distortion caused by the downsampling process and the upsampling process. Characterized in that the information indicating that the information is held in the locally decoded shape signal Shape coding method that. 17. An object shape signal indicating a shape of an object included in an image display signal forming an image space including an object having an arbitrary shape is input, and the input object shape signal is converted to the image space. A shape encoding method for encoding for each unit area of a predetermined size by dividing the object shape signal, the object shape signal corresponding to the target unit area to be encoded is subjected to a compression process at a predetermined compression rate to perform a compression shape. A signal is generated, and at this time, a plurality of compression / decompression processes having different compression ratios are performed as a compression / decompression process of performing a decompression process corresponding to the compression process on the compressed shape signal to generate a locally decoded shape signal. By analyzing the locally decoded shape signal obtained by each compression / decompression process having a different compression ratio, it is confirmed that the target unit area included in the input object shape signal is not located outside the object. The compression / expansion processing with the highest compression ratio, in which the information is held in the local decoded shape signal, is selected from the plurality of compression / expansion processings, and the compression shape obtained by the selected compression / expansion processing is selected. A shape encoding method comprising encoding a signal and outputting the encoded signal as a shape encoded signal. 18. The shape encoding method according to claim 17, wherein the compression processing in the compression / decompression processing with different compression ratios for the object shape signal corresponding to the target unit area includes a predetermined down-conversion for the object shape signal. The decompression process in each of the compression and decompression processes includes an upsampling process corresponding to the downsampling process on the compressed shape signal, and a plurality of compressions having different compression rates. One compression / decompression process selected from the decompression process is such that the target unit area included in the input object shape signal is located outside the object, regardless of the encoding distortion caused by the downsampling process and the upsampling process. Information indicating that the local decoding shape signal is not stored. Shape coding method comprising and. 19. An object shape signal indicating a shape of an object included in an image display signal forming an image space including an object having an arbitrary shape is input, and the input object shape signal is converted to the image space. A shape encoding apparatus that encodes a unit area of a predetermined size by dividing the object shape signal corresponding to the target unit area to be encoded by performing a compression process at a predetermined compression rate. A plurality of compression / expansion units having different compression ratios as compression / expansion means for generating a signal and performing a decompression process corresponding to the compression process on the compressed shape signal to generate a locally decoded shape signal; By analyzing the locally decoded shape signal output from each of the compression / decompression devices having different rates, information indicating that the target unit area is located outside the object, which is included in the input object shape signal. Analysis means for specifying a compression / expansion unit which performs compression / expansion processing with the highest compression ratio, which is held in the locally decoded shape signal, and receives an output of the analysis means, A signal selecting means for selecting a compression shape signal as an output of the compression / decompression device designated by the analysis means; and a shape encoding by coding the compression shape signal from the compression / decompression device selected by the signal selection means. A shape encoding device comprising: an encoder that outputs a signal. 20. The shape encoding apparatus according to claim 19, wherein each of the compression and decompression units includes a part for the object shape signal as a part of the compression processing.
Performing a down-sampling process having a predetermined down-sampling rate, as part of the decompression process, for the compressed shape signal, configured to perform up-sampling process corresponding to the down-sampling process, the analysis means, Regardless of the encoding distortion caused by the down-sampling process and the up-sampling process, the information included in the input object shape signal and indicating that the target unit area is located outside the object is the local decoding. A shape encoding device configured to designate a compression / expansion unit that performs a compression / expansion process held by an encoded shape signal. 21. An object shape signal indicating the shape of an object included in an image display signal forming an image space including an object having an arbitrary shape is input, and the input object shape signal is converted to the image space. A shape encoding device that encodes a unit area of a predetermined size by dividing the object shape signal corresponding to a target unit area to be encoded by performing a compression process at a predetermined compression rate. A plurality of compression / expansion units having different compression ratios as compression / expansion means for generating a signal and performing a decompression process corresponding to the compression process on the compressed shape signal to generate a locally decoded shape signal; By analyzing the locally decoded shape signal output from each compression / decompression device having a different rate, information indicating that the target unit area is not located outside the object, which is included in the input object shape signal. Analysis means for specifying a compression / expansion device which performs compression / expansion processing with the highest compression ratio, which is held in the locally decoded shape signal, and receives the output of the analysis means, and outputs the outputs of the plurality of compression / expansion processing. A signal selecting means for selecting a compression shape signal as an output of the compression / decompression device designated by the analysis means; and a shape encoding by coding the compression shape signal from the compression / decompression device selected by the signal selection means. A shape encoding device comprising: an encoder that outputs a signal. 22. The shape encoding apparatus according to claim 21, wherein each of the compression / decompression units includes:
Performing a down-sampling process having a predetermined down-sampling rate, and as a part of the decompression process, for the compressed shape signal, configured to perform an up-sampling process corresponding to the down-sampling process, Regardless of the encoding distortion caused by the down-sampling process and the up-sampling process, the information included in the input object shape signal and indicating that the target unit area is not located outside the object is the local decoding. A shape encoding apparatus characterized in that it is configured to designate a compression / expansion unit that performs compression / expansion processing held by a shape signal. 23. An object shape signal indicating the shape of an object included in an image display signal forming an image space including an object having an arbitrary shape is input, and the input object shape signal is converted to the image space. Is a data storage medium storing a program for causing a computer to perform shape encoding processing for encoding for each unit area of a predetermined size by dividing the program, wherein the program is an object of the encoding. The object shape signal corresponding to the target unit area is subjected to a compression process at a predetermined compression ratio to generate a compressed shape signal. A first process of performing a plurality of compression / expansion processes having different compression ratios as a compression / expansion process for generating a signal; By the analysis of the shape signal, information indicating that the target unit area is located outside the object, which is included in the input object shape signal, is held in the locally decoded shape signal. A second process for selecting a compression / expansion process having a high rate from among the plurality of compression / expansion processes; Third
A data storage medium configured to perform the above-described processing. 24. An object shape signal indicating the shape of an object included in an image display signal forming an image space including an object having an arbitrary shape is input, and the input object shape signal is converted into the image space. Is a data storage medium storing a program for causing a computer to perform shape encoding processing for encoding for each unit area of a predetermined size by dividing the program, wherein the program is an object of the encoding. The object shape signal corresponding to the target unit area is subjected to a compression process at a predetermined compression ratio to generate a compressed shape signal. A first process of performing a plurality of compression / expansion processes having different compression ratios as a compression / expansion process for generating a signal; a local decoding obtained by each of the compression / expansion processes having different compression ratios By analyzing the shape signal, information indicating that the target unit area is not located outside the object, which is included in the input object shape signal, is held in the locally decoded shape signal. A second process of selecting a high compression / expansion process from among the plurality of compression / expansion processes; and 3
A data storage medium configured to perform the above-described processing.