WO1998031151A1 - Image processing method, image processing device, and data recording medium - Google Patents
Image processing method, image processing device, and data recording medium Download PDFInfo
- Publication number
- WO1998031151A1 WO1998031151A1 PCT/JP1998/000040 JP9800040W WO9831151A1 WO 1998031151 A1 WO1998031151 A1 WO 1998031151A1 JP 9800040 W JP9800040 W JP 9800040W WO 9831151 A1 WO9831151 A1 WO 9831151A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- resolution
- image
- low
- signal
- block
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/30—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using hierarchical techniques, e.g. scalability
- H04N19/33—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using hierarchical techniques, e.g. scalability in the spatial domain
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T9/00—Image coding
- G06T9/20—Contour coding, e.g. using detection of edges
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
- H04N19/17—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
- H04N19/176—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/20—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using video object coding
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/30—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using hierarchical techniques, e.g. scalability
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
- H04N19/51—Motion estimation or motion compensation
- H04N19/513—Processing of motion vectors
- H04N19/517—Processing of motion vectors by encoding
Definitions
- the present invention relates to an image processing method, an image processing apparatus, and a data recording medium, and more particularly to a hierarchical encoding process for recording or transmitting an image signal with a smaller number of bins and without deteriorating image quality.
- the present invention relates to a ⁇ -layer decoding process and a recording medium storing a program for realizing the tier coding process or the ⁇ -layer decoding process.
- the image signal In order to enable image processing to be performed for each individual object (object) displayed on the display screen, the image signal must be represented by the shape of the object in addition to the normal luminance signal and color difference signal.
- a shape signal is required.
- an image signal including a shape signal in addition to the luminance signal and the color difference signal, that is, an image signal having shape information of an object is simply referred to as an image signal.
- Such an image signal is suitable for multimedia that simultaneously associates image information, audio information, and the like from the viewpoint that it can be handled on an object-by-object basis.
- Standardization activities are being carried out by the MPEG (Moving Picture Experts Group) 4, a working group of the International Organization for Standardization / International Electrotechnical (OMRA) Joint Technical Commission.
- FIG. 22 shows a display surface (hereinafter, also referred to as a frame) corresponding to image signals having different resolutions.
- LF is a display screen corresponding to a low-resolution image signal (FIG. 22 (a))
- HF is a display screen corresponding to a high-resolution image signal (FIG. 22 (b)).
- Lob is an image of one object displayed on the frame LF by a low-resolution image signal
- Hob is displayed on the frame HF by a high-resolution image signal.
- An image of two objects, and a dot display area is an area inside each object.
- a rectangular area corresponding to each object and including the object is set on one frame, and the rectangular area is further blocked ( In the MPEG 4 evaluation model, it is divided into 16 x 16 pixel square blocks.
- the encoding process of the image signal for each object is performed in units of blocks constituting the rectangular area.
- a low-resolution rectangular area LR corresponding to the object L0b is set on the low-resolution frame LF, and the As shown in Fig. 2 (d), it is necessary to set a ⁇ resolution rectangular area HR corresponding to the object Hob on the high resolution frame HF.
- FIG. 23 is a block diagram for explaining a conventional hierarchical image encoding device.
- the conventional image encoding apparatus 200a is configured to receive an image input signal and perform a hierarchical encoding process using the image input signal as a high-resolution image signal HSg.
- the hierarchical image coding device 20a includes a sub-sampler 2 that sub-samples the high-resolution image signal HSg to generate a low-resolution image signal LSg, and a sub-sampler 2 that generates the low-resolution image signal LSg.
- a low-resolution encoding unit 201L that performs an encoding process on the image data to generate a low-resolution encoded signal LEg.
- the hierarchical image encoding device 20 O a includes a decoder 9 a for decoding the low-resolution encoded signal L £ g, and an up-sample for up-sampling the output L dg of the decoder 9 a.
- a high-resolution unit that performs encoding processing on the high-resolution image signal HS g based on the output LA g of the sampler 10 a and the up-sampler 10 a to output a high-resolution encoded signal HE g Encoding unit 201H.
- the low-resolution encoding unit 201L detects information such as the position and size of the low-resolution rectangular area LR corresponding to each object on the low-resolution frame LF based on the low-resolution surface image signal LSg. Then, based on the rectangular signal LRg, an area detector 3 that outputs the information as a rectangular signal LRg, and an image signal LOg corresponding to the rectangular area LR is extracted from the low-resolution image signal LSg. And a region extractor 5. Further, the encoding unit 2 OIL converts the image signal (object-corresponding image signal) L ⁇ g corresponding to the rectangular area LR into a block MB composed of 16 ⁇ 16 pixels that partition the rectangular area.
- a block generator 6 that divides and outputs an image signal (blocked image signal) LBg corresponding to each block, and encodes the blocked image signal LBg to generate a low-resolution encoded signal LEg And an encoder 7 for outputting. Further, based on the high-resolution image signal HSg, the high-resolution encoding unit 201H transmits information such as the position and size of the high-resolution rectangular area HR corresponding to each object on the high-resolution frame HF. And outputs the information as a rectangular signal HRg. An image signal HOg corresponding to the rectangular area HR is obtained from the high-resolution surface image signal HS based on the rectangular signal HRg and the rectangular signal HRg.
- the encoding unit 201H divides the image signal (object-corresponding image signal) HO g corresponding to the rectangular area HR so as to correspond to a block MB consisting of 16 ⁇ 16 pixels that divides the rectangular area. Then, a block generator 15 that outputs an image signal (blocked image signal) HBg corresponding to each block, and encodes the blocked image signal HBg to generate a high-resolution encoded signal HEg And an encoder 16 for output.
- the sub-sampler 2 converts the high-resolution image signal HS g into a low-resolution image signal LS g by the sub-sample.
- the area detector 3 of the low-resolution encoding unit 201L performs processing on a low-resolution frame: LF as shown in FIG. 22 (c).
- the range of the rectangular area LR including the object L ob is detected, and the rectangular area LR is detected. And outputs information such as the position and size of the rectangular signal LR g.
- the region extractor 5 of the encoding unit 201L extracts the object-related image signal LOg corresponding to the rectangular region LR from the low-resolution image signal LSg based on the rectangular signal LRg.
- the block generator 6 of the encoding unit 201 L as shown in FIG.
- the object-corresponding image signal L ⁇ g includes a plurality of blocks that divide the low-resolution rectangular area LR. It is divided so as to correspond to each MB, and is output to the encoder 7 as a block image signal LB g corresponding to each block MB. Then, the encoder 7 performs an encoding process on the block image signal LBg, and outputs a low-resolution encoded signal LEg from the encoding unit 201L.
- the low-resolution coded signal LE g is converted into a low-resolution decoded signal L dg by a decoding process in a decoder 9a, and the decoded signal L dg is converted by an up-sampler 10a.
- the signal is converted into an interpolated decoded signal LAg having the same spatial resolution as the high-resolution image signal, and is output to the encoder 16 of the high-resolution encoding unit 201H.
- the high-resolution encoding unit 201H performs substantially the same processing as the low-resolution encoding unit 201L.
- the area detector 12 of the encoding unit 201H determines the processing target on the high-resolution frame HF as shown in FIG. 22 (d). A range of a rectangular area HR including the object Hob is detected, and information such as the position and size of the rectangular area HR is output as a rectangular signal HRg. Then, based on the rectangular signal HRg, the region extractor 14 of the encoding unit 201H converts the high-resolution image signal HSg into an object-corresponding image corresponding to the rectangular region HR. g is extracted. Further, in the blocker 15 of the encoding unit 201H, as shown in FIG.
- the object-corresponding image signal HO g divides the high-resolution rectangular area HR into each of a plurality of blocks MB. And output to the encoder 16 as a block image signal HBg corresponding to each block MB. Then, in the encoder 16, the encoding processing of the block image signal HB g is performed based on the interpolation decoded signal LA g, and the encoding section 201 H outputs the high-resolution encoded signal HE g is output.
- Low resolution encoding encoded by the hierarchical image encoding device 200a as described above The signal LE g can be subjected to a decoding process for generating a decoded signal corresponding to the low-resolution image signal LS g based on the signal LE g and the rectangular signal LR g.
- a decoding process for generating a decoded signal corresponding to the high-resolution image signal HSg can be performed.
- the high resolution image signal is obtained by using the correlation of the pixel values between the two image signals LSg and HSg with reference to the low resolution image signal LSg.
- HS g can be encoded with a smaller number of bits than in the case where it is independently encoded.
- FIG. 24 is a block diagram for explaining a conventional hierarchical image decoding apparatus.
- the hierarchical image decoding device 200b receives the low-resolution encoded signal LEg and the high-resolution encoded signal HEg encoded by the conventional image encoding device 200a shown in FIG. The decoding process is performed.
- the hierarchical image decoding device 200 b includes a low-resolution decoding unit 202 L that performs a decoding process on the low-resolution encoded signal LE g to generate a low-resolution reproduction signal LC g,
- the up-sampler 10 b interpolating the signal LD g in the middle of the decoding process in the section 202 L by up-sampling, and the high-resolution encoded signal HE g based on the output AD g of the amplifier sampler 10 b.
- a high-resolution decoding unit 202H that performs a decoding process to generate a high-resolution reproduction signal HC.
- the low-resolution decoding unit 202L performs a decoding process on the low-resolution encoded signal LE g to generate a low-resolution decoded signal LD g corresponding to each block, and a decoder 9;
- a deblocking unit 20 that integrates the resolution decoded signal LD g to generate a decoded integrated signal LI g corresponding to the rectangular area LR, and a decoded integrated signal LI g corresponding to the rectangular area so that the rectangular region LR on the position on the low-resolution frame LF illustrated by rectangle signal LR g from the encoding device 200 a is disposed, perforated and an area combiner 21 for combining with other image signal for forming a frame are doing.
- the high-resolution decoding section 202H performs a decoding process on the high-resolution coded signal HE g based on the output AD g of the up-sampler 10b, and performs high-resolution decoding corresponding to each block.
- the decoded integrated signal HI g corresponding to the rectangular area HR is converted into one frame so that the rectangular area HR is arranged at a position on the frame HF indicated by the rectangular signal HR g from the encoding device 200a.
- a region synthesizer 32 for synthesizing the image signal with other image signals forming the image signal.
- the decoding of the low-resolution coded signal LE g is the low-resolution decoding section 202 L
- the decoder 9 performs a decoding process to generate a low-resolution decoded signal LD g.
- the low-resolution decoded signal LD g is interpolated by the up-sampler 10 b by the up-sampling process and converted into an interpolated decoded signal AD g having a spatial resolution corresponding to the high resolution.
- the low-resolution decrypted signal LD g is further integrated by the deblocking unit 20 to generate a decoded integrated signal LI g corresponding to the rectangular area LR.
- the decoded integrated signal LI g is synthesized by the area synthesizer 21 with other image signals forming one frame based on the rectangular signal LR g from the encoding device 200a, and the low-resolution reproduction is performed. Output as signal LC g.
- an image of the rectangular area LR based on the decoded integrated signal LI g is arranged at a position on the frame LF indicated by the rectangular signal LR g.
- the high-resolution coded signal HE g is subjected to decoding processing by the decoder 30 of the high-resolution decoding section 202 H based on the output AD g of the up-sampler 10 b, and A resolution decoded signal HD g is generated.
- the high-resolution decoded signal HD g is further integrated by the deblocking unit 31 to generate a decoded integrated signal HI g corresponding to the image in the rectangular area HR.
- the decoded integrated signal HI g is combined with another image signal corresponding to one frame by the area combiner 32 based on the rectangular signal HR g from the encoding device 200a, and is subjected to high-resolution reproduction.
- Faith No. Output as HC g By this combining processing, the image of the rectangular area HR based on the decoded integrated signal HI g is arranged at the position on the frame HF indicated by the rectangular signal HR g.
- the rectangular region LR is arranged at a predetermined position in the frame LF. Since the low-resolution decoding signal Ig corresponding to the rectangular area LR is subjected to area synthesis processing, the low-resolution decoding processing performed on the rectangular area LR corresponding to each object in the frame FL ⁇ The encoded signal LE g can be decoded.
- the high-resolution coded signal HE g is decoded with reference to the low-resolution decoded signal LD g to generate a high-resolution decoded signal HD g, and then the decoded signal HD g is inverted.
- the high-resolution decoded signal HI g corresponding to the rectangular area HR is subjected to area synthesis processing so that the rectangular area HR is arranged at a predetermined position in the frame HF.
- high resolution coded signal HE g obtained by hierarchical coding processing for the rectangular region HR, corresponding to the object can be correctly decoded.
- the process of detecting the range of the rectangular region LR in the low-resolution frame LF ⁇ and the process of detecting the range of the rectangular region HR in the high-resolution frame HF are performed. Therefore, as shown in FIGS. 22 (c) and 22 (d), the low-resolution image L of each block MB that partitions the low-resolution rectangular area LR The space position with respect to 0b does not match the spatial position with respect to the high-resolution image Hob of each block MB that partitions the high-resolution rectangular area HR.
- the present invention has been made to solve the above-described problem.
- a high-level encoding method that refers to a low-resolution image signal is performed.
- An object of the present invention is to obtain a data recording medium storing a program capable of realizing an encoding process and a hierarchical decoding process by software. Disclosure of the invention
- An image processing method includes at least a hierarchical image signal that forms a plurality of image spaces having different spatial resolutions based on an image input signal having shape information of an object.
- a low-resolution image signal and a high-resolution image signal are generated, and the high-resolution image signal is divided so as to correspond to a high-resolution block including a predetermined number of pixels in a high-resolution image space.
- the coding process is performed on a high-resolution image signal forming a target high-resolution block to be processed, which is divided so as to correspond to a low-resolution block consisting of a predetermined number of pixels in the low-resolution image space.
- a hierarchical image encoding method that sequentially performs with reference to a low-resolution image signal that forms a low-resolution block corresponding to a resolution block.
- the reference low-resolution block to be referred to is determined by determining the space position in the low-resolution image space corresponding to the reference low-resolution block and the low resolution associated with the spatial position in the high-resolution image space of the target high-resolution block. It is a block.
- the image processing method when encoding a high-resolution image signal, a space having a correlation with a spatial position of a target high-resolution block to be encoded in a low-resolution image space.
- the image signal of the low-resolution block located at the position is referred to, and it is possible to perform the hierarchical encoding process on the image signal having the shape information of the object without deteriorating the encoding efficiency.
- each pixel constituting the high-resolution plane image space is replaced with an image of the low-resolution image space. This is a one-to-one correspondence with each pixel in a resolution-converted image space whose spatial resolution is equal to the high-resolution image space obtained by the resolution conversion.
- all of the plurality of pixels in the high-resolution block correspond to predetermined pixels in the resolution conversion block obtained by converting the resolution of the low-resolution block. Efficiency can be further improved.
- the present invention (Claim 3) is the image processing method according to Claim 1, wherein the number of pixels constituting the reference low-resolution block and the target high-resolution block are constituted. This is the same as the number of pixels.
- the image processing method having such a configuration it is possible to share the block encoder and the encoder between the high-resolution image signal and the low-resolution image signal, so that the circuit configuration can be made compact.
- the present invention (claim 4) is the image processing method according to claim 1, wherein a relative position of the reference low-resolution block in the low-resolution image space and a high-resolution image of the target high-resolution block are set. It is the same as the relative position in space.
- the pixel value of each pixel constituting the high-resolution block and the low-resolution block are configured.
- the difference from the pixel value of each pixel to be processed does not become unnecessarily large, and the hierarchical coding process can be performed with high coding efficiency.
- the present invention (Claim 5) is the image processing method according to Claim 1, wherein the reference low-resolution block corresponds to an encoding mode for identifying an encoding processing method.
- the encoding method of the mode signal indicating the encoding mode for identifying the encoding processing method for the target high-resolution block is switched.
- the mode signal indicating the encoding mode of the high resolution image signal is assigned. The number of coded bits in the coding process can be reduced.
- the present invention is the image processing method according to Claim 5, wherein the encoding mode is set such that the boundary of the shape of the object displayed on the image space is set to the pair. This is an encoding mode indicating whether or not the block is included in the elephant high resolution block.
- the mode signal indicating the encoding mode of the high-resolution image signal is used.
- the encoding mode is set by sequentially encoding an image signal corresponding to the reference low-resolution block for each pixel. This is a coding mode that indicates whether the coding process to be performed is performed along the scanning direction of the horizontal scanning direction or the vertical scanning direction.
- a high-resolution image signal is encoded in a scanning direction in which the correlation between pixel values constituting the high-resolution image signal is large.
- a short code is assigned to the mode signal indicating the coding mode of the high-resolution image signal, thereby coding the coding mode signal. It is possible to reduce the number of coding bits required for the coding.
- the present invention (claim 8) is the image processing method according to claim 1, wherein the motion corresponding to the reference low-resolution bronk and indicating the motion of an object in the low-resolution image space.
- a coding method of the motion information indicating the motion of the object in the high resolution image space corresponding to the target high resolution block is switched.
- the motion between the high-resolution block and the corresponding low-resolution block is reduced.
- the vectors match, by assigning a short code to the mode signal indicating the motion vector (coding mode) of the high-resolution image signal, the number of coding bins necessary for coding the motion vector is obtained. Can be reduced.
- the present invention provides the image processing method according to Claim 1.
- the encoding method of the motion information indicating the motion of the object in the high-resolution image space corresponding to the target high-resolution block is switched.
- the prediction vector is obtained from the motion vector of the processed high-resolution block corresponding to the target high-resolution block and the motion vector of the low-resolution block corresponding to the target high-resolution block. Since the motion vector of the target high-resolution block is encoded based on the prediction vector, the image signal has a correlation between pixel values between screens, and the high-resolution image signal and the low-resolution image Since there is a large correlation between the signal and the pixel value, the error between the motion vector of the target high-resolution block and the predicted motion vector is reduced, and the motion vector of the high-resolution image signal is coded. Thus, the number of coding bits required for the operation can be reduced.
- An image processing method includes at least two block-coded hierarchical coded signals obtained by performing a hierarchical coding process on an image signal having shape information of an object.
- the low-resolution coded signal is decoded to generate a low-resolution decoded signal corresponding to a low-resolution block composed of a predetermined number of pixels in the low-resolution image space, and the low-resolution decoded signal is integrated.
- To generate a low-resolution integrated signal corresponding to the low-resolution image space and decode the high-resolution coded signal of the two block-coded hierarchical coded signals with reference to the corresponding low-resolution decoded signal.
- a hierarchical image decoding method for generating a high-resolution integrated signal wherein a reference low-resolution block referred to in the decoding process of the high-resolution coded signal is referred to as a spatial position in the low-resolution image space.
- the spatial position of the target high-resolution block to be decoded is set in the low-resolution image space.
- the decoded signal of the low-resolution block located at the spatial position correlated with the position is referred to, and it corresponds to the hierarchical coding processing of the image signal having the shape information of the object, which suppresses the deterioration of the coding efficiency.
- Hierarchical decoding processing can be realized.
- the invention is the image processing method according to claim 10, wherein each pixel constituting the high-resolution image space is converted into a resolution of the low-resolution image space. And has a one-to-one correspondence with individual pixels in a resolution-converted image space having a spatial resolution equal to the high-resolution image space.
- a plurality of pixels in a high-resolution block all correspond to predetermined pixels in a resolution conversion block obtained by resolution conversion of a low-resolution block.
- the number of surface elements constituting the reference low-resolution block and the target high-resolution block are determined. This is the same as the number of pixels.
- a decoder and an inverse blocker can be shared between a high-resolution coded signal and a low-resolution coded signal, and the circuit configuration can be made compact. Can be.
- the present invention (claim 13) is the image processing method according to claim 10, wherein a relative position of the reference low-resolution block in the low-resolution image space and a high-resolution image of the target high-resolution block are set. This is the same as the relative position in the air question.
- the pixel value of each pixel constituting the high-resolution block and the low-resolution block are calculated. It is possible to realize a hierarchical decoding process corresponding to a hierarchical encoding process with high encoding efficiency in which the difference between the pixel value of each constituent pixel and the pixel value is not unnecessarily large.
- the present invention (claim 14) is the image processing method according to claim 10, wherein the reference low-resolution block has a coding mode for identifying a decoding processing method.
- Decoding method for the target high-resolution block This is to switch the decoding method of the mode coded signal indicating the coding mode for identifying the data.
- the coding mode of the high-resolution plane image signal is assigned by assigning a short code to the mode signal indicating the coding mode of the high-resolution block that matches the coding mode for the low-resolution block.
- the present invention (claim 15) is the image processing method according to claim 14, wherein the encoding mode is set such that the boundary of the shape of the object displayed in the image space is the target height. This is an encoding mode that indicates whether or not it is included in the resolution block.
- the mode signal indicating the encoding mode of the high-resolution image signal is used.
- the present invention (claim 16) is the image processing method according to claim 14, wherein the encoding mode is sequentially decoded for each pixel of an image signal corresponding to a reference low-resolution block.
- the encoding mode indicates whether the decoding process is performed along the scanning direction of the horizontal scanning direction or the vertical scanning direction.
- the mode signal indicating the encoding mode of the high-resolution image signal By assigning a short code to the code and performing coding, it is possible to realize a hierarchical decoding process corresponding to the hierarchical coding process in which the number of code S-coded bits is reduced.
- the invention is the image processing method according to Claim 10, wherein the motion of the object in the low-resolution dual image space corresponding to the reference low-resolution block is indicated.
- a method of decoding the motion information indicating the motion of the object in the high-resolution image space corresponding to the target high-resolution block is switched.
- the image processing method having such a configuration when the motion vector between the high-resolution block and the corresponding low-resolution block matches, the motion vector of the high-resolution image signal (encoding mode By assigning a short code to the mode signal indicating), it is possible to realize a hierarchical decoding process corresponding to the hierarchical encoding process in which the number of encoded bits is reduced.
- the present invention (claim 18) is the image processing method according to claim 10, wherein the processed high-resolution block, which has already been subjected to the decoding processing, is an object in a high-resolution image space.
- the high-resolution image corresponding to the target high-resolution block with reference to the motion information indicating the motion of the object and the motion information indicating the motion of the object in the low-resolution image space corresponding to the reference low-resolution block. It switches the method of decoding motion information indicating the motion of an object in space.
- the hierarchical decoding processing corresponding to the hierarchical coding processing in which the error between the motion vector of the target high-resolution block and the prediction motion vector is small and the number of coding bits is reduced is reduced. Can be realized.
- An image processing apparatus receives an image input signal having shape information of an object and encodes a hierarchical image signal forming a plurality of image spaces having different spatial resolutions.
- a hierarchical image encoding device comprising: a sub-sampling means for sub-sampling the image input signal to generate a low-resolution image signal; and converting the low-resolution image signal into a low-resolution image signal comprising a predetermined number of pixels.
- a first blocking means for blocking to correspond to the resolution block, and a first code for sequentially performing an encoding process on a low-resolution image signal forming a low-resolution block to be subjected to the encoding process Means for converting the image input signal as a high-resolution image signal into blocks corresponding to a high-resolution block consisting of a predetermined number of pixels in a high-resolution image space.
- a second low-resolution block referred to at the time of the above-described encoding processing, and a spatial position in the low-resolution plane image space of the second low-resolution block corresponds to the reference low-resolution block.
- the image processing apparatus when encoding a high-resolution image signal, a space having a correlation with a spatial position of a target high-resolution block to be encoded in a low-resolution image space.
- the image signal of the low-resolution block located at the position is referred to, and it is possible to perform the hierarchical encoding process on the image signal having the shape information of the object without deteriorating the encoding efficiency.
- An image processing apparatus includes at least two block-coded hierarchical coded signals obtained by performing a hierarchical coding process on an image signal having shape information of an object.
- a hierarchical image decoding device that decodes a low-resolution coded signal of the two block-coded hierarchical coded signals to obtain a predetermined number of pixels in a low-resolution image space.
- the first decoding means for generating a low-resolution decoding signal corresponding to the low-resolution block consisting of: and the low-resolution decoding signal corresponding to each of the low-resolution blocks are integrated into the low-resolution image space.
- First deblocking means for generating a corresponding low-resolution integrated signal; and referencing the high-resolution coded signal of the two broken hierarchical coded signals with reference to the corresponding low-resolution decoded signal.
- Decrypt Second decoding means for generating a high-resolution decoded signal corresponding to a high-resolution block composed of a predetermined number of pixels in a high-resolution image space; and high-resolution decoding corresponding to each of the high-resolution blocks.
- Second deblocking means for integrating the decoded signals to generate a high-resolution integrated signal corresponding to the high-resolution image space, and for decoding the high-resolution encoded signal.
- the reference low-resolution block to be referenced is identified by the spatial position in the low-resolution image space of the target high-resolution block to be subjected to the decoding process corresponding to the reference low-resolution block. This is a low-resolution block associated with a position.
- the image processing device when decoding a high-resolution image signal, a space having a correlation with a spatial position of a target high-resolution block to be decoded in a low-resolution image space.
- the image decoding signal of the low-resolution block located at the position is referred to, and the layer decoding corresponding to the hierarchical coding processing of the image signal having the shape information of the object with the deterioration of the coding efficiency suppressed. Processing can be realized.
- the data recording medium according to the present invention (claim 21) is stored on a computer.
- a data recording medium storing a program for performing a layer image encoding process, wherein the computer is configured to perform the layer image encoding process according to the image processing method according to claim 1. .
- a data recording medium having such a configuration when encoding a high-resolution image signal, a space having a correlation with the spatial position of a target high-resolution block to be encoded in a low-resolution image space. Since the image signal of the low-resolution block located at the position is referred to, it is possible to realize by a computer hierarchical coding processing on the image signal having the shape information of the object without deteriorating the coding efficiency.
- a data recording medium according to the present invention (Claim 22) is a data recording medium storing a program for causing a computer to perform a hierarchical image decoding process.
- a hierarchical image decoding process according to the image processing method described in 10 is configured to be performed by a computer.
- a data recording medium having such a configuration, when decoding a high-resolution image signal, a space having a correlation with a spatial position of a target high-resolution block to be decoded in a low-resolution image space.
- the image decoding signal of the low-resolution block located at the position is referred to, and the hierarchical decoding processing corresponding to the hierarchical encoding processing of the image signal having the shape information of the object, which suppresses the deterioration of the encoding efficiency, is performed. It can be realized by a computer.
- FIG. 1 is a block diagram for explaining a hierarchical image encoding device as an image processing device according to Embodiment 1 of the present invention.
- 2 (a) to 2 (d) are diagrams for explaining the operation of the hierarchical image coding device according to the first embodiment.
- FIG. 3 is a block diagram for explaining a hierarchical image encoding device as an image processing device according to Embodiment 2 of the present invention.
- FIGS. 4 (a) to 4 (f) are diagrams for explaining the operation of the hierarchical image encoding device according to the second embodiment of the present invention.
- FIG. 5 shows a hierarchical image encoding device as an image processing device according to Embodiment 3 of the present invention.
- FIG. 2 is a block diagram for explaining the arrangement.
- 6 (a) to 6 (d) are diagrams for explaining the operation of the hierarchical image encoding device according to Embodiment 3 of the present invention.
- FIG. 7 is a block diagram for explaining a hierarchical image encoding device as an image processing device according to Embodiment 4 of the present invention.
- FIG. 8 is a block diagram for explaining a hierarchical image decoding device as an image processing device according to Embodiment 5 of the present invention.
- 9 (a) and 9 (b) are block diagrams for explaining a hierarchical image encoding device as an image processing device according to Embodiment 6 of the present invention.
- FIGS. 10 (a) and 10 (b) are block diagrams for explaining a hierarchical image decoding device as an image processing device according to Embodiment 7 of the present invention.
- FIGS. 11 (a) and 11 (b) are block diagrams for explaining a hierarchical image encoding device as an image processing device according to Embodiment 8 of the present invention.
- FIGS. 12 (3) to 12 (d) are diagrams for explaining the operation of the hierarchical image coding apparatus according to the eighth embodiment.
- FIGS. 13 (a) and 13 (b) are block diagrams illustrating a hierarchical image decoding device as an image processing device according to Embodiment 9 of the present invention.
- FIGS. 14 (a) and 14 (b) are block diagrams for explaining a hierarchical image encoding device as an image processing device according to Embodiment 10 of the present invention.
- FIGS. 15 (a) and 15 (b) are pronk diagrams for explaining a hierarchical image decoding device as an image processing device according to Embodiment 11 of the present invention.
- FIGS. 16 (a) to 16 (c) are block diagrams for explaining a hierarchical image encoding device as an image processing device according to Embodiment 12 of the present invention.
- FIGS. 17 ( 3 ) to 17 (c) are block diagrams for explaining a hierarchical image decoding device as an image processing device according to Embodiment 13 of the present invention.
- FIGS. 18 (a) to 18 (c) are block diagrams for explaining a hierarchical image encoding device as an image processing device according to Embodiment 14 of the present invention.
- FIG. 19 (3) to FIG. 19 (d) are diagrams for explaining the operation of the hierarchical image coding device of the embodiment 14 above.
- FIG. 20 (a) to FIG. 20 (c) are broken diagrams for explaining a hierarchical image decoding device as an image processing device according to Embodiment 15 of the present invention.
- FIGS. 21 (a) to 21 (c) show data storing a program for realizing a hierarchical encoding process or a hierarchical decoding process by the image processing apparatus of each of the above-described embodiments by a computer system.
- FIG. 3 is a diagram illustrating a recording medium.
- FIG. 22 is a diagram for explaining a conventional hierarchical image encoding process.
- FIG. 23 is a block diagram for explaining a hierarchical image encoding device as a conventional image processing device.
- FIG. 24 is a block diagram for explaining a hierarchical image decoding device as a conventional image processing device.
- Embodiment 1 is a diagrammatic representation of the present invention.
- FIG. 1 is a block diagram for explaining an image processing device (hierarchical image coding device) according to Embodiment 1 of the present invention.
- 2 (a) to 2 (d) are schematic diagrams for explaining an encoding process performed by the hierarchical image encoding device according to the first embodiment.
- LF is a display screen corresponding to a low-resolution image signal (see Fig. 2 (a))
- HF is a display screen corresponding to a high-resolution image signal (see Fig. 2 (b)).
- Lob is a low-resolution image, an image of one object displayed on the frame (display screen) LF by a symbol
- Hob is displayed on the frame (display screen) HF by a high-resolution image signal.
- This is an image of one object, and the dot display area of each image L ob and H ob is an area inside each object.
- the hierarchical image encoding device 101 is configured to receive an image input signal and perform a hierarchical encoding process on the image input signal as a high-resolution image signal HSg.
- the hierarchical image encoding device 101 includes a sub-sampler 2 for sub-sampling the high-resolution image signal HSg, Low-resolution encoding low-resolution image signal LS g Image encoding unit 101L, a decoder 9a for decoding the low-resolution encoded signal LEg that is the output thereof, and an amplifier sample for up-sampling the output Ldg of the decoder 9a And a high-resolution encoding unit 101H for encoding the high-resolution image signal HSg based on the output LAg of the up-sampler 10a.
- each of the encoding units 101 L and 101 H has an area detector 30 a and 12, an area extractor 5 and 14, and a block extractor, respectively, as in the conventional hierarchical image encoding apparatus 200 a. It has a zipper 6 and 15 and an encoder 7 and 16a.
- the area detector 30a has a different configuration from that of the conventional hierarchical image coding apparatus 200a. That is, the area detector 30a refers to the rectangular signal HRg indicating the range of the rectangular area HR corresponding to the individual object Hob on the high-resolution frame HF (see FIG. 2 (b)). It is configured to detect the range of the rectangular area corresponding to each object L ob on the resolution frame LF (see Fig. 2 (a)). Specifically, the area detector 30a detects the spatial position HRp of the ⁇ resolution rectangular area HR from the high resolution rectangular signal HR g, and the low resolution image signal: an interpolated rectangular area obtained by up-sampling LS g. The spatial position LR of the low-resolution rectangular area LR is determined such that the AR spatial position ARp matches the spatial position HRp of the high-resolution rectangular area HR.
- the spatial position LRp of the low-resolution rectangular region LR is detected by determining whether the spatial position ARp of the interpolated rectangular region AR is higher than the spatial position LRp of the high-resolution rectangular region HR. It is only necessary to perform the position detection so as to match the position ⁇ HR p, and it is necessary to detect the position of the above-mentioned region so that the spatial position LR p of the low-resolution rectangular region LR and the spatial position HR p of the high-resolution rectangular region HR match. There is no. Also, the encoder 16a has exactly the same configuration as the encoder 6 in the conventional hierarchical image encoding device 200a.
- the sub-sampler 2 converts the high-resolution image signal HS g into a low-resolution image signal LS g by the sub-sample. Is converted to Then, in each of the encoding units 101 H and 101 L, the high-resolution image signal HS g and the low-resolution image signal LS g is encoded.
- the high-resolution encoding unit 101H performs exactly the same image signal processing as the high-resolution encoding unit 201H in the conventional hierarchical image encoding device 200a.
- the low-resolution encoding unit 101L except for the position detection processing of the low-resolution rectangular area LR on the frame LF by the area detector 30a, the low-resolution encoding in the conventional hierarchical image encoding apparatus 200a is performed. Processing exactly the same as that of the unit 201L is performed. Therefore, the following describes mainly the operation of the area detector 30a in the encoding process by the low-resolution encoding unit 101L.
- the rectangular area HR of the corresponding object H0b is detected by the area detector 12 of the high-resolution encoding unit 101H based on the high-resolution image signal HSg (second (b)
- the spatial detector LR detects the spatial position LRp of the rectangular region LR of the low-resolution image signal LSg based on the spatial position HRp of the high-resolution rectangular region HR (see FIG. ) See figure).
- the reference position HRp of the high-resolution rectangular area HR is shifted by a distance ( ⁇ ) in the horizontal direction and a distance (HAy) in the vertical direction from the reference position HFp of the high-resolution frame HF. If they are far apart, the low-resolution rectangular area LR has a horizontal distance (L x) and a vertical distance (L Ay) between its reference position LR p and the reference position LF p of the low-resolution frame LF. So that it is placed on the low-resolution frame LF.
- the distance ( ⁇ ) and the distance (LA x) and the distance ( ⁇ y) / distance (LA y) are the ratio of the spatial resolution of the high-resolution image to the spatial resolution of the low-resolution image (here, 2Z1)).
- the spatial position (reference position) of the low-resolution rectangular area LR on the low-resolution frame LF can be changed on the high-resolution frame HF. Can be matched to the spatial position (reference position) of the high-resolution rectangular area HR.
- the image signal LOg corresponding to the low-resolution rectangular area LR set on the low-resolution frame LF as described above is divided by the blocking unit 6 so as to correspond to each block LMB that divides the rectangular area LR. Is done. Further, the block-blocked image ⁇ signal LB s is encoded by the encoder 7 and output as a low-resolution encoded signal LE g. Is forced.
- the block LMB is an image space composed of 16 ⁇ 16 pixels.
- the low-resolution coded signal LE g is converted to a low-resolution decoded signal L dg by a decoder 9a, and this signal L dg is interpolated by an up-sampler 10a to obtain a spatial resolution of high resolution. It is converted to a supplementary image signal LA g equal to the image signal HS g.
- the process AMB that divides the interpolated rectangular area AR corresponding to the interpolated image signal LA g obtained by the up-sampling of the low-resolution encoded signal LE g has an image space consisting of 32 ⁇ 32 pixels. Become.
- the 2 (c) the figure correspond to the interpolation image signal LA g, interpolated frame AF with ⁇ resolution frame HF same spatial resolution and, and in the frame AF inter ⁇ , interpolation corresponding to the interpolation image signal LA g
- the image A ob is shown, and the relative position of the interpolated rectangular area AR corresponding to the interpolated image with respect to the interpolated frame AF is the relative position of the high-resolution rectangular area HR with respect to the high-resolution frame HF. Match.
- the region extractor 14 corresponds to the high-resolution rectangular region HR based on the rectangular signal HR g from the region detector 12.
- An image signal HO g to be extracted is extracted, and further, the image signal H ⁇ g is divided by a blocker 15 so as to correspond to each of the blocks HMB 1 that divides the rectangular area HR, and is divided into each block HMB 1.
- the corresponding image signal HB g is output.
- the block image signal HB g is converted into a high-resolution encoded signal HE g by an encoding process with reference to the interpolated image signal LA g.
- the block HMB 1 is an image space composed of 32 ⁇ 32 pixels.
- the relative position of the interpolation rectangular area AR on the frame AF coincides with the relative position of the high-resolution rectangular area HR on the frame HF.
- the difference value of the blocked image signal can be easily calculated in block units of each rectangular area, and the difference value between the image value of the pixel forming the high-resolution image and the pixel value of the pixel forming the interpolation image is obtained. Can be easily encoded.
- FIG. 2 (d) shows a difference image D corresponding to the difference value (error) D on a high-resolution frame HF.
- the coding processing is performed without lowering the coding efficiency by referring to the low resolution image signal LS g. be able to.
- FIG. 3 is a block diagram for explaining an image processing device (hierarchical image coding device) according to Embodiment 2 of the present invention.
- the hierarchical image encoding apparatus has a configuration in which a difference value between an interpolated image signal obtained by performing resolution conversion on a low-resolution image signal and a ⁇ -resolution image signal is encoded as an error.
- the image signal is a binary image signal, that is, a binary shape signal, there is a more effective method than directly encoding the block difference value.
- the hierarchical image encoding device 102 is configured to encode the difference value between the interpolated image signal obtained by converting the resolution of the low-resolution image signal and the ⁇ -resolution image signal according to the first embodiment.
- an encoder 16b that encodes the magnitude of the shift between the boundary of the image obtained from the low-resolution image signal and the boundary of the image obtained from the high-resolution image signal is used.
- the other configuration is the same as that of the hierarchical image encoding device 101 of the first embodiment.
- the low-resolution encoding unit 102 L of the hierarchical image encoding device 102 of the second embodiment has the same configuration as that of the first embodiment.
- the high-resolution encoding unit 102H constituting O2 differs from that of Embodiment 1 only in the configuration of the encoder 16b.
- FIG. 4 is a schematic diagram for explaining an encoding process performed by the hierarchical image encoding device according to the second embodiment.
- FIG. 4 (a) shows a frame LF corresponding to a low-resolution image signal
- FIG. (b) shows the frame HF corresponding to the high resolution image signal
- Fig. 4 (c) shows the frame AF corresponding to the resolution converted image signal (interpolated image signal).
- Fig. 4 (d) shows the boundary HB of the object H ob in the rectangular area HR set on the frame HF
- Fig. 4 ( ⁇ ) shows the object A ob in the rectangular area AR set on the frame AF.
- FIG. 4 (f) shows the boundary HB of the object Hog corresponding to the high-resolution image signal and the boundary AB of the object A ob corresponding to the interpolated image signal on the frame HF. .
- the boundary of the object is the position of a pixel where the value of the binary shape signal constituting the image signal changes spatially.
- the rectangular area HR of the corresponding object is detected by the area detector 12 of the high-resolution encoding unit 102H based on the high-resolution image signal HSg (the 4 (b)), the low-resolution encoding unit 102 L detects the rectangular area LR of the low-resolution image signal LSg by the area detector 30a so that the rectangular area HR matches the spatial position.
- the spatial position LRp is detected (see Fig. 4 (a)).
- the low-resolution frame By setting the position LR p of the rectangular area LR on the low-resolution frame LF in this manner, the low-resolution frame: the position LRP of the low-resolution rectangular area LR on the LF is changed to the high-resolution frame on the high-resolution frame HF. It can match the position HRp of the resolution rectangular area HR.
- the image signal LOg corresponding to the low-resolution rectangular area set on the low-resolution frame as described above is blocked by the block converter 6, and the blocked image signal LBg is encoded. It is encoded by the unit 7 and output as a low-resolution encoded signal LE g.
- the low-resolution coded signal LE g is subjected to a decoding process by the decoder 3 a and an interpolation process by the up-sampler 10 a to perform the interpolation image signal LA having the same spatial resolution as the high-resolution image signal. Be converted to g (see Fig. 4 (c)).
- the region extractor 14 extracts an image signal HO g corresponding to the high-resolution rectangular region HR from the high-resolution image signal HS g, and the image signal HO g is blocked by the blocker 15. You.
- the encoder 16b calculates a boundary position HB of the object H ob obtained from the blocked image signal HB g and a boundary position AB of the object A ob obtained from the interpolation image signal LA g. Based on this, the difference ⁇ B between the two is encoded and output as a high-resolution encoded signal HEg.
- the boundary of the interpolated image obtained from the low-resolution image signal An encoder 16b that encodes the magnitude of the shift between the field and the boundary of the image obtained from the high-resolution image signal is provided, so that the hierarchical encoding process is performed when the image signal is a binary image signal. Encoding can be performed efficiently.
- the hierarchical encoding process for encoding the deviation between the boundary HB of the high-resolution image Hob and the boundary AB of the interpolated image Aob has been described.
- the encoding table used for the encoding processing of the high-resolution image signal may be switched for each pixel according to the interpolated image signal obtained from the low-resolution image signal.
- FIG. 5 is a block diagram for explaining an image processing apparatus (hierarchical image coding apparatus) according to Embodiment 3 of the present invention.
- the position of each block in the interpolation rectangular area obtained by converting the resolution of the low-resolution rectangular area and the position of each block in the high-resolution rectangular area completely match. Even when an integrated block obtained by integrating a plurality of blocks in a rectangular area matches one block in the interpolation rectangular area, hierarchical coding processing should be performed while avoiding a decrease in coding efficiency, as in the first embodiment. Can be done.
- the hierarchical image coding apparatus 1 0 3 of the third embodiment blocking the Interpolation image signal and Proc of the high resolution image signal and by comparing the configuration for obtaining the difference value encoder 1 6 a
- the integrated signal obtained by integrating the blocked high-resolution image signals is compared with the blocked interpolated image signal to obtain a difference value between the integrated signal and the interpolated image signal, and the difference value is encoded.
- an encoder 16c an encoder 16c.
- the high-resolution block HMB 2 (see FIG. 6 (b)) that forms the high-resolution rectangular area HR is replaced with the low-resolution block LMB (the sixth form) that forms the low-resolution rectangular area LR.
- the block generator 15 is configured to perform the block processing of the high-resolution image signal in units of the Bronk HMB 2 composed of 16 ⁇ 16 pixels.
- the third hierarchical image encoding device 103 is the same as the hierarchical image encoding device 101 of the first embodiment.
- the low-resolution encoding unit 103L included in the hierarchical image encoding device 103 according to the third embodiment has the same configuration as that in the first embodiment.
- the high-resolution encoding section 103H differs from that of the first embodiment only in the configuration of the encoder 16c.
- FIG. 6 is a schematic diagram for explaining an encoding process performed by the hierarchical image encoding device according to the third embodiment.
- FIG. 6 (a) shows a display screen (frame) corresponding to a low-resolution image signal. ) LF
- Fig. 6 (b) shows the display screen (frame) corresponding to the high resolution image signal
- FIG. 6 (c) shows a display screen (frame) AF corresponding to an interpolated image signal obtained by resolution-converting a low-resolution image signal.
- the rectangular area HR of the object Hob corresponding to the high-resolution image signal HSg is detected by the area detector 12 based on the high-resolution image signal HSg (see FIG. 6 (b)).
- the spatial position HRp of the rectangular area HR and the interpolated rectangular area A are detected by the detector 30a.
- the spatial position LRp of the rectangular area LR of the low-resolution image signal Lsg is detected so that the spatial position ARp of R matches (see Fig. 6 (a)).
- the spatial position LR p of the rectangular area LR on the low-resolution frame LF can be set on the high-resolution frame HF.
- R p can be substantially matched.
- the image signal LO g corresponding to the low-resolution rectangular area set on the low-resolution frame LF as described above is blocked by the blocker 6, and the blocked image signal LB g is further obtained.
- the low-resolution coded signal LEg is subjected to the decoding process by the decoder 9a and the interpolation process by the up-sampler 10a, and the interpolation image signal having the same spatial resolution as the high-resolution image signal is obtained. Converted to LA g (No. 6
- the region extractor 14 converts the high-resolution image signal HSg into a rectangular region HR.
- the corresponding image signal HO g is extracted, and the image signal HO g is divided into blocks by the blocker 15 so as to correspond to the block HMB 2 composed of 16 ⁇ 16 pixels.
- one block AM B constituting the rectangular area AR of the interpolation image A ob coincides with an area obtained by integrating the four blocks HMB 2 constituting the rectangular area HR of the high-resolution image H ob
- the encoder 16c refers to the complementary image signal corresponding to one block AMB, and refers to the image signal HBg corresponding to the four blocks HMB2 of the high-resolution rectangular area corresponding to the block AMB Is encoded. That is, the difference value between the image signal HO g corresponding to the VA blocks HMB 2 and the complementary image signal LA g corresponding to the one block AMB is encoded and output as a high-resolution encoded signal HE g.
- the high-resolution image signal HO g corresponding to the integrated area obtained by integrating the four high-resolution blocks HMB 2 and the complementary image corresponding to one interpolation block AMB that matches the integrated area Since the difference signal from the signal AL g is encoded, even if the high-resolution block and the interpolation block do not correspond one-to-one, the pixels in the high-resolution block and the pixels in the interpolation block can be correlated.
- the high-resolution image signal can be encoded with reference to the low-resolution image signal.
- the correspondence between the high-resolution block and the interpolation block can be established as described above, predictive coding of information in block units (such as coding mode information) can be easily realized, and furthermore, between the two blocks. Since the spatial position can be adjusted, the coding efficiency is also improved.
- the encoder 7 in the low-resolution encoding unit and the high-resolution rectangular area have the same size, the encoder 7 in the low-resolution encoding unit and the high-resolution The configuration of the encoder 17 in the encoder is almost the same, and sharing of hardware resources in the encoder can be easily realized by time division processing or the like.
- FIG. 7 is a block diagram for explaining an image processing device (hierarchical image coding device) according to Embodiment 4 of the present invention.
- the hierarchical image encoding device 104 of fi 4 is a low-resolution image encoding device according to Embodiment 1.
- the region detector 30a corresponding to the image signal is omitted, and instead of the region extractor 5 of the first embodiment, a low resolution is obtained based on the output HRg of the region detector 12 of the first embodiment.
- It is provided with an area extractor 31d that extracts a low-resolution rectangular area LR located at a spatial position corresponding to the space position of the high-resolution rectangular area HR on the frame LF. That is, the low-resolution encoding unit 104 L of the hierarchical image encoding device gl 04 of the fourth embodiment is blocked by the region extractor 31 d and the block generator 6 that blocks its output.
- the high-resolution encoding unit 104 H of the hierarchical image encoding device 104 is the same as the high-resolution encoding unit 101 H of the first embodiment, and has a region detector 12 and an area extraction unit. It consists of an encoder 14, a blocker 15 and an encoder 16a.
- the shape of the object corresponding to the interpolated image signal obtained by converting the resolution of the low-resolution image signal is always the size of the object corresponding to the high-resolution image signal. If it is known that it will not be larger than this, the spatial position HR p (see FIG. 2 (b)) corresponding to the high-resolution rectangular area HR indicated by the rectangular signal HR g is converted to the corresponding low-resolution rectangular area.
- the spatial position LR p of the LR see FIG. 2 (a)
- the object L 0b of the low-resolution image signal can be completely included in the low-resolution rectangular area LR.
- the high-resolution image signal HS g which is the image input signal, the high-resolution coding section 1 0 4 H and shaped state 1 described above at all Similar processing is performed.
- the low-resolution image signal LSg obtained from the sub-sample of the high-resolution image signal HSg is coded by the low-resolution coding unit 104L by substantially the same processing as in the first embodiment. You. S At this time, in the region extractor 3 I d, the low resolution image receives the signal LS g, output HR g of the high-resolution coding section 1 0 4 H region detector 1 2 Then, the position of the rectangular area on the low-resolution frame is determined, and the image signal LO g corresponding to the rectangular area is output to the block converter 6.
- the image signal LO g is blocked by the block converter 6, and the blocked image signal LB g is encoded by the encoder 7, and the low-resolution encoder 104 L corresponds to each block.
- the encoded low-resolution encoded signal LE g is output.
- the low-resolution coded signal LE g is decoded by the decoder 9a, and the output L dg of the decoder 9a is output from the up-sampler 10a to the same spatial resolution as the high-resolution image signal by the up-sampler 10a. It is converted to a complementary image signal LAg having a degree and output to the encoder 16a of the high-resolution encoder 104H.
- the region extractor 31 d of the low-resolution encoding unit 104 L outputs the rectangular signal HR output from the region detector 12 of the high-resolution encoding unit 104 H. Since the position of the rectangular area LR on the low-resolution frame is determined based on g, the low-resolution encoding unit 104 L does not need a circuit configuration for area extraction, and the image coding of the first embodiment is not required. The same effect as the conversion device can be realized with a simpler configuration.
- FIG. 8 is a block diagram for explaining an image processing device (hierarchical image decoding device) according to Embodiment 5 of the present invention.
- the hierarchical image decoding device 105 includes a low-resolution encoded signal LEg and a high-resolution code encoded by the hierarchical image encoding device 104 according to the fourth embodiment shown in FIG.
- hierarchical decoding processing is performed using the decoded signal HE g as an input signal. That is, like the conventional hierarchical image decoding device 200b, the hierarchical image decoding device 105 performs a decoding process on the low-resolution coded signal LEg to generate the low-resolution reproduction signal LCg.
- a low-resolution decoding unit 105L that generates g
- an amplifier sampler 10b that interpolates the decoded signal LDg in the decoding unit 105L by upsampling
- an upsampler 10b an upsampler 10b.
- a resolution decoding unit that decodes the high-resolution coded signal HE g based on the output LA g of b to generate a high-resolution reproduction signal HC g
- each of the decoding units 105 L and 105 H is similar to the conventional hierarchical image decoding device 200 b, and the decoders 9 and 30, the deblocking device 20 and 3 1, Region synthesizers 34 and 32 are provided.
- the area combiner 34 refers to the low-resolution rectangular signal HR g from the hierarchical image encoding device 104 of the fourth embodiment, and The image signal L 1 g corresponding to the rectangular area is combined with another image signal corresponding to the frame so that the low-resolution rectangular area is arranged at the position indicated by the rectangular signal HR g. This point is different from the conventional hierarchical image decoding device 200b shown in FIG.
- the low-resolution encoded signal LE g and the high-resolution encoded signal HE g output from the hierarchical image encoding device 104 of the fourth embodiment are input to the hierarchical image decoding device 105,
- the low-resolution coded signal LE g is decoded by the decoder 9 in the low-resolution decoding unit 105L, and the output LD g of the decoder 9 is integrated by the inverse blocker 20. Further, based on the output LI g of the deblocker 20 and the high-resolution rectangular signal HR g, it is synthesized with an image signal corresponding to a low-resolution frame.
- the output LDg of the decoder 9 is interpolated by the upsampler 10b and output to the high-resolution decoding unit 105H.
- the high-resolution coded signal HE g is output from the up-sampler 10 b by the high-resolution coding unit 105 H in exactly the same manner as the conventional hierarchical image decoding device 200 b. Decoding is performed based on the interpolated image signal AD g and the high-resolution rectangular signal HR g.
- the coded signals LE g and HE g and the rectangular signal g from the hierarchical image coding device 104 of the fourth embodiment are received, and the By the same operation as that of the region extractor 30d in the fourth embodiment, the low-resolution rectangular region can be combined with the spatial position indicated by the high-resolution rectangular signal HRg on the low-resolution frame.
- 9 (a) and 9 (b) are block diagrams for explaining an image processing device (hierarchical image coding device) according to Embodiment 6 of the present invention.
- the hierarchical image encoding device 106 of the sixth embodiment is similar to the hierarchical image encoding device 101 of the first embodiment shown in FIG. 6L, a decoder 9a, an upsampler 10a, and a high-resolution encoder 106H.
- components denoted by the same reference numerals as those in the first embodiment have exactly the same configurations as those in the first embodiment.
- low-resolution encoding section 106 L for encoding low-resolution image signal LS g is replaced with encoder 7 in low-resolution encoding section 101 L of the first embodiment.
- a coding mode signal Mg indicating the coding mode of each block image signal LBg is generated together with the low-resolution coding signal LEg corresponding to the blocked low-resolution image signal LBg.
- the configuration is the same as that of the first embodiment.
- the high-resolution encoding unit 106H for encoding the high-resolution image signal HSg is the same as the encoder in the high-resolution encoding unit 101H of the first embodiment.
- code upsampling unit 1 0 a output LA g just not also based on the coding mode signal M g of performing sign-treatment of the high-resolution image signal HB g which is proc of
- the other configuration is completely the same as that of the first embodiment.
- the encoder 16 f receives the above-blocked high-resolution image signal HBg, determines the coding mode, and outputs a high-resolution block coding mode signal MD.
- Mode encoding that encodes the encoding mode signal MD and the encoding mode of the low-resolution image signal LBg into blocks based on the symbol Mg and outputs the mode encoding signal EMg vessel 5 1, and a coder 5 3 and 5 coding methods are different first and second against the high-resolution Bronk coded image signal HB g (Chapter 9 (b) see Figure).
- the first encoder 53 refers to the interpolation signal LA g obtained by up-sampling the low-resolution decoded signal L dg, and encodes the high-resolution block-coded image signal HB g
- the second encoder 54 is configured to encode the high-resolution block image signal HBg without referring to the interpolation signal LAg. ing. Further, based on the high-resolution block image signal HB, the mode determination unit 50 converts the block image signal HB g into an interpolation signal LA g corresponding to the low-resolution image signal. , Or whether to encode the image signal HB g without referring to the interpolation signal LA g.
- the encoder 16 f converts the high-resolution blocked image signal HB g into the first and second encoding signals in accordance with the output (encoding mode signal) MD of the mode determination unit 50.
- the first and second encoders 53, 54 according to the preceding switch 52 supplied to one of the encoders 53, 54 and the output MD of the mode decision unit 50.
- a subsequent switch 55 for selecting one of the outputs HE1 or HEg2, and a multiplexer 56 for multiplexing the output SHEg of the subsequent switch 55 and the mode coded signal EMg. Have.
- the sub-sampler 2 converts the low-resolution image signal LSg by subsampling the high-resolution image signal HSg.
- the generated signal LSg is encoded by the low-resolution encoding unit 106L.
- the encoder 7f of the encoding unit 106L outputs an encoding mode signal Mg corresponding to the image signal LBg when encoding the low-resolution blocked image signal LBg.
- the high-resolution encoding unit 106H performs an area detection process and an area extraction process on the high-resolution image signal HSg in the same manner as in the first embodiment, so that the image signal HO g corresponding to the high-resolution rectangular region HR is obtained. Is generated, and a blocked image signal HBg corresponding to the high-resolution block HMB1 is generated from the image signal H ⁇ g by the blocking process. Then, the broken image signal H B g is encoded by the encoder 16 # based on the up sample output LAg and the encoding mode signal Mg.
- the encoding mode signal is obtained from the encoder 7 # of the low-resolution encoding unit 106 L. It is output to the encoder 16 f of the high-resolution encoding unit 106 H, and the encoder 16 f refers to the encoding mode M g of the low-resolution blocked image signal LB g, and This is different from the encoding processing of the first embodiment in that the encoded high-resolution image signal HB g is encoded.
- the encoding method for the image signal HBg is determined based on the blocked high-resolution image signal HBg by the mode determination in the mode determiner 50. That is, it is determined whether to encode the image signal HBg with reference to the interpolation signal ALG or to encode the image signal HBg without referring to the interpolation signal ALG. Then, an encoding mode signal MD corresponding to this mode judgment result is output from the mode judging device 50, and each of the switches 52 and 55 receives the encoding mode signal MD according to the encoding mode signal MD. One of the first and second encoders 53 and 54 is selected. As a result, the high-resolution image signal HBg is encoded by the selected encoder 53 or 54 and output to the multiplexer 56.
- the mode encoder 51 the encoding process of the encoding mode signal MD is performed based on the encoding mode Mg of the low-resolution block. Then, the output SHE g of the switch 55 and the output EM g of the mode encoder 51 are multiplexed and output as a high-resolution coded signal HE g. As described above, in the sixth embodiment, the encoder 16 f performs the encoding processing of the high-resolution block with reference to the encoding mode of the low-resolution block. An encoding process can be realized.
- the coding mode of the high-resolution block has a correlation with the coding mode of the corresponding low-resolution block. For example, if the low-resolution block is located at the boundary of the object, the corresponding high-resolution block is also likely to be located at the boundary, and similarly if the low-resolution block is located outside or inside the object. The corresponding high-resolution block is also likely to be located outside or inside the object.
- the coding mode of the encoder I 6 f in the high-resolution coding unit 106 H causes This encoding mode is referred to by referring to the encoding mode signal Mg of the resolution block.
- Performing predictive coding by assigning a short code to the high-resolution block coding mode signal whose mode is the same as that of the low-resolution block coding mode signal Mg does not refer to the low-resolution block coding mode signal Mg.
- the number of coding bits can be reduced.
- a configuration is shown in which the encoding method of the high-resolution image signal is switched according to the encoding mode of the low-resolution block, and the encoding method of the encoding mode of the high-resolution block is switched.
- the high-resolution image signal coding method and the high-resolution block coding method may be switched according to the low-resolution block coding mode.
- the hierarchical image encoding device 106 the hierarchical encoding process in the hierarchical image encoding device 101 of the first embodiment shown in FIG.
- FIG. 6 The hierarchical coding process according to the third embodiment described above may be configured to be performed based on a low-resolution image signal and a low-resolution block coding mode.
- the encoding process in the encoding mode corresponding to (1) is a single interpolation block AMB consisting of 32 x 32 pixels obtained by the resolution conversion process of the low-resolution block LMB (see Fig. 6 (c)). This is performed in units of an integrated area consisting of four high-resolution blocks HMB2 corresponding to the following.
- information corresponding to a unit smaller than the low-resolution block LMB that is, the prediction error of the pixel value, etc.
- the error (difference) between the pixel value corresponding to the high-resolution frame HF (Fig. 6 (b)) and the pixel value corresponding to the resolution conversion (interpolation) frame AF (Fig. 6 (c)) Image D) is calculated, and the error signal is encoded in small blocks of 16 ⁇ 16 pixels.
- FIG. 3 is a block diagram for explaining a layer image decoding device.
- the hierarchical image decoding device 107 includes a low-resolution decoding unit 107L, an up-sampler 10b, and a high-resolution decoding device, similarly to the conventional hierarchical image decoding device 200b shown in FIG. It has a resolution decoding section 107H.
- the components denoted by the same reference numerals as those in the conventional multilayer image decoding device 200b have the same configuration as that of the conventional hierarchical image decoding device 200b.
- the hierarchical image decoding device 107 according to 7 performs decoding of the high-resolution coded signal HE g with reference to the coding mode Mg output from the low-resolution decoder 9 g. This is different from the conventional hierarchical image decoding device 200b shown in FIG.
- low-resolution decoding section 107L for decoding low-resolution coded signal LEg performs decoding in low-resolution coding section 202L of conventional hierarchical image decoding apparatus 200b.
- the decoder 9g for outputting the encoding mode signal Mg together with the low-resolution decoded signal LDg is provided.Other configurations are the same as those of the conventional hierarchical image decoding device 200b. Exactly the same.
- high-resolution decoding section 107 # for decoding high-resolution coded signal HEg performs decoding in high-resolution coding section 201H of conventional hierarchical image decoding apparatus 200b.
- the decoder 40g includes a separator 60 that separates and extracts the mode coded signal EMg from the high-resolution coded signal HEg, and a mode decoding that decodes the separated mode coded signal EMg. And a first and second decoders 63 and 64 that use different decoding methods for the high-resolution coded signal SHEg separated from the mode coded signal EMg.
- the decoder 40g converts the high-resolution coded signal SHEg into the first and second signals based on the coded mode signal DMg decoded by the mode decoder 61.
- a first-stage switch 62 for supplying one of the two decoders 63 and 64; Based on the encoded mode signal DMg decoded by the mode decoder 61, the output HD g1 or HDg of any of the first and second decoders 63 and 64 and a subsequent switch 65 for selecting g2 and outputting it as a high-resolution decoded signal HDg.
- the first decoder 63 performs decoding of the high-resolution encoded signal HE g with reference to the interpolation signal AD g obtained by up-sampling the low-resolution decoded signal LD g.
- the second decoder 64 is configured to perform a decoding process on the high-resolution encoded signal HEg without referring to the interpolation signal ADg. Next, the operation and effect will be described.
- the low-resolution coder 107L When the high-resolution coded signal HEg and the low-resolution coded signal LEg are input to the hierarchical image decoding device 107, the low-resolution coder 107L performs decoding on the coded signal LEg.
- the decoding process and the deblocking process are sequentially performed, and the low-resolution decoded signal LIg corresponding to the predetermined rectangular region integrated by the deblocking process is processed by the region synthesizer 21 to correspond to the frame. Is synthesized with the image signal.
- the decoding mode 9 g outputs the encoding mode signal M g together with the low-resolution decoded signal LD g, and the low-resolution decoded signal LD g is output from the up-sampler 1 g.
- the spatial resolution is converted to an interpolated decoded signal ADg whose spatial resolution is equal to the high-resolution decoded signal LDg.
- the decoding process of the high-resolution encoded signal HE g by the decoder 40 g is performed by the encoding mode signal M g and the interpolated decoded signal AD g Further, the output HD g of the decoder 40 g is subjected to a reverse blocking process by the reverse blocking device 31. Then, the high-resolution decoded signal HI g corresponding to the predetermined rectangular area integrated by the inverse block processing is synthesized by the area synthesizer 32 with other image signals corresponding to the frame.
- the high-resolution coded signal HE g is separated into a code part EMg corresponding to the coding mode and another code part S HE g by the separator 60,
- the code part EMg corresponding to the coding mode is output to the mode decoder 61, and the other code part S HEg is output to the preceding switch 62.
- the mode The encoder 61 decodes the mode encoded signal corresponding to the high-resolution block with reference to the encoded mode signal M g of the low-resolution block.
- any one of the first decoder 63 and the second decoder 64 is referred to with reference to the decoded coding mode signal DMg of the high-resolution block. One of them is selected, and the output HD g1 or HD g2 of one of the decoders is output as the output HD g of the decoder 40 g.
- the decoding process of the high-resolution coded signal LE g is performed by not only the interpolated decoded signal AD g obtained by interpolating the low-resolution decoded signal LD g by up-sampling but also the low-resolution block
- the decoding is performed by the decoder 40 g with reference to the coding mode signal M g, and a decoding process corresponding to the coding process by the encoder 16 f in the sixth embodiment is performed. be able to. For this reason, the low-resolution coded signal LEg and the low-resolution coded signal HEg coded by the hierarchical image coding apparatus according to the sixth embodiment can be correctly decoded.
- the decoding method of the high-resolution encoded signal and the decoding method of the high-resolution encoding mode are switched according to the encoding mode of the low-resolution block. Only one of the high-resolution coded signal decoding method and the high-resolution block coding mode decoding method may be switched according to the low-resolution block coding mode.
- FIG. 11 is a diagram for explaining an image processing device (hierarchical image encoding device) according to Embodiment 8 of the present invention.
- FIG. 11 (a) is a block diagram showing the overall configuration
- FIG. FIG. 11 (b) is a block diagram showing a detailed configuration of an encoder constituting the hierarchical image encoding device.
- the hierarchical image coding device 108 of the eighth embodiment is different from the hierarchical image decoding device 106 of the sixth embodiment shown in FIG. This is provided with an encoder 16h that switches the encoding method in the same way, and the other configuration is the same as that of the hierarchical image encoding device 106 of the sixth embodiment.
- the encoder 16 h of the eighth embodiment receives the blocked high-resolution encoded signal HB g and determines whether or not a block corresponding to this image signal includes a boundary of an object. And outputs a determination signal BD corresponding to the determination result, and outputs a determination signal BD 1 indicating that the boundary of the object is included in the block.
- a mode judgment that receives the signal HB g and the judgment signal BD 1 and outputs an identification signal MD 1 indicating whether or not to refer to the low-resolution image signal LS g when the block includes an object boundary.
- the encoder 16h encodes the decision signal BD and the identification signal MD1 based on the encoding mode signal Mg of the low-resolution encoded signal LEg, and outputs a mode encoded signal EMg.
- Mode encoder 71, first and second arbitrary shape encoders 73a and 73b different in the method of arbitrary shape encoding processing for high-resolution image signal HSg, and high-resolution image signal A fixed shape encoder 74 for performing a fixed shape encoding process on HS g.
- the first arbitrary shape encoder 73 a is configured to perform arbitrary shape encoding of the high-resolution image signal HS g with reference to an interpolation signal LA g obtained by up-sampling the low-resolution decoded signal L dg.
- the second arbitrary shape encoder 73 b generates a high-resolution image signal without referring to an interpolation signal LA g obtained by up-sampling the low-resolution decoded signal. It is configured to perform HS g arbitrary shape encoding processing.
- the encoder 16h converts the high-resolution surface image signal HSg into the first and second signals based on the output BD of the boundary determiner 70 and the output MD1 of the mode determiner 75.
- the front-stage switch 52 to be supplied to any one of the arbitrary shape encoders 7 3a and 7 3b and the fixed shape encoder 74, and the output MD and the mode determiner of the boundary determiner 70
- a subsequent switch 5 5 for selecting one of the outputs of the first and second arbitrary shape encoders 73 a, 73 b and the fixed shape encoder 74 based on the output MD 1 of 75
- a multiplexer 56 for multiplexing the output of the post-stage switch 55 and the mode coded signal EMg.
- the arbitrary-shape encoder 73a or 73b performs an arbitrary-shape encoding process according to the object shape, and the high-resolution block is located outside the object.
- a fixed shape decoder 74 that can efficiently encode the image signal (fixed shape) in the block is encoded.
- the object boundary is included in the low-resolution block, the object boundary is also included in the corresponding high-resolution block, and if the low-resolution block is located inside or outside the object, the corresponding high-resolution block is used.
- the resolution block is also likely to be located inside and outside the object. Therefore, performing the encoding process on the high-resolution block with reference to the encoding mode for the low-resolution block (that is, the positional relationship between the low-resolution block and the boundary of the object) increases the encoding efficiency. Will be effective.
- FIG. 12 is a diagram showing whether the low-resolution rectangular area and the high-resolution rectangular area are inside the object or outside the object.
- Fig. 12 (a) shows whether or not each of the broken LMBs constituting the low-resolution rectangular area LR shown in Fig. 2) is inside the object Lob.
- ) Indicates whether or not each block HMB 1 constituting the high-resolution rectangular area HR shown in the figure is within the object H 0 b. It shows whether or not the block AM B constituting the interpolated rectangular area AR shown in the figure is inside the object A ob.
- FIG. 12 (d) shows the high resolution rectangular area HR shown in FIG. 6 (d). It is shown whether each block HMB 2 formed is in the object H ob or not.
- the block with the symbol] is a block located inside the object
- the block with the symbol O is Blocks located outside the object
- blocks with the symbol IO are blocks located at the object boundary.
- the interpolation rectangular area AR (FIG. 12 (c)) corresponding to the resolution image signal LA g of the low-resolution image signal LS g is converted into the high-resolution image signal HS g Compared to the rectangular region HR (Fig. 12 (b) or Fig. 12 (d)) corresponding to, whether the low-resolution block LMB is located inside the object, outside the object, or at the object boundary?
- the boundary determiner 70 determines whether or not the boundary of the object is included in the high-resolution block.
- the mode encoder 71 When encoding is performed by the mode encoder 71 with reference to the encoded mode signal Mg, if the presence or absence of an object boundary matches between the low-resolution block and the corresponding high-resolution block, the high-resolution block By assigning a code having a short code length to the coding mode signals BD and BD 1 of the first embodiment, the number of coding bits can be saved.
- the image signal HBg is subjected to an arbitrary shape encoding process or a fixed shape code. It is determined whether or not to perform the conversion process.
- the mode determiner 75 converts the low-resolution decryption signal L dg based on the determination signal BD 1 corresponding to the determination result from the boundary determiner 70 and the blocked high-resolution image signal HB g.
- the first and second arbitrary shape encoders 73a and 73b, and the fixed shape encoding are performed according to the determination results in the determiners 70 and 7S.
- One of the containers 74 is selected.
- the high-resolution image signal HB g is encoded by the selected encoder 73 a, 73 b, or 74 and output to the multiplexer 56.
- the encoding process of the determination signal BD of the boundary determiner 70 and the determination signal MD1 of the mode determiner 75 is performed by the encoding module KM g of the low-resolution block.
- the output SHEg of the switch 56 and the output EMg of the mode encoder 71 are multiplexed to the multiplexer 56 and output as a high-resolution encoded signal HEg. You.
- the encoding mode of the high-resolution block is encoded by the encoder 16h with reference to the encoding mode Mg of the low-resolution block.
- the positional relationship between the high-resolution block and the object is determined, and the encoding method for the high-resolution block is determined.
- the arbitrary shape code that refers to the low-resolution image signal is determined.
- An arbitrary shape encoding process that does not refer to the low-resolution image signal, or a fixed shape encoding process is applied to the high-resolution image signal, so that the hierarchical encoding process with even higher encoding efficiency can be performed.
- Embodiment 8 the configuration in which the encoding method of the high-resolution image signal is switched and the encoding method of the encoding mode of the high-resolution block is switched according to the encoding mode of the low-resolution block is described. According to the low-resolution block encoding mode, only one of the high-resolution image signal encoding method and the high-resolution block encoding method may be switched.
- FIG. 13 is a diagram for explaining an image processing device (hierarchical image decoding device) according to Embodiment 9 of the present invention.
- FIG. 13 (a) is a block diagram showing the overall configuration
- FIG. FIG. 13 (b) is a block diagram showing a detailed configuration of a decoder constituting the hierarchical image decoding device.
- the hierarchical image decoding apparatus 109 of the ninth embodiment includes a low-resolution encoded signal and a high-resolution code encoded by the hierarchical image encoding apparatus 108 of the eighth embodiment shown in FIG. It is intended to decode the signal, in place of the decoder 4 0 g in the hierarchical image coding apparatus 1 0 7 of the seventh embodiment shown in 1 0 Figure, according to the coding mode signal obtained by decoding And a decryption device 40 i for switching the decryption method.
- Other configurations of the hierarchical image encoding device 109 are the same as those of the hierarchical image encoding device 107 of the seventh embodiment.
- the decoder 40 i includes a separator 60 that separates and extracts the mode coded signal EM g from the high-resolution coded signal HE g, the coded signal EM g, first a mode decryption device 7 5 a for decoding with reference to the coding mode signal M g of the low-resolution image signal, the decoding method for the high-resolution coded signal HS g is different 1 And a second arbitrary shape decoder 76a and 76b, and a fixed shape decoder 77.
- the first arbitrary shape decoder 7 6 a is referring to high-resolution code Cassin interpolation signal the low-resolution decoded signal to up-sampling
- the second arbitrary-shape decoder 76b does not refer to the interpolated signal obtained by up-sampling the low-resolution decoded signal, and The decryption process is performed.
- the decoder 40 i based on the encoded mode signal DM g decoded by the mode decoder 75 a, converts the ⁇ resolution encoded signal HE g into the first and second encoded signals HE g. 2 and the former-stage switch 62, which is supplied to one of the arbitrary shape decoders 76a and 76b and the fixed shape decoder 77, and the mode decoder 75a. Based on the coding mode signal DMg, one of the outputs of the first and second arbitrary-shape decoders 76a and 76b and the fixed-shape decoder 77 is selected to perform high-resolution decoding. And a rear-stage switch 65 that outputs the converted signal HDg.
- the operation other than the decoder 40 i is exactly the same as that of the hierarchical image decoding apparatus 107 according to the seventh embodiment. Describes only the operation of the decoder 40 i.
- the mode decoder 75a of the decoder 401 the mode decoder 75a of the high-resolution image is referred to by referring to the encoded mode signal Mg of the low-resolution image (that is, whether it is inside or outside the object). Decode the encoded signal EM g. In the former-stage switch 62 and the latter-stage switch 65, one of the three decoders is selected according to the encoding mode of the decoded high-resolution image.
- the image-encoded signal having the arbitrary shape encoded is decoded by the first or second arbitrary-shape decoder 76, and the image signal having the fixed shape encoded is decoded by the fixed-shape decoder 77. Decrypted.
- FIG. 14 is a diagram for explaining an image encoding device (hierarchical image encoding device) according to Embodiment 10 of the present invention
- FIG. 14 (a) is a block diagram showing the overall configuration thereof
- FIG. 14 (b) is a block diagram showing a detailed configuration of an encoder that constitutes the hierarchical image encoding device.
- the hierarchical image encoding device 110 of the tenth embodiment is different from the hierarchical image encoding device 106 of the sixth embodiment shown in FIG. This is provided with an encoder 16 j for switching the encoding method in accordance with the scanning direction in which is larger, and the other configuration is the same as that of the hierarchical image encoding device 106 of the sixth embodiment.
- the encoder 16 j receives the block-converted high-resolution coded signal HBg, and determines a scanning direction in which the correlation between image values in each block is large.
- a mode encoder that encodes the signal SD based on a mode signal M g indicating a scanning direction having a large correlation between pixel values and the low-resolution encoded signal LE g and outputs a mode encoded signal EM g 8 1
- a horizontal scan encoder 83 that performs horizontal scan encoding on the high-resolution image signal HSg; and a vertical encoder that performs vertical encoding on the high-resolution image signal HSg. 8 and 4.
- the encoder 16 j converts the high-resolution image signal HS g into the horizontal scan encoder 83 and the vertical scan encoder 84 based on the output SD of the scan direction determiner 80. Either the horizontal scan encoder 83 or the vertical scan encoder 84 based on the pre-switch 52 supplied to any one of the above and the output SD of the scan direction determiner 80. And a multiplexer 56 for multiplexing the output of the subsequent switch 55 and the mode coded signal EMg.
- the operation other than the encoder 16 j is performed in exactly the same way as the hierarchical image encoding device 106 of the sixth embodiment, Only the operation related to j will be described.
- the encoding efficiency changes depending on the scanning direction. That is, in an image signal in which the horizontal correlation of pixel values is large, by sequentially encoding the pixel values of each pixel along the horizontal scanning direction, it is possible to perform encoding using the horizontal correlation effectively. In an image signal in which the vertical correlation of pixel values is large, by sequentially encoding the pixel values of each pixel along the vertical scanning direction, it is possible to perform encoding using the vertical correlation of pixel values effectively. is there.
- the scanning direction determiner 80 determines the scanning direction in which the correlation of the pixel values is large, and switches either the horizontal scan encoder 83 or the vertical scan encoder 84 based on the determination result.
- the encoding mode signal indicating the scanning direction in which the pixel value in the high-resolution image has a large correlation is used as the pixel value in the low-resolution image.
- the mode encoder 81 with reference to the encoding mode signal Mg indicating the scanning direction having a large correlation of the scanning direction in which the pixel value between the low-resolution image and the high-resolution image has a large correlation If they match, a code having a short code length is assigned to the encoding mode signal corresponding to the high-resolution image signal. As a result, the number of coding bits required for coding the coding mode signal can be further reduced.
- Embodiment 11 1.
- FIG. 15 is a diagram for explaining an image processing device (hierarchical image decoding device) according to Embodiment 11 of the present invention.
- FIG. 15 (a) is a block diagram showing the entire configuration thereof.
- FIG. 15 (b) is a block diagram showing a detailed configuration of a decoder constituting the hierarchical image decoding device.
- the hierarchical image decoding device 1 11 of the embodiment 11 includes a low-resolution coded signal and a high-resolution coded by the hierarchical image coding device 110 of the embodiment 10 shown in FIG. This is for decoding a coded signal.
- a decoder 40 k for switching a decoding method in accordance with a decoded coding mode signal is used. Have.
- Other configurations of the hierarchical image encoding device 110 are the same as those of the hierarchical image encoding device 107 of the seventh embodiment shown in FIG.
- the decoder 40 k of Embodiment 1] outputs the high-resolution encoded signal H O 98/31151
- the decoder 4 Ok converts the high-resolution coded signal HEg into the horizontal scanning decoder 8 based on the coded mode signal DMg decoded by the mode decoder 85. 6 and the vertical scanning decoder 87, and the horizontal scanning decoding based on the encoding mode signal DMg decoded by the mode decoder 85 and the pre-switch 62 supplied to any of the above.
- a post-stage switch 65 for selecting one of the outputs of the decoder 86 and the vertical scanning decoder 87 and outputting it as a high-resolution decoded signal HDg.
- the mode encoding of the high-resolution image is performed by referring to the encoding mode Mg of the low-resolution image (that is, the scanning direction in which the correlation between the pixel values is large).
- the horizontal scan encoding process is performed by switching the switches 62 and 65 according to the DMg, which is the encoding mode signal of the decoded high-resolution image (a signal indicating the scanning direction in which the pixel value has a large correlation).
- the resulting image signal is decoded by a horizontal scanning decoder 86, and the image signal that has been subjected to the vertical scanning encoding process is decoded by a vertical scanning decoding device 87.
- the eleventh embodiment it is possible to correctly decode a coded signal that has been subjected to the horizontal scan coding process or the vertical scan coding process according to the shape of an object.
- Embodiment 1 2.
- FIG. 16 is a diagram for explaining an image processing device (hierarchical image coding device) according to Embodiment 12 of the present invention.
- FIG. 16 (a) is a Bronk diagram showing the entire configuration thereof
- FIG. 16 (b) is a block diagram showing a detailed configuration of a porcelain encoder 16m constituting the hierarchical image encoding device.
- FIG. 16 (c) is a block diagram showing the encoder 16m.
- FIG. 28 is a block diagram illustrating a specific configuration of a second encoder 54 m that is configured.
- the hierarchical image encoding device 112 of the embodiment 12 is different from the encoder 16 of the above-described embodiment 6 in that the encoder 16 refers to a motion vector of a low-resolution image,
- the encoder is provided with an encoder 16 m for encoding an image signal, and the other configuration is the same as that of the hierarchical image encoding device 106 of the sixth embodiment shown in FIG.
- This encoder 16 m is configured such that the second encoder 54 in the encoder 16 f of the sixth embodiment performs inter-picture predictive encoding as necessary.
- the other configuration is the same as that of the encoder 6f.
- the second encoder 54 m constituting the encoder 16 m in the present embodiment encodes the high-resolution image signal HBg based on the prediction signal Pc to perform high-resolution encoding.
- the second encoder 54 ⁇ receives the high-resolution image signal HB g, and receives the high-resolution local decoding S-decoded Hd g stored in the memory 94 and the low-resolution image.
- a motion detector 90 that detects a motion vector HMV corresponding to a high-resolution image by referring to the motion vector LMV as the encoding mode Mg of the motion vector MV, and a motion vector HMV based on the motion vector HMV.
- the motion compensator 91 for extracting the prediction signal P c from the memory 94 and the motion vector HMV of the high-resolution image are encoded based on the motion vector LMV of the low-resolution image.
- a motion encoder 95 that outputs a motion vector coded signal H MV c, and a multiplexing unit that multiplexes and outputs the motion vector coded signal HM V c and the high-resolution coded signal HE g Vessel 56 m.
- Image ⁇ The symbol has a correlation in the time direction, that is, a correlation between pixel values between the previous and next frames. Therefore, it is known that coding efficiency is improved by performing motion compensation using a motion vector.
- the motion detector 90 of the encoder 16 m calculates the motion of the high-resolution image based on the decoded image signal H dg stored in the memory 94 and the blocked high-resolution image signal HB g.
- the vector HMV is detected, and the motion compensator 91 generates a motion-compensated image (prediction signal) Pc based on the motion vector HMV of the detected high-resolution image.
- the encoder 92 performs an encoding process on the block image signal HBg of the high-resolution image with reference to the motion compensation image Pc.
- the high-resolution coded signal HEg obtained by this processing is decoded by the decoder 93 and stored in the memory 94 as a high-resolution locally-decoded signal Hdg.
- the motion vector between the low resolution image motion vector LMV and the high resolution image motion vector HMV is also There is a correlation.
- the motion encoder 95 refers to the encoding mode signal corresponding to the motion vector LMV of the low-resolution image, and refers to the motion vector of the high-resolution image detected by the motion detector 90.
- the motion encoder 95 When encoding HMV, if the motion vector matches between the low-resolution image and the high-resolution image, assign a code with a short code length to the motion vector HMV of the high-resolution image. I have to. As a result, the number of coding bits required for coding a motion vector of a high-resolution image can be saved.
- Embodiment 1 3.
- FIG. 17 is a diagram for explaining an image processing device (hierarchical image decoding device) according to Embodiment 13 of the present invention.
- FIG. 17 (a) is a block diagram showing the entire configuration thereof
- FIG. 17 (b) is a block diagram showing a detailed configuration of a decoder 40n constituting the hierarchical image decoding device.
- FIG. 17 (c) is a block diagram showing the decoder 40n.
- FIG. 28 is a block diagram showing a specific configuration of a second decoder 64 n to be configured.
- the hierarchical image decoding apparatus 1 13 of the embodiment 13 converts the encoded signal encoded by the hierarchical image encoding apparatus 1 12 of the embodiment 12 shown in FIG. It is becoming something.
- This hierarchical image decoding apparatus 113 has the structure shown in FIG. 7. Instead of the decoder 40 g in FIG. 7, a decoder 40 n for decoding a high-resolution coded signal with reference to a motion vector of a low-resolution image is provided. Is the same as the hierarchical image decoding apparatus 107 of the seventh embodiment shown in FIG.
- the decoder 64 n constituting the decoder 40 n may be replaced by the second decoder 64 in the decoder 107 of the seventh embodiment, if necessary.
- the configuration is such that inter-picture prediction decoding is performed, and the other configuration is the same as that of the decoder 40 g of the seventh embodiment.
- the second decoder 64 n constituting the decoder 40 n decodes the high-resolution coded signal HE g based on the prediction signal P d to perform high-resolution decoding.
- the second decoder 64 n refers to the coding mode (motion vector) LMV of the low-resolution image and performs decoding processing on the separated mode coded signal EMg.
- a motion decoder 96 for reproducing the motion vector HMV of the high-resolution image, and a high-resolution decoded signal HD g stored in the memory 94 b based on the reproduced motion vector HMV. 2 and a motion compensator 91b for extracting the prediction signal Pd from the second.
- the separator 60 in the decoder 40n separates the code portion corresponding to the mode information (motion vector information) EMg from the high-resolution coded signal HEg.
- the decoder 64 n refers to the encoding mode M g of the low-resolution image (that is, the motion vector L MV), and determines the motion vector of the high-resolution image from the code of the mode information separated by the separator 60.
- the HMV is decoded, and the motion vector HMV is supplied to the motion compensator 91b.
- the decoded high-resolution image stored in the memory 94b is obtained.
- the motion compensation is performed by referring to the image signal Pd, and the decoder 93 b refers to the output Pd of the motion compensator 91 b to obtain a signal other than the mode information of the high-resolution coded signal HEg.
- the encoded part is decoded, and a high-resolution decoded signal HDg2 is output.
- the decoded signal HDg2 is stored in the memory 94b, and is referred to when a subsequent block is decoded.
- Embodiment 13 of the present invention the high-resolution coded signal coded by referring to the motion vector LMV of the low-resolution image in the hierarchical image coding apparatus 112 of Embodiment 12 is described.
- HE g can be correctly decoded.
- Embodiment 1 4.
- FIG. 18 is a diagram for explaining an image processing device (hierarchical image coding device) according to Embodiment 14 of the present invention.
- FIG. 18 (a) is a block diagram showing the entire configuration thereof
- FIG. 18 (b) is a block diagram showing a detailed configuration of an encoder 16p constituting the hierarchical image encoding device
- FIG. 18 (c) is a block diagram showing the encoder 16p.
- FIG. 21 is a block diagram showing a specific configuration of a second encoder 54 p that is configured.
- the hierarchical image encoding device 114 of the embodiment 14 is different from the encoder 16 f of the embodiment 6 shown in FIG. 9 in that the motion vector of the low-resolution image and the encoded
- the encoder is provided with an encoder 16p for encoding a high-resolution image signal by referring to a prediction vector predicted from a motion vector of a high-resolution image. This is the same as the hierarchical image encoding device 106 of mode 6.
- This encoder 16 p is configured such that the second encoder 54 in the encoder 16 f of the sixth embodiment performs screen interpolative coding as necessary.
- the other configuration is the same as that of the encoder 16f.
- the second encoder 54 p constituting the encoder 16 p in the embodiment 14 constitutes the encoder 16 m in the embodiment 12 shown in FIG.
- the motion vector coded signal corresponding to the low-resolution image and the motion vector coded signal corresponding to the coded high-resolution image A motion vector predictor 97 for generating a prediction motion vector P MV based on the motion vector, and the other configurations are the same as those of the encoder 16 m in the above-described Embodiment 12 It is the same as 2 encoder 54 m. Next, the operation and effect will be described.
- the difference between the second encoder 54p in Embodiment 14 and the second encoder 54m in Embodiment 12 is the same as that of Embodiment 12 shown in FIG.
- the encoder 54m encodes the motion vector of the high-resolution image with reference to the motion vector of the low-resolution image
- the encoder of the embodiment 14 shown in FIG. 18 encodes the motion vector of the high-resolution image.
- the quantizer 54 p refers to the motion vector of the high-resolution image by referring to the motion vector predicted and generated from the motion vector of the low-resolution image and the motion vector of the encoded high-resolution image. The point is to encode the torque.
- Fig. 19 is an explanatory diagram for referring to the motion vector.
- Fig. 19 (a) shows the motion vector LMV of the low-resolution image
- Fig. 19 (b) shows the motion vector HMV of the high-resolution image
- Fig. 19 (c) shows the resolution of the low-resolution image.
- the converted interpolation motion vector AMV is shown.
- the motion vector of the coded block B X for the high-resolution image shown in Fig. 19 (d) is converted into the motion vector L MV (Fig. 19 ( a)) and the motion vectors MMV1 to HMV3 of the coded high-resolution image (see Fig. 19 (b)). )
- the coding efficiency of the motion vector can be improved as compared to the case where the motion vector of the high-resolution image is simply referred to only by referring to the motion vector of the low-resolution image at the same spatial position.
- the motion estimator 97 uses the motion vector HMV (HMV 1 to HMV 3) coded signal HE mv of the high-resolution image, which is encoded by the motion encoder 95, and the motion of the low-resolution image.
- a motion vector prediction value PMV of a high-resolution image is generated with reference to the coded signal M g (LE mv) of the vector, and the motion coder 95 generates the predicted motion vector P MV.
- the motion vector HMV of both high-resolution images detected by the motion detector 90 is encoded.
- the predicted motion vector PE m V encoded by the motion encoder 95 is multiplexed by the multiplexer 56 m with the high-resolution encoded signal output from the encoder 92. Then, it is output as the output HD g2 of the second encoder 54 p.
- the motion vector of the low-resolution image and the motion vector of the high-resolution image are different. Since the motion vector of the high-resolution image is encoded with reference to the prediction motion vector predicted and generated from the motion vector of the image, the motion vector of the high-resolution image is Greater savings in the number of coded bits for a vector.
- the motion detector 90 uses the locally decoded signal stored in the memory 94 to correspond to the target high-resolution block to be subjected to the encoding process.
- a prediction signal prediction region
- a motion vector having a value as close as possible to the motion vector of a low-resolution image is selected as a motion vector of a high-resolution image.
- the encoding method of the high-resolution image signal and the encoding method of the encoding mode of the high-resolution image signal are switched according to the encoding mode of the low-resolution image.
- the encoding mode of the low-resolution image only one of the encoding method of the high-resolution image signal and the encoding method of the encoding mode may be switched.
- FIG. 20 is a diagram for explaining an image processing device (hierarchical image decoding device) according to Embodiment 15 of the present invention
- FIG. 20 (a) is a pronk diagram showing the overall configuration thereof
- FIG. 20 (b) is a block diagram showing a detailed configuration of a decoder 40q constituting the hierarchical image decoding device.
- FIG. 20 (c) is a block diagram showing the decoder 40q.
- FIG. 27 is a block diagram showing a specific configuration of a second decryption unit 64 q that constitutes FIG.
- the hierarchical image decoding device 1 15 of the embodiment 15 is adapted to hierarchically decode the coded signal encoded by the hierarchical image coding device 114 of the embodiment 14 shown in FIG. Is what you do.
- This hierarchical image decoding apparatus 115 is different from the encoder 40 g in the seventh embodiment shown in FIG. 10 in that a motion vector of a low-resolution image and a motion vector of an encoded high-resolution image are replaced.
- a decoder 40 q that decodes a high-resolution image signal with reference to a prediction vector predicted from a vector is provided. It is the same as the decoding device 107.
- This decoder 40 ⁇ is the second code in the decoder 40 g of the seventh embodiment.
- the encoder 64 is configured to perform screen interpolative prediction encoding as necessary, and the other configuration is the same as that of the decoder 40g.
- the second decoder 64 q constituting the decoder 40 q in this embodiment is the same as that of the second decoder 64 n in the embodiment 13 shown in FIG.
- a motion vector LMV corresponding to the low-resolution image and a motion vector HMV corresponding to the encoded high-resolution surface image are used to generate a predicted motion vector PMV.
- the motion vector predictor is provided with a motion vector predictor 98, and the other configuration is the same as that of the second encoder 64n in Embodiment 13 described above.
- the operation of the decoder 40 q other than the second decoder 64 q is the same as that of the hierarchical image decoding apparatus of the above embodiment 13. Since the operation is performed in exactly the same way as the decoding device 113, only the operation of the decoder 40q regarding the second decoder 64q will be described.
- the motion estimator 98 in the second decoder 64q the motion vector HMV of the high-resolution image decoded by the motion decoder 96 and the low resolution supplied from the mode decoder 61 With reference to the motion vector L MV (DM g) of the image, a predicted value P MV of the motion vector of the high-resolution image block is generated. Then, the motion decoder 96 decodes the encoded signal E Mg of the motion vector for the high-resolution image with reference to the motion vector P MV generated by prediction. Other operations are the same as those of the hierarchical image decoding apparatus 113 of the embodiment 13 in FIG.
- the motion vector coded signal encoded with reference to the motion vector of the encoded high-resolution image and the motion vector of the low-resolution image Can be correctly decoded.
- FIG. 21 shows a hierarchical image encoding process or a hierarchical image decoding process by the image processing apparatus according to each of the above-described embodiments, in which programs corresponding to these image processes are stored.
- FIG. 8 is a diagram for explaining a case in which the present invention is implemented by a computer system using a floppy disk.
- Fig. 21 (b) shows the external appearance, cross-sectional structure, and the main body of the floppy disk FD as a recording medium when viewed from the front of the floppy disk FD
- Fig. 21 (a) shows the physical format of the main body D of the floppy disk.
- the floppy disk main body D is housed in a case F.
- On the surface of the disk main body D a plurality of tracks Tr are formed concentrically from the outer periphery to the outer periphery, and each track is formed in an angular direction. Is divided into sectors Se. Accordingly, in the floppy disk main body D storing the program, data as the program is recorded in an area allocated on the floppy disk main body D.
- FIG. 21 (c) shows a configuration for recording and reproducing the above program on a floppy disk FD.
- the above program is recorded on the floppy disk FD
- data as the above program is written from the computer system Cs to the floppy disk FD via the floppy disk drive F DD.
- the program is read from the floppy disk FD by the floppy disk drive FDD, and the computer system Cs is read. Transfer to
- image processing by a computer system using a floppy disk as a data recording medium has been described.
- this image processing can be similarly performed using an optical disk.
- the recording medium is not limited to this, and the above-described image processing can be similarly performed as long as the recording medium can record a program such as an IC card or a ROM cassette.
- the mode determiner 50 (FIG. 9 (b)), the boundary determiner 70 (the 11th (b) ) And the scanning direction determiner 80 (FIG. 14 (b)) are configured to determine the coding method based on the externally input high-resolution image signal before the coding process.
- the determinator may determine the coding method, that is, the coding mode, by comparing the results (coded signals) obtained by coding with a plurality of coding methods.
- the image processing method, the image processing apparatus, and the data recording medium according to the present invention can improve the coding efficiency in the compression processing of the image signal, and can improve the image processing in the system for transmitting and storing the image signal. It is extremely useful for realizing encoding and image decoding, and is particularly suitable for compression and decompression of moving images conforming to MPEG4 and other standards.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
- Editing Of Facsimile Originals (AREA)
- Compression Of Band Width Or Redundancy In Fax (AREA)
Description
Claims
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP98900183A EP0971545A4 (en) | 1997-01-10 | 1998-01-09 | IMAGE PROCESSING METHOD AND DEVICE, AND DATA RECORDING MEDIUM |
US09/341,472 US6690724B1 (en) | 1997-01-10 | 1998-01-09 | Image processing method, image processing device, and data recording medium |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP265997 | 1997-01-10 | ||
JP9/2659 | 1997-01-10 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO1998031151A1 true WO1998031151A1 (en) | 1998-07-16 |
Family
ID=11535475
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP1998/000040 WO1998031151A1 (en) | 1997-01-10 | 1998-01-09 | Image processing method, image processing device, and data recording medium |
Country Status (6)
Country | Link |
---|---|
US (1) | US6690724B1 (ja) |
EP (1) | EP0971545A4 (ja) |
KR (1) | KR100341079B1 (ja) |
CN (1) | CN1243635A (ja) |
TW (1) | TW395131B (ja) |
WO (1) | WO1998031151A1 (ja) |
Families Citing this family (29)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4034380B2 (ja) * | 1996-10-31 | 2008-01-16 | 株式会社東芝 | 画像符号化/復号化方法及び装置 |
US6259738B1 (en) | 1996-10-31 | 2001-07-10 | Kabushiki Kaisha Toshiba | Video encoding apparatus and video decoding apparatus |
TWI249144B (en) * | 2003-02-21 | 2006-02-11 | Via Tech Inc | Single level MIP filtering algorithm for anisotropic texturing |
JP3772846B2 (ja) * | 2003-03-24 | 2006-05-10 | ソニー株式会社 | データ符号化装置およびデータ符号化方法、並びにデータ出力装置およびデータ出力方法 |
EP1608164A1 (en) * | 2003-03-24 | 2005-12-21 | Sony Corporation | Data encoding apparatus, data encoding method, data output apparatus, data output method, signal processing system, signal processing apparatus, signal processing method, data decoding apparatus, and data decoding method |
JP2005142654A (ja) * | 2003-11-04 | 2005-06-02 | Matsushita Electric Ind Co Ltd | 映像送信装置および映像受信装置 |
WO2006008681A1 (en) * | 2004-07-13 | 2006-01-26 | Koninklijke Philips Electronics N.V. | Method of spatial and snr picture compression |
US20080267291A1 (en) * | 2005-02-18 | 2008-10-30 | Joseph J. Laks Thomson Licensing Llc | Method for Deriving Coding Information for High Resolution Images from Low Resolution Images and Coding and Decoding Devices Implementing Said Method |
WO2006087319A2 (en) | 2005-02-18 | 2006-08-24 | Thomson Licensing | Method for deriving coding information for high resolution pictures from low resoluton pictures and coding and decoding devices implementing said method |
KR100891662B1 (ko) | 2005-10-05 | 2009-04-02 | 엘지전자 주식회사 | 비디오 신호 디코딩 및 인코딩 방법 |
KR20070038396A (ko) | 2005-10-05 | 2007-04-10 | 엘지전자 주식회사 | 영상 신호의 인코딩 및 디코딩 방법 |
EP1969853A1 (en) * | 2006-01-05 | 2008-09-17 | Thomson Licensing | Inter-layer motion prediction method |
EP1879399A1 (en) * | 2006-07-12 | 2008-01-16 | THOMSON Licensing | Method for deriving motion data for high resolution pictures from motion data of low resolution pictures and coding and decoding devices implementing said method |
US8374247B2 (en) * | 2008-01-14 | 2013-02-12 | Broadcom Corporation | Method and system for hierarchical motion estimation with multi-layer sub-pixel accuracy and motion vector smoothing |
JP5147597B2 (ja) * | 2008-08-13 | 2013-02-20 | キヤノン株式会社 | 画像形成装置、画像形成方法およびプログラム |
JP5081113B2 (ja) | 2008-09-17 | 2012-11-21 | キヤノン株式会社 | 画像符号化装置及び画像復号装置、並びにそれらの制御方法 |
JP5173873B2 (ja) | 2008-11-20 | 2013-04-03 | キヤノン株式会社 | 画像符号化装置及びその制御方法 |
JP2011244210A (ja) * | 2010-05-18 | 2011-12-01 | Sony Corp | 画像処理装置および方法 |
EP2645716A4 (en) | 2010-11-24 | 2016-04-13 | Panasonic Ip Corp America | MOTION VECTOR CALCULATION METHOD, IMAGE ENCODING METHOD, IMAGE DECODING METHOD, MOTION VECTOR CALCULATING DEVICE, AND IMAGE ENCODING / DECODING DEVICE |
JP2012124653A (ja) * | 2010-12-07 | 2012-06-28 | Canon Inc | 符号化装置、符号化方法およびプログラム |
EP2665271A4 (en) * | 2011-01-12 | 2014-06-18 | Panasonic Corp | VIDEO ENCODING METHOD AND VIDEO DECODING METHOD |
JP6108309B2 (ja) | 2011-02-22 | 2017-04-05 | サン パテント トラスト | 動画像符号化方法、動画像符号化装置、動画像復号方法、および、動画像復号装置 |
JP5358746B2 (ja) | 2011-03-03 | 2013-12-04 | パナソニック株式会社 | 動画像符号化方法、動画像符号化装置及びプログラム |
JP5492139B2 (ja) | 2011-04-27 | 2014-05-14 | 富士フイルム株式会社 | 画像圧縮装置、画像伸長装置、方法、及びプログラム |
US20120300844A1 (en) * | 2011-05-26 | 2012-11-29 | Sharp Laboratories Of America, Inc. | Cascaded motion compensation |
KR101756841B1 (ko) * | 2011-06-07 | 2017-07-11 | 삼성전자주식회사 | 블록 기반 영상의 해상도 변환 방법 및 장치 |
CN105981051B (zh) * | 2014-10-10 | 2019-02-19 | 北京旷视科技有限公司 | 用于图像解析的分层互连多尺度卷积网络 |
CN106296578B (zh) * | 2015-05-29 | 2020-04-28 | 阿里巴巴集团控股有限公司 | 一种图像处理方法及装置 |
CN110915196B (zh) * | 2017-07-24 | 2023-01-10 | 索尼公司 | 成像设备、图像处理方法和程序 |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH07212761A (ja) * | 1994-01-17 | 1995-08-11 | Toshiba Corp | 階層符号化装置及び階層復号化装置 |
Family Cites Families (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE69322769T2 (de) | 1992-03-03 | 1999-07-22 | Toshiba Kawasaki Kk | Koder für zeitveränderliche bilder |
JPH06197334A (ja) * | 1992-07-03 | 1994-07-15 | Sony Corp | 画像信号符号化方法、画像信号復号化方法、画像信号符号化装置、画像信号復号化装置及び画像信号記録媒体 |
JP3135692B2 (ja) | 1992-08-28 | 2001-02-19 | 株式会社日立製作所 | 階層的動画像信号符号化装置及び方法 |
NL9201594A (nl) * | 1992-09-14 | 1994-04-05 | Nederland Ptt | Systeem omvattende ten minste één encoder voor het coderen van een digitaal signaal en ten minste één decoder voor het decoderen van een gecodeerd digitaal signaal, en encoder en decoder voor toepassing in het systeem. |
JP2979900B2 (ja) * | 1993-05-26 | 1999-11-15 | 日本ビクター株式会社 | 記録媒体 |
CA2127151A1 (en) * | 1993-09-21 | 1995-03-22 | Atul Puri | Spatially scalable video encoding and decoding |
WO1996008114A1 (en) * | 1994-09-02 | 1996-03-14 | David Sarnoff Research Center, Inc. | Method and apparatus for global-to-local block motion estimation |
US6023301A (en) * | 1995-07-14 | 2000-02-08 | Sharp Kabushiki Kaisha | Video coding device and video decoding device |
JP3210862B2 (ja) * | 1996-06-27 | 2001-09-25 | シャープ株式会社 | 画像符号化装置及び画像復号装置 |
JP3191922B2 (ja) * | 1997-07-10 | 2001-07-23 | 松下電器産業株式会社 | 画像復号化方法 |
-
1998
- 1998-01-09 CN CN98801758.XA patent/CN1243635A/zh active Pending
- 1998-01-09 TW TW087100232A patent/TW395131B/zh active
- 1998-01-09 KR KR1019997006218A patent/KR100341079B1/ko not_active IP Right Cessation
- 1998-01-09 WO PCT/JP1998/000040 patent/WO1998031151A1/ja not_active Application Discontinuation
- 1998-01-09 US US09/341,472 patent/US6690724B1/en not_active Expired - Fee Related
- 1998-01-09 EP EP98900183A patent/EP0971545A4/en not_active Withdrawn
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH07212761A (ja) * | 1994-01-17 | 1995-08-11 | Toshiba Corp | 階層符号化装置及び階層復号化装置 |
Also Published As
Publication number | Publication date |
---|---|
EP0971545A1 (en) | 2000-01-12 |
KR20000070003A (ko) | 2000-11-25 |
US6690724B1 (en) | 2004-02-10 |
TW395131B (en) | 2000-06-21 |
EP0971545A4 (en) | 2003-08-13 |
CN1243635A (zh) | 2000-02-02 |
KR100341079B1 (ko) | 2002-06-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO1998031151A1 (en) | Image processing method, image processing device, and data recording medium | |
US6198768B1 (en) | Video encoding apparatus and video decoding apparatus | |
CN1535024B (zh) | 视频编码设备和方法及视频解码设备和方法 | |
US6766061B2 (en) | Method and apparatus for digital image coding and decoding, and data recording media | |
KR19990013748A (ko) | 화상 전송용 데이터 구조, 화상 부호화 방법 및 화상 복호화 방법 | |
US6259738B1 (en) | Video encoding apparatus and video decoding apparatus | |
KR100256859B1 (ko) | 동화상을 코딩/인코딩하는 장치와 방법 그리고동화상을 저장하는 저장매체 | |
JPH10257505A (ja) | 画像処理方法及び画像処理装置、並びにデータ記録媒体 | |
JPH06351000A (ja) | 画像信号符号化装置と画像信号復号装置 | |
JPH10276437A (ja) | 階層的動画像信号符号化/復号化方法及び装置 | |
JPH104549A (ja) | 画像符号化装置および画像復号化装置 | |
JP3222876B2 (ja) | 画像符号化方法および画像符号化装置並びに画像符号化プログラムを記録した記録媒体 | |
JP2008228336A (ja) | 画像符号化方法及び装置 | |
JP2001160971A (ja) | デジタル画像符号化方法及びデジタル画像符号化装置,デジタル画像復号化方法及びデジタル画像復号化装置,並びにデータ記憶媒体 | |
JP2000324501A (ja) | 画像符号化装置、画像復号化装置および動きベクトル検出装置 | |
JP2008011564A (ja) | 画像符号化装置および画像復号化装置 | |
JP2007110748A (ja) | 動き補償予測方法、映像信号復号方法、映像信号復号化装置 | |
JP2001069509A (ja) | 画像処理装置及びその方法 | |
JP2001086498A (ja) | 符号化装置及び符号化方法 | |
JP2000069470A (ja) | 画像復号化装置および方法、画像符号化・復号化システム |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 98801758.X Country of ref document: CN |
|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): CN KR US |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): AT BE CH DE DK ES FI FR GB GR IE IT LU MC NL PT SE |
|
DFPE | Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101) | ||
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWE | Wipo information: entry into national phase |
Ref document number: 1019997006218 Country of ref document: KR |
|
WWE | Wipo information: entry into national phase |
Ref document number: 1998900183 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 09341472 Country of ref document: US |
|
WWP | Wipo information: published in national office |
Ref document number: 1998900183 Country of ref document: EP |
|
WWP | Wipo information: published in national office |
Ref document number: 1019997006218 Country of ref document: KR |
|
WWG | Wipo information: grant in national office |
Ref document number: 1019997006218 Country of ref document: KR |
|
WWW | Wipo information: withdrawn in national office |
Ref document number: 1998900183 Country of ref document: EP |