Coding method for coding moving image

1. A computer-implemented coding method, comprising:

extracting a background image from a target frame included in a moving image;

extracting, from the target frame, a plurality of kinds of pieces of metadata indicating a feature of at least one object included in the target frame;

disposing hierarchically the plurality of kinds of pieces of metadata with respect to each kind; and

generating coded information including background image information specifying the background image and meta-information, which is information about an estimation technique for estimating at least a part of all the pieces of metadata corresponding to a selected hierarchy.

2. The computer-implemented decoding method according to claim 1 ,

wherein, in deriving the at least one kind of metadata, whether a correction parameter exists in the coded information is determined, and the metadata is derived by correcting the estimated metadata obtained by estimating the metadata based on the correction parameter when it is determined that the correction parameter exists in the coded information.

3. The computer-implemented decoding method according to claim 1 ,

wherein, in estimating the metadata, a coordinate of the metadata is estimated by performing interpolation or extrapolation on a coordinate of the identical object in at least a reference frame included in the coded moving image.

4. The computer-implemented decoding method according to claim 3 ,

wherein, in estimating the metadata, motion model information indicating changes in spatial position and size of the object in the coded moving image is extracted from the coded information, and

the coordinate of the metadata and the size of another kind of metadata in the at least one kind of metadata are estimated based on the interpolation or extrapolation and the motion model information.

5. A computer-implemented decoding method, comprising:

specifying a background image of a target frame from background image information included in coded information corresponding to a frame in a coded moving image;

deriving at least one kind of metadata from meta-information included in the coded information, the metadata indicating a feature of one object included in the target frame and being hierarchically disposed with respect to each kind; and

generating decoded image by superimposing the one object represented by the at least one kind of metadata on the specified background image;

wherein the metadata indicates the feature of the object with respect to at least the one object included in the target frame, and

in deriving the at least one kind of metadata, the metadata is derived by estimating the metadata using information about an estimation technique, the information being included as the meta-information.

6. The computer-implemented decoding method according to claim 5 ,

wherein the decoded image is generated by superimposing the object having a different abstraction level in each hierarchy of the metadata on the background image.

7. The computer-implemented decoding method according to claim 5 ,

wherein the meta-information further includes a total number of objects included in the target frame.

8. The computer-implemented decoding method according to claim 5 ,

wherein the kind belonging to a first hierarchy in the coded information is a coordinate of the object in the target frame.

9. The computer-implemented decoding method according to claim 8 ,

wherein the kind belonging to a second hierarchy in the coded information is a size of the object in the target frame.

10. The computer-implemented decoding method according to claim 9 ,

wherein the kind belonging to a third hierarchy in the coded information is a classification name of a subject that is projected as the object in the target frame.

11. The computer-implemented decoding method according to claim 10 , further comprising:

determining whether a difference image exists in the coded information,

wherein, in generating the decoded image, a reconstructed image is generated by superimposing the at least one object on the background image, and the decoded image is generated by adding the difference image to the reconstructed image.

12. The computer-implemented decoding method according to claim 11 ,

wherein, in generating the reconstructed image, the coordinate of the object is specified using the metadata belonging to the first hierarchy,

the size of the object is specified using the metadata belonging to the second hierarchy,

an associated image associated with the classification name of the subject is specified using the metadata belonging to the third hierarchy,

the background image is specified using the background image information, and

the reconstructed image is generated by superimposing the specified associated image having the specified size on the specified coordinate in the specified background image.

13. The computer-implemented decoding method according to claim 12 ,

wherein, in generating the reconstructed image, the associated image is acquired from a server through a network.

14. A decoding device comprising:

a processor; and

a memory having a program stored thereon, the program causing the processor to execute operations including

specifying a background image of a target frame from background image information included in coded information corresponding to a frame in a coded moving image;

deriving at least one kind of metadata from meta-information included in the coded information, the metadata indicating a feature of one object included in the target frame and being hierarchically disposed with respect to each kind; and

generating decoded image by superimposing the one object represented by the at least one kind of metadata on the specified background image;

wherein the metadata indicates the feature of the object with respect to at least the one object included in the target frame, and

in deriving the at least one kind of metadata, the metadata is derived by estimating the metadata using information about an estimation technique, the information being included as the meta-information.