Method for encoding immersive image and method for decoding immersive image
Disclosed herein is a method for encoding an immersive image. The method includes detecting a non-diffuse surface in a first texture image of a first view, generating an additional texture image from the first texture image based on the detected non-diffuse surface, performing pruning on the additional texture image based on a second texture image of a second view, generating a texture atlas based on the pruned additional texture image, and encoding the texture atlas.
1. A method for encoding an immersive image, comprising:
detecting a non-diffuse surface in a first texture image of a first view;
generating an additional texture image from the first texture image for the non-diffuse surface in the first texture image;
performing pruning on the additional texture image based on a second texture image of a second view;
generating at least one texture atlas by packing the pruned additional texture image; and
encoding the at least one texture atlas,
wherein the additional texture image corresponds to a partial region of the first texture image in which the non-diffuse surface is included,
wherein the pruned additional texture image is generated by removing a redundant region, with the second texture image, from the additional texture image, and
wherein difference information indicates a difference between the additional texture image and the second texture image for the redundant region and is further encoded as metadata.
2. The method of claim 1 , wherein:
the difference information comprises at least one of a difference in an average value in a histogram distribution for the redundant region between the additional texture image and the second texture image, a difference in a standard deviation for the redundant region between the additional texture image and the second texture image, or a ratio of color values of the additional texture image and the second texture image for the redundant region.
3. The method of claim 1 , further comprising:
generating an additional depth image, corresponding to the additional texture image, from a first depth image of the first view,
wherein the additional depth image corresponds to a partial region of the first depth image in which the non-diffuse surface is included.
4. The method of claim 3 , further comprising:
performing pruning on the additional depth image based on a second depth image of the second view; and
generating at least one depth atlas by packing the pruned additional depth image.
5. The method of claim 4 , further comprising:
encoding information on a pruning priority for the additional depth image.
6. The method of claim 4 , further comprising:
encoding scaling-related information pertaining to a patch extracted from the pruned additional depth image as metadata.
7. The method of claim 1 , wherein:
a view identifier assigned to the additional texture image is different from a view identifier of the first texture image, and
a first view identifier indicating the view identifier assigned to the first texture image and a second view identifier indicating the view identifier assigned to the additional texture image are encoded, respectively.
8. A method for decoding an immersive image, comprising:
decoding a texture atlas; and
synthesizing a viewport image based on at least one patch extracted from the texture atlas,
wherein in response to the viewport image comprising a non-diffuse surface, a region, included in the non-diffuse surface, is modified based on difference information decoded from a bitstream,
wherein the difference information indicates a difference for a redundant region between an additional texture image and a second texture image, and
wherein the additional texture image corresponds to a partial region of a first texture image in which the non-diffuse surface is included.
9. The method of claim 8 , wherein the difference information comprises at least one of a difference in an average value in a histogram distribution for the redundant region between the additional texture image and the second texture image, a difference in a standard deviation for the redundant region between the additional texture image and the second texture image, or a ratio of color values of the additional texture image and the second texture image for the redundant region.
10. The method of claim 8 , further comprising:
decoding a depth atlas,
wherein the depth atlas comprises patches extracted from an additional depth image, and
wherein the additional depth image corresponds to a partial region of a first depth image in which the non-diffuse surface is included.
11. The method of claim 10 , further comprising:
decoding scaling-related information pertaining to the patch for the non-diffuse surface.