Methods and devices for encoding / reconstructing 3D points

Abstract

To provide a method for encoding three-dimensional points whose geometry is represented by a geometry image and whose attributes are represented by attribute images.SOLUTION: There is provided a method for checking whether a depth value of a pixel in a first geometry image D0 and a depth value of a pixel located at the same location in a second geometry image D1 are not the same (not identical values). A module M3 assigns (encodes) an attribute of a three-dimensional point defined from the two-dimensional spatial coordinates of the pixel in the first geometry image and the depth value of the pixel located at the same location in the second geometry image when the depth value of the pixel in the first geometry image and the depth value of the pixel located at the same location in the second geometry image are not the same.SELECTED DRAWING: Figure 1

Claims

[Claim 1] It is a method, The method includes encoding a 3D point in which the geometry is represented by a first geometry image and a second geometry image, and the attributes are represented by a first attribute image associated with the first geometry image and a second attribute image associated with the second geometry image, wherein encoding the 3D point is Encoding syntactic elements related to a group of point cloud frames, wherein a first value of the syntactic element indicates that the depth value of a pixel in the first geometry image and the depth value of a pixel located at the same location in the second geometry image are compared before a 3D point is reconstructed using the depth value of the pixel located at the same location in the second geometry image, and a second value of the syntactic element indicates that the depth value of the pixel in the first geometry image and the depth value of the pixel located at the same location in the second geometry image are not compared. Encoding the first geometric image and the second geometric image, wherein the first geometric image and the second geometric image store the depth values ​​of the 3D points of the point cloud frame projected onto the projection plane. Methods that include... [Claim 2] An apparatus for encoding a 3D point in which the geometry is represented by a first geometry image and a second geometry image, and the attributes are represented by a first attribute image associated with the first geometry image and a second attribute image associated with the second geometry image, The processor comprises, Encoding syntactic elements related to a group of point cloud frames, wherein a first value of the syntactic element indicates that the depth value of a pixel in the first geometry image and the depth value of a pixel located at the same location in the second geometry image are compared before a 3D point is reconstructed using the depth value of the pixel located at the same location in the second geometry image, and a second value of the syntactic element indicates that the depth value of the pixel in the first geometry image and the depth value of the pixel located at the same location in the second geometry image are not compared. Encoding the first geometric image and the second geometric image, wherein the first geometric image and the second geometric image store the depth values ​​of the 3D points of the point cloud frame projected onto the projection plane. A device configured to perform the following actions. [Claim 3] Encoding a first attribute image that stores the attribute values ​​of the 3D points projected onto the projection surface, wherein the depth value of the 3D points is stored in the first geometry image, and the first attribute value of the first 3D points is stored in a pixel located at the same location in the first attribute image. Encoding a second attribute image that stores the attribute values ​​of the 3D points projected onto the projection plane, wherein, in response to a determination that a first depth value of a pixel in the first geometry image and a second depth value of a pixel located at the same location in the second geometry image are the same, a dummy attribute value is assigned to the pixel located at the same location in the second attribute image. The method according to claim 1, further comprising: [Claim 4] The method according to claim 3, wherein the dummy attribute value is the attribute of a pixel located at the same location in another attribute image. [Claim 5] The method according to claim 3, wherein the dummy attribute value is the average of the attributes associated with neighboring pixels located around the pixel located at the same location. [Claim 6] The method according to claim 1, wherein the first geometric image, the second geometric image, and the syntactic element are encoded in a bitstream carrying encoded attributes of 3D points configured as a number of blocks, patches of blocks, and frames of patches, and in response to a determination that the syntactic element is equal to the first value, another syntactic element is signaled at the frame level, patch level, or block level, the other syntactic element indicating the same as the syntactic element. [Claim 7] The aforementioned processor, The method involves encoding a first attribute image that stores the attribute values ​​of a 3D point projected onto the projection surface, wherein the depth value of the 3D point is stored in the first geometry image, and the first attribute value of the first 3D point is stored in a pixel located at the same location within the first attribute image. Encoding the second attribute image which stores the attribute values ​​of the 3D points projected onto the projection plane, wherein, in response to a determination that the first depth value of a pixel in the first geometry image and the second depth value of a pixel located at the same location in the second geometry image are the same, a dummy attribute value is assigned to the pixel located at the same location in the second attribute image. The apparatus according to claim 2, further configured to perform the following: [Claim 8] The apparatus according to claim 7, wherein the dummy attribute value is the attribute of a pixel located at the same location in another attribute image. [Claim 9] The apparatus according to claim 7, wherein the dummy attribute value is the average of the attributes associated with neighboring pixels located around the pixel located at the same location. [Claim 10] The apparatus according to claim 2, wherein the first geometric image, the second geometric image, and the syntactic element are encoded in a bitstream carrying encoded attributes of 3D points configured as a number of blocks, patches of blocks, and frames of patches, and in response to a determination that the syntactic element is equal to the first value, another syntactic element is signaled at the frame level, patch level, or block level, the other syntactic element indicating the same as the syntactic element. [Claim 11] It is a method, Decoding a syntactic element associated with a group of point cloud frames, wherein the first value of the syntactic element indicates that the depth value of a pixel in a first geometry image and the depth value of a pixel located at the same location in a second geometry image are compared before the 3D point is reconstructed using the depth value of the pixel located at the same location in the second geometry image, and the second value of the syntactic element indicates that the depth value of the pixel in the first geometry image and the depth value of the pixel located at the same location in the second geometry image are not compared. Decoding the first geometric image and the second geometric image, wherein the first geometric image and the second geometric image store the depth values ​​of the 3D points of the point cloud frame projected onto the projection plane. Reconstructing a first 3D point from a first 2D coordinate in the first geometry image and a first depth value of a first pixel in the first geometry image at the first 2D coordinate, In response to the determination that the decoded syntactic element is equal to the first value, it is determined whether the first depth value of the first pixel and the second depth value of a pixel located at the same location in the second geometry image are the same. In response to the determination that the first depth value and the second depth value are not the same, the second 3D point is reconstructed from the first 2D coordinates and the second depth value, Methods that include... [Claim 12] An apparatus for reconstructing a 3D point from a first geometry image and a second geometry image representing the geometry of the 3D point, The processor comprises, Decoding a syntactic element associated with a group of point cloud frames, wherein a first value of the syntactic element indicates that the depth value of a pixel in the first geometry image and the depth value of a pixel located at the same location in the second geometry image are compared before the 3D point is reconstructed using the depth value of the pixel located at the same location in the second geometry image, and a second value of the syntactic element indicates that the depth value of the pixel in the first geometry image and the depth value of the pixel located at the same location in the second geometry image are not compared. Decoding the first geometric image and the second geometric image, wherein the first geometric image and the second geometric image store the depth values ​​of the 3D points of the point cloud frame projected onto the projection plane. Reconstructing a first 3D point from a first 2D coordinate in the first geometry image and a first depth value of a first pixel in the first geometry image at the first 2D coordinate, In response to the determination that the decoded syntactic element is equal to the first value, it is determined whether the first depth value of the first pixel and the second depth value of a pixel located at the same location in the second geometry image are the same. In response to the determination that the first depth value and the second depth value are not the same, the second 3D point is reconstructed from the first 2D coordinates and the second depth value, A device configured to perform the following actions. [Claim 13] The method according to claim 11, wherein a bitstream carrying encoded attributes of 3D points is composed of a number of blocks, patches of blocks, and frames of patches, and in response to a determination that the syntactic element is equal to the first value, another syntactic element is signaled at the frame level, patch level, or block level, and the other syntactic element indicates that it is identical to the syntactic element. [Claim 14] The method according to any one of claims 1, 3 to 6, 11, or 13, wherein the attribute of the 3D point is a color value or a texture value. [Claim 15] The method according to claim 11, wherein, in response to the determination that the first depth value and the second depth value are the same, the reconstruction of the second 3D point from the 2D spatial coordinates of the pixels located at the same location in the second geometry image and the second depth value is skipped. [Claim 16] The apparatus according to claim 12, wherein a bitstream carrying encoded attributes of 3D points is composed of a number of blocks, patches of blocks, and frames of patches, and in response to a determination that the syntactic element is equal to the first value, another syntactic element is signaled at the frame level, patch level, or block level, and the other syntactic element is identical to the syntactic element. [Claim 17] The apparatus according to claim 2, 7 to 10, 12, or 16, wherein the attribute of the 3D point is a color value or a texture value. [Claim 18] The apparatus according to claim 12, wherein, in response to the determination that the first depth value and the second depth value are the same, the reconstruction of the second 3D point from the 2D spatial coordinates of the pixels located at the same location in the second geometric image and the second depth value is skipped. [Claim 19] A computer program comprising program code instructions, wherein when the computer program is executed on a computer, the computer program comprises program code instructions for performing the method according to any one of claims 1, 3 to 6, 11, 13 to 15. [Claim 20] A non-temporary storage medium that carries instructions for program code, wherein when the program code is executed on a computing device, the non-temporary storage medium carries instructions for the program code to execute the method according to any one of claims 1, 3 to 6, 11, 13 to 15.

Images