Point cloud encoding and decoding method and device based on two-dimensional regularization plane projection
Disclosed are a point cloud encoding and decoding method and device based on a two-dimensional regularization plane projection. The encoding method includes: acquiring original point cloud data; performing two-dimensional regularization plane projection on the original point cloud data to obtain a two-dimensional projection plane structure; obtaining a plurality of pieces of two-dimensional image information according to the two-dimensional projection plane structure; and encoding the plurality of pieces of two-dimensional image information to obtain code stream information. According to the present disclosure, the strong correlation representation of a point cloud on the two-dimensional projection plane structure is obtained, so that the spatial correlation of the point cloud is better reflected, and the encoding efficiency of the point cloud is improved. Moreover, a placeholder information map is used for assisting in encoding the depth information map, so that the encoding efficiency is further improved.
1 . A point cloud encoding method, the method comprising:
acquiring original point cloud data;
performing two-dimensional regularization plane projection on the original point cloud data to obtain a two-dimensional projection plane structure;
obtaining a plurality of pieces of two-dimensional image information according to the two-dimensional projection plane structure, wherein the plurality of pieces of two-dimensional image information comprise a depth information map, and wherein the depth information map represents a distance between a corresponding point of each occupied pixel in the two-dimensional projection plane structure and a coordinate origin; and
encoding the plurality of pieces of two-dimensional image information to obtain code stream information.
2 . The method according to claim 1 , wherein the encoding the plurality of pieces of two-dimensional image information to obtain the code stream information comprises:
encoding the depth information map to obtain a depth information code stream.
3 . The method according to claim 2 , wherein the encoding the depth information map to obtain the depth information code stream comprises:
performing prediction on a pixel in the depth information map based on a placeholder information map to obtain a predicted residual; or
performing prediction on a pixel in the depth information map based on reconstructed depth information of encoded pixels to obtain a predicted residual, and encoding the predicted residual to obtain the depth information code stream.
4 . The method according to claim 3 , wherein before the encoding the depth information map, the method further comprises:
traversing the depth information map in a preset order, and performing, in a case that a current pixel is an empty pixel, depth information filling on the empty pixel.
5 . The method according to claim 3 , wherein the performing prediction on the pixel in the depth information map based on the placeholder information map to obtain the predicted residual comprises:
traversing pixels in the depth information map in a scanning order; and
determining whether a current pixel is non-empty according to the placeholder information map, and predicting depth information of a current non-empty pixel by using the reconstructed depth information of encoded non-empty pixels, to obtain the predicted residual.
6 . The method according to claim 3 , wherein performing prediction on the pixel in the depth information map based on the reconstructed depth information of encoded pixels to obtain the predicted residual comprises:
performing prediction on the pixel in the depth information map by selecting an optimal prediction mode among a plurality of preset prediction modes to obtain the predicted residual.
7 . The method according to claim 6 , wherein the plurality of preset prediction modes comprises two or more of:
direct mode,
to-the-left prediction mode,
upward prediction mode,
upper left prediction mode,
upper right prediction mode, or
a mode of use pixels on the left, above, on the upper left, and on the upper right as reference pixels.
8 . The method according to claim 3 , wherein encoding the predicted residual to obtain the depth information code stream comprises:
determining whether the predicted residual of the depth information of the pixel is 0, and based on the predicted residual being 0, encoding the predicted residual 0.
9 . The method according to claim 8 , determining whether the predicted residual of the depth information of the pixel is 0, and based on the predicted residual not being 0:
determining whether the predicted residual of the depth information of the pixel is 1, based on the predicted residual being 1, encoding the predicted residual 1 .
10 . A point cloud encoding device, the device comprising:
a processor; and
a memory storing instruction that, when executed by the processor, cause the device to perform operations comprising:
acquiring original point cloud data;
performing two-dimensional regularization plane projection on the original point cloud data to obtain a two-dimensional projection plane structure;
obtaining a plurality of pieces of two-dimensional image information according to the two-dimensional projection plane structure, wherein the plurality of pieces of two-dimensional image information comprise a depth information map, and wherein the depth information map represents a distance between a corresponding point of each occupied pixel in the two-dimensional projection plane structure and a coordinate origin; and
encoding the plurality of pieces of two-dimensional image information to obtain code stream information.
11 . The point cloud encoding device according to claim 10 , wherein the encoding the plurality of pieces of two-dimensional image information to obtain the code stream information comprises:
encoding the depth information map to obtain a depth information code stream.
12 . The point cloud encoding device according to claim 11 , wherein the encoding the depth information map to obtain the depth information code stream comprises:
performing prediction on a pixel in the depth information map based on a placeholder information map to obtain a predicted residual; or
performing prediction on a pixel in the depth information map based on reconstructed depth information of encoded pixels to obtain a predicted residual, and encoding the predicted residual to obtain the depth information code stream.
13 . The point cloud encoding device according to claim 12 , wherein before the encoding the depth information map, the device is caused to perform operations further comprising:
traversing the depth information map in a preset order, and performing, in a case that a current pixel is an empty pixel, depth information filling on the empty pixel.
14 . The point cloud encoding device according to claim 12 , wherein the performing prediction on the pixel in the depth information map based on the placeholder information map to obtain the predicted residual comprises:
traversing pixels in the depth information map in a scanning order; and
determining whether a current pixel is non-empty according to the placeholder information map, and predicting depth information of a current non-empty pixel by using the reconstructed depth information of encoded non-empty pixels, to obtain the predicted residual.
15 . The point cloud encoding device according to claim 12 , wherein performing prediction on the pixel in the depth information map based on the reconstructed depth information of encoded pixels to obtain the predicted residual comprises:
performing prediction on the pixel in the depth information map by selecting an optimal prediction mode among a plurality of preset prediction modes to obtain the predicted residual.
16 . The point cloud encoding device according to claim 15 , wherein the plurality of preset prediction modes comprises two or more of:
direct mode,
to-the-left prediction mode,
upward prediction mode,
upper left prediction mode,
upper right prediction mode, or
a mode of use pixels on the left, above, on the upper left, and on the upper right as reference pixels.
17 . A point cloud decoding method, the method comprising:
acquiring code stream information;
decoding the code stream information to obtain parsed data;
reconstructing a plurality of pieces of two-dimensional image information according to the parsed data, wherein reconstructing the plurality of pieces of two-dimensional image information according to the parsed data comprises reconstructing, according to predicted residuals of a depth information map in the parsed data, a reconstructed depth information map, wherein the reconstructed depth information map represents a distance between a corresponding point of each occupied pixel in a two-dimensional projection plane structure and a coordinate origin;
obtaining a two-dimensional projection plane structure according to the plurality of pieces of two-dimensional image information; and
reconstructing a point cloud based on the two-dimensional projection plane structure.
18 . The method according to claim 17 , wherein decoding the code stream information to obtain the parsed data comprises:
determining whether the predicted residual of a depth information of a pixel is 0, and based on the predicted residual being 0, the predicted residual of the pixel decoded is 0.
19 . The method according to claim 17 , wherein decoding the code stream information to obtain the parsed data comprises:
using a placeholder information map and reconstructed depth information of encoded pixels to perform linear fitting and filling on depth information of a current empty pixel.
20 . The method according to claim 17 , wherein reconstructing, according to the predicted residuals of the depth information map in the parsed data, the depth information map comprises:
predicting a predicted value of the depth information of a current pixel to be reconstructed according to a placeholder information map and reconstructed depth information of encoded pixels; and
reconstructing the depth information of the current pixel according to the predicted value and a predicted residual of the current pixel.