Point cloud attribute encoding method and apparatus, decoding method and apparatus, and related device
A point cloud attribute encoding method and apparatus, a decoding method and apparatus, and a related device are disclosed. The point cloud attribute encoding method includes sorting all point cloud data to be encoded to acquire sorted point cloud data, the point cloud data to be encoded being point cloud data with attributes to be encoded; grouping the sorted point cloud data based on correlation between the sorted point cloud data to acquire groups to be encoded; and performing point cloud attribute encoding based on the groups to be encoded. It is beneficial for enhancing the correlation between point cloud data within the group, improving the efficiency of decorrelation during intra group transform after grouping, and improving coding efficiency.
1 . A point cloud attribute encoding method, comprising:
sorting all point cloud data to be encoded to acquire sorted point cloud data, wherein the point cloud data to be encoded are point cloud data with attributes to be encoded;
grouping the sorted point cloud data based on correlation between the sorted point cloud data to acquire groups to be encoded; and
performing point cloud attribute encoding based on the groups to be encoded;
wherein the grouping the sorted point cloud data based on correlation between the sorted point cloud data to acquire groups to be encoded, comprises:
acquiring a target size; dividing a target space based on the target size to obtain a plurality of space blocks, wherein the target space is a space where all the sorted point cloud data are located, and a size of each space block is the same as the target size; for each space block, grouping all the sorted point cloud data in the space block into one group to be encoded; or
acquiring a target bit number; comparing target codes of the sorted point cloud data; grouping the sorted point cloud data to obtain the groups to be encoded by grouping the sorted point cloud data with the same last bits of the target bit number in the target codes into one group to be encoded; or
acquiring a target bit number; comparing target codes of the sorted point cloud data; grouping the sorted point cloud data based on bits before the last bits of the target bit number in the target codes to acquire the groups to be encoded.
2 . The point cloud attribute encoding method according to claim 1 , wherein the correlation between the sorted point cloud data comprises a spatial position relationship of the sorted point cloud data.
3 . The point cloud attribute encoding method according to claim 1 , wherein the sorting all point cloud data to be encoded to acquire sorted point cloud data, comprises:
acquiring target codes corresponding to the point cloud data to be encoded, wherein the target codes are Morton codes or Hilbert codes; and
sorting the point cloud data to be encoded according to an ascending order of the target codes to acquire the sorted point cloud data.
4 . The point cloud attribute encoding method according to claim 1 , wherein the acquiring a target size comprises:
acquiring a points number of the sorted point cloud data and a maximum side length corresponding to all the sorted point cloud data, wherein the maximum side length is a length of the longest side of the minimum cuboid bounding box corresponding to all the sorted point cloud data;
acquiring a target mean points number of the groups to be encoded; and
calculating the target size, wherein S=2 L/3 ,
L
=
3
*
(
log
2
(
max
Size
)
-
log
2
(
N
K
mean
)
2
)
,
S is the target size, L is a target bit number, maxSize is the maximum side length, N is the points number of the sorted point cloud data, and K mean is the target mean points number.
5 . The point cloud attribute encoding method according to claim 1 , wherein the acquiring a target bit number comprises:
acquiring a points number of the sorted point cloud data and a maximum side length corresponding to all the sorted point cloud data, wherein the maximum side length is a length of the longest side of the minimum cuboid bounding box corresponding to all the sorted point cloud data;
acquiring a target mean points number of the groups to be encoded; and
calculating the target bit number, wherein
L
=
3
*
(
log
2
(
max
Size
)
-
log
2
(
N
K
mean
)
2
)
,
L is the target bit number, maxSize is the maximum side length, N is the points number of the sorted point cloud data, and K mean is the target mean points number.
6 . The point cloud attribute encoding method according to claim 1 , wherein the grouping the sorted point cloud data based on the bits before the last bits of the target bit number in the target codes to acquire the groups to be encoded, comprises:
grouping the sorted point cloud data to obtain the groups to be encoded by grouping the sorted point cloud data with the same bits before the last bits of the target bit number in the target codes into one group to be encoded.
7 . The point cloud attribute encoding method according to claim 1 , wherein the acquiring a target bit number comprises:
dynamically adjusting the target bit number during encoding based on a statistical characteristic of a points number of the sorted point cloud data in encoded groups.
8 . The point cloud attribute encoding method according to claim 7 , wherein the dynamically adjusting the target bit number comprises:
determining an initial target bit number Ld;
presetting variable parameters BN1 and BN2, wherein BN1<BN2, BN1 and BN2 are positive integers greater than 0; and presetting variable parameters KN and KM, wherein KN<KM, KN and KM are positive integers greater than 1; and
during encoding, calculating a mean points number BN of the sorted point cloud data in KN encoded groups; if BN is less than BN1, then Ld=Ld+1; if BN is greater than BN2, then Ld=Ld−1; otherwise, Ld is unchanged; adjusting once after every KM groups are encoded.
9 . The point cloud attribute encoding method according to claim 1 , wherein the performing point cloud attribute encoding based on the groups to be encoded, comprises:
acquiring a maximum points number;
for each group to be encoded, if a points number of the sorted point cloud data in the group to be encoded is greater than the maximum points number, taking the group to be encoded as a group to be subdivided; otherwise, taking the group to be encoded as a qualified group to be encoded;
for each group to be subdivided, subdividing the sorted point cloud data in the group to be subdivided to acquire a plurality of target subdivided groups to be encoded, wherein the points number of the sorted point cloud data in each of the target subdivided groups to be encoded is not greater than the maximum points number;
taking the qualified groups to be encoded and the target subdivided groups to be encoded as target groups to be encoded; and
performing point cloud attribute encoding based on the target groups to be encoded.
10 . The point cloud attribute encoding method according to claim 9 , wherein among the plurality of target subdivided groups to be encoded corresponding to each group to be subdivided, the points number of the sorted point cloud data in one target subdivided group to be encoded is not greater than the maximum points number, the points number of the sorted point cloud data in other target subdivided groups to be encoded is equal to the maximum points number.
11 . The point cloud attribute encoding method according to claim 9 , wherein the performing point cloud attribute encoding based on the target groups to be encoded, comprises:
performing forward discrete cosine transform on each of the target groups to be encoded based on the points number of the sorted point cloud data in each of the target groups to be encoded to acquire transform coefficients of each of the target groups to be encoded; and
performing quantization and entropy encoding on the transform coefficients of each of the target groups to be encoded.
12 . A point cloud attribute decoding method, comprising:
sorting all point cloud data to be decoded to acquire sorted point cloud data to be decoded, wherein the point cloud data to be decoded are point cloud data with attributes to be decoded;
grouping the sorted point cloud data to be decoded based on correlation between the sorted point cloud data to be decoded to acquire groups to be decoded;
performing entropy decoding and inverse quantization on each of the groups to be decoded to obtain transform coefficients of each of the groups to be decoded; and
performing inverse discrete cosine transform on the transform coefficients of each of the groups to be decoded based on the points number of the sorted point cloud data to be decoded in each of the groups to be decoded to obtain attribute reconstruction values corresponding to each of the groups to be decoded.
13 . The point cloud attribute decoding method according to claim 12 , wherein the correlation between the sorted point cloud data to be decoded comprises a spatial position relationship of the sorted point cloud data to be decoded.
14 . An intelligent terminal, comprising a memory, a processor, and a point cloud attribute decoding program stored in the memory and executable on the processor, wherein the point cloud attribute decoding program, when executed by the processor, is configured to perform a point cloud attribute decoding method, comprising:
sorting all point cloud data to be decoded to acquire sorted point cloud data to be decoded, wherein the point cloud data to be decoded are point cloud data with attributes to be decoded;
grouping the sorted point cloud data to be decoded based on correlation between the sorted point cloud data to be decoded to acquire groups to be decoded;
performing entropy decoding and inverse quantization on each of the groups to be decoded to obtain transform coefficients of each of the groups to be decoded; and
performing inverse discrete cosine transform on the transform coefficients of each of the groups to be decoded based on the points number of the sorted point cloud data to be decoded in each of the groups to be decoded to obtain attribute reconstruction values corresponding to each of the groups to be decoded.