Large-scale point cloud-oriented two-dimensional regularized planar projection and encoding and decoding method
A large-scale point cloud-oriented two-dimensional regularized planar projection method and an encoding and decoding method are disclosed. The two-dimensional regularized planar projection method includes: obtaining original point cloud data; initializing a two-dimensional projection plane structure of a point cloud; and determining a mapping relationship between the original point cloud data and the two-dimensional projection plane structure to obtain a two-dimensional regularized projection plane structure corresponding to the point cloud. The encoding and decoding method includes: predicting based on the two-dimensional regularized projection plane structure of the point cloud to obtain to-be-encoded data; dividing the to-be-encoded data into a first type of to-be-encoded data and a second type of to-be-encoded data; and encoding different data in the first type of to-be-encoded data separately in a preset encoding manner, and correspondingly encoding the second type of to-be-encoded data, to obtain a geometric information bitstream.
1 . A large-scale point cloud-oriented two-dimensional regularized planar projection method, comprising:
obtaining original point cloud data;
initializing a two-dimensional projection plane structure of a point cloud by using regularization parameters; and
determining a mapping relationship between the original point cloud data and the two-dimensional projection plane structure to obtain a two-dimensional regularized projection plane structure of the point cloud;
wherein the initializing the two-dimensional projection plane structure of the point cloud by using the regularization parameters comprises:
initializing the two-dimensional projection plane structure of the point cloud based on a number of laser scanners and one of (i) a sampling angular resolution Δφ in a horizontal azimuth, or (ii) a number of sampling points of a laser scanner in the regularization parameters, that is,
M
=
laserNum
;
and
N
=
360
°
Δφ
or
N
=
pointNumPerLaser
,
wherein M and N respectively represent a resolution in a vertical direction and a resolution in a horizontal direction in the two-dimensional projection plane structure, laserNum represents the number of laser scanners, Δφ represents the sampling angular resolution of the laser scanner in the horizontal azimuth, and pointNumPerLaser represents the number of sampling points of the laser scanner.
2 . The large-scale point cloud-oriented two-dimensional regularized planar projection method according to claim 1 , wherein the regularization parameters comprise a calibration parameter of a lidar.
3 . The large-scale point cloud-oriented two-dimensional regularized planar projection method according to claim 1 , wherein the determining the mapping relationship between the original point cloud data and the two-dimensional projection plane structure comprises:
seeking a solution through the regularization parameters and calibration formulas to obtain a corresponding position of a point in the original point cloud data in the two-dimensional projection plane structure through calculation to determine the mapping relationship between the original point cloud data and the two-dimensional projection plane structure, wherein the calibration formulas are represented as:
r
=
x
2
+
y
2
;
i
=
arg
min
i
=
1
…
laserNum
❘
"\[LeftBracketingBar]"
z
-
V
o
-
r
×
tan
θ
o
❘
"\[RightBracketingBar]"
θ
i
=
θ
0
;
x
=
r
·
sin
(
ϕ
j
-
α
)
-
H
o
·
cos
(
ϕ
j
-
α
)
;
y
=
r
·
cos
(
ϕ
j
-
α
)
+
H
o
·
sin
(
ϕ
j
-
α
)
;
z
=
r
·
tan
θ
i
+
V
o
;
and
j
=
(
ϕ
j
+
180
°
)
/
Δφ
,
wherein (x, y, z) represents Cartesian coordinates of the point in the original point cloud data, r is a cylindrical coordinate component of the point representing a distance between the point and a coordinate origin of a lidar, θ i and φ j are a pitch angle and an azimuth of a corresponding pixel of the point in the two-dimensional projection plane structure, (i, j) represents a position of the corresponding pixel of the point in the two-dimensional projection plane structure, and θ 0 , V o , H o , and α are the regularization parameters.
4 . The large-scale point cloud-oriented two-dimensional regularized planar projection method according to claim 1 , wherein the determining the mapping relationship between the original point cloud data and the two-dimensional projection plane structure further comprises:
determining a cylindrical coordinate component of a current point in the original point cloud data;
determining a search region of the current point in the two-dimensional projection plane structure;
traversing pixels in the search region, calculating a position of a current pixel in a Cartesian coordinate system, and calculating a spatial distance between the position and the current point;
selecting a pixel with a shortest spatial distance as a corresponding pixel of the current point in the two-dimensional projection plane structure; and
repeating the foregoing steps until corresponding pixels in the two-dimensional projection plane structure are found for each point in the original point cloud data.
5 . The large-scale point cloud-oriented two-dimensional regularized planar projection method according to claim 4 , wherein the determining the search region of the current point in the two-dimensional projection plane structure comprises one of:
(i) determining the search region of the current point in the two-dimensional projection plane structure through a pitch angle θ and an azimuth φ of the current point in a cylindrical coordinate system; or
(ii) determining the search region of the current point in the two-dimensional projection plane structure through the regularization parameters and calibration formulas; or
(iii) determining the search region of the current point according to a-priori information.
6 . The large-scale point cloud-oriented two-dimensional regularized planar projection method according to claim 4 , wherein calculation formulas for the position of the current pixel in the Cartesian coordinate system are:
θ
i
=
θ
0
;
ϕ
j
=
-
180
°
+
j
×
Δφ
;
r
=
x
2
+
y
2
;
xl
=
r
·
sin
(
ϕ
j
-
α
)
-
H
o
·
cos
(
ϕ
j
-
α
)
;
yl
=
r
·
cos
(
ϕ
j
-
α
)
+
H
o
·
sin
(
ϕ
j
-
α
)
;
and
zl
=
r
·
tan
θ
i
+
V
o
,
wherein (i, j) represents a position of the current pixel in the two-dimensional projection plane structure, a pitch angle and an azimuth that correspond to the current pixel are θ i and φ j , x and y are Cartesian coordinate components of the current point in the original point cloud data, r is a cylindrical coordinate component of the current point in the original point cloud data, (xl, yl, zl) represents the position of the current pixel in the Cartesian coordinate system, and θ 0 , V o , H o and α are the regularization parameters.
7 . The large-scale point cloud-oriented two-dimensional regularized planar projection method according to claim 1 , wherein the regularization parameters comprise a parameter obtained through optimized estimation or data fitting.
8 . A large-scale point cloud-oriented encoding method, comprising:
obtaining original point cloud data;
initializing a two-dimensional projection plane structure of a point cloud by using regularization parameters, wherein the initializing the two-dimensional projection plane structure of the point cloud by using the regularization parameters comprises:
initializing the two-dimensional projection plane structure of the point cloud based on a number of laser scanners and one of (i) a sampling angular resolution Δφ in a horizontal azimuth, or (ii) a number of sampling points of a laser scanner in the regularization parameters, that is,
M
=
laserNum
;
and
N
=
360
°
Δφ
or
N
=
pointNumPerLaser
,
wherein M and N respectively represent a resolution in a vertical direction and a resolution in a horizontal direction in the two-dimensional projection plane structure, laserNum represents the number of laser scanners, Δφ represents the sampling angular resolution of the laser scanner in the horizontal azimuth, and pointNumPerLaser represents the number of sampling points of the laser scanner;
obtaining a two-dimensional regularized projection plane structure of a point cloud based on a mapping relationship between the original point cloud data and the two-dimensional projection plane structure;
predicting, based on the two-dimensional regularized projection plane structure of the point cloud, to-be-encoded data;
dividing the to-be-encoded data into a first type of to-be-encoded data and a second type of to-be-encoded data, wherein the first type of to-be-encoded data is two-dimensional regularized projection plane structure data, and the second type of to-be-encoded data is remaining to-be-encoded data other than the two-dimensional regularized projection plane structure data; and
encoding different data in the first type of to-be-encoded data separately in a preset encoding manner, and encoding the second type of to-be-encoded data, to obtain a geometric information bitstream.
9 . The large-scale point cloud-oriented encoding method according to claim 8 , wherein the predicting the to-be-encoded data comprises:
setting different prediction modes for different data according to the two-dimensional regularized projection plane structure of the point cloud; and
predicting, according to a corresponding prediction mode, to obtain the to-be-encoded data.
10 . An electronic device, comprising:
a memory configured to store computer-executable instructions; and
a processor configured to execute the computer-executable instructions to cause the electronic device to:
obtain original point cloud data;
initialize a two-dimensional projection plane structure of a point cloud by using regularization parameters; and
determine a mapping relationship between the original point cloud data and the two-dimensional projection plane structure to obtain a two-dimensional regularized projection plane structure of the point cloud;
wherein the initializing the two-dimensional projection plane structure of the point cloud by using the regularization parameters comprises:
initializing the two-dimensional projection plane structure of the point cloud based on a number of laser scanners and one of (i) a sampling angular resolution Δφ in a horizontal azimuth, or (ii) a number of sampling points of a laser scanner in the regularization parameters, that is,
M
=
laserNum
;
and
N
=
360
°
Δφ
or
N
=
pointNumPerLaser
,
wherein M and N respectively represent a resolution in a vertical direction and a resolution in a horizontal direction in the two-dimensional projection plane structure, laserNum represents the number of laser scanners, Δφ represents the sampling angular resolution of the laser scanner in the horizontal azimuth, and pointNumPerLaser represents the number of sampling points of the laser scanner.
11 . The electronic device according to claim 10 , wherein the regularization parameters comprise a calibration parameter of a lidar.
12 . The electronic device according to claim 10 , wherein the regularization parameters comprise a parameter obtained through optimized estimation or data fitting.
13 . The electronic device according to claim 10 , wherein the determining the mapping relationship between the original point cloud data and the two-dimensional projection plane structure comprises:
seeking a solution through the regularization parameters and calibration formulas to obtain a corresponding position of a point in the original point cloud data in the two-dimensional projection plane structure through calculation to determine the mapping relationship between the original point cloud data and the two-dimensional projection plane structure, wherein the calibration formulas are represented as:
r
=
x
2
+
y
2
;
i
=
arg
min
i
=
1
…
laserNum
❘
"\[LeftBracketingBar]"
z
-
V
o
-
r
×
tan
θ
o
❘
"\[RightBracketingBar]"
θ
i
=
θ
0
;
x
=
r
·
sin
(
ϕ
j
-
α
)
-
H
o
·
cos
(
ϕ
j
-
α
)
;
y
=
r
·
cos
(
ϕ
j
-
α
)
+
H
o
·
sin
(
ϕ
j
-
α
)
;
z
=
r
·
tan
θ
i
+
V
o
;
and
j
=
(
ϕ
j
+
180
°
)
/
Δφ
,
wherein (x, y, z) represents Cartesian coordinates of the point in the original point cloud data, r is a cylindrical coordinate component of the point representing a distance between the point and a coordinate origin of a lidar, θ i and φ j are a pitch angle and an azimuth of a corresponding pixel of the point in the two-dimensional projection plane structure, (i, j) represents a position of the corresponding pixel of the point in the two-dimensional projection plane structure, and θ 0 , V o , H o , and α are the regularization parameters.
14 . The electronic device according to claim 10 , wherein the determining the mapping relationship between the original point cloud data and the two-dimensional projection plane structure further comprises:
determining a cylindrical coordinate component of a current point in the original point cloud data;
determining a search region of the current point in the two-dimensional projection plane structure;
traversing pixels in the search region, calculating a position of a current pixel in a Cartesian coordinate system, and calculating a spatial distance between the position and the current point;
selecting a pixel with a shortest spatial distance as a corresponding pixel of the current point in the two-dimensional projection plane structure; and
repeating the foregoing steps until corresponding pixels in the two-dimensional projection plane structure are found for each point in the original point cloud data.
15 . The electronic device according to claim 14 , wherein the determining the search region of the current point in the two-dimensional projection plane structure comprises one of:
(i) determining the search region of the current point in the two-dimensional projection plane structure through a pitch angle θ and an azimuth φ of the current point in a cylindrical coordinate system; or
(ii) determining the search region of the current point in the two-dimensional projection plane structure through the regularization parameters and calibration formulas; or
(iii) determining the search region of the current point according to a-priori information.
16 . The electronic device according to claim 14 , wherein calculation formulas for the position of the current pixel in the Cartesian coordinate system are:
θ
i
=
θ
0
;
ϕ
j
=
-
180
°
+
j
×
Δφ
;
r
=
x
2
+
y
2
;
xl
=
r
·
sin
(
ϕ
j
-
α
)
-
H
o
·
cos
(
ϕ
j
-
α
)
;
yl
=
r
·
cos
(
ϕ
j
-
α
)
+
H
o
·
sin
(
ϕ
j
-
α
)
;
and
zl
=
r
·
tan
θ
i
+
V
o
,
wherein (i, j) represents a position of the current pixel in the two-dimensional projection plane structure, a pitch angle and an azimuth that correspond to the current pixel are θ i and φ j , x and y are Cartesian coordinate components of the current point in the original point cloud data, r is a cylindrical coordinate component of the current point in the original point cloud data, (xl, yl, zl) represents the position of the current pixel in the Cartesian coordinate system, and θ 0 , V o , H o and α are the regularization parameters.