CN115841515A - Environment sensing method based on vehicle body pose - Google Patents

Abstract

The invention discloses an environment sensing method based on a vehicle body pose, which is characterized in that vehicle body pose parameters of a vehicle are integrated, the deviation of a plane of a vehicle body and a real plane is considered during coordinate mapping, a coordinate conversion model between coordinate systems is calculated according to the vehicle body pose parameters and external parameters of a sensor, and sensing coordinates under a sensor coordinate system can be mapped into the coordinate system of the real plane according to the determined coordinate conversion model, so that accurate environment sensing is realized. The error of the mapping position caused by neglecting the deviation between the real plane and the ideal plane of the vehicle in the prior art is eliminated, and the accuracy of environment perception is improved.

Description

Environment sensing method based on vehicle body pose

Technical Field

The invention relates to the technical field of vehicle control, in particular to an environment perception method based on a vehicle body pose.

Background

In the driving support function, it is necessary to sense the environment by the position of a ground mark such as a lane line or a vehicle line. The vehicle usually uses fixed mapping calculation to obtain the perception position of an object according to the position of the object obtained from sensors around the vehicle body, and the mapping calculation in the calculation process is performed based on the condition that the vehicle is on a stable and non-inclined ground.

However, in the actual use process, when sensing and identifying are performed, the vehicle is not necessarily in a stable and non-inclined ground condition, the inclination of the ground can cause deviation of mapping calculation of the sensing position, the position of the vehicle environment sensing and identifying is not accurate, and the implementation of the driving assistance function is affected.

Disclosure of Invention

In order to solve the problems, the invention provides an environment sensing method based on the position and the posture of the vehicle body, which can determine the real position of an object sensed by the environment based on the position and the posture of the vehicle body and improve the accuracy of environment sensing.

The embodiment of the invention provides an environment perception method based on a vehicle body pose, which comprises the following steps:

constructing a coordinate conversion model between a preset ideal horizontal coordinate system, a sensor coordinate system and a real plane coordinate system established by a real plane where the vehicle is located by acquiring vehicle body pose parameters of the vehicle and sensor external parameters calibrated by a vehicle sensor;

and calculating the mapping coordinate of the to-be-positioned point in the real plane coordinate system according to the sensing coordinate of the to-be-positioned point in the sensor coordinate system acquired by the sensor and the coordinate conversion model.

Preferably, the construction process of the real plane coordinate system specifically includes:

taking a foot of a preset position of the vehicle on a horizontal plane as a coordinate origin of the ideal horizontal coordinate system;

establishing a vehicle body coordinate system of a plane where a vehicle body is located relative to the ideal horizontal coordinate system according to the vehicle body pose parameters;

and taking the original point position of the vehicle body coordinate system as a center, and correspondingly rotating the ideal horizontal coordinate system according to the yaw angle and the pitch angle of the real plane of the vehicle body in the vehicle body position and posture parameters relative to the horizontal plane to obtain a real plane coordinate system of the real plane of the current vehicle.

As a preferable scheme, the coordinate transformation model building process includes:

determining an ideal-body coordinate transformation matrix from the ideal horizontal coordinate system to a body coordinate system where a body plane is located and a body-real coordinate transformation matrix from the body coordinate system to the real plane coordinate system according to the body pose parameters;

determining a sensor-body coordinate transformation matrix from the sensor coordinate system to the body coordinate system according to the sensor external parameters;

and taking the ideal-vehicle body coordinate transformation matrix, the vehicle body-real coordinate transformation matrix and the sensor-vehicle body coordinate transformation matrix as the coordinate transformation model.

Further, the ideal-body coordinate transformation matrix is

The body-real coordinate transformation matrix is

The sensor-car body coordinate transformation matrix is T _BC ，T _BC The sensor is externally involved;

and the translation matrix t = [0 h ], h is the height difference between the vehicle body coordinate system and the plane coordinate system, R is a rotation matrix of the vehicle body coordinate system relative to the real plane coordinate system, and is determined by a yaw angle and a pitch angle in the vehicle body pose parameters.

Preferably, the calculating the mapping coordinate of the to-be-positioned point in the real plane coordinate system according to the sensing coordinate of the to-be-positioned point in the sensor coordinate system acquired by the sensor and the coordinate conversion model specifically includes:

determining two ideal positioning lines of the to-be-positioned point under the ideal horizontal coordinate system; the two ideal positioning lines are respectively a first ideal positioning line connected with the origin of coordinates of the ideal horizontal coordinate system and the point to be positioned, a sensor optical center of the ideal horizontal coordinate system and a second ideal positioning line connected with the point to be positioned;

converting the first ideal positioning line into the real plane coordinate system as a first real positioning line according to the coordinate conversion model;

converting the second ideal positioning line into the real plane coordinate system as a second real positioning line according to the coordinate conversion model;

determining an intersection of the first real positioning line and the second real positioning line as the mapping coordinate.

As an improvement of the above solution, the conversion process of the first real positioning line specifically includes:

and rotating a first ideal positioning line connecting the origin of coordinates of the ideal horizontal coordinate system and the sensing coordinates to a first real positioning line passing through the origin of coordinates in the real plane coordinate system according to the rotation angle from the ideal horizontal coordinate system to the real plane coordinate system.

Preferably, the determination process of the second real positioning line comprises:

calculating a sensor-real coordinate transformation matrix from the sensor coordinate system to the real planar coordinate system according to the coordinate transformation model

Sensing coordinates P 'of the to-be-positioned point in a sensor coordinate system according to the sensor-real coordinate transformation matrix' _C Is converted into a first converted coordinate P in the real planar coordinate system _Gr ＝T _GrC P′ _C ；

According to the vehicle body-real coordinate transformation matrix T in the coordinate transformation model _GrB The optical center of the sensor is positioned at the coordinate C of the vehicle body coordinate system _B Is converted into a second conversion coordinate C in the real plane coordinate system _Gr ＝T _GrB C _B ；

Converting the first conversion coordinate P _Gr And the second conversion coordinate C _Gr The connecting line of (a) is taken as the second real positioning line;

wherein an ideal-sensor coordinate transformation matrix of the ideal horizontal coordinate system to the sensor coordinate system

Ideal-real coordinate transformation matrix T of the ideal horizontal coordinate system to the real planar coordinate system _GrG ＝T _GrB T _BG ；T _BC Is a stand forSensor-body coordinate transformation matrix from the sensor coordinate system to the body coordinate system, T _BG Is an ideal-body coordinate transformation matrix, T, of the ideal horizontal coordinate system to the body coordinate system in which the body plane is located _GrB A body-to-real coordinate transformation matrix from the body coordinate system to the real planar coordinate system. />

As a parallel implementation of the above solution, the determining process of the second real positioning line includes:

calculating a sensor-real coordinate transformation matrix from the sensor coordinate system to the real planar coordinate system according to the coordinate transformation model

Sensing coordinates P 'of the to-be-positioned point in a sensor coordinate system according to the sensor-real coordinate transformation matrix' _C Is converted into a first converted coordinate P in the real planar coordinate system _Gr ＝T _GrC P′ _C ；

Calculating perceptual coordinate P 'of the to-be-positioned point' _C Normalized coordinates P of a normalized plane in the sensor coordinate system _C ；

Transforming the matrix T according to the sensor-real coordinates _GrC The normalized coordinate P is measured _C To conversion to normalized conversion coordinates P in said real planar coordinate system _Cr ＝T _GrC P _C ；

Converting the normalized coordinates P _Cr And the first conversion coordinate P _Gr The connecting line of (a) is taken as the second real positioning line;

wherein, P' _C ＝[x y z] ^T ，P _C ＝[x/z y/z 1] ^T An ideal-sensor coordinate transformation matrix of the ideal horizontal coordinate system to the sensor coordinate system

Ideal-real coordinate transformation matrix T of the ideal horizontal coordinate system to the real planar coordinate system _GrG ＝T _GrB T _BG ；T _BC A sensor-body coordinate transformation matrix, T, for said sensor coordinate system to said body coordinate system _BG Is an ideal-body coordinate transformation matrix, T, of the ideal horizontal coordinate system to the body coordinate system in which the body plane is located _GrB A body-to-real coordinate transformation matrix from the body coordinate system to the real planar coordinate system.

As a parallel implementation of the above solution, the determining process of the second real positioning line includes:

calculating a sensor-real coordinate transformation matrix from the sensor coordinate system to the real planar coordinate system according to the coordinate transformation model

Calculating the normalized coordinate P of the perception coordinate of the to-be-positioned point on the normalized plane of the sensor coordinate system _C ；

Transforming the matrix T according to the sensor-real coordinates _GrC The normalized coordinate P is measured _C To conversion to normalized conversion coordinates P in said real planar coordinate system _Cr ＝T _GrC P _C ；

According to the vehicle body-real coordinate transformation matrix T in the coordinate transformation model _GrB The optical center of the sensor is positioned at the coordinate C of the vehicle body coordinate system _B Is converted into a second conversion coordinate C in the real plane coordinate system _Gr ＝T _GrB C _B ；

Converting the normalized coordinates P _Cr And the second conversion coordinate C _Gr The connecting line of (a) is taken as the second real positioning line;

wherein, P' _C ＝[x y z] ^T ，P _C ＝[x/z y/z 1] ^T An ideal-sensor coordinate transformation matrix of the ideal horizontal coordinate system to the sensor coordinate system

The ideal horizontal coordinate system to the real plane coordinateIdeal-true coordinate transformation matrix T of system _GrG ＝T _GrB T _BG ；T _BC A sensor-body coordinate transformation matrix, T, for said sensor coordinate system to said body coordinate system _BG Is an ideal-body coordinate transformation matrix, T, of the ideal horizontal coordinate system to the body coordinate system in which the body plane is located _GrB A body-to-real coordinate transformation matrix of the body coordinate system to the real planar coordinate system.

Another embodiment of the present invention further provides an environment sensing method based on a vehicle body pose, where the method includes:

respectively calculating sensing coordinates of N to-be-positioned points in the sensor coordinate system to corresponding N mapping coordinates in the real plane coordinate system according to the coordinate conversion model to obtain a point pair group consisting of point pairs of the N sensing coordinates and the corresponding mapping coordinates, wherein N is an integer not less than 4;

calculating a coordinate mapping equation from the sensor coordinate system to the real plane coordinate system according to the point pair group;

and mapping the image acquired by the sensor into a real plane according to the coordinate mapping equation.

Preferably, calculating a coordinate mapping equation from the sensor coordinate system to the real plane coordinate system according to the point pair group specifically includes:

constructing a transformation matrix of 8 degrees of freedom for the normalized coordinate transformation from the sensor coordinate system to the real plane coordinate system;

solving the freedom degree parameter in the conversion matrix according to the point pair group;

calculating the conversion matrix according to the calculated freedom degree parameters to obtain the coordinate mapping equation;

wherein the coordinate mapping equation is

In the sensor coordinate systemNormalized coordinate->

-transforming to normalized coordinates in the real planar coordinate system; the conversion matrix is

The matrix form of the degree of freedom parameter of the transformation matrix is expressed as

Corresponding parameter (u) in matrix form of degree of freedom parameter ₁ ，v ₁ ，1)(u ₂ ，v ₂ ，1)、(u ₃ ，v ₃ ，1)(u ₄ ，v ₄ ，1)、(u ₅ ，v ₅ ，1)(u ₆ ，v ₆ 1) and (u) ₇ ，v ₇ ，1)(u ₈ ，v ₈ And 1) four groups of normalized coordinates in the point pair group respectively>

s is a scale factor of two-dimensional homogeneous coordinate normalization, h ₉ Are normalized parameters.

Preferably, the method further comprises:

calculating coordinate mapping equations of different sensors of the vehicle configuration;

respectively mapping the images acquired by the corresponding sensors into real images under a real plane according to coordinate mapping equations of different sensors;

and splicing the plurality of mapped real images to obtain a 360-degree panoramic image under a real plane.

As a preferable scheme, the method further comprises:

mapping the coordinate position of the ground semantic information acquired by a sensor configured by the vehicle into real ground semantic information under a real plane according to the coordinate mapping equation;

the ground semantic information comprises a vehicle position line and/or a lane line.

Compared with the prior art, the invention integrates the vehicle body position and attitude parameters of the vehicle, considers the deviation between the plane of the vehicle body and the real plane during the coordinate mapping, calculates the coordinate conversion model between the coordinate systems according to the vehicle body position and attitude parameters and the sensor external parameters, and can map the perception coordinates under the sensor coordinate system into the real plane coordinate system according to the determined coordinate conversion model, thereby realizing accurate environment perception. The method overcomes the defect that the deviation of a real plane and an ideal plane of the vehicle is ignored in the prior art, so that the deviation of the mapping position occurs, and the accuracy of environment perception is improved.

Drawings

FIG. 1 is a schematic flow chart of an environment sensing method based on a vehicle body pose provided by an embodiment of the invention;

FIG. 2 is a schematic diagram of a location of a coordinate system provided by an embodiment of the present invention;

FIG. 3 is a schematic diagram illustrating a position of an ideal positioning line of a point to be positioned according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of the position of a real positioning line of a point to be positioned according to the present invention;

FIG. 5 is a schematic flow chart of an environment sensing method based on a vehicle body pose in another embodiment of the present invention;

fig. 6 is a schematic view of a perceived image perceived by a vehicle environment according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Referring to fig. 1, a schematic flow diagram of an environment sensing method based on a vehicle body pose according to an embodiment of the present invention is shown, where the method includes steps S1 to S2:

s1, constructing a coordinate conversion model between a preset ideal horizontal coordinate system, a sensor coordinate system and a real plane coordinate system established by a real plane where a vehicle is located by acquiring vehicle body pose parameters of the vehicle and sensor external parameters calibrated by a vehicle sensor;

and S2, calculating the mapping coordinate of the to-be-positioned point in the real plane coordinate system according to the sensing coordinate of the to-be-positioned point in the sensor coordinate system acquired by the sensor and the coordinate conversion model.

When the embodiment is implemented, when the vehicle senses the environment, the vehicle body pose parameter of the vehicle needs to be recognized according to a pose sensing module configured for the vehicle, the pose sensing module comprises a positioning system, an inertia measuring unit and the like, the current position and the current posture of the vehicle, namely the current real plane of the vehicle and the inclination information of the vehicle relative to the real plane can be measured, and the coordinate conversion between a vehicle body coordinate system of the vehicle body plane and a real plane coordinate system constructed by an ideal horizontal coordinate system and the real plane of the vehicle can be determined according to the vehicle body pose parameter;

the vehicle is also provided with a sensor for sensing an external locating point, the sensor is arranged at the position of the vehicle body and corresponds to the sensor external parameter of the vehicle at the angle, and the coordinate conversion between the sensor coordinate system and the vehicle body coordinate system of the vehicle can be constructed according to the sensor external parameter;

the sensor includes various sensors for environment detection, including a camera, a radar sensor, and the like.

According to the vehicle body pose parameters and the sensor external parameters, a preset ideal horizontal coordinate system, a sensor coordinate system and a coordinate conversion model between real plane coordinate systems established by a real plane where the vehicle is located are established; and coordinate conversion calculation for each coordinate system in the coordinate conversion model.

It should be noted that, in this embodiment, the specific construction mode of the ideal horizontal coordinate system, the real planar coordinate system, the sensor coordinate system and the vehicle body coordinate system defines the position by the coordinate system, the ideal horizontal coordinate system is unfolded by taking the ideal horizontal plane as the xy plane, the real planar coordinate system is unfolded by taking the real plane as the xy plane, the vehicle body coordinate system is unfolded by taking the plane where the vehicle chassis is located as the xy plane, and the sensor coordinate system is unfolded by taking the plane perpendicular to the center of the sensor detection direction as the xy plane; the four coordinate systems can be established by taking four different positions of the vehicle body as coordinate origins respectively;

the specific implementation of the scheme is not influenced by the construction position of the coordinate system, and a coordinate conversion model of the coordinate system can be determined according to the pose parameters of the vehicle body and the external parameters of the sensor after the coordinate system is constructed;

the method comprises the steps that a sensor configured by a vehicle acquires a sensing coordinate of a to-be-positioned point in a sensor coordinate system, in the prior art, the sensing coordinate in the sensor coordinate system is converted into a coordinate in a specific coordinate system by adopting a fixed coordinate mapping model to serve as a real coordinate, wherein the specific coordinate system is determined by an ideal horizontal coordinate system, and the deviation of a real plane and an ideal plane of the vehicle is ignored by the method, so that the deviation of a mapped position is caused;

therefore, according to the scheme, the sensing coordinate in the sensor coordinate system can be mapped according to a predetermined coordinate conversion model, and the mapping coordinate of the to-be-positioned point in the real plane coordinate system is calculated.

By integrating the vehicle body position and attitude parameters of the vehicle, the deviation between the plane of the vehicle body and the real plane is considered during coordinate mapping, a coordinate conversion model between coordinate systems is calculated according to the vehicle body position and attitude parameters and the sensor external parameters, and the sensing coordinate under the sensor coordinate system can be mapped into the real plane coordinate system according to the determined coordinate conversion model, so that accurate environment sensing is realized.

In another embodiment of the present invention, the constructing of the real planar coordinate system includes:

establishing a vehicle body coordinate system of a plane where a vehicle body is located relative to the ideal horizontal coordinate system according to the vehicle body pose parameters;

and correspondingly rotating the ideal horizontal coordinate system by taking the original point position of the vehicle body coordinate system as a center according to the yaw angle and the pitch angle of the real plane of the vehicle body in the vehicle body position and posture parameters relative to the horizontal plane to obtain the real plane coordinate system of the real plane of the current vehicle.

In the specific implementation of the present embodiment, refer to fig. 2, which is a schematic position diagram of a coordinate system provided in the embodiment of the present invention;

establishing an ideal horizontal coordinate system by taking a foot G from a rear axle center B of the vehicle to a horizontal plane pi as a coordinate origin of the ideal horizontal coordinate system;

taking the rear axle center B of the vehicle as the origin of coordinates of a vehicle body coordinate system, and taking a vehicle body ground plane as an xy plane to establish the vehicle body coordinate system;

in the embodiment, the vehicle body coordinate system is determined by the fact that the vehicle body does not incline relative to the horizontal plane, and the ideal horizontal coordinate system is translated upwards according to the length h of the vertical line of the vehicle from the horizontal plane in the vehicle body pose parameter, so as to determine the vehicle body coordinate system where the vehicle body plane is located;

determining the inclination angle of a real plane where a vehicle body is located relative to a horizontal plane according to the position and posture parameters of the vehicle body, wherein the inclination angle comprises a yaw angle and a pitch angle, rotating the ideal horizontal coordinate system according to the inclination angle by taking the rear axle center B of the vehicle as a rotating center to obtain a real plane coordinate system, the origin of coordinates of the real plane coordinate system is Gr, and the xy plane of the real plane coordinate system is the real plane pi where the vehicle is located _r ；

The sensor coordinate system is determined by taking the optical center of the sensor as a coordinate origin C, taking the positive detection direction of the sensor as a z coordinate axis, and the normalization plane of the sensor coordinate system is pi _c According to the installation position and the detection direction of the sensor, the coordinate transformation of a sensor coordinate system and a vehicle body coordinate system can be determined, namely the external parameters of the sensor are calibrated.

It should be noted that fig. 2 provided in this embodiment is a cross-sectional view of the coordinate system position on a vertical plane, and only shows a two-dimensional distribution of the coordinate system and the vehicle, and the coordinate system is a three-dimensional coordinate system in the actual implementation process.

It should be noted that, this embodiment provides a preferred embodiment of constructing a coordinate system, and in other embodiments, the position where the coordinate system is established may be determined according to actual situations, and the determining the position of the coordinate system does not affect the implementation of this solution.

The calculation process of the subsequent coordinate conversion can be simplified by rotating the ideal horizontal coordinate system to obtain the actual plane coordinate system.

In another embodiment of the present invention, the coordinate transformation model building process includes:

determining an ideal-body coordinate transformation matrix from the ideal horizontal coordinate system to a body coordinate system where a body plane is located and a body-real coordinate transformation matrix from the body coordinate system to the real plane coordinate system according to the body pose parameters;

determining a sensor-body coordinate transformation matrix from the sensor coordinate system to the body coordinate system according to the sensor external parameters;

and taking the ideal-vehicle body coordinate transformation matrix, the vehicle body-real coordinate transformation matrix and the sensor-vehicle body coordinate transformation matrix as the coordinate transformation model.

In the specific implementation of the embodiment, the coordinate transformation model comprises an ideal-body coordinate transformation matrix, a body-real coordinate transformation matrix and a sensor-body coordinate transformation matrix;

the ideal-vehicle body coordinate transformation matrix is a coordinate transformation matrix from an ideal horizontal coordinate system to a vehicle body coordinate system, the vehicle body pose parameters comprise inclination parameters of a vehicle body plane relative to an ideal horizontal plane, and the ideal-vehicle body coordinate transformation matrix from the ideal horizontal coordinate system to the vehicle body coordinate system is calculated according to the offset of the vehicle body coordinate system and the ideal horizontal coordinate system when the coordinate system is established;

the vehicle body-real coordinate transformation matrix is a coordinate transformation matrix from a vehicle body coordinate system to a real plane coordinate system, the vehicle body pose parameters comprise the inclination parameters of a vehicle body plane relative to the real plane where the vehicle body is located, and the vehicle body-real coordinate transformation matrix from the vehicle body coordinate system to the real plane coordinate system is calculated according to the offset of the vehicle body coordinate system and the real plane coordinate system when the coordinate system is established;

the sensor-vehicle body coordinate transformation matrix is a coordinate transformation matrix from a sensor coordinate system to a vehicle body coordinate system, and when the sensor is installed, the coordinate transformation of the sensor coordinate system and the vehicle body coordinate system can be determined according to the installation position and the detection direction of the sensor relative to the vehicle body, namely the external parameter of the sensor, namely the vehicle body-real coordinate transformation matrix from the vehicle body coordinate system to a real plane coordinate system, is calibrated.

The ideal-vehicle body coordinate transformation matrix, the vehicle body-real coordinate transformation matrix and the sensor-vehicle body coordinate transformation matrix are used as coordinate transformation models, transformation matrixes among four coordinate systems can be deduced according to reversibility of the transformation matrixes, calculation is carried out according to the transformation matrixes, and mapping coordinates of sensing coordinates acquired by the sensor under a real horizontal coordinate system can be determined.

In yet another embodiment of the present invention, the ideal-body coordinate transformation matrix is

The body-real coordinate transformation matrix is

The sensor-car body coordinate transformation matrix is T _BC ，T _BC The sensor is externally involved;

and the translation matrix t = [0 h ], h is the height difference between the vehicle body coordinate system and the plane coordinate system, R is a rotation matrix of the vehicle body coordinate system relative to the real plane coordinate system, and is determined by a yaw angle and a pitch angle in the vehicle body pose parameters.

In the embodiment, referring to fig. 2, in the coordinate system position established in fig. 2, there is only the offset h in the z-axis between the vehicle body coordinate system and the ideal horizontal coordinate system, so that the ideal-vehicle body coordinate transformation matrix can be determined as

Vehicle bodyThe pure rotation of the coordinate system (namely the ideal plane coordinate system) after the coordinate system of the vehicle body is shifted to the z axis by h around the coordinate origin of the coordinate system of the vehicle body as the center between the coordinate system and the real plane coordinate system, thereby determining the transformation matrix of the vehicle body and the real coordinate as

R is a rotation matrix of the vehicle body coordinate system relative to the real plane coordinate system, is determined by a yaw angle and a pitch angle in the vehicle body pose parameters and is used for representing the rotation process of the real plane coordinate system relative to the vehicle body coordinate system; translation matrix t = [0 h =]The translation process of the real plane coordinate system relative to the vehicle body coordinate system is represented;

the sensor-body coordinate transformation matrix, i.e. the position of the sensor relative to the body, can be determined from the installation position and the detection angle of the sensor, i.e. the pre-calibrated sensor external parameter T _BC 。

And a coordinate conversion matrix between coordinate systems can be deduced through the constructed coordinate system model.

In another embodiment provided by the present invention, the step S2 specifically includes:

determining two ideal positioning lines of the to-be-positioned point under the ideal horizontal coordinate system; the two ideal positioning lines are respectively a first ideal positioning line connected with the origin of coordinates of the ideal horizontal coordinate system and the point to be positioned, a sensor optical center of the ideal horizontal coordinate system and a second ideal positioning line connected with the point to be positioned.

And converting the first ideal positioning line into the real plane coordinate system as a first real positioning line according to the coordinate conversion model.

And converting the second ideal positioning line into the real plane coordinate system as a second real positioning line according to the coordinate conversion model.

Determining an intersection of the first real positioning line and the second real positioning line as the mapping coordinate.

In the specific implementation of the present embodiment, refer to fig. 3, which is a schematic position diagram of an ideal positioning line of a to-be-positioned point provided in the embodiment of the present invention.

The specific coordinate conversion process is described by taking the example that the ideal coordinate P of the ideal horizontal coordinate system of the point to be located is on an ideal horizontal plane.

And determining two ideal positioning lines of the point to be positioned under an ideal horizontal coordinate system, wherein the two ideal positioning lines are a first ideal positioning line and a second ideal positioning line respectively.

In this embodiment, the first ideal positioning line is a connection line between the ideal coordinate P and the origin of coordinates G of the ideal horizontal coordinate system, and the ideal coordinate P of the ideal horizontal coordinate system at the point to be positioned is in the ideal horizontal plane, so the first ideal positioning line is in the horizontal plane pi.

The second ideal positioning line is the connecting line CP between the ideal coordinate P and the optical center C of the sensor in the ideal horizontal coordinate system.

And the intersection point of the first ideal positioning line and the second ideal positioning line is the position of the point to be positioned in the ideal horizontal coordinate system.

In the coordinate conversion process, converting the first ideal positioning line into the real plane coordinate system as a first real positioning line according to the coordinate conversion model; converting the second ideal positioning line into the real plane coordinate system as a second real positioning line according to the coordinate conversion model; and determining the mapping position of the point to be positioned in the real plane coordinate system according to the intersection point of the first real positioning line and the second real positioning line.

It should be noted that, in this embodiment, two positioning lines are used to implement mapping of coordinates, and in other embodiments, other coordinate mapping methods may be used, for example, a coordinate mapping matrix is listed, and the coordinate mapping matrix is solved in combination with a coordinate transformation matrix in the coordinate transformation model, so as to determine the coordinate mapping method.

And respectively converting the two positioning lines, and accurately realizing coordinate mapping according to the intersection point of the positioning lines in the real plane coordinate system.

In another embodiment provided by the present invention, the converting process of the first real positioning line specifically includes:

and rotating a first ideal positioning line connecting the origin of coordinates of the ideal horizontal coordinate system and the sensing coordinates to a first real positioning line passing through the origin of coordinates in the real plane coordinate system according to the rotating angle from the ideal horizontal coordinate system to the real plane coordinate system.

In the specific implementation of the present embodiment, refer to fig. 4, which is a schematic position diagram of a real positioning line of a to-be-positioned point provided by the implementation of the present invention:

the first ideal positioning line GP is a line segment passing through the origin of coordinates of the ideal horizontal coordinate system, so that the first real positioning also passes through the origin of coordinates of the real planar coordinate system after coordinate conversion;

and because the ideal horizontal coordinate system and the real plane coordinate system are pure rotations taking the coordinate origin of the vehicle body coordinate system as the rotation center, the first ideal positioning line can be rotated according to the rotation angles of the ideal horizontal coordinate system and the real plane coordinate system, and the first real positioning line G passing through the coordinate origin in the real plane coordinate system is determined _r P’。

The first ideal location line passes through the origin of coordinates, so that when the transformation is performed, a point in the first real location line can be determined as the origin of coordinates, and the first real location line can be determined by rotation.

It should be noted that, in this embodiment, the first ideal positioning line is located in the horizontal plane of the ideal horizontal coordinate system, so the rotated first real positioning line is a line in the real plane of the real plane coordinate system, in other embodiments, the first ideal positioning line is not located in the horizontal plane of the ideal horizontal coordinate system, and the specific calculation method is the same.

In another embodiment of the present invention, the determining process of the second real positioning line includes:

calculating a sensor-real coordinate transformation matrix from the sensor coordinate system to the real planar coordinate system according to the coordinate transformation model

According to the sensor-real coordinate transformation matrixPerception coordinate P 'of to-be-positioned point in sensor coordinate system' _C To a first transformed coordinate P in said real planar coordinate system _Gr ＝T _GrC P′ _C ；

According to the vehicle body-real coordinate transformation matrix T in the coordinate transformation model _GrB The optical center of the sensor is positioned at the coordinate C of the vehicle body coordinate system _B Is converted into a second conversion coordinate C in the real plane coordinate system _Gr ＝T _GrB C _B ；

Converting the first conversion coordinate P _Gr And the second conversion coordinate C _Gr The connecting line of (a) is taken as the second real positioning line;

wherein an ideal-sensor coordinate transformation matrix of the ideal horizontal coordinate system to the sensor coordinate system

An ideal-to-real coordinate transformation matrix T of the ideal horizontal coordinate system to the real planar coordinate system _GrG ＝T _GrB T _BG ；T _BC A sensor-body coordinate transformation matrix, T, for said sensor coordinate system to said body coordinate system _BG Is an ideal-body coordinate transformation matrix, T, of the ideal horizontal coordinate system to the body coordinate system in which the body plane is located _GrB A body-to-real coordinate transformation matrix from the body coordinate system to the real planar coordinate system. />

In this embodiment, the determining process of the second positioning line specifically includes:

ideal-sensor coordinate transformation matrix from ideal horizontal coordinate system to sensor coordinate system

And an ideal-to-real coordinate transformation matrix T of the ideal horizontal coordinate system to the real planar coordinate system _GrG ＝T _GrB T _BG Sensor-real coordinate transformation matrix for calculating sensor coordinate system to real plane coordinate system

Transforming the matrix T according to the sensor-real coordinates _GrC Sensing coordinate P 'of the to-be-positioned point in a sensor coordinate system' _C Is converted into a first converted coordinate P in the real planar coordinate system _Gr ＝T _GrC P′ _C ；

Transforming the matrix T according to the body-real coordinates _GrB Converting the coordinates of the sensor optical center in the vehicle body coordinate system into second conversion coordinates C in the real plane coordinate system _Gr ＝T _GrB C _B ；

Converting the first conversion coordinate P _Gr And the second conversion coordinate C _Gr The connecting line of (a) is taken as the second real positioning line;

the first real positioning line G _r P' and a second true location line P _Gr C _Gr The coordinate of the intersection point Pr between the two points is determined as a mapping coordinate after the mapping of the to-be-positioned point;

wherein the content of the first and second substances,

C _Gr ＝[x ₁ y ₁ z ₁ ] ^T ，P _Gr ＝[x ₂ y ₃ z ₄ ] ^T ，T _BC a sensor-body coordinate transformation matrix, T, for said sensor coordinate system to said body coordinate system _BG An ideal-to-body coordinate transformation matrix, T, for said ideal horizontal coordinate system to the body coordinate system in which the body plane lies _GrB A body-to-real coordinate transformation matrix from the body coordinate system to the real planar coordinate system.

And determining a mapping coordinate after the to-be-positioned point is mapped according to the first real positioning line and the second real positioning line, thereby realizing real environment sensing positioning.

In another embodiment of the present invention, the determining process of the second real positioning line includes:

calculating a sensor-real coordinate transformation matrix from the sensor coordinate system to the real planar coordinate system according to the coordinate transformation model

Sensing coordinates P 'of the to-be-positioned point in a sensor coordinate system according to the sensor-real coordinate transformation matrix' _C To a first transformed coordinate P in said real planar coordinate system _Gr ＝T _GrC P′ _C ；

Calculating perceptual coordinate P 'of the to-be-positioned point' _C Normalized coordinates P of a normalized plane in the sensor coordinate system _C ；

Transforming the matrix T according to the sensor-real coordinates _GrC The normalized coordinate P is measured _C To conversion to normalized conversion coordinates P in said real planar coordinate system _Cr ＝T _GrC P _C ；

Converting the normalized coordinates P _Cr And the first conversion coordinate P _Gr The connecting line of (a) is taken as the second real positioning line;

wherein, P' _C ＝[x y z] ^T ，P _C ＝[x/z y/z 1] ^T An ideal-sensor coordinate transformation matrix of the ideal horizontal coordinate system to the sensor coordinate system

An ideal-to-real coordinate transformation matrix T of the ideal horizontal coordinate system to the real planar coordinate system _GrG ＝T _GrB T _BG ；T _BC A sensor-body coordinate transformation matrix, T, for said sensor coordinate system to said body coordinate system _BG Is an ideal-body coordinate transformation matrix, T, of the ideal horizontal coordinate system to the body coordinate system in which the body plane is located _GrB A body-to-real coordinate transformation matrix from the body coordinate system to the real planar coordinate system.

When the present embodiment is embodied, it is regarded as a parallel embodiment of the above-mentioned embodimentThe points on the second ideal locating line also include, with reference to fig. 2, the normalized coordinates P of the point to be located on the normalized plane of the sensor _C Thus, the normalized coordinate P may be used in performing the conversion of the second ideal locating line _C Positioning a second real positioning line;

in particular, an ideal-sensor coordinate transformation matrix from an ideal horizontal coordinate system to a sensor coordinate system

And an ideal-to-real coordinate transformation matrix T of the ideal horizontal coordinate system to the real planar coordinate system _GrG ＝T _GrB T _BG Sensor-real coordinate transformation matrix for calculating sensor coordinate system to real plane coordinate system

Transforming the matrix T according to the sensor-real coordinates _GrC Sensing coordinate P 'of the to-be-positioned point in a sensor coordinate system' _C Is converted into a first converted coordinate P in the real planar coordinate system _Gr ＝T _GrC P′ _C ；

Calculating perceptual coordinate P 'of the to-be-positioned point' _C Normalized coordinates P of a normalized plane in the sensor coordinate system _C The coordinate of the to-be-positioned point under the sensor coordinate system is P' _C ＝T _CG P＝[x y z] ^T Thus the normalized coordinate P after normalization on the normalization plane _C ＝[x/z y/z 1] ^T ；

Transforming the matrix T according to the sensor-real coordinates _GrC The normalized coordinate P is measured _C To conversion to normalized conversion coordinates P in said real planar coordinate system _Cr ＝T _GrC P _C ；

Converting the normalized coordinates P _Cr And the first conversion coordinate P _Gr The connecting line of (a) is taken as the second real positioning line;

and mapping the to-be-positioned point in the real plane coordinate system according to the second real positioning line and the first real positioning line.

Two points are needed for determining the second real positioning line, and therefore, the coordinates of the points on the second ideal positioning line in the real plane coordinate system can be used for determining the second real positioning line.

In another embodiment of the present invention, the determining process of the second real positioning line includes:

calculating a sensor-real coordinate transformation matrix from the sensor coordinate system to the real planar coordinate system according to the coordinate transformation model

Calculating the normalized coordinate P of the perception coordinate of the to-be-positioned point on the normalized plane of the sensor coordinate system _C ；

Transforming the matrix T according to the sensor-real coordinates _GrC The normalized coordinate P is measured _C To conversion to normalized conversion coordinates P in said real planar coordinate system _Cr ＝T _GrC P _C ；

According to the vehicle body-real coordinate transformation matrix T in the coordinate transformation model _GrB The optical center of the sensor is positioned at the coordinate C of the vehicle body coordinate system _B Converted into a second converted coordinate C in the real planar coordinate system _Gr ＝T _GrB C _B ；

Converting the normalized coordinates P _Cr And the second conversion coordinate C _Gr The connecting line of (a) is taken as the second real positioning line;

wherein, P' _C ＝[x y z] ^T ，P _C ＝[x/z y/z 1] ^T An ideal-sensor coordinate transformation matrix of the ideal horizontal coordinate system to the sensor coordinate system

An ideal-to-real coordinate transformation matrix T of the ideal horizontal coordinate system to the real planar coordinate system _GrG ＝T _GrB T _BG ；T _BC Is said sensingSensor-body coordinate transformation matrix from machine coordinate system to said body coordinate system, T _BG Is an ideal-body coordinate transformation matrix, T, of the ideal horizontal coordinate system to the body coordinate system in which the body plane is located _GrB A body-to-real coordinate transformation matrix from the body coordinate system to the real planar coordinate system.

In the embodiment of the present invention, as a parallel implementation of the above embodiment, referring to fig. 2, the point on the second ideal positioning line further includes a normalized coordinate P of the point to be positioned on the normalized plane of the sensor _C Thus, the normalized coordinate P may be used in performing the conversion of the second ideal locating line _C Positioning a second real positioning line;

in particular, the matrix T is transformed according to the body-real coordinates _GrB Converting the coordinates of the sensor optical center in the vehicle body coordinate system into second conversion coordinates C in the real plane coordinate system _Gr ＝T _GrB C _B ；

Calculating perceptual coordinates P 'of the to-be-positioned point' _C Normalized coordinates P of a normalized plane in the sensor coordinate system _C The coordinate of the to-be-positioned point under the sensor coordinate system is P' _C ＝T _CG P＝[x y z] ^T Thus normalized coordinates P after normalization in the normalization plane _C ＝[x/z y/z 1] ^T ；

Transforming the matrix T according to the sensor-real coordinates _GrC The normalized coordinate P is measured _C To convert to a transformed coordinate P normalized in said real planar coordinate system _Cr ＝T _GrC P _C ；

Converting the normalized coordinates P _Cr And the second conversion coordinate C _Gr The connecting line of (a) is taken as the second real positioning line;

and mapping the to-be-positioned point in the real plane coordinate system according to the second real positioning line and the first real positioning line.

Referring to fig. 4, the normalized transformed coordinates P for determining that the point of the second real positioning line is included in the real planar coordinate system _Cr The second conversionCoordinate C _Gr First transformed coordinate P _Gr (ii) a Two points are needed for determining the second real positioning line, and therefore, the coordinates of the points on the second ideal positioning line in the real plane coordinate system can be used for determining the second real positioning line.

By integrating the vehicle body position and attitude parameters of the vehicle, the deviation between the plane of the vehicle body and the real plane is considered during coordinate mapping, a coordinate conversion model between coordinate systems is calculated according to the vehicle body position and attitude parameters and the sensor external parameters, and the sensing coordinate under the sensor coordinate system can be mapped into the real plane coordinate system according to the determined coordinate conversion model, so that accurate environment sensing is realized. The method overcomes the defect that the deviation of a real plane and an ideal plane of the vehicle is ignored in the prior art, so that the deviation of the mapping position occurs, and the accuracy of environment perception is improved.

Referring to fig. 5, a schematic flow chart of an environment sensing method based on a vehicle body pose in another embodiment of the present invention is shown, where the method includes steps S501 to S503:

s501, constructing a coordinate conversion model among a preset ideal horizontal coordinate system, a sensor coordinate system and a real plane coordinate system established by a real plane where a vehicle is located by acquiring vehicle body pose parameters of the vehicle and sensor external parameters calibrated by a vehicle sensor;

s502, respectively calculating sensing coordinates of N to-be-positioned points in the sensor coordinate system to corresponding N mapping coordinates in the real plane coordinate system according to the coordinate conversion model to obtain a point pair group consisting of the N sensing coordinates and point pairs of the corresponding mapping coordinates, wherein N is an integer not less than 4;

s503, calculating a coordinate mapping equation from the sensor coordinate system to the real plane coordinate system according to the point pair group, and mapping the image acquired by the sensor to the real plane according to the coordinate mapping equation.

When the embodiment is implemented, when the vehicle senses the environment, the vehicle body position and attitude parameters of the vehicle need to be recognized according to a position and attitude sensing module configured for the vehicle, wherein the position and attitude sensing module comprises a positioning system, an inertia measuring unit and the like, and can measure the current position and attitude of the vehicle, namely the current real plane of the vehicle and the inclination information of the vehicle relative to the real plane, and the coordinate conversion between a vehicle body coordinate system of a vehicle body plane and a real plane coordinate system constructed by an ideal horizontal coordinate system and the real plane of the vehicle can be determined according to the vehicle body position and attitude parameters;

the vehicle is also provided with a sensor for sensing an external locating point, the sensor is arranged at the position of the vehicle body and corresponds to the sensor external parameter of the vehicle at the angle, and the coordinate conversion between the sensor coordinate system and the vehicle body coordinate system of the vehicle can be constructed according to the sensor external parameter;

the sensor includes various sensors for environment detection, including a camera, a radar sensor, and the like.

According to the vehicle body pose parameters and the sensor external parameters, a preset ideal horizontal coordinate system, a sensor coordinate system and a coordinate conversion model between real plane coordinate systems established by a real plane where the vehicle is located are established; and calculating the coordinate conversion of each coordinate system in the coordinate conversion model.

It should be noted that, in this embodiment, the specific construction mode of the ideal horizontal coordinate system, the real plane coordinate system, the sensor coordinate system and the vehicle body coordinate system defines the position by the coordinate system, the ideal horizontal coordinate system is developed by taking the ideal horizontal plane as the xy plane, the real plane coordinate system is developed by taking the real plane as the xy plane, the vehicle body coordinate system is developed by taking the plane where the vehicle chassis is located as the xy plane, the sensor coordinate system is developed by taking the plane perpendicular to the center of the sensor detection direction as the xy plane, and the origin points of the three coordinate systems are all set around the vehicle for calculation; it should be noted that the construction position of the coordinate system does not affect the specific implementation of the scheme, and the coordinate transformation model of the coordinate system can be determined according to the pose parameters of the vehicle body and the external parameters of the sensor after the coordinate system is constructed.

As a matrix needs to be transformed into an 8-degree-of-freedom matrix in the coordinate mapping calculation process from the sensor coordinate system to the real plane coordinate system, the sensing coordinates of at least 4 to-be-positioned points in the sensor coordinate system need to be pre-calculated to the corresponding at least 4 mapping coordinates in the real plane coordinate system, and a point pair group consisting of point pairs of the 4 sensing coordinates and the corresponding mapping coordinates is obtained;

and calculating a coordinate mapping equation from the sensor coordinate system to the real plane coordinate system according to the point pair group, and mapping the image acquired by the sensor into a real plane according to the coordinate mapping equation.

The method comprises the steps of calculating mapping point pairs of a coordinate system through a mapping coordinate calculation method, solving a coordinate mapping equation of the coordinate system according to coordinates of a point pair group formed by at least four point pairs, and directly mapping the coordinates according to the coordinate mapping equation to transform the coordinates in an original image acquired by a sensor and correct the image.

The method for calculating the mapping coordinate point in this embodiment adopts the method for calculating the mapping coordinate based on the environment sensing method of the vehicle body pose provided in any one of the embodiments, and the implementation process of the mapping coordinate in the embodiments is specifically described, which is not described herein again.

In another embodiment provided by the present invention, the step S503 specifically includes:

constructing a transformation matrix of 8 degrees of freedom for the normalized coordinate transformation from the sensor coordinate system to the real plane coordinate system;

solving the freedom degree parameter in the conversion matrix according to the point pair group;

calculating the conversion matrix according to the calculated freedom degree parameters to obtain the coordinate mapping equation;

wherein the coordinate mapping equation is

Is a normalized coordinate pickin the sensor coordinate system>

Converting to normalized coordinates in the real planar coordinate system; the conversion matrix is

The matrix form of the degree of freedom parameter of the transformation matrix is expressed as ^ 4>

Corresponding parameters (u) in the form of a matrix of degree of freedom parameters ₁ ，v ₁ ，1)(u ₂ ，v ₂ ，1)、(u ₃ ，v ₃ ，1)(u ₄ ，v ₄ ，1)、(u ₅ ，v ₅ ，1)(u ₆ ，v ₆ 1) and (u) ₇ ，v ₇ ，1)(u ₈ ，v ₈ 1) four sets of normalized coordinates in the respective pair of point pairs, based on the mean value of the normalized coordinates in the respective pair of point pairs>

s is a scale factor of two-dimensional homogeneous coordinate normalization, h ₉ Are normalized parameters.

In the embodiment, the normalized coordinates in the sensor coordinate system are used

Normalized coordinate transformed into the real planar coordinate system->

The linear mapping of (a) can be described by an 8-degree-of-freedom matrix, specifically:

wherein h is ₁ ～h ₉ In order to convert the parameters of the matrix,

s is a scale factor of two-dimensional homogeneous coordinate normalization, h ₉ Are normalized parameters.

Calculating the above calculation formula to obtain a coordinate mapping equation

Wherein H ₁ ～H ₈ The degree of freedom parameter of (2) needs to be solved and determined according to four groups of normalized coordinates in the point pair group, specifically:

thus the parameters of degree of freedom

According to four pairs of point pairs (u) ₁ ，v ₁ ，1)(u ₂ ，v ₂ ，1)、(u ₃ ，v ₃ ，1)(u ₄ ，v ₄ ，1)、(u ₅ ，v ₅ ，1)(u ₆ ，v ₆ 1) and (u) ₇ ，v ₇ ，1)(u ₈ ，v ₈ 1) calculating to determine a coordinate mapping equation;

according to the coordinate mapping equation, the points in the image can be directly subjected to mapping calculation, and the method can be used for directly performing the mapping calculation of the image.

In another embodiment provided by the present invention, the method further comprises:

calculating coordinate mapping equations of different sensors of the vehicle configuration;

respectively mapping the images acquired by the corresponding sensors into real images under a real plane according to coordinate mapping equations of different sensors;

and splicing the plurality of mapped real images to obtain a 360-degree panoramic image under a real plane. When the method is implemented specifically, sensor images in different angle ranges around the vehicle can be acquired through the plurality of sensors, each sensor corresponds to different sensor external parameters, and a coordinate mapping equation of each sensor is calculated;

carrying out mapping calculation on images acquired by different sensors according to different coordinate mapping equations, and mapping the images acquired by different sensors into a plurality of real images under a real plane;

and splicing the plurality of mapped real images to obtain a 360-degree panoramic image under a real plane.

According to coordinate mapping equations of different sensors, images acquired by the different sensors are mapped to a real plane and spliced to obtain a 360-degree panoramic image of a vehicle body, and the accuracy of the 360-degree panoramic image of the vehicle is improved.

In another embodiment provided by the invention, the coordinate position of the ground semantic information acquired by the vehicle-configured sensor is mapped into real ground semantic information under a real plane according to the coordinate mapping equation;

the ground semantic information comprises a vehicle position line and/or a lane line.

In the specific implementation of the present embodiment, referring to fig. 6, it is a schematic view of a perception image for vehicle environment perception provided in the embodiment of the present invention;

the method includes the steps that ground semantic information is acquired through a sensor configured for a vehicle, the ground semantic information includes a vehicle position line and/or a lane line, in this embodiment, the lane line is specifically described as a lane line, the lane line acquired by the sensor is converted into a first lane line a in an ideal horizontal coordinate system, a second lane line is the position of the lane line of the vehicle in an ideal horizontal state, but the second lane line is influenced by position and posture and is inaccurate;

and mapping the lane line acquired by the sensor to a real plane through a coordinate mapping equation of the sensor to obtain a second lane line B.

The obtained ground semantic information can be directly and independently mapped through a coordinate mapping equation, so that the lane lines and the parking space lines can be efficiently extracted and mapped, and the accurate positioning of ground semantics can be realized.

It should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of the present invention, and such improvements and modifications are also considered to be within the scope of the present invention.

Claims (13)

1. An environment sensing method based on a vehicle body pose is characterized by comprising the following steps:

constructing a coordinate conversion model between a preset ideal horizontal coordinate system, a sensor coordinate system and a real plane coordinate system established by a real plane where the vehicle is located by acquiring vehicle body pose parameters of the vehicle and sensor external parameters calibrated by a vehicle sensor;

and acquiring the sensing coordinate of the to-be-positioned point in the sensor coordinate system according to the sensor, and calculating the mapping coordinate of the to-be-positioned point in the real plane coordinate system by using the coordinate conversion model.

2. The vehicle body pose-based environment sensing method according to claim 1, wherein the construction process of the real plane coordinate system specifically comprises:

taking the vertical feet of the preset position of the vehicle on the horizontal plane as the origin of coordinates of the ideal horizontal coordinate system;

establishing a vehicle body coordinate system where a vehicle body plane is located relative to the ideal horizontal coordinate system according to the vehicle body pose parameters;

and correspondingly rotating the ideal horizontal coordinate system by taking the original point position of the vehicle body coordinate system as a center according to the yaw angle and the pitch angle of the real plane of the vehicle body in the vehicle body position and posture parameters relative to the horizontal plane to obtain the real plane coordinate system of the real plane of the current vehicle.

3. The environment perception method based on the vehicle body pose as claimed in claim 1, wherein the coordinate transformation model building process comprises:

determining an ideal-body coordinate transformation matrix from the ideal horizontal coordinate system to a body coordinate system where a body plane is located and a body-real coordinate transformation matrix from the body coordinate system to the real plane coordinate system according to the body pose parameters;

determining a sensor-body coordinate transformation matrix from the sensor coordinate system to the body coordinate system according to the sensor external parameters;

and taking the ideal-vehicle body coordinate transformation matrix, the vehicle body-real coordinate transformation matrix and the sensor-vehicle body coordinate transformation matrix as the coordinate transformation model.

4. The vehicle body pose-based environment sensing method according to claim 3, wherein the ideal-vehicle body coordinate transformation matrix is

The body-real coordinate transformation matrix is

The sensor-car body coordinate transformation matrix is T _BC ，T _BC The sensor is externally involved;

and the translation matrix t = [0 h ], h is the height difference between the vehicle body coordinate system and the plane coordinate system, R is a rotation matrix of the vehicle body coordinate system relative to the real plane coordinate system, and is determined by a yaw angle and a pitch angle in the vehicle body pose parameters.

5. The environment sensing method based on the vehicle body pose as claimed in claim 1, wherein the calculating of the mapping coordinates of the to-be-positioned point in the real plane coordinate system according to the sensing coordinates of the to-be-positioned point obtained by the sensor in the sensor coordinate system and the coordinate transformation model specifically comprises:

determining two ideal positioning lines of the to-be-positioned point under the ideal horizontal coordinate system; the two ideal positioning lines are respectively a first ideal positioning line connected with the origin of coordinates of the ideal horizontal coordinate system and the point to be positioned, and a second ideal positioning line connected with the sensor optical center of the ideal horizontal coordinate system and the point to be positioned;

converting the first ideal positioning line into the real plane coordinate system as a first real positioning line according to the coordinate conversion model;

converting the second ideal positioning line into the real plane coordinate system as a second real positioning line according to the coordinate conversion model;

determining an intersection of the first real positioning line and the second real positioning line as the mapping coordinate.

6. The environment sensing method based on the vehicle body pose as claimed in claim 5, wherein the transformation process of the first real positioning line specifically comprises:

and rotating a first ideal positioning line connecting the origin of coordinates of the ideal horizontal coordinate system and the sensing coordinates to a first real positioning line passing through the origin of coordinates in the real plane coordinate system according to the rotation angle from the ideal horizontal coordinate system to the real plane coordinate system.

7. The vehicle body pose-based environment sensing method according to claim 5, wherein the determination of the second true positioning line comprises:

calculating a sensor-real coordinate transformation matrix from the sensor coordinate system to the real planar coordinate system according to the coordinate transformation model

Sensing coordinates P 'of the to-be-positioned point in a sensor coordinate system according to the sensor-real coordinate transformation matrix' _C To a first transformed coordinate P in said real planar coordinate system _Gr ＝T _GrC P′ _C ；

According to the vehicle body-real coordinate transformation matrix T in the coordinate transformation model _GrB The optical center of the sensor is positioned at the coordinate C of the vehicle body coordinate system _B Is converted into a second conversion coordinate C in the real plane coordinate system _Gr ＝T _GrB C _B ；

Converting the first conversion coordinate P _Gr And the second conversion coordinate C _Gr The connecting line of (a) is taken as the second real positioning line;

wherein the ideal horizontal coordinate systemIdeal-sensor coordinate transformation matrix to said sensor coordinate system

An ideal-to-real coordinate transformation matrix T of the ideal horizontal coordinate system to the real planar coordinate system _GrG ＝T _GrB T _BG ；T _BC A sensor-body coordinate transformation matrix, T, for said sensor coordinate system to said body coordinate system _BG Is an ideal-body coordinate transformation matrix, T, of the ideal horizontal coordinate system to the body coordinate system in which the body plane is located _GrB A body-to-real coordinate transformation matrix from the body coordinate system to the real planar coordinate system.

8. The vehicle body pose-based environment sensing method according to claim 5, wherein the determination of the second true positioning line comprises:

calculating a sensor-real coordinate transformation matrix from the sensor coordinate system to the real planar coordinate system according to the coordinate transformation model

Sensing coordinates P 'of the to-be-positioned point in a sensor coordinate system according to the sensor-real coordinate transformation matrix' _C Is converted into a first converted coordinate P in the real planar coordinate system _Gr ＝T _GrC P′ _C ；

Calculating perceptual coordinate P 'of the to-be-positioned point' _C Normalized coordinate P of a normalized plane in the sensor coordinate system _C ；

Transforming the matrix T according to the sensor-real coordinates _GrC The normalized coordinate P is measured _C To conversion to normalized conversion coordinates P in said real planar coordinate system _Cr ＝T _GrC P _C ；

Converting the normalized coordinates P _Cr And the first conversion coordinate P _Gr The connecting line of (a) is taken as the second real positioning line;

wherein, P' _C ＝[x y z] ^T ，P _C ＝[x/z y/z 1] ^T An ideal-sensor coordinate transformation matrix of the ideal horizontal coordinate system to the sensor coordinate system

An ideal-to-real coordinate transformation matrix T of the ideal horizontal coordinate system to the real planar coordinate system _GrG ＝T _GrB T _BG ；T _BC A sensor-body coordinate transformation matrix, T, for said sensor coordinate system to said body coordinate system _BG Is an ideal-body coordinate transformation matrix, T, of the ideal horizontal coordinate system to the body coordinate system in which the body plane is located _GrB A body-to-real coordinate transformation matrix from the body coordinate system to the real planar coordinate system.

9. The vehicle body pose-based environment sensing method according to claim 5, wherein the determining process of the second real positioning line comprises:

calculating a sensor-real coordinate transformation matrix from the sensor coordinate system to the real planar coordinate system according to the coordinate transformation model

Calculating the normalized coordinate P of the perception coordinate of the to-be-positioned point on the normalized plane of the sensor coordinate system _C ；

Transforming the matrix T according to the sensor-real coordinates _GrC The normalized coordinate P is measured _C To conversion to normalized conversion coordinates P in said real planar coordinate system _Cr ＝T _GrC P _C ；

According to the vehicle body-real coordinate transformation matrix T in the coordinate transformation model _GrB The optical center of the sensor is positioned at the coordinate C of the vehicle body coordinate system _B Is converted into a second conversion coordinate C in the real plane coordinate system _Gr ＝T _GrB C _B ；

Converting the normalized coordinates P _Cr And the second conversion coordinate C _Gr The connecting line of (a) is taken as the second real positioning line;

wherein, P' _C ＝[x y z] ^T ，

Ideal-sensor coordinate transformation matrix of the ideal horizontal coordinate system to the sensor coordinate system

An ideal-to-real coordinate transformation matrix T of the ideal horizontal coordinate system to the real planar coordinate system _GrG ＝T _GrB T _BG ；T _BC A sensor-body coordinate transformation matrix, T, for said sensor coordinate system to said body coordinate system _BG Is an ideal-body coordinate transformation matrix, T, of the ideal horizontal coordinate system to the body coordinate system in which the body plane is located _GrB A body-to-real coordinate transformation matrix of the body coordinate system to the real planar coordinate system.

10. An environment sensing method based on a vehicle body pose is characterized by comprising the following steps:

constructing a coordinate conversion model between a preset ideal horizontal coordinate system, a sensor coordinate system and a real plane coordinate system established by a real plane where the vehicle is located by acquiring vehicle body pose parameters of the vehicle and sensor external parameters calibrated by a vehicle sensor;

respectively calculating sensing coordinates of N to-be-positioned points in the sensor coordinate system to corresponding N mapping coordinates in the real plane coordinate system according to the coordinate conversion model to obtain a point pair group consisting of point pairs of the N sensing coordinates and the corresponding mapping coordinates, wherein N is an integer not less than 4;

and calculating a coordinate mapping equation from the sensor coordinate system to the real plane coordinate system according to the point pair group, and mapping the image acquired by the sensor into a real plane according to the coordinate mapping equation.

11. The vehicle body pose-based environment sensing method according to claim 10, wherein calculating a coordinate mapping equation of the sensor coordinate system to the real plane coordinate system according to the point pair group comprises:

constructing a transformation matrix of 8 degrees of freedom for the normalized coordinate transformation from the sensor coordinate system to the real plane coordinate system;

solving the freedom degree parameter in the conversion matrix according to the point pair group;

calculating the conversion matrix according to the calculated freedom degree parameters to obtain the coordinate mapping equation;

wherein the coordinate mapping equation is

For normalized coordinates in the sensor coordinate system

Converting to normalized coordinates in the real planar coordinate system; the conversion matrix is

The matrix form of the degree of freedom parameter of the transformation matrix is expressed as

Corresponding parameter (u) in matrix form of degree of freedom parameter ₁ ，v ₁ ，1)(u ₂ ，v ₂ ，1)、(u ₃ ，v ₃ ，1)(u ₄ ，v ₄ ，1)、(u ₅ ，v ₅ ，1)(u ₆ ，v ₆ 1) and (u) ₇ ，v ₇ ，1)(u ₈ ，v ₈ And 1) the point pair groups respectivelyFour sets of normalized coordinates of (a) are,

s is a scale factor of two-dimensional homogeneous coordinate normalization, h ₉ Are normalized parameters.

12. The vehicle body pose-based environment sensing method according to claim 10, further comprising:

calculating coordinate mapping equations of different sensors of the vehicle configuration;

respectively mapping the images acquired by the corresponding sensors into real images under a real plane according to coordinate mapping equations of different sensors;

and splicing the plurality of mapped real images to obtain a 360-degree panoramic image under a real plane.

13. The vehicle body pose-based environment perception method according to claim 10, further comprising:

mapping the coordinate position of the ground semantic information acquired by a sensor configured by the vehicle into real ground semantic information under a real plane according to the coordinate mapping equation;

the ground semantic information comprises a vehicle position line and/or a lane line.

Images