Laser odometer method, system and equipment based on reflectivity characteristics and medium

Abstract

The embodiment of the invention provides a laser odometer method, system and device based on reflectivity characteristics and a medium. The method comprises the steps that the initial pose of a vehicle is obtained according to vehicle inertia information; registering the target laser point cloud with the geometric feature point map according to the initial pose; determining a target geometric feature point in the target laser point cloud; constructing a geometric residual matrix and a first Jacobian matrix; registering the target laser point cloud with the luminosity feature point map according to the initial pose; extracting a target luminosity feature point in the target laser point cloud; constructing a luminosity residual matrix and a second jacobian matrix; constructing a target residual matrix and a target Jacobian matrix; performing iterative optimization on the initial pose based on the target residual matrix and the target Jacobian matrix to obtain a target pose; and updating a geometric feature point map and a luminosity feature point map based on the target pose and the target laser point cloud. The invention aims to ensure the positioning precision of the laser odometer in a tunnel environment.

Description

Technical Field

[0001] The present invention relates to the field of laser radar detection, and in particular to a laser odometer method, system, equipment and medium based on reflectivity characteristics. Background Art

[0002] With the rapid development of vehicle positioning technology, laser odometry, as a key sensor technology, has been widely used in autonomous driving and vehicle navigation systems. It provides real-time and accurate pose estimation for vehicles by quickly and accurately acquiring three-dimensional point cloud data of the environment. However, in some environmental conditions with degraded geometric features, such as tunnel environments, the point cloud data returned by laser scanning often show a high degree of similarity and redundancy, lacking significant geometric features such as straight lines, planes or edges, making it difficult for laser odometry to accurately extract and match feature points in tunnel environments, resulting in inaccurate pose estimation and affecting positioning accuracy. Summary of the invention

[0003] In view of this, the embodiments of the present application provide a laser odometer method, system, device and medium based on reflectivity characteristics, aiming to ensure the positioning accuracy of the laser odometer in a tunnel environment.

[0004] A first aspect of an embodiment of the present application provides a laser odometer method based on reflectivity characteristics, the method comprising:

[0005] According to the collected vehicle inertia information, the initial position and posture of the vehicle is obtained through integration operation;

[0006] According to the initial posture, registering the collected target laser point cloud with the geometric feature point map to obtain a first registration result;

[0007] Determine a target geometric feature point in the target laser point cloud according to the first registration result, wherein the target geometric feature point is a laser point with a planar feature in the target laser point cloud;

[0008] Construct the geometric residual matrix and the first Jacobian matrix according to the target geometric feature points, the geometric feature point map and the initial pose;

[0009] According to the initial posture, registering the target laser point cloud with the photometric feature point map to obtain a second registration result;

[0010] Extracting target photometric feature points with the highest reflectivity in each area of ​​the target laser point cloud;

[0011] Constructing a photometric residual matrix and a second Jacobian matrix according to the second registration result, the target photometric feature points, the photometric feature point map and the initial pose;

[0012] Constructing a target residual matrix based on the geometric residual matrix and the photometric residual matrix, and constructing a target Jacobian matrix based on the first Jacobian matrix and the second Jacobian matrix;

[0013] Iteratively optimizing the initial pose based on the target residual matrix and the target Jacobian matrix to obtain a target pose;

[0014] Based on the target pose and the target laser point cloud, a geometric feature point map and a photometric feature point map are updated.

[0015] Optionally, determining target geometric feature points in the target laser point cloud according to the first registration result includes:

[0016] According to the first registration result, extracting a matching point group corresponding to the laser point in the target laser point cloud from the geometric feature point map, wherein the matching point group is a plurality of points in the geometric feature point map that are closest to the laser point;

[0017] Fitting points in the matching point group to obtain corresponding matching planes, and calculating distances from points in the matching point group to the matching planes;

[0018] Determine a target matching point group among all matching point groups according to the distances from the points in each matching point group to the corresponding matching planes, and the matching plane corresponding to the target matching point group is the target matching plane;

[0019] The laser points in the target laser point cloud corresponding to the target matching point group are determined as target geometric feature points.

[0020] Optionally, a geometric residual matrix and a first Jacobian matrix are constructed according to the target geometric feature points, the geometric feature point map and the initial pose, including:

[0021] Determine the distance function from the target geometric feature point to the corresponding target matching plane in the geometric feature point map as a geometric residual function;

[0022] Based on each geometric residual function, a geometric residual matrix is ​​constructed;

[0023] According to each geometric residual function and the initial pose, the first Jacobian matrix is ​​constructed.

[0024] Optionally, according to the initial posture, registering the target laser point cloud with the photometric feature point map to obtain a second registration result includes:

[0025] The target laser point cloud is converted into a two-dimensional laser point map by a coordinate system conversion function;

[0026] According to the initial posture, the two-dimensional laser point image is aligned with the photometric feature point map to obtain a second alignment result.

[0027] Optionally, extracting target photometric feature points with the highest reflectivity in each area of ​​the target laser point cloud includes:

[0028] Taking the reflectivity of each point in the two-dimensional laser point image corresponding to the target laser point cloud as a grayscale value, converting the two-dimensional laser point image into a two-dimensional reflectivity grayscale image;

[0029] According to the grayscale gradient of each point in the reflectivity grayscale two-dimensional map, the highest grayscale value point of each local area is extracted, and the point in the laser point two-dimensional map corresponding to the highest grayscale value point is used as the target photometric feature point.

[0030] Optionally, constructing a photometric residual matrix and a second Jacobian matrix according to the second registration result, the target photometric feature points, the photometric feature point map and the initial pose, including:

[0031] Taking the target photometric feature point as the center, a target window of preset size is constructed in the two-dimensional image of the laser point;

[0032] According to the second registration result, a matching window corresponding to the target window is constructed in the photometric feature point map, wherein the matching window is a window in the photometric feature point map having the same size and position as the target window;

[0033] Determine a reflectivity difference function between a point in the target window and a point in the corresponding matching window as a photometric residual function;

[0034] According to each photometric residual function, a photometric residual matrix is ​​constructed;

[0035] The second Jacobian matrix is ​​constructed according to various photometric residual functions, coordinate system conversion functions and initial poses.

[0036] Optionally, based on the target pose and the target laser point cloud, updating the geometric feature point map and the photometric feature point map includes:

[0037] Based on the target posture, the target laser point cloud is registered with the geometric feature point map, and new target geometric feature points in the target laser point cloud are determined according to the registration result;

[0038] Adding the new target geometric feature point to the geometric feature point map, and adjusting the position of each geometric feature point in the geometric feature point map according to the target posture;

[0039] Based on the target posture, the target laser point cloud is registered with the photometric feature point map, the target photometric feature points are added to the photometric feature point map according to the registration result, and the position of each photometric feature point in the photometric feature point map is adjusted according to the target posture.

[0040] A second aspect of an embodiment of the present application provides a laser odometer system based on reflectivity characteristics, the system comprising:

[0041] The first computing module is used to obtain the initial position and posture of the vehicle through integration operation according to the collected vehicle inertia information;

[0042] A first registration module is used to register the collected target laser point cloud with the geometric feature point map according to the initial posture to obtain a first registration result;

[0043] A first feature point determination module, used to determine a target geometric feature point in a target laser point cloud according to the first registration result, wherein the target geometric feature point is a laser point with a planar feature in the target laser point cloud;

[0044] A first construction module is used to construct a geometric residual matrix and a first Jacobian matrix according to target geometric feature points, a geometric feature point map and an initial pose;

[0045] A second registration module is used to register the target laser point cloud with the photometric feature point map according to the initial posture to obtain a second registration result;

[0046] A second feature point determination module is used to extract the target photometric feature points with the highest reflectivity in each area of ​​the target laser point cloud;

[0047] A second construction module is used to construct a photometric residual matrix and a second Jacobian matrix according to the second registration result, the target photometric feature points, the photometric feature point map and the initial pose;

[0048] A third construction module is used to construct a target residual matrix based on the geometric residual matrix and the photometric residual matrix, and to construct a target Jacobian matrix based on the first Jacobian matrix and the second Jacobian matrix;

[0049] A first optimization module, configured to iteratively optimize the initial pose based on the target residual matrix and the target Jacobian matrix to obtain a target pose;

[0050] The first updating module is used to update the geometric feature point map and the photometric feature point map based on the target posture and the target laser point cloud.

[0051] A third aspect of an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, a laser odometer method based on reflectivity features as described in the first aspect of the present application is implemented.

[0052] A fourth aspect of an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, a laser odometer method based on reflectivity features as described in the first aspect of the present application is implemented.

[0053] Compared with the prior art, this application has the following advantages:

[0054] The embodiments of the present application provide a laser odometer method, system, device and medium based on reflectivity features, which establish geometric constraints for initial pose optimization through geometric features, and establish photometric constraints for initial pose optimization through photometric features based on reflectivity. The geometric constraints and photometric constraints are integrated and used together for initial pose optimization, thereby avoiding the problem of decreased pose estimation accuracy due to insufficient geometric features in tunnel environments and ensuring the positioning accuracy of the laser odometer.

[0055] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments of the present application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor.

[0057] Figure 1 A flow chart of a laser odometer method based on reflectivity features is shown as an embodiment of the present application;

[0058] Figure 2 A structural block diagram of a laser odometer system based on reflectivity characteristics is shown as an embodiment of the present application. DETAILED DESCRIPTION

[0059] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0060] refer to Figure 1 , Figure 1The following is a flow chart of a laser odometer method based on reflectivity characteristics according to an embodiment of the present application. Figure 1 As shown, a laser odometer method based on reflectivity characteristics provided in an embodiment of the present application includes steps S1 to S10:

[0061] Step S1: Based on the collected vehicle inertia information, the initial position and posture of the vehicle is obtained through integration operation.

[0062] In this embodiment, the vehicle is equipped with an on-board inertial measurement unit (IMU), including a three-axis gyroscope and a three-axis accelerometer. The on-board inertial measurement unit collects the acceleration and angular velocity of the vehicle in three directions, and obtains a total of six component data, namely the vehicle inertial information. By integrating the three angular velocity components, the attitude change of the vehicle is estimated, and by integrating the three acceleration components, the position change of the vehicle is estimated. According to the change of the attitude and position of the vehicle, the initial posture of the vehicle is finally determined. Due to the presence of factors such as hardware error, random error and integral cumulative error, there is a certain difference between the initial posture obtained by the on-board inertial measurement unit and the actual vehicle posture.

[0063] Step S2: According to the initial posture, the collected target laser point cloud is registered with the geometric feature point map to obtain a first registration result.

[0064] In this embodiment, the vehicle is equipped with a vehicle-mounted laser radar (LiDAR), which emits a laser beam and receives the reflected signal to form a laser point cloud, and obtains the three-dimensional information and reflectivity information of the surrounding environment through the laser point cloud. Since the laser point cloud is obtained in the form of frames, when the vehicle-mounted laser radar is working, the vehicle itself is also constantly moving, resulting in distortion of each frame of the collected laser point cloud, affecting the reliability of the data. Therefore, the laser point cloud of each frame is subjected to distortion compensation through the initial posture to eliminate the laser point cloud distortion caused by the vehicle movement. The laser point cloud after distortion compensation is the target laser point cloud.

[0065] The geometric feature point map contains geometric feature points (such as edge lines, plane points, etc.) that can reflect the structure of the vehicle's external environment. These geometric feature points are obtained through feature point extraction from each frame of laser point cloud collected by the on-board lidar, and are continuously updated to the geometric feature point map through an incremental geometric feature point update process.

[0066] Specifically, the three-dimensional coordinates of each point in the target laser point cloud are adjusted according to the initial posture, so that the position state of each point in the target laser point cloud returns to the position state reflected by the current geometric feature point map (that is, the moment when the geometric feature points in the previous frame of laser point cloud are updated to the geometric feature point map), and the target laser point cloud is spatially overlapped with the geometric feature point map to obtain the first alignment result.

[0067] Step S3: determining target geometric feature points in the target laser point cloud according to the first registration result, wherein the target geometric feature points are laser points with planar features in the target laser point cloud.

[0068] In this embodiment, the target laser point cloud and the geometric feature point map are spatially overlapped. After obtaining the first alignment result, each geometric feature point in the geometric feature point map is matched with the laser point in the target laser point cloud that is closest to it. The matched laser points are the target geometric feature points. The target geometric feature points reflect the three-dimensional positions of the geometric feature points in the geometric feature point map at the moment when the target laser point cloud is collected.

[0069] Step S4: construct a geometric residual matrix and a first Jacobian matrix according to the target geometric feature points, the geometric feature point map and the initial pose.

[0070] In this embodiment, the target geometric feature points are obtained by aligning the target laser point cloud with the geometric feature point map according to the initial pose. Therefore, there is a distance difference caused by the initial pose error between the target geometric feature points and the geometric feature points in the geometric feature point map. The distance difference between each target geometric feature point and the corresponding geometric feature point in the geometric feature point map is used as the geometric residual of each target geometric feature point, and a geometric residual matrix containing all geometric residuals is constructed. According to the geometric residual calculation function of each target geometric feature point, the partial derivative of each geometric residual calculation function with respect to each component in the initial pose is calculated to construct the first Jacobian matrix, which reflects how the geometric residual of each target geometric feature point changes when each component in the initial pose changes.

[0071] Step S5: According to the initial posture, the target laser point cloud is registered with the photometric feature point map to obtain a second registration result.

[0072] The photometric feature point map contains photometric feature points that can reflect the locations of objects with high reflectivity in the vehicle's external environment (such as road signs, reflectors, etc.). These photometric feature points are obtained through feature point extraction from each frame of laser point cloud collected by the on-board lidar, and are continuously updated to the photometric feature point map through an incremental photometric feature point update process.

[0073] Specifically, the three-dimensional coordinates of each point in the target laser point cloud are adjusted according to the initial posture, so that the position state of each point in the target laser point cloud returns to the position state reflected by the current photometric feature point map (that is, the moment when the photometric feature points in the previous frame of laser point cloud are updated to the photometric feature point map), and the target laser point cloud is spatially overlapped with the photometric feature point map to obtain a second registration result.

[0074] Step S6: extracting the target photometric feature points with the highest reflectivity in each area of ​​the target laser point cloud.

[0075] In this embodiment, the target laser point cloud contains the reflectivity information of each laser point. By comparing the reflectivity of each laser point in a local area, the laser point with the highest reflectivity in the area is extracted as the target photometric feature point of the area. The target laser point cloud is divided into several local areas, and the laser point with the highest reflectivity in each local area is extracted as the target photometric feature point of the target laser point cloud.

[0076] Step S7: constructing a photometric residual matrix and a second Jacobian matrix according to the second registration result, the target photometric feature points, the photometric feature point map and the initial pose.

[0077] In this embodiment, according to the second registration result of the target laser point cloud and the photometric feature point map, the target photometric feature point is matched with the photometric feature point in the photometric feature point map closest to the target photometric feature point, and the reflectivity difference between each target photometric feature point and the photometric feature point in the matched photometric feature point map is used as the photometric residual of each target photometric feature point, and a photometric residual matrix containing all photometric residuals is constructed. According to the photometric residual calculation function of each target photometric feature point, the partial derivative of each photometric residual calculation function with respect to each component in the initial pose is calculated, and a second Jacobian matrix is ​​constructed, which reflects how the photometric residual of each target photometric feature point changes when each component in the initial pose changes.

[0078] Step S8: construct a target residual matrix based on the geometric residual matrix and the photometric residual matrix, and construct a target Jacobian matrix based on the first Jacobian matrix and the second Jacobian matrix.

[0079] In this embodiment, the geometric constraint and the photometric constraint are used together to determine the accurate posture of the vehicle. Therefore, the geometric residual matrix and the photometric residual matrix after the normalized parameter constraint are vertically merged to obtain the merged target residual matrix, which is specifically calculated in the following way:

[0080]

[0081] Where e is the target residual matrix, e geois the geometric residual matrix, e photo is the photometric residual matrix, and α is the normalization parameter.

[0082] The first Jacobian matrix and the second Jacobian matrix after normalization parameter constraints are vertically merged to obtain the merged target Jacobian matrix, which is calculated in the following way:

[0083]

[0084] Where J is the target residual matrix, J geo is the geometric residual matrix, J photo is the photometric residual matrix.

[0085] Step S9: Based on the target residual matrix and the target Jacobian matrix, the initial pose is iteratively optimized to obtain the target pose.

[0086] In this embodiment, the initial pose is iteratively optimized by iterative error state Kalman filtering to obtain the target pose. The iterative error state Kalman filtering continuously adjusts the components of the initial pose through iterative optimization technology, and updates the estimated value of the target residual matrix according to the target Jacobian matrix until the residual estimate is minimized. At this time, the pose obtained is the optimized target pose.

[0087] Step S10: Based on the target pose and the target laser point cloud, update the geometric feature point map and the photometric feature point map.

[0088] In this embodiment, the target posture is obtained after iteratively optimizing the initial posture through iterative error state Kalman filtering, which can accurately reflect the actual posture of the vehicle when the target laser point cloud is collected.

[0089] According to the target posture, the target laser point cloud is realigned with the geometric feature point map, and new target geometric feature points are re-extracted, the new target geometric feature points are added to the geometric feature point map, and the three-dimensional coordinates of the geometric feature point map are adjusted according to the target posture, and the position state of each geometric feature point in the geometric feature point map is updated to the position state corresponding to the moment when the target laser point cloud is collected.

[0090] According to the target posture, the target laser point cloud is realigned with the photometric feature point map, and new target photometric feature points are re-extracted, the new target photometric feature points are added to the photometric feature point map, and the three-dimensional coordinates of the photometric feature point map are adjusted according to the target posture, and the position state of each photometric feature point in the photometric feature point map is updated to the position state corresponding to the moment when the target laser point cloud is collected.

[0091] An embodiment of the present application provides a laser odometry method based on reflectivity features, which establishes geometric constraints for initial pose optimization through geometric features, and establishes photometric constraints for initial pose optimization through photometric features based on reflectivity. The geometric constraints and photometric constraints are integrated and used together for initial pose optimization, thereby ensuring the accuracy of pose estimation and the positioning accuracy of the laser odometry.

[0092] In combination with the above embodiments, in one implementation, the embodiment of the present application further provides a laser odometer method based on reflectivity features. In the laser odometer method based on reflectivity features, in step S3, according to the first registration result, the target geometric feature points in the target laser point cloud are determined, including steps S31 to S34:

[0093] Step S31: extracting a matching point group corresponding to the laser point in the target laser point cloud from the geometric feature point map according to the first registration result, wherein the matching point group is a plurality of points in the geometric feature point map that are closest to the laser point.

[0094] In this embodiment, the target laser point cloud and the geometric feature point map are spatially overlapped. After obtaining the first registration result, for each laser point in the target laser point cloud, several (e.g., 10) geometric feature points in the geometric feature point map that are closest to the laser point are used as matching points corresponding to the laser point to form a matching point group corresponding to the laser point.

[0095] Step S32: fitting the points in the matching point group to obtain a corresponding matching plane, and calculating the distance between the points in the matching point group and the matching plane.

[0096] In this embodiment, after extracting the matching point group corresponding to the laser point from the geometric feature point map, a matching plane is fitted according to the coordinates of each point in the matching point group so that the sum of the squares of the distance differences from each point in the matching point group to the matching plane is minimized, and the sum of the squares of the distance differences is calculated. The method of fitting the matching plane can be used, for example, to construct a plane equation group according to the coordinates of each point, and then use the least squares method to approximate the solution, etc., which is not limited here.

[0097] Step S33: determining a target matching point group in all matching point groups according to the distances from points in each matching point group to their corresponding matching planes, and the matching plane corresponding to the target matching point group is the target matching plane.

[0098] In this embodiment, after calculating the sum of the squares of the distance differences between the points in each matching point group and the corresponding matching plane, the matching point group whose sum of the squares of the distance differences is less than a preset threshold is taken as the target matching point group. At this time, all the points in the target matching point group can be regarded as belonging to the same plane, and the matching plane corresponding to the target matching point group is taken as the target matching plane containing all the points in the target matching point group.

[0099] Step S34: Determine the laser points in the target laser point cloud corresponding to the target matching point group as target geometric feature points.

[0100] In this embodiment, after the target matching point group is determined, the laser points in the target laser point cloud corresponding to the target matching point group are the target geometric feature points, and the target geometric feature points reflect the three-dimensional positions of each geometric feature point in the geometric feature point map at the moment when the target laser point cloud is collected.

[0101] A laser odometer method based on reflectivity features is provided in an embodiment of the present application. The method uses several geometric feature points in a geometric feature point map that are closest to a laser point in a laser point cloud as matching points of the laser point, and judges whether several geometric feature points are located in the same plane based on the distance difference between a matching plane fitted by the several geometric feature points and the several geometric feature points. Furthermore, the method judges whether the laser points in the laser point cloud corresponding to the several geometric feature points are target geometric feature points, thereby improving the accuracy of extracting target geometric feature points from the target laser point cloud.

[0102] In combination with the above embodiments, in one implementation, the present application embodiment further provides a laser odometer method based on reflectivity characteristics. In the laser odometer method based on reflectivity characteristics, step S4 includes steps S41 to S43:

[0103] Step S41: Determine the distance function from the target geometric feature point to the corresponding target matching plane in the geometric feature point map as a geometric residual function.

[0104] In this embodiment, the method for determining the target geometric feature point and the target matching plane is as described in steps S31 to S34. After determining the target geometric feature point and the corresponding target matching plane, the distance function from the target geometric feature point to the corresponding target matching plane is used as the geometric residual function of the target geometric feature point, which is specifically calculated in the following manner:

[0105] cost geo =n T *p+d

[0106] Where n T is the transposed matrix of the normal vector of the target matching plane, p is the world coordinate of the target geometric feature point, and d is the intercept of the target matching plane equation.

[0107] Step S42: construct a geometric residual matrix based on each geometric residual function.

[0108] In this embodiment, after obtaining the geometric residual function of each target geometric feature point, the three-dimensional coordinates of the corresponding target geometric feature point are substituted into the function, the geometric residual of each target geometric feature point is calculated, and a multi-row and single-column geometric residual matrix is ​​constructed.

[0109] Step S43: construct a first Jacobian matrix according to each geometric residual function and the initial pose.

[0110] In this embodiment, the chain derivative rule is used to construct the first Jacobian matrix according to each geometric residual function and the initial pose, which is specifically calculated in the following way:

[0111]

[0112] In the formula, is the partial derivative of the geometric residual function with respect to the coordinates of the target geometric feature points, It is the partial derivative of the target geometric feature point coordinates with respect to each component in the initial pose.

[0113] A laser odometer method based on reflectivity features is provided in an embodiment of the present application. The distance function from the target geometric feature point to the corresponding target matching plane is used as the geometric residual function of the target geometric feature point. According to the geometric residual function and the initial pose, a geometric residual matrix and a first Jacobian matrix are constructed, thereby improving the accuracy of constructing the geometric constraints of the laser odometer.

[0114] In combination with the above embodiments, in one implementation, the embodiment of the present application further provides a laser odometer method based on reflectivity characteristics. In the laser odometer method based on reflectivity characteristics, step S5 includes steps S51 to S52:

[0115] Step S51: converting the target laser point cloud into a two-dimensional laser point map through a coordinate system conversion function.

[0116] In this embodiment, the three-dimensional coordinates of each laser point in the target laser point cloud are converted into two-dimensional coordinates through a coordinate system conversion function to obtain a two-dimensional map of the laser points, which is specifically calculated in the following way:

[0117]

[0118] Where u and v represent the coordinates of the laser point in the two-dimensional image, x, y, and z represent the three-dimensional coordinates of the laser point, w is the width of the two-dimensional image of the laser point, and h is the height of the two-dimensional image of the laser point. is the depth value of the laser point, and θ is the vertical field of view angle of the laser radar.

[0119] Step S52: According to the initial posture, the two-dimensional laser point image is registered with the photometric feature point map to obtain a second registration result.

[0120] In this embodiment, the photometric feature point map is a two-dimensional map. The two-dimensional coordinates of each point in the laser point two-dimensional map are adjusted according to the initial posture, so that the position state of each point in the laser point two-dimensional map returns to the position state reflected by the current photometric feature point map (that is, the moment when the photometric feature points in the previous frame of laser point cloud are updated to the photometric feature point map), and the laser point two-dimensional map is planarly overlapped with the photometric feature point map to obtain a second alignment result.

[0121] A laser odometer method based on reflectivity features is provided in an embodiment of the present application. By converting a target laser point cloud into a two-dimensional laser point map, and aligning the two-dimensional laser point map with a photometric feature point map according to an initial pose, calculation conditions are provided for the laser odometer to construct a photometric constraint based on reflectivity.

[0122] In combination with the above embodiments, in one implementation, the embodiment of the present application further provides a laser odometer method based on reflectivity characteristics. In the laser odometer method based on reflectivity characteristics, step S6 includes steps S61 to S62:

[0123] Step S61: taking the reflectivity of each point in the two-dimensional laser point image corresponding to the target laser point cloud as a grayscale value, and converting the two-dimensional laser point image into a two-dimensional reflectivity grayscale image.

[0124] In this embodiment, the method for generating a two-dimensional laser point map is as described in step S51, and the grayscale value of each laser point is determined according to the reflectivity of each laser point in the two-dimensional laser point map to generate a grayscale map. The position of each point on the grayscale map corresponds one-to-one to the position of each laser point in the two-dimensional laser point map, and the grayscale value of each point on the grayscale map is consistent with the reflectivity of the corresponding laser point in the two-dimensional laser point map. The grayscale map is a reflectivity grayscale two-dimensional map.

[0125] Step S62: extract the local highest grayscale value points according to the grayscale gradient of each point in the reflectivity grayscale two-dimensional map, and use the points in the laser point two-dimensional map corresponding to the highest grayscale value points as target photometric feature points.

[0126] In this embodiment, by calculating the grayscale gradient of each point in the reflectivity grayscale two-dimensional map in both the horizontal and vertical directions, the points whose grayscale gradients in both directions are greater than the preset threshold are screened out. The point screened out by this method is the point with the largest grayscale in the local area of ​​the reflectivity grayscale two-dimensional map where the point is located, and the laser point in the laser point two-dimensional map corresponding to the point is the target photometric feature point with the largest reflectivity in the local area of ​​the laser point two-dimensional map where the point is located. For example, the target laser point cloud contains the laser point returned by the rectangular plane of the rectangular sign with high reflectivity. Then, after converting the target laser point cloud into a laser point two-dimensional map and then into a reflectivity grayscale two-dimensional map, by screening the points whose grayscale gradients in both directions in the reflectivity grayscale two-dimensional map are greater than the preset threshold, the corresponding target photometric feature points with the largest reflectivity in the laser point two-dimensional map are the four vertices of the rectangular plane of the rectangular sign.

[0127] A laser odometer method based on reflectivity features is provided in an embodiment of the present application. A reflectivity grayscale two-dimensional map is generated according to the grayscale values ​​of each laser point in the laser point two-dimensional map, and corresponding target photometric feature points are obtained from the laser point two-dimensional map by extracting points whose grayscale gradient is greater than a preset threshold in the reflectivity grayscale two-dimensional map, thereby providing calculation conditions for the laser odometer to construct photometric constraints based on reflectivity.

[0128] In combination with the above embodiments, in one implementation, the embodiment of the present application further provides a laser odometer method based on reflectivity characteristics. In the laser odometer method based on reflectivity characteristics, step S7 includes steps S71 to S75:

[0129] Step S71: construct a target window of a preset size in the two-dimensional laser point map with the target photometric feature point as the center.

[0130] In this embodiment, the method for obtaining the target photometric feature point is as described in step S61 to step S62. With each obtained target photometric feature point as the center, each square target window with a preset side length is constructed in the laser point two-dimensional map, and the target window includes a plurality of laser points in the laser point two-dimensional map in addition to the target photometric feature point.

[0131] Step S72: Based on the second registration result, a matching window corresponding to the target window is constructed in the photometric feature point map, where the matching window is a window in the photometric feature point map that has the same size and position as the target window.

[0132] In this embodiment, the method for obtaining the second registration result is as described in step S51 to step S52. According to the initial posture, after the two-dimensional laser point image and the photometric feature point map are plane-overlapped, each matching window of the same size and position is constructed in the photometric feature point map according to each square target window constructed in the two-dimensional laser point image.

[0133] Step S73: Determine that the reflectivity difference function between the point in the target window and the point in the corresponding matching window is a photometric residual function.

[0134] In this embodiment, after constructing various matching windows of the same size and position in the photometric feature point map according to the various square target windows constructed in the two-dimensional laser point map, the average reflectivity of all laser points contained in the target window is used as the reflectivity value of the target window, and the average reflectivity of all photometric feature points contained in the matching window is used as the reflectivity value of the matching window. The function of calculating the reflectivity difference between the target window and the corresponding matching window is used as the photometric residual function of the target photometric feature points contained in the target window.

[0135] Step S74: construct a photometric residual matrix according to each photometric residual function.

[0136] In this embodiment, after obtaining the photometric residual function of each target photometric feature point, the reflectivity of the laser point contained in the target window where the corresponding target photometric feature point is located and the reflectivity of the photometric feature point contained in the corresponding matching window are substituted into the function to calculate the photometric residual of each target photometric feature point, and construct a multi-row and single-column photometric residual matrix.

[0137] Step S75: construct a second Jacobian matrix according to various photometric residual functions, coordinate system conversion functions and initial poses.

[0138] In this embodiment, the chain derivation rule is used to construct the second Jacobian matrix according to each photometric residual function, coordinate system conversion function and initial pose, which is specifically calculated in the following way:

[0139] in

[0140] In the formula, is the partial derivative of the photometric residual function with respect to the target window position, is the partial derivative of the target window position with respect to the target photometric feature point coordinates, is the partial derivative of the target photometric feature point coordinates with respect to each component in the initial pose, f x and f y is the normalized focal length when generating the two-dimensional image of the laser point.

[0141] A laser odometer method based on reflectivity features is provided in an embodiment of the present application. A target window is constructed with a target photometric feature point as the center, and a corresponding matching window is constructed in a photometric feature point map. The reflectivity difference function between the point in the target window and the point in the corresponding matching window is used as a photometric residual function. A second Jacobian matrix is ​​constructed according to each photometric residual function, a coordinate system conversion function and an initial pose, thereby improving the accuracy of constructing the photometric constraints of the laser odometer.

[0142] In combination with the above embodiments, in one implementation, the present application embodiment further provides a laser odometer method based on reflectivity characteristics. In the laser odometer method based on reflectivity characteristics, step S10 includes steps S101 to S103:

[0143] Step S101: Based on the target posture, the target laser point cloud is registered with the geometric feature point map, and new target geometric feature points in the target laser point cloud are determined according to the registration result.

[0144] In this embodiment, the initial pose is iteratively optimized by iterative error state Kalman filtering, and after the target pose is obtained, the target laser point cloud is re-aligned with the geometric feature point map according to the target pose, and new target geometric feature points are re-extracted according to the alignment result. The method for re-extracting new target geometric feature points in the target laser point cloud according to the alignment result is as described in steps S31 to S34, which will not be repeated here.

[0145] Step S102: adding the new target geometric feature point to the geometric feature point map, and adjusting the position of each geometric feature point in the geometric feature point map according to the target posture.

[0146] In this embodiment, after determining the new target geometric feature points, according to the alignment results, the new target geometric feature points located at each coordinate position in the target laser point cloud are added to the same coordinate positions in the geometric feature point map, and the three-dimensional coordinates of the geometric feature point map are adjusted according to the target posture, and the position state of each geometric feature point in the geometric feature point map is updated to the position state corresponding to the moment when the target laser point cloud is collected.

[0147] Step S103: Based on the target posture, the target laser point cloud is registered with the photometric feature point map, the target photometric feature points are added to the photometric feature point map according to the registration result, and the position of each photometric feature point in the photometric feature point map is adjusted according to the target posture.

[0148] In this embodiment, the initial posture is iteratively optimized by iterative error state Kalman filtering. After the target posture is obtained, the target laser point cloud and the photometric feature point map are re-aligned according to the target posture. The method is as described in steps S51 to S52, which will not be repeated here. According to the alignment result, the target photometric feature points at each coordinate position in the two-dimensional laser point map are added to the same coordinate positions in the photometric feature point map, and the two-dimensional coordinates of the geometric feature point map are adjusted according to the target posture, and the position state of each photometric feature point in the photometric feature point map is updated to the position state corresponding to the moment when the target laser point cloud is collected.

[0149] A laser odometer method based on reflectivity features is provided in an embodiment of the present application. Geometric constraints and photometric constraints are integrated and used together for initial pose optimization. Geometric feature points and photometric feature points in the target laser point cloud are added to their corresponding feature point maps according to the optimized target pose. The positions of feature points in each feature point map are updated in time according to the target pose, thereby ensuring the positioning accuracy of the laser odometer.

[0150] Based on the same inventive concept, the second aspect of the embodiment of the present application provides a laser odometer system based on reflectivity characteristics, such as Figure 2 As shown, the system includes: a first operation module, a first registration module, a first feature point determination module, a first construction module, a second registration module, a second feature point determination module, a second construction module, a third construction module, a first optimization module, and a first update module;

[0151] The first computing module is used to obtain the initial position and posture of the vehicle through integration operation according to the collected vehicle inertia information;

[0152] The first registration module is used to register the collected target laser point cloud with the geometric feature point map according to the initial posture to obtain a first registration result;

[0153] The first feature point determination module is used to determine a target geometric feature point in the target laser point cloud according to the first registration result, wherein the target geometric feature point is a laser point with a planar feature in the target laser point cloud;

[0154] The first construction module is used to construct a geometric residual matrix and a first Jacobian matrix according to the target geometric feature points, the geometric feature point map and the initial pose;

[0155] The second registration module is used to register the target laser point cloud with the photometric feature point map according to the initial posture to obtain a second registration result;

[0156] The second feature point determination module is used to extract the target photometric feature points with the highest reflectivity in each area of ​​the target laser point cloud;

[0157] The second construction module is used to construct a photometric residual matrix and a second Jacobian matrix according to the second registration result, the target photometric feature points, the photometric feature point map and the initial pose;

[0158] The third construction module is used to construct a target residual matrix based on the geometric residual matrix and the photometric residual matrix, and to construct a target Jacobian matrix based on the first Jacobian matrix and the second Jacobian matrix;

[0159] The first optimization module is used to iteratively optimize the initial pose based on the target residual matrix and the target Jacobian matrix to obtain a target pose;

[0160] The first updating module is used to update the geometric feature point map and the photometric feature point map based on the target posture and the target laser point cloud.

[0161] In combination with the above embodiments, in one implementation, the embodiment of the present application further provides a laser odometer system based on reflectivity features. In the laser odometer system based on reflectivity features, it also includes: a first extraction module, a first fitting module, a first determination module, a second determination module, a third determination module, a fourth construction module, a fifth construction module, a first conversion module, a third registration module, a second conversion module, a second extraction module, a sixth construction module, a seventh construction module, a fourth determination module, an eighth construction module, a ninth construction module, a fourth registration module, a first addition module, a fifth registration module, and a second addition module;

[0162] The first extraction module is used to extract a matching point group corresponding to the laser point in the target laser point cloud from the geometric feature point map according to the first registration result, wherein the matching point group is a plurality of points in the geometric feature point map that are closest to the laser point;

[0163] The first fitting module is used to fit the points in the matching point group to obtain the corresponding matching plane, and calculate the distance from the points in the matching point group to the matching plane;

[0164] The first determination module is used to determine a target matching point group among all matching point groups according to the distances from points in each matching point group to the matching planes corresponding to each of them, and the matching plane corresponding to the target matching point group is the target matching plane;

[0165] The second determination module is used to determine the laser points in the target laser point cloud corresponding to the target matching point group as target geometric feature points;

[0166] The third determination module is used to determine that the distance function from the target geometric feature point to the corresponding target matching plane in the geometric feature point map is a geometric residual function;

[0167] The fourth construction module is used to construct a geometric residual matrix based on each geometric residual function;

[0168] The fifth construction module is used to construct a first Jacobian matrix according to each geometric residual function and the initial pose;

[0169] The first conversion module is used to convert the target laser point cloud into a two-dimensional laser point map through a coordinate system conversion function;

[0170] The third registration module is used to register the laser point two-dimensional image with the photometric feature point map according to the initial posture to obtain a second registration result;

[0171] The second conversion module is used to use the reflectivity of each point in the two-dimensional laser point image corresponding to the target laser point cloud as a grayscale value, and convert the two-dimensional laser point image into a two-dimensional reflectivity grayscale image;

[0172] The second extraction module is used to extract the highest grayscale value points of each local area according to the grayscale gradient of each point in the reflectance grayscale two-dimensional map, and use the points in the laser point two-dimensional map corresponding to the highest grayscale value points as the target photometric feature points;

[0173] The sixth construction module is used to construct a target window of a preset size in the two-dimensional laser point map with the target photometric feature point as the center;

[0174] The seventh construction module is used to construct a matching window corresponding to the target window in the photometric feature point map according to the second registration result, wherein the matching window is a window in the photometric feature point map having the same size and position as the target window;

[0175] The fourth determination module is used to determine that the reflectivity difference function of the point in the target window and the point in the corresponding matching window is a photometric residual function;

[0176] The eighth construction module is used to construct a photometric residual matrix according to each photometric residual function;

[0177] The ninth construction module is used to construct a second Jacobian matrix according to each photometric residual function, the coordinate system conversion function and the initial pose;

[0178] The fourth registration module is used to register the target laser point cloud with the geometric feature point map based on the target posture, and determine new target geometric feature points in the target laser point cloud according to the registration result;

[0179] The first adding module is used to add the new target geometric feature point to the geometric feature point map, and adjust the position of each geometric feature point in the geometric feature point map according to the target posture;

[0180] The second adding module is used to align the target laser point cloud with the photometric feature point map based on the target posture, add the target photometric feature points to the photometric feature point map according to the alignment result, and adjust the position of each photometric feature point in the photometric feature point map according to the target posture.

[0181] Based on the same inventive concept, the third aspect of an embodiment of the present application provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the computer program is executed by the processor, a laser odometer method based on reflectivity features as described in the first aspect of the present application is implemented.

[0182] Based on the same inventive concept, the fourth aspect of an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, a laser odometer method based on reflectivity features as described in the first aspect of the present application is implemented.

[0183] As for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0184] It should be noted that, for the method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present application are not limited by the described order of actions, because according to the embodiments of the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present application.

[0185] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0186] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the embodiments of the present application may adopt the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the embodiments of the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0187] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0188] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0189] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0190] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present application.

[0191] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0192] The above is a detailed introduction to the laser odometer method, system, device and medium based on reflectivity characteristics provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A laser odometer method based on reflectivity characteristics, characterized in that: Applied to a vehicle-mounted laser odometer, the method comprises: According to the collected vehicle inertia information, the initial position and posture of the vehicle is obtained through integration operation; According to the initial posture, registering the collected target laser point cloud with the geometric feature point map to obtain a first registration result; Determine a target geometric feature point in the target laser point cloud according to the first registration result, wherein the target geometric feature point is a laser point with a planar feature in the target laser point cloud; Construct the geometric residual matrix and the first Jacobian matrix according to the target geometric feature points, the geometric feature point map and the initial pose; According to the initial posture, registering the target laser point cloud with the photometric feature point map to obtain a second registration result; Extracting target photometric feature points with the highest reflectivity in each area of ​​the target laser point cloud; Constructing a photometric residual matrix and a second Jacobian matrix according to the second registration result, the target photometric feature points, the photometric feature point map and the initial pose; Constructing a target residual matrix based on the geometric residual matrix and the photometric residual matrix, and constructing a target Jacobian matrix based on the first Jacobian matrix and the second Jacobian matrix; Iteratively optimizing the initial pose based on the target residual matrix and the target Jacobian matrix to obtain a target pose; Based on the target pose and the target laser point cloud, a geometric feature point map and a photometric feature point map are updated.

2. The laser odometer method based on reflectivity characteristics according to claim 1, characterized in that: Determining target geometric feature points in the target laser point cloud according to the first registration result includes: According to the first registration result, extracting a matching point group corresponding to the laser point in the target laser point cloud from the geometric feature point map, wherein the matching point group is a plurality of points in the geometric feature point map that are closest to the laser point; Fitting points in the matching point group to obtain corresponding matching planes, and calculating distances from points in the matching point group to the matching planes; Determine a target matching point group among all matching point groups according to the distances from the points in each matching point group to the corresponding matching planes, and the matching plane corresponding to the target matching point group is the target matching plane; The laser points in the target laser point cloud corresponding to the target matching point group are determined as target geometric feature points.

3. The laser odometer method based on reflectivity characteristics according to claim 2, characterized in that: According to the target geometric feature points, geometric feature point map and initial pose, the geometric residual matrix and the first Jacobian matrix are constructed, including: Determine the distance function from the target geometric feature point to the corresponding target matching plane in the geometric feature point map as a geometric residual function; Based on each geometric residual function, a geometric residual matrix is ​​constructed; According to each geometric residual function and the initial pose, the first Jacobian matrix is ​​constructed.

4. The laser odometer method based on reflectivity characteristics according to claim 1, characterized in that: According to the initial pose, the target laser point cloud is registered with the photometric feature point map to obtain a second registration result, including: The target laser point cloud is converted into a two-dimensional laser point map by a coordinate system conversion function; According to the initial posture, the two-dimensional laser point image is aligned with the photometric feature point map to obtain a second alignment result.

5. The laser odometer method based on reflectivity characteristics according to claim 4, characterized in that: Extracting target photometric feature points with the highest reflectivity in each area of ​​the target laser point cloud, including: Taking the reflectivity of each point in the two-dimensional laser point image corresponding to the target laser point cloud as a grayscale value, converting the two-dimensional laser point image into a two-dimensional reflectivity grayscale image; According to the grayscale gradient of each point in the reflectivity grayscale two-dimensional map, the highest grayscale value point of each local area is extracted, and the point in the laser point two-dimensional map corresponding to the highest grayscale value point is used as the target photometric feature point.

6. The laser odometer method based on reflectivity characteristics according to claim 5, characterized in that: According to the second registration result, the target photometric feature points, the photometric feature point map and the initial pose, a photometric residual matrix and a second Jacobian matrix are constructed, including: Taking the target photometric feature point as the center, a target window of preset size is constructed in the two-dimensional image of the laser point; According to the second registration result, a matching window corresponding to the target window is constructed in the photometric feature point map, wherein the matching window is a window in the photometric feature point map having the same size and position as the target window; Determine a reflectivity difference function between a point in the target window and a point in the corresponding matching window as a photometric residual function; According to each photometric residual function, a photometric residual matrix is ​​constructed; The second Jacobian matrix is ​​constructed according to various photometric residual functions, coordinate system conversion functions and initial poses.

7. The laser odometer method based on reflectivity characteristics according to claim 1, characterized in that: Based on the target pose and the target laser point cloud, a geometric feature point map and a photometric feature point map are updated, including: Based on the target posture, the target laser point cloud is registered with the geometric feature point map, and new target geometric feature points in the target laser point cloud are determined according to the registration result; Adding the new target geometric feature point to the geometric feature point map, and adjusting the position of each geometric feature point in the geometric feature point map according to the target posture; Based on the target posture, the target laser point cloud is registered with the photometric feature point map, the target photometric feature points are added to the photometric feature point map according to the registration result, and the position of each photometric feature point in the photometric feature point map is adjusted according to the target posture.

8. A laser odometer system based on reflectivity characteristics, characterized in that: The system comprises: The first computing module is used to obtain the initial position and posture of the vehicle through integration operation according to the collected vehicle inertia information; A first registration module is used to register the collected target laser point cloud with the geometric feature point map according to the initial posture to obtain a first registration result; A first feature point determination module, used to determine a target geometric feature point in a target laser point cloud according to the first registration result, wherein the target geometric feature point is a laser point with a planar feature in the target laser point cloud; A first construction module is used to construct a geometric residual matrix and a first Jacobian matrix according to target geometric feature points, a geometric feature point map and an initial pose; A second registration module is used to register the target laser point cloud with the photometric feature point map according to the initial posture to obtain a second registration result; A second feature point determination module is used to extract the target photometric feature points with the highest reflectivity in each area of ​​the target laser point cloud; A second construction module is used to construct a photometric residual matrix and a second Jacobian matrix according to the second registration result, the target photometric feature points, the photometric feature point map and the initial pose; A third construction module is used to construct a target residual matrix based on the geometric residual matrix and the photometric residual matrix, and to construct a target Jacobian matrix based on the first Jacobian matrix and the second Jacobian matrix; A first optimization module, configured to iteratively optimize the initial pose based on the target residual matrix and the target Jacobian matrix to obtain a target pose; The first updating module is used to update the geometric feature point map and the photometric feature point map based on the target posture and the target laser point cloud.

9. An electronic device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, a laser odometer method based on reflectivity features as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the laser odometer method based on reflectivity features as described in any one of claims 1 to 7 is implemented.

Images