A multi-sensor fusion odometer generation method and system based on surface element fitting

Abstract

The present invention relates to the field of autonomous driving technology, and discloses a multi-sensor fusion odometer generation method and system based on facet fitting, the method comprising: obtaining laser radar sensor data, visual sensor data and IMU sensor data; pre-integrating the IMU data to obtain the relative transformation relationship within the posture update cycle; performing coordinate transformation on the laser radar sensor data and the visual sensor data to obtain the laser point cloud disparity map and the visual disparity map of adjacent frames, and eliminating dynamic obstacles; extracting facets from the laser point cloud data and extracting key points from the visual image data, and constructing the constraint relationship between frames; according to the relative transformation relationship within the posture update cycle, the facet information and the constraint relationship between frames are jointly optimized to obtain the posture output information; performing smoothing based on the RTS smoother to obtain a stable posture result. The present invention improves the efficiency and accuracy of the odometer by eliminating dynamic obstacles and facet fitting.

Description

Technical Field

[0001] The present application belongs to the field of autonomous driving technology and relates to a multi-sensor fusion odometer generation method and system based on face element fitting. Background Art

[0002] With the popularity of cars and the increase in the frequency of use, road congestion, environmental pollution and traffic accidents have become problems that people have to face. Developing unmanned vehicles and researching the autonomous driving system of vehicles is currently the best choice for achieving safe and efficient transportation. In the development of unmanned vehicles, the ability to accurately control the distance and speed of the vehicle has always been one of the key factors affecting the intelligent control ability of the vehicle, and obtaining accurate real-time vehicle distance and speed parameters is an important basis for achieving accurate control of the distance and speed of the vehicle.

[0003] In the field of unmanned driving, vehicles will collect data from the three-dimensional world by acquiring external information, and accurately obtain the current vehicle's posture information by fusing multiple sensors, and obtain the vehicle's posture, speed and orientation in the unmanned driving system, helping the vehicle to drive smoothly to the destination. In the process of multi-source sensor fusion, in order to reduce processing time, improve processing efficiency and achieve real-time processing, feature extraction is generally used to pre-process data for each module, so as to achieve the purpose of real-time output of posture.

[0004] The current solution for sensor feature extraction is mainly reflected in dimensionality reduction operations, such as point-line feature and point-surface feature extraction for lidar point cloud data, and key pixel point extraction for visual sensors. This method achieves a balance between features and computational efficiency by downsampling the original large-scale observation data. At the same time, when processing point cloud operations, most of them only perform de-distortion processing without removing dynamic obstacles. Dynamic obstacles will cause greater interference in the actual odometer calculation and affect the accuracy of the odometer. Summary of the invention

[0005] In order to provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. The summary is not an extensive review, nor is it intended to identify key / critical elements or delineate the scope of protection of these embodiments, but rather serves as a prelude to the detailed description that follows.

[0006] In order to solve the problems existing in the related art, the embodiments of the present disclosure provide a multi-sensor fusion odometer generation method and system based on facet fitting, which is used to solve the problem of insufficient accuracy caused by the failure to eliminate dynamic obstacles in the evaluation odometer of the driverless car in the prior art.

[0007] In some embodiments, a method is provided for an unmanned vehicle, wherein the vehicle includes multiple sensors, wherein the multiple sensors are a laser radar system, a visual sensor, and an IMU sensor, and the method includes:

[0008] S10, acquiring laser radar sensor data, visual sensor data, and IMU sensor data;

[0009] S20, pre-integrating the IMU data to obtain the relative transformation relationship within the posture update period;

[0010] S30, performing coordinate transformation on the laser radar sensor data and the visual sensor data according to the relative transformation relationship within the posture update cycle to obtain the laser point cloud disparity map and the visual disparity map of adjacent frames, and removing dynamic obstacles by jointly processing the disparity maps;

[0011] S40, extracting facets from the laser point cloud data, wherein the facets include three-dimensional coordinate points x, y, z, normal vectors nx, ny, nz and radius r, and constructing a plane by selecting four non-collinear adjacent points for each point to obtain corresponding facet information;

[0012] S50, extracting key points from the visual image data and building constraint relationships between frames;

[0013] S60, performing joint optimization according to the relative transformation relationship within the pose update period, the face information and the constraint relationship between frames to obtain pose output information;

[0014] S70, the posture output information is smoothed based on the RTS smoother to obtain a stable posture result.

[0015] Preferably, S20, before performing pre-integration processing on the IMU data to obtain the relative transformation relationship within the posture update period, also includes:

[0016] Preprocess the lidar sensor data, visual sensor data, and IMU sensor data, including time synchronization and calibration.

[0017] Preferably, the calibration used in the calibration process is a joint calibration, which constructs the residual information of the lidar sensor, the visual sensor and the IMU sensor after preprocessing, and builds a joint optimization equation to achieve the estimation of the external parameters and the sensor time difference.

[0018] Preferably, S20, before performing pre-integration processing on the IMU data to obtain the relative transformation relationship within the posture update period, also includes:

[0019] Initialize the IMU and obtain the offset of gravity acceleration.

[0020] Preferably, extracting facets from the laser point cloud data in S40 includes:

[0021] The small facets are fitted based on the union-find method, and the obtained planes are fitted into large facets.

[0022] In some embodiments, a system is disclosed, comprising:

[0023] A multi-sensor fusion odometer generation system based on surface element fitting, comprising:

[0024] A data alignment module, configured to acquire lidar sensor data, visual sensor data, and IMU sensor data;

[0025] The IMU pre-integration module is configured to perform pre-integration processing on the IMU data to obtain the relative transformation relationship within the posture update cycle;

[0026] The feature extraction module is configured to transform the laser radar sensor data and the visual sensor data according to the relative transformation relationship within the posture update cycle to obtain the laser point cloud disparity map and the visual disparity map of the adjacent frames, remove dynamic obstacles through the joint processing of the disparity map, extract the surface element of the laser point cloud data, where the surface element includes the three-dimensional coordinate point x, y, z, the normal vector nx, ny, nz and the radius r, and construct a plane by selecting four adjacent non-collinear points for each point to obtain the corresponding surface element information, extract the key points of the visual image data, and construct the constraint relationship between frames;

[0027] The pose smoothing module is configured to perform joint optimization based on the relative transformation relationship within the pose update period, the surface element information and the constraint relationship between frames to obtain the pose output information;

[0028] The pose estimation module is configured to smooth the pose output information based on the RTS smoother to obtain a stable pose result.

[0029] Preferably, the system further comprises:

[0030] The data preprocessing module is configured to preprocess the lidar sensor data, the visual sensor data and the IMU sensor data, including time synchronization processing and calibration processing.

[0031] Preferably, the calibration used in the calibration process is a joint calibration, which constructs the residual information of the lidar sensor, the visual sensor and the IMU sensor after preprocessing, and builds a joint optimization equation to achieve the estimation of the external parameters and the sensor time difference.

[0032] Preferably, the system further includes an IMU initialization module configured to initialize the IMU and obtain a bias of gravity acceleration.

[0033] In some embodiments, an unmanned vehicle is disclosed, which is applied to the above-mentioned multi-sensor fusion odometer generation method based on surface element fitting.

[0034] The embodiment of the present disclosure provides a multi-sensor fusion odometer generation method and system based on facet fitting, which can achieve the following technical effects:

[0035] In the disclosed embodiment, dynamic obstacle removal and facet fitting are implemented by processing lidar sensors and visual sensors to improve the efficiency and accuracy of the odometer. Compared with other existing solutions, this solution has great advantages in reducing computing resource consumption and improving odometer accuracy. At the same time, it can output smooth and stable posture information, thereby improving the robustness of the odometer.

[0036] The above general description and the following description are exemplary and explanatory only and are not intended to limit the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] One or more embodiments are exemplarily described by corresponding drawings, which do not limit the embodiments. Elements with the same reference numerals in the drawings are shown as similar elements, and the drawings do not constitute a scale limitation, and wherein:

[0038] Figure 1 is a schematic diagram of an unmanned vehicle provided by an embodiment of the present disclosure;

[0039] Figure 2 It is a flow chart of a multi-sensor fusion odometer generation method based on facet fitting provided by an embodiment of the present disclosure;

[0040] Figure 3 is a schematic diagram of facet fitting provided by an embodiment of the present disclosure;

[0041] Figure 4 It is a schematic diagram of the smoothing effect of an RTS smoother on an original trajectory provided by an embodiment of the present disclosure;

[0042] Figure 5 A multi-sensor fusion odometer generation system based on face element fitting is provided in an embodiment of the present disclosure;

[0043] Figure 6 It is a data flow diagram of a multi-sensor fusion odometer generation system based on face element fitting provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0044] In order to be able to understand the features and technical contents of the embodiments of the present disclosure in more detail, the implementation of the embodiments of the present disclosure is described in detail below in conjunction with the accompanying drawings. The attached drawings are for reference only and are not used to limit the embodiments of the present disclosure. In the following technical description, for the convenience of explanation, a full understanding of the disclosed embodiments is provided through multiple details. However, one or more embodiments can still be implemented without these details. In other cases, to simplify the drawings, well-known structures and systems can be simplified for display.

[0045] The following description and accompanying drawings fully illustrate specific embodiments of the present invention so that those skilled in the art can practice them. Other embodiments may include structural, logical, electrical, process and other changes. The examples represent possible changes only. Unless explicitly required, separate components and functions are optional, and the order of operations may vary. The parts and features of some embodiments may be included in or replace the parts and features of other embodiments. The scope of the embodiments of the present invention includes the entire scope of the claims, and all available equivalents of the claims. In this article, each embodiment may be represented individually or generally by the term "invention", which is only for convenience, and if more than one invention is disclosed in fact, it is not intended to automatically limit the scope of the application to any single invention or inventive concept. 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, without requiring or implying any actual relationship or order between these entities or operations. Moreover, the terms "comprises", "includes" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method or automobile including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method or automobile. In the absence of further restrictions, an element defined by the sentence "comprises a..." does not exclude the presence of other identical elements in the process, method or automobile including the element. The various embodiments herein are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same and similar parts between the various embodiments can be referred to each other. For the methods, products, etc. disclosed in the embodiments, since they correspond to the method part disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0046] The current solution for sensor feature extraction is mainly reflected in dimensionality reduction operations, such as point-line feature and point-surface feature extraction for lidar point cloud data, and key pixel point extraction for visual sensors. This method achieves a balance between features and computational efficiency by downsampling the original large-scale observation data. At the same time, when processing point cloud operations, most of them only perform de-distortion processing without removing dynamic obstacles. Dynamic obstacles will cause greater interference in the actual odometer calculation and affect the accuracy of the odometer.

[0047] In order to solve the problems existing in the related art, the embodiments of the present disclosure provide a multi-sensor fusion odometer generation method and system based on facet fitting, which is used to solve the problem of insufficient accuracy caused by the failure to eliminate dynamic obstacles in the evaluation odometer of the driverless car in the prior art.

[0048] The present disclosure discloses a method for generating a multi-sensor fusion odometer based on surface element fitting. Figure 1 The method is applied to an unmanned vehicle, wherein the vehicle includes multiple sensors, wherein the multiple sensors are a laser radar system, a visual sensor, and an IMU sensor. Figure 2 , a multi-sensor fusion odometer generation method based on surface element fitting, comprising:

[0049] S10, acquiring lidar sensor data, visual sensor data, and IMU sensor data.

[0050] In a preferred embodiment, the method further includes: preprocessing the lidar sensor data, the visual sensor data and the IMU sensor data, including time synchronization processing and calibration processing.

[0051] It should be understood that time synchronization processing is performed on the lidar sensor, visual sensor and IMU sensor, and a smaller sensor time difference is obtained through hardware-level time synchronization. Subsequently, through calibration processing, the external parameters and time differences between the sensors are obtained for subsequent processing.

[0052] The calibration used is a joint calibration, which constructs the residual information of the lidar sensor, visual sensor and IMU sensor after preprocessing, and builds a joint optimization equation to achieve the estimation of external parameters and sensor time differences.

[0053] Preferably, the optional sensors also include any one or more of a wheel speed meter, UWB and RTK, and the sensors can be adjusted according to actual needs.

[0054] S20, pre-integrate the IMU data to obtain the relative transformation relationship within the posture update period.

[0055] A preferred embodiment, before pre-integration processing of IMU data, further includes: initializing the IMU to obtain the offset of gravity acceleration. By integrating the linear acceleration and angular velocity of the IMU data within a certain time period, the posture transformation relationship in the current time period can be calculated. In this way, a more accurate pre-integration result is obtained.

[0056] S30, coordinate transformation is performed on the laser radar sensor data and the visual sensor data according to the relative transformation relationship within the posture update cycle to obtain the laser point cloud disparity map and the visual disparity map of adjacent frames, and dynamic obstacles are eliminated by jointly processing the disparity maps.

[0057] It should be understood that dynamic obstacles are objects that move in the environment, such as walking pedestrians, moving vehicles, etc. When using lidar or visual sensors for feature extraction and pose estimation in the field of autonomous driving, they generally rely on static reference objects between frames, such as high-rise buildings in three-dimensional space. These objects will not change, and the pose transformation between two adjacent frames can be solved based on the changes in objects in different frames. However, dynamic reference objects will also move when the current carrier moves, so it is impossible to use the reference objects in adjacent frames for pose solution. If they are not removed, the solution results will be negatively optimized, resulting in failure to obtain correct results.

[0058] S40, extracting facets from the laser point cloud data, wherein the facets include three-dimensional coordinate points x, y, z, normal vectors nx, ny, nz and radius r, and constructing a plane by selecting four non-collinear adjacent points for each point to obtain the corresponding facet information.

[0059] Specifically, the acquired small facets are fitted based on a union-find method, and the obtained planes are fitted into large facets.

[0060] It should be noted that the laser point cloud data comes from the lidar sensor data.

[0061] See also Figure 3 , which is a schematic diagram of facet fitting in an embodiment of the present disclosure.

[0062] S50, extract key points from the visual image data and build constraint relationships between frames.

[0063] It should be noted that the visual image data comes from the visual sensor data.

[0064] S60, jointly optimize the pre-integrated relative transformation relationship, the face information and the key point constraint relationship between frames within the acquired posture update period to obtain posture output information.

[0065] It should be noted that the facet information is the lidar constraint relationship between adjacent frames, the key point constraint is the visual sensor constraint relationship between adjacent frames, and the pre-integrated relative transformation relationship is the IMU sensor constraint relationship between adjacent frames. By jointly optimizing the constraint relationship of these three sensors, the sensor data can be better utilized to obtain a more accurate pose.

[0066] S70, the posture output information is smoothed based on the RTS smoother to obtain a stable posture result.

[0067] It should be understood that facet fitting is to extract facets from laser point cloud data, thereby improving the efficiency of point cloud registration without downsampling the point cloud. Dynamic obstacle removal is to obtain disparity maps using depth maps based on visual and laser sensor data to obtain dynamic obstacles, and remove these obstacles in advance to reduce interference in multi-source sensor pose optimization, thereby improving the accuracy of the odometer. Finally, RTS smoothing is used to perform trajectory fitting to obtain relatively smooth pose output information, providing stable and reliable pose output for autonomous driving.

[0068] See also Figure 4 , which is a schematic diagram of the smoothing effect of an RTS smoother on the original trajectory in an embodiment of the present disclosure. It can be seen from the figure that when the original data oscillates, a relatively smooth and stable output result can be obtained based on the RTS smoother.

[0069] Figure 5 A multi-sensor fusion odometer generation system based on surface element fitting according to an embodiment of the present disclosure is shown, and the system includes:

[0070] A data alignment module, configured to acquire lidar sensor data, visual sensor data, and IMU sensor data;

[0071] The IMU pre-integration module is configured to perform pre-integration processing on the IMU data to obtain the relative transformation relationship within the posture update cycle;

[0072] The feature extraction module is configured to transform the laser radar sensor data and the visual sensor data according to the relative transformation relationship within the posture update cycle to obtain the laser point cloud disparity map and the visual disparity map of the adjacent frames, remove dynamic obstacles through the joint processing of the disparity map, extract the surface element of the laser point cloud data, where the surface element includes the three-dimensional coordinate point x, y, z, the normal vector nx, ny, nz and the radius r, and construct a plane by selecting four adjacent non-collinear points for each point to obtain the corresponding surface element information, extract the key points of the visual image data, and construct the constraint relationship between frames;

[0073] The pose smoothing module is configured to perform joint optimization based on the relative transformation relationship within the pose update period, the surface element information and the constraint relationship between frames to obtain the pose output information;

[0074] The pose estimation module is configured to smooth the pose output information based on the RTS smoother to obtain a stable pose result.

[0075] See also Figure 6 , which is a data flow diagram of a multi-sensor fusion odometer generation system based on facet fitting provided in an embodiment of the present disclosure.

[0076] Preferably, the system further comprises:

[0077] The data preprocessing module is configured to preprocess the lidar sensor data, the visual sensor data and the IMU sensor data, including time synchronization processing and calibration processing.

[0078] Preferably, the calibration used in the calibration process is a joint calibration, which constructs the residual information of the lidar sensor, the visual sensor and the IMU sensor after preprocessing, and builds a joint optimization equation to achieve the estimation of the external parameters and the sensor time difference.

[0079] Preferably, the system further includes an IMU initialization module configured to initialize the IMU and obtain a bias of gravity acceleration.

[0080] In some embodiments, an unmanned vehicle is disclosed, which is applied to the above-mentioned multi-sensor fusion odometer generation method based on surface element fitting.

[0081] The above description and the accompanying drawings fully illustrate the embodiments of the present disclosure so that those skilled in the art can practice them. Other embodiments may include structural, logical, electrical, process and other changes. The embodiments represent only possible changes. Unless explicitly required, separate components and functions are optional, and the order of operations may vary. Parts and features of some embodiments may be included in or replace parts and features of other embodiments. Moreover, the words used in this application are only used to describe the embodiments and are not used to limit the claims. As used in the description of the embodiments and the claims, unless the context clearly indicates, the singular forms of "a", "an" and "the" are intended to include plural forms as well. Similarly, the term "and / or" as used in this application refers to any and all possible combinations of one or more associated listings. In addition, when used in this application, the term "comprise" and its variants "comprises" and / or comprising refer to the presence of stated features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or groups thereof. In the absence of further restrictions, the elements defined by the sentence "including one..." do not exclude the presence of other identical elements in the process, method or automobile including the elements. In this article, each embodiment may focus on the differences from other embodiments, and the same and similar parts between the embodiments may refer to each other. For the methods, products, etc. disclosed in the embodiments, if they correspond to the method part disclosed in the embodiments, then the relevant parts can refer to the description of the method part.

[0082] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software may depend on the specific application and design constraints of the technical solution. The technicians may use different methods for each specific application to implement the described functions, but such implementations should not be considered to exceed the scope of the embodiments of the present disclosure. The technicians may clearly understand that, for the convenience and simplicity of description, the specific working processes of the systems, systems and units described above may refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here.

[0083] The flowchart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to the embodiment of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of the code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. In some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. In the description corresponding to the flowchart and the block diagram in the accompanying drawings, the operations or steps corresponding to different boxes can also occur in a different order from the order disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. Each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented by a dedicated hardware-based system that performs the specified functions or actions, or may be implemented by a combination of dedicated hardware and computer instructions.

Claims

1. A multi-sensor fusion odometer generation method based on surface element fitting, characterized in that: Applied to an unmanned vehicle, the vehicle includes multiple sensors, the multiple sensors are a laser radar system, a visual sensor and an IMU sensor, and the method includes: S10, acquiring laser radar sensor data, visual sensor data, and IMU sensor data; S20, pre-integrating the IMU data to obtain the relative transformation relationship within the posture update period; S30, performing coordinate transformation on the laser radar sensor data and the visual sensor data according to the relative transformation relationship within the posture update cycle to obtain the laser point cloud disparity map and the visual disparity map of adjacent frames, and removing dynamic obstacles by jointly processing the disparity maps; S40, extracting facets from the laser point cloud data, wherein the facets include three-dimensional coordinate points x, y, z, normal vectors nx, ny, nz and radius r, and constructing a plane by selecting four adjacent non-collinear points for each point to obtain corresponding facet information; S50, extracting key points from the visual image data and building constraint relationships between frames; S60, performing joint optimization according to the relative transformation relationship within the pose update period, the face information and the constraint relationship between frames to obtain pose output information; S70, the posture output information is smoothed based on the RTS smoother to obtain a stable posture result.

2. The multi-sensor fusion odometer generation method based on surface element fitting according to claim 1 is characterized in that: S20, before performing pre-integration processing on the IMU data to obtain the relative transformation relationship within the posture update period, also includes: Preprocess the lidar sensor data, visual sensor data, and IMU sensor data, including time synchronization and calibration.

3. The multi-sensor fusion odometer generation method based on surface element fitting according to claim 2 is characterized in that: The calibration used in the calibration process is a joint calibration, which constructs the residual information of the lidar sensor, the visual sensor and the IMU sensor after preprocessing, and builds a joint optimization equation to achieve the estimation of the external parameters and the sensor time difference.

4. The multi-sensor fusion odometer generation method based on surface element fitting according to claim 1 is characterized in that: S20, before performing pre-integration processing on the IMU data to obtain the relative transformation relationship within the posture update period, also includes: Initialize the IMU and obtain the offset of gravity acceleration.

5. The multi-sensor fusion odometer generation method based on surface element fitting according to claim 1 is characterized in that: The extraction of facets from laser point cloud data in S40 includes: The small facets are fitted based on the union-find method, and the obtained planes are fitted into large facets.

6. A multi-sensor fusion odometer generation system based on surface element fitting, characterized in that: include: A data alignment module, configured to acquire lidar sensor data, visual sensor data, and IMU sensor data; The IMU pre-integration module is configured to perform pre-integration processing on the IMU data to obtain the relative transformation relationship within the posture update cycle; The feature extraction module is configured to transform the laser radar sensor data and the visual sensor data according to the relative transformation relationship within the posture update cycle to obtain the laser point cloud disparity map and the visual disparity map of the adjacent frames, remove dynamic obstacles through the joint processing of the disparity map, extract the surface element of the laser point cloud data, where the surface element includes the three-dimensional coordinate point x, y, z, the normal vector nx, ny, nz and the radius r, and construct a plane by selecting four adjacent non-collinear points for each point to obtain the corresponding surface element information, extract the key points of the visual image data, and construct the constraint relationship between frames; The pose smoothing module is configured to perform joint optimization based on the relative transformation relationship within the pose update period, the surface element information and the constraint relationship between frames to obtain the pose output information; The pose estimation module is configured to smooth the pose output information based on the RTS smoother to obtain a stable pose result.

7. The multi-sensor fusion odometer generation system based on surface element fitting according to claim 6, characterized in that: Also includes: The data preprocessing module is configured to preprocess the lidar sensor data, the visual sensor data and the IMU sensor data, including time synchronization processing and calibration processing.

8. The multi-sensor fusion odometer generation system based on surface element fitting according to claim 7, characterized in that: The calibration used in the calibration process is a joint calibration, which constructs the residual information of the lidar sensor, the visual sensor and the IMU sensor after preprocessing, and builds a joint optimization equation to achieve the estimation of the external parameters and the sensor time difference.

9. The multi-sensor fusion odometer generation system based on surface element fitting according to claim 6, characterized in that: Also includes: The IMU initialization module is configured to initialize the IMU and obtain the offset of gravity acceleration.

10. An unmanned vehicle, characterized in that: Applicable to the multi-sensor fusion odometer generation method based on surface element fitting as described in any one of claims 1 to 5.

Images