Parking mapping method and system based on combination of laser and imu, and vehicle

Abstract

The application provides a parking mapping method and system based on laser and IMU combination and a vehicle. The method comprises the following steps: acquiring observation data of a self vehicle pose in real time based on lidar observation data; acquiring prediction data of the self vehicle pose in real time based on IMU odometry data; constructing a pose optimization objective function by minimizing the difference between the observation data of the self vehicle pose and the prediction data of the self vehicle pose; obtaining optimal pose data by optimizing and solving the pose optimization objective function based on the observation data of the self vehicle pose and the prediction data of the self vehicle pose; optimizing the lidar observation data based on the obtained optimal pose data to obtain a local subgraph; and constructing a global map based on the obtained local subgraph.

Description

Technical Field

[0001] This invention belongs to the field of intelligent driving assistance technology, and in particular relates to a parking mapping method, system and vehicle based on the combination of laser and IMU. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] Parking mapping typically refers to the process of constructing a parking environment map in parking scenarios such as parking lots using various sensors (such as cameras and ultrasonic sensors). This map provides key environmental information for the vehicle's automatic parking system, helping the vehicle to better plan parking paths and perform operations such as parking space detection.

[0004] The inventors discovered that traditional parking mapping methods typically use vision combined with IMU (Inertial Measurement Unit) odometry for trajectory estimation, and ultrasonic waves for space parking search, corner point updates, and obstacle avoidance. This method has poor visual capture capabilities in low-light areas such as underground parking garages, and can only rely on IMU odometry for parking mapping. Furthermore, using single data for parking mapping is highly susceptible to the accuracy of the data acquisition equipment and its built-in algorithms, resulting in poor accuracy of parking mapping and a higher risk of parking accidents. Summary of the Invention

[0005] This invention provides a parking mapping method, system, and vehicle based on the combination of laser and IMU, to solve the problems of traditional solutions having poor visual capture capabilities in low-light areas such as underground parking garages, relying solely on IMU odometers for parking mapping; and using single data for parking mapping is easily affected by the accuracy of the data acquisition equipment and its built-in algorithm, resulting in poor accuracy of parking mapping and high parking risk.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] In a first aspect, the present invention provides a parking mapping method based on the combination of laser and IMU, which is applied to vehicles equipped with lidar, the method comprising:

[0008] Based on lidar observation data, real-time observation data of vehicle position and pose is acquired;

[0009] Based on IMU odometry data, predictive data of vehicle position and orientation are obtained in real time;

[0010] A pose optimization objective function is constructed to minimize the difference between the observed pose data and the predicted pose data of the vehicle.

[0011] Based on the observation data and prediction data of the vehicle's pose, the optimal pose data is obtained by optimizing the pose optimization objective function.

[0012] The optimal pose data is used to optimize the lidar observation data to obtain a local sub-image;

[0013] The global map is constructed based on the obtained local subgraphs.

[0014] Furthermore, the real-time acquisition of vehicle pose observation data based on lidar observation data specifically involves: the lidar continuously scanning the surrounding environment to generate several point cloud data frames arranged in chronological order; each point cloud data frame reflects the three-dimensional spatial information around the vehicle at the corresponding moment; and the vehicle pose observation data is determined by inter-frame matching between adjacent point cloud data frames.

[0015] Furthermore, the real-time acquisition of vehicle position prediction data based on IMU odometer data specifically involves: acquiring vehicle acceleration data in real time based on the IMU odometer; and determining the vehicle position prediction data based on the acquired acceleration data through integral processing.

[0016] Furthermore, the pose optimization objective function is specifically expressed as follows:

[0017] minF(x)=min∑e i (x) T Ω i e i (x)

[0018] Among them, e i (x) represents the difference between the lidar observation and the IMU prediction, Ω i e i (x) is the variance of a normal distribution that (x) follows.

[0019] Furthermore, the optimization solution of the pose optimization objective function is specifically as follows: the pose optimization objective function is subjected to a first-order Taylor approximation transformation, and the optimization solution is performed based on the least squares method to obtain the optimal pose data.

[0020] Furthermore, the optimization of the lidar observation data using the obtained optimal pose data specifically involves: performing a translation transformation on the lidar observation data based on the obtained optimal pose data, thereby achieving coordinate transformation of the lidar observation data.

[0021] Furthermore, the construction of the global map based on the obtained local subgraphs specifically involves: obtaining a global map by stitching together several local subgraphs obtained at different times during the parking process.

[0022] Furthermore, during the construction of the local sub-map, loop closure detection is performed based on whether the features in the current lidar observation data are consistent with one or several frames of historical point cloud data, and the local sub-map is optimized based on the loop closure detection results.

[0023] Secondly, the present invention provides a parking mapping system based on a combination of laser and IMU, comprising:

[0024] The pose observation data acquisition unit is used to acquire the vehicle's pose observation data in real time based on lidar observation data.

[0025] The pose prediction data acquisition unit is used to acquire the vehicle's pose prediction data in real time based on the integration of IMU odometry data.

[0026] The objective function construction unit is used to construct a pose optimization objective function by minimizing the difference between the observed data and the predicted data of the vehicle pose.

[0027] The pose optimization unit is used to obtain the optimal pose data by optimizing the pose optimization objective function based on the observation data and prediction data of the vehicle pose.

[0028] The local subgraph construction unit is used to optimize the lidar observation data based on the obtained optimal pose data to obtain a local subgraph.

[0029] The global map construction unit constructs a global map based on the obtained local submaps.

[0030] Thirdly, the present invention provides a vehicle that employs the aforementioned parking mapping method based on a combination of laser and IMU.

[0031] The above one or more technical solutions have the following beneficial effects:

[0032] This invention provides a parking mapping method, system, and vehicle based on the combination of laser and IMU. The solution uses a combination of laser radar and IMU odometer to achieve parking mapping, which can effectively address the problem that traditional solutions rely on only a single data source (i.e., IMU) for parking mapping in dark environments (such as underground parking garages), resulting in poor accuracy.

[0033] The solution described in this invention optimizes the vehicle pose by minimizing the difference between the observed data and the predicted data of the vehicle pose. Based on the optimized vehicle pose, the lidar observation data is updated and optimized to obtain an optimized local sub-map. Simultaneously, the solution performs secondary optimization on the local sub-map through loop closure detection, further ensuring the accuracy of the local sub-map and thus improving the effectiveness of the parking mapping results.

[0034] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0035] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0036] Figure 1 This is an overall flowchart of a parking mapping method based on the combination of laser and IMU as described in an embodiment of the present invention. Detailed Implementation

[0037] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0038] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.

[0039] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0040] Terminology Explanation:

[0041] Local submap: A local submap is a map representation of the local environment in which the vehicle is located. It is extracted and constructed from environmental data surrounding the vehicle, such as point cloud data from LiDAR scanning or image data captured by cameras. Local submaps can take various forms, such as raster maps and feature maps, and are used to describe information such as obstacles, road boundaries, and lane lines within a certain range around the vehicle.

[0042] In one or more embodiments, such as Figure 1 As shown, this application provides a parking mapping method based on the combination of laser and IMU, which is applied to vehicles equipped with lidar. The method includes the following processing steps:

[0043] Step 1: Based on lidar observation data, acquire real-time observation data of the vehicle's position and pose; based on the integration of IMU odometry data, acquire real-time prediction data of the vehicle's position and pose.

[0044] In practical implementation, the vehicle pose data obtained from lidar observation data is as follows:

[0045]

[0046] Assume the noise data (non-quantizable) of the lidar is as follows:

[0047]

[0048] The vehicle's pose data obtained from the integration of the IMU are as follows:

[0049]

[0050] The IMU's noise data (non-quantizable) is as follows:

[0051]

[0052] The observation equation for lidar is:

[0053]

[0054] Where R is the rotation matrix and T is the translation matrix. It is the pose matrix in the laser coordinate system. This is the pose matrix transformed to vehicle coordinates. The data observed by the LiDAR is translated and rotated before being mapped to vehicle coordinates. Due to the different vector dimensions, it is necessary to... Perform homogeneous transformation to right Perform homogeneous transformation to

[0055] Step 2: Construct a pose optimization objective function by minimizing the difference between the observed pose data and the predicted pose data of the vehicle.

[0056] In practical implementation, it is assumed that the difference between the observed value of the lidar and the predicted value of the IMU is e. i ,

[0057] e i =X i -Z i

[0058] Among them, X i The row vectorization matrix of Zl

[0059]

[0060] Z i The row vectorized matrix of Zi

[0061]

[0062] Among them, e i It follows a mean of 0 and a variance of Ω. i It follows a normal distribution. That is, e i ~N(0,Ω) i ).

[0063] The error e is obtained from Bayesian probability estimation. i The joint probability distribution is:

[0064]

[0065] To minimize the error, the Gaussian distribution should be as large as possible. Since the joint probability distribution function remains unchanged, the problem is transformed into an optimization problem. Taking the logarithm of both sides of the equation, we get:

[0066]

[0067] because Since it is independent of the variable x, the problem of finding the maximum value of lnG(ei(x)) is transformed into minimizing... The problem.

[0068] Finally, the objective function is expressed as: minF(x)=min∑e i (x) T Ω i e i (x).

[0069] Step 3: Based on the observed and predicted vehicle pose data, the optimal pose data is obtained by optimizing the pose optimization objective function.

[0070] Based on the objective function described above, we can see that F(x) is a convex function. We can use gradient descent to find the minimum point of F(x). However, since F(x) is nonlinear, we cannot directly use the least squares method. Therefore, we perform a first-order Taylor approximation transformation on the original expression:

[0071] e i (x+Δx)=e i (x)+J i (x)Δx

[0072] Among them, J i (x) is the partial derivative of the mapping function F(x) with respect to the state vector x, that is:

[0073]

[0074] F(x + Δx) = ∑e i (x + Δx) T Ω i e i (x + Δx)

[0075] = ∑(e i (x) + J i Δx) T Ω i e i (x) + J i Δx

[0076] = ∑(e i T Ω i e i + e i T Ω i J i Δx + e i T Δx T J i T Ω i e i + Δx T J i T Ω i J i Δx)

[0077] = ∑(e i T Ω i e i + 2e i T Ω i J i Δx + Δx T J i T Ω i J i Δx)

[0078] = ∑c i + ∑(2b i Δx + Δx T H i Δx)

[0079] = ∑c i + ∑2b i Δx + Δx T ∑H i Δx

[0080] Combining the above equations, it can be seen that F(x + Δx) is a quadratic function of Δx.

[0081] Then take the partial derivative with respect to F(x+Δx), when At that time, we find:

[0082] Δx=-H- 1 b

[0083] Therefore, the Δx corresponding to the point where the derivative is 0 is the minimum iteration value of gradient descent. Using this value, gradient descent can eventually converge to obtain the optimal pose data.

[0084] Step 4: Optimize the lidar observation data with the obtained optimal pose data to obtain a local sub-image;

[0085] In a real-time scenario, the optimization of the lidar observation data based on the obtained optimal pose data involves performing a translation transformation on the lidar observation data to achieve coordinate transformation of the lidar observation data.

[0086] Specifically, lidar observation data is typically constructed using its own location as the origin, with the coordinate axes set conventionally. For example, the x-axis points directly in front of the lidar, the y-axis points to the left, and the z-axis points upwards. The point cloud data acquired by the lidar is initially based on this coordinate system to represent the positional relationship of each point relative to the lidar.

[0087] Given the position coordinates in the optimal pose data, assuming the vehicle's position is (x0, y0, z0) in the target reference coordinate system (taking the world coordinate system as an example), for any point cloud data point P(x0, y0, z0) in the lidar coordinate system... l ,y l ,z l First, a translation transformation must be performed to convert the coordinates from the lidar coordinate system to the target reference coordinate system. The translated P′(x) w ,y w , z w The coordinates of a point are calculated using the following formula:

[0088]

[0089] Through the above addition operation, the lidar point cloud data can be converted in terms of position to the target reference coordinate system.

[0090] In other implementations, data filtering and noise processing can also be performed based on the point cloud data after coordinate system transformation. Specifically:

[0091] Data Filtering: After coordinate transformation, the optimized LiDAR observation data can be filtered according to actual application needs. For example, in some data application scenarios that only focus on specific areas (such as a certain range in front of a vehicle for obstacle avoidance), point cloud data points that are too far away or in directions that do not require attention can be removed, reducing the amount of data while improving subsequent processing efficiency and focusing on key information. Data filtering can be performed by setting conditions such as distance thresholds and angle ranges.

[0092] Noise Processing: Although lidar itself has high precision, the acquired data may still contain noisy points due to environmental factors (such as strong light reflection, smoke, etc.) or the equipment itself. Filtering techniques can be used to process the noise in the optimized point cloud data. Common filtering methods include statistical filtering (removing noisy points that significantly deviate from the statistical distribution of surrounding points) and radius filtering (eliminating isolated points with too few points within a certain radius). These methods further improve data quality, making the optimized lidar observation data cleaner, more reliable, and more conducive to subsequent analysis and application.

[0093] In more implementations, during the construction of local sub-graphs, loop closure detection is performed based on whether the features in the current lidar observation data are consistent with one or several frames of historical point cloud data, and the local sub-graphs are optimized based on the loop closure detection results.

[0094] Specifically, the optimization of the local subgraph based on the loop closure detection results includes:

[0095] (1) Error correction based on loop closure:

[0096] Once loop closures are identified, the loop closure information can be used to correct errors in the local submap. Because accumulated errors exist during the construction of the local submap and previous map stitching processes, the accurate position and attitude of the vehicle at the loop closure point can be recalculated using the matched features. For example, based on the coordinate differences between matched features, the position and angle of the local submap in the global coordinate system can be adjusted, redistributing and correcting the previously accumulated errors, making the local submap more accurately reflect the actual environmental conditions.

[0097] At the same time, the point cloud data in the local sub-graph can be optimized to remove unreasonable points caused by errors or to refit and correct some blurry features, thereby further improving the quality of the local sub-graph.

[0098] (2) Local subgraph merging and updating:

[0099] For multiple local subgraphs related to loop closures, the fusion of them is optimized based on the accurate positional relationships determined by loop closure detection. If there were inaccuracies in the previous stitching of these local subgraphs, the loop closure information is used for re-stitching and fusion to ensure a natural transition between local subgraphs and accurately reflect the continuity of the environment.

[0100] In addition, by combining current lidar observation data, the local sub-map is updated with parts that may change, such as newly detected obstacles or changes in parking space occupancy, so that the local sub-map always reflects the latest state of the actual environment.

[0101] Step 5: Construct the global map based on the obtained local subgraphs.

[0102] In specific implementation, the construction of the global map based on the obtained local subgraphs is as follows: based on several local subgraphs obtained at different times during the parking process, the global map is obtained by stitching together the obtained local subgraphs.

[0103] In one or more embodiments, corresponding to the above method, this embodiment provides a parking mapping system based on a combination of laser and IMU, including:

[0104] The pose observation data acquisition unit is used to acquire the vehicle's pose observation data in real time based on lidar observation data.

[0105] The pose prediction data acquisition unit is used to acquire the vehicle's pose prediction data in real time based on the integration of IMU odometry data.

[0106] The objective function construction unit is used to construct a pose optimization objective function by minimizing the difference between the observed data and the predicted data of the vehicle pose.

[0107] The pose optimization unit is used to obtain the optimal pose data by optimizing the pose optimization objective function based on the observation data and prediction data of the vehicle pose.

[0108] The local subgraph construction unit is used to optimize the lidar observation data based on the obtained optimal pose data to obtain a local subgraph.

[0109] The global map construction unit constructs a global map based on the obtained local submaps.

[0110] It is understood that the system described in this embodiment corresponds to the method embodiment described above, and its technical details have been described in detail in the method embodiment, so they will not be repeated here.

[0111] In further embodiments, this embodiment provides a vehicle that employs the aforementioned parking mapping method based on a combination of laser and IMU.

[0112] In specific implementations, the vehicle may further include RF (Radio Frequency) circuitry, a memory including one or more computer-readable storage media, an input unit, a display unit, sensors, audio circuitry, a WiFi (Wireless Fidelity) module, a processor including one or more processing cores, and a power supply, among other components. Those skilled in the art will understand that the above-described components do not constitute a limitation on the vehicle, and may include more or fewer components than illustrated, or combine certain components, or have different component arrangements. Wherein:

[0113] RF circuits are used for receiving and transmitting signals during information transmission or calls. Specifically, they receive downlink information from the base station and process it through one or more processors; additionally, they transmit uplink data to the base station. Typically, RF circuits include, but are not limited to, antennas, at least one amplifier, tuner, one or more oscillators, a Subscriber Identity Module (SIM) card, transceiver, coupler, LNA (Low Noise Amplifier), and duplexer. Furthermore, RF circuits can communicate wirelessly with networks and other devices. This wireless communication can use any communication standard or protocol, including but not limited to GSM (Global System for Mobile communication), GPRS (General Packet Radio Service), CDMA (Code Division Multiple Access), WCDMA (Wideband Code Division Multiple Access), LTE (Long Term Evolution), email, and SMS (Short Messaging Service).

[0114] Memory can be used to store software programs and modules. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory. Memory can mainly include a program storage area and a data storage area. The program storage area can store the operating system, application programs required for at least one function (such as sound playback, image playback, etc.), etc.; the data storage area can store data created based on vehicle usage (such as audio data, phone books, etc.). Furthermore, memory can include high-speed random access memory and non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage devices. Accordingly, memory can also include a memory controller to provide access to the memory for the processor and input units.

[0115] The input unit can be used to receive input numerical or character information, and to generate keyboard, mouse, joystick, optical, or trackball signal inputs related to user settings and function control. Specifically, the input unit may include a touch-sensitive surface and other input devices. A touch-sensitive surface, also known as a touch display or touchpad, can collect user touch operations on or near it (such as user operations using fingers, styluses, or any suitable object or accessory on or near the touch-sensitive surface) and drive corresponding connected devices according to a pre-set program. Optionally, the touch-sensitive surface may include two parts: a touch detection device and a touch controller. The touch detection device detects the user's touch location and the signal generated by the touch operation, transmitting the signal to the touch controller; the touch controller receives touch information from the touch detection device, converts it into touch point coordinates, sends it to the processor, and can receive and execute commands from the processor. Furthermore, various types of touch-sensitive surfaces, such as resistive, capacitive, infrared, and surface acoustic wave, can be used. In addition to the touch-sensitive surface, the input unit may also include other input devices. Specifically, other input devices may include, but are not limited to, one or more of the following: physical keyboard, function keys (such as volume control buttons, power buttons, etc.), trackball, mouse, joystick, etc.

[0116] The display unit can be used to display information input by the user or information provided to the user, as well as various graphical user interfaces (GUIs) of the vehicle. These GUIs can be composed of graphics, text, icons, video, and any combination thereof. The display unit may include a display panel, optionally configured as an LCD (Liquid Crystal Display), OLED (Organic Light-Emitting Diode), or similar display panel. Furthermore, a touch-sensitive surface may cover the display panel. When the touch-sensitive surface detects a touch operation on or near it, it transmits the information to the processor to determine the type of touch event. Subsequently, the processor provides corresponding visual output on the display panel based on the type of touch event. The touch-sensitive surface and the display panel are implemented as two separate components to achieve input and output functions; however, in some embodiments, the touch-sensitive surface and the display panel can be integrated to achieve both input and output functions.

[0117] The vehicle may also include at least one sensor, such as a light sensor, a motion sensor, and other sensors. Specifically, the light sensor may include an ambient light sensor and a proximity sensor. The ambient light sensor can adjust the brightness of the display panel according to the ambient light level, and the proximity sensor can turn off the display panel and / or backlight when the vehicle is moved close to the ear. As a type of motion sensor, a gravity acceleration sensor can detect the magnitude of acceleration in various directions (generally three axes). When stationary, it can detect the magnitude and direction of gravity and can be used for applications that recognize the phone's posture (such as landscape / portrait switching, related games, magnetometer posture calibration), vibration recognition functions (such as pedometers, taps), etc. Other sensors that the vehicle may also be equipped with, such as gyroscopes, barometers, hygrometers, thermometers, and infrared sensors, will not be elaborated here.

[0118] Audio circuitry, speakers, and microphones provide an audio interface between the user and the vehicle. The audio circuitry converts received audio data into electrical signals, which are then transmitted to the speakers, where they are converted into sound signals for output. Conversely, the microphone converts collected sound signals into electrical signals, which are received by the audio circuitry, converted back into audio data, and then processed by a processor. The audio data is then transmitted via RF circuitry to, for example, another vehicle, or output to memory for further processing. The audio circuitry may also include an earphone jack to facilitate communication between external headphones and the vehicle.

[0119] WiFi is a short-range wireless transmission technology. Vehicles using WiFi modules can help users send and receive emails, browse web pages, and access streaming media, providing wireless broadband internet access. Although a WiFi module is shown, it is understood that it is not an essential component of the vehicle and can be omitted as needed without altering the essence of the invention.

[0120] The processor is the control center of the vehicle, connecting various parts of the phone via various interfaces and lines. It executes software programs and / or modules stored in memory, and calls data stored in memory to perform various vehicle functions and process data, thereby providing overall monitoring of the phone. Optionally, the processor may include one or more processing cores; preferably, the processor may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, while the modem processor mainly handles wireless communication. It is understood that the modem processor may also not be integrated into the processor.

[0121] The vehicle also includes a power supply (such as a battery) that powers various components. Preferably, the power supply can be connected to the processor logic through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. The power supply may also include one or more DC or AC power sources, a recharging system, a power fault detection circuit, a power converter or inverter, a power status indicator, and any other components.

[0122] Although not shown, the vehicle may also include cameras, Bluetooth modules, etc., which will not be described in detail here. Specifically, in this embodiment, the vehicle's display unit is a touch screen display, and the vehicle also includes a memory and one or more programs, wherein one or more programs are stored in the memory and configured to be executed by one or more processors. The one or more programs include methods for performing the methods shown in the above embodiments.

[0123] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A parking mapping method based on the combination of laser and IMU, applied to vehicles equipped with lidar, characterized in that, The method includes: Based on lidar observation data, real-time observation data of vehicle position and pose is acquired; Based on IMU odometry data, predictive data of vehicle position and orientation are obtained in real time; A pose optimization objective function is constructed to minimize the difference between the observed pose data and the predicted pose data of the vehicle. The pose optimization objective function is specifically expressed as follows: in, This represents the difference between the lidar observations and the IMU predictions. express The variance corresponding to a normal distribution; Based on the observation data and prediction data of the vehicle's pose, the optimal pose data is obtained by optimizing the pose optimization objective function. The optimization solution of the pose optimization objective function is specifically as follows: the pose optimization objective function is subjected to a first-order Taylor approximation transformation, and the optimization solution is performed based on the least squares method to obtain the optimal pose data; The optimal pose data is used to optimize the lidar observation data to obtain a local sub-image; The optimization of the lidar observation data based on the obtained optimal pose data specifically involves: performing a translation transformation on the lidar observation data based on the obtained optimal pose data to achieve coordinate transformation of the lidar observation data; The global map is constructed based on the obtained local subgraphs.

2. The parking mapping method based on laser and IMU combination as described in claim 1, characterized in that, The method of acquiring real-time observation data of vehicle pose based on lidar observation data specifically involves: LiDAR continuously scanning the surrounding environment to generate several point cloud data frames arranged in chronological order; each point cloud data frame reflects the three-dimensional spatial information around the vehicle at the corresponding moment; and the observation data of vehicle pose is determined by inter-frame matching between adjacent point cloud data frames.

3. The parking mapping method based on laser and IMU combination as described in claim 1, characterized in that, The method of obtaining real-time prediction data of vehicle position and orientation based on IMU odometer data specifically involves: obtaining real-time acceleration data of the vehicle based on IMU odometer data; and determining the prediction data of vehicle position and orientation based on the obtained acceleration data through integral processing.

4. The parking mapping method based on laser and IMU combination as described in claim 1, characterized in that, The construction of the global map based on the obtained local subgraphs specifically involves: using several local subgraphs obtained at different times during the parking process, and stitching them together to obtain the global map.

5. The parking mapping method based on laser and IMU combination as described in claim 1, characterized in that, During the construction of the local sub-map, loop closure detection is performed based on whether the features in the current lidar observation data are consistent with one or several frames of historical point cloud data, and the local sub-map is optimized based on the loop closure detection results.

6. A parking mapping system based on a combination of laser and IMU, characterized in that, include: The pose observation data acquisition unit is used to acquire the vehicle's pose observation data in real time based on lidar observation data. The pose prediction data acquisition unit is used to acquire the vehicle's pose prediction data in real time based on the integration of IMU odometry data. The objective function construction unit is used to construct a pose optimization objective function by minimizing the difference between the observed data and the predicted data of the vehicle pose. The pose optimization objective function is specifically expressed as follows: in, This represents the difference between the lidar observations and the IMU predictions. express The variance corresponding to a normal distribution; The pose optimization unit is used to obtain the optimal pose data by optimizing the pose optimization objective function based on the observation data and prediction data of the vehicle pose. The optimization solution of the pose optimization objective function is specifically as follows: the pose optimization objective function is subjected to a first-order Taylor approximation transformation, and the optimization solution is performed based on the least squares method to obtain the optimal pose data; The local subgraph construction unit is used to optimize the lidar observation data based on the obtained optimal pose data to obtain a local subgraph; the optimization of the lidar observation data based on the obtained optimal pose data specifically involves: performing a translation transformation on the lidar observation data based on the obtained optimal pose data to achieve coordinate transformation of the lidar observation data. The global map construction unit constructs a global map based on the obtained local submaps.

7. A vehicle, characterized in that, The parking mapping method based on the combination of laser and IMU as described in any one of claims 1-5 is adopted.

Images