Positioning method and device of collaborative work vehicle, computer device and storage medium

Abstract

The application relates to the technical field of vehicle positioning, and discloses a positioning method and device for a cooperative work vehicle, computer equipment and a storage medium, the method comprising the following steps: acquiring point cloud data containing a target object on the cooperative work vehicle in real time; dividing the point cloud data into a plurality of sliding window regions based on a preset sliding window, and determining the probability density corresponding to the point cloud data in each sliding window region; determining a region of interest according to the probability density corresponding to the point cloud data in each sliding window region; clustering the point cloud data in the region of interest to determine a center point set of the target object; and determining a positioning result of the cooperative work vehicle based on the center point set of the target object. The application quickly positions the region of interest where the target object is located by calculating the probability density corresponding to the point cloud data, does not need to train a large amount of data, is less affected by data transmission errors between devices, has high positioning precision, is time-saving and efficient, and has strong robustness.

Description

Technical Field

[0001] This invention relates to the field of vehicle positioning technology, and more specifically to positioning methods, devices, computer equipment, and storage media for collaborative work vehicles. Background Technology

[0002] In the collaborative operation of engineering vehicles, especially in collaborative operation scenarios such as transportation, loading and unloading, it is necessary to determine the relative positional relationship between the vehicle and the collaborative operation vehicles, and continuously adjust the positions of the two vehicles to ensure smooth operation.

[0003] Currently, there are two main methods for multi-vehicle cooperative positioning. One is to deploy multiple GPS devices and information interaction devices to achieve high-precision positioning of cooperative vehicles, but the cost is too high and cannot be adopted on a large scale. The other is to acquire vehicle information by deploying sensor devices on the vehicle itself, and to obtain the relative position of the cooperative vehicles by performing target detection through methods such as clustering segmentation or deep learning. However, this requires a large amount of data labeling and training, which is time-consuming and labor-intensive, and sensor errors have a significant impact on the positioning results. Summary of the Invention

[0004] In view of this, the present invention provides a positioning method, device, computer equipment and storage medium for collaborative operation vehicles to solve the problems of long time consumption and low positioning accuracy of existing positioning methods.

[0005] In a first aspect, the present invention provides a method for locating a collaborative work vehicle, the method comprising:

[0006] Real-time acquisition of point cloud data containing target objects on collaborative work vehicles;

[0007] The point cloud data is divided into multiple sliding window regions based on a preset sliding window, and the probability density of the point cloud data in each sliding window region is determined.

[0008] The region of interest is determined based on the probability density of the point cloud data within each sliding window region;

[0009] Cluster the point cloud data within the region of interest to determine the set of center points of the target object;

[0010] Based on the set of center points of the target object, the positioning results of the collaborative operation vehicles are determined.

[0011] Therefore, by acquiring point cloud data containing target objects on collaborative work vehicles in real time, and dividing the point cloud data into multiple sliding window regions based on a preset sliding window, the probability density of the point cloud data in each sliding window region is determined, thereby quickly filtering out the region of interest where the target object is located. Then, the point cloud data in the region of interest is clustered to determine the set of center points of the target object, and the target center point is determined based on the set of center points. High-precision positioning of collaborative work vehicles is achieved without the need for training on a large amount of data, which is time-saving and efficient, and is less affected by data transmission errors between devices, and has strong robustness.

[0012] In one optional implementation, determining the probability density of point cloud data within each sliding window region includes:

[0013] For each sliding window region, the probability distribution parameters corresponding to the point cloud data within the sliding window region are calculated based on the pre-stored target object probability model.

[0014] For each sliding window region, the probability density of the point cloud data within the sliding window region is calculated based on the probability distribution parameters of the point cloud data within the sliding window region.

[0015] Therefore, by using a pre-stored target probability model to determine the probability distribution parameters of point cloud data within the sliding window area, the probability density of the point cloud data within the sliding window area can be obtained. This eliminates the need for real-time analysis of the spatial distribution characteristics of the point cloud data collected by the lidar, resulting in higher efficiency and shorter processing time.

[0016] In one optional implementation, the probability density of the point cloud data within the sliding window area is calculated using the following formula:

[0017]

[0018] Among them, P ij X′ represents the probability density of the point cloud data within the sliding window region. k Let represent the point cloud data within the sliding window region, k represent the number of point cloud data within the sliding window region, q represent the mean of the point cloud data within the sliding window region, and ∑ represent the covariance of the point cloud data within the sliding window region.

[0019] Therefore, by calculating the probability density of point cloud data within each sliding window area, the region of interest where the target object is located can be quickly filtered out in the subsequent process without the need to analyze a large amount of data, which is more efficient and improves the efficiency of collaborative operation vehicle positioning.

[0020] In one optional implementation, the positioning result of the collaborative operation vehicle is determined based on the set of center points of the target object, including:

[0021] The target center point is obtained by filtering the set of center points;

[0022] The coordinates of the target center point are transformed to determine its world coordinates, thus obtaining the positioning results of the collaborative operation vehicle.

[0023] Therefore, by filtering and screening the point cloud data within the region of interest, it is possible to further determine which points may be the center point of the target object, thereby improving the positioning accuracy of the positioning method.

[0024] In one optional implementation, filtering the set of center points to obtain the target center point includes:

[0025] Based on the point cloud coordinates of each center point in the center point set, calculate the Euclidean distance of each center point's coordinates at the current time relative to its coordinates at the previous time.

[0026] Based on the Euclidean distances of each center point, determine the minimum Euclidean distance and take the center point corresponding to the minimum Euclidean distance as the target center point.

[0027] Therefore, by combining the motion relationship of the target object at the current moment with that of the previous moment to determine the target center point, compared with directly using GPS data and radar laser point cloud data, errors that may be caused by sensors or the external environment are eliminated, and the positioning accuracy of the positioning method is further improved.

[0028] In one optional implementation, the point cloud data includes the reflection intensity of the point cloud. Clustering the point cloud data within the region of interest to determine the set of center points of the target object includes:

[0029] Based on the reflection intensity of the point cloud within the region of interest, the point cloud data within the region of interest is filtered to obtain the first set of point clouds;

[0030] The first point cloud set is clustered according to the preset density clustering algorithm to obtain the second point cloud set;

[0031] The second point cloud set is filtered based on the pre-stored geometric distribution model of the target object to determine the set of center points of the target object.

[0032] Therefore, the points in the first point cloud set are clustered using a preset density clustering algorithm, and the centroids of the multiple clusters are used as the points in the second point cloud set to further filter out the point cloud data reflected by the target object.

[0033] In one optional implementation, the region of interest is determined based on the probability density of the point cloud data within each sliding window region, including:

[0034] The maximum probability density is determined based on the probability density of the point cloud data within each sliding window region;

[0035] The region of interest is obtained by expanding the sliding window region corresponding to the maximum probability density.

[0036] Therefore, by analyzing the probabilistic characteristics of point cloud data, the region of interest where the target object is located can be quickly determined without the need to collect a large amount of data for annotation and training, which is more efficient.

[0037] Secondly, the present invention provides a positioning device for a collaborative work vehicle, the device comprising:

[0038] The acquisition module is used to acquire point cloud data containing target objects on collaborative operation vehicles in real time;

[0039] The first processing module is used to divide point cloud data into multiple sliding window regions based on a preset sliding window, and determine the probability density of point cloud data in each sliding window region.

[0040] The second processing module is used to determine the region of interest based on the probability density of the point cloud data within each sliding window area;

[0041] The third processing module is used to cluster the point cloud data within the region of interest to determine the set of center points of the target object;

[0042] The fourth processing module is used to determine the positioning results of the collaborative operation vehicles based on the set of center points of the target object.

[0043] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the positioning method for cooperative work vehicles described in the first aspect or any corresponding embodiment thereof.

[0044] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the positioning method for a cooperative working vehicle according to the first aspect or any corresponding embodiment described above. Attached Figure Description

[0045] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0046] Figure 1 This is a flowchart illustrating a positioning method for collaborative work vehicles according to an embodiment of the present invention;

[0047] Figure 2 This is a schematic diagram of a mesh division according to an embodiment of the present invention;

[0048] Figure 3 This is a flowchart illustrating another method for locating collaborative work vehicles according to an embodiment of the present invention;

[0049] Figure 4 This is a flowchart illustrating another method for positioning collaborative work vehicles according to an embodiment of the present invention;

[0050] Figure 5 This is a schematic diagram illustrating the result of a clustering process according to an embodiment of the present invention;

[0051] Figure 6 This is a schematic diagram of the motion relationship of a target object according to an embodiment of the present invention;

[0052] Figure 7 This is a structural block diagram of a positioning device for a collaborative work vehicle according to an embodiment of the present invention;

[0053] Figure 8 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0055] Collaborative operations of engineering vehicles require determining the relative positions of the vehicle and target vehicles to ensure smooth workflow, especially in scenarios such as loading and unloading, where continuous adjustments to the positions of the two vehicles are necessary for successful operation. Therefore, high accuracy in the relative positions of the two vehicles is crucial. In the field of autonomous driving, multi-vehicle collaborative operations not only require GPS devices to determine the vehicle's own position but also various sensors to determine its relative position with other collaborating vehicles. This provides reliable data input for vehicle path planning and control, ensuring the smooth progress of engineering operations. Therefore, research on multi-vehicle collaborative positioning methods is of great significance.

[0056] Currently, there are two main technologies: one is to deploy multiple GPS devices and information interaction devices to collect and share location information of different vehicles. This method can achieve high-precision positioning of collaborative vehicles, but the cost is too high and it is not easy to adopt on a large scale. The other is to obtain the relative position of collaborative vehicles through sensor devices deployed on the vehicle itself, such as cameras and lidar. This method can be divided into two steps: target detection and target position calculation. The latter depends on the former. Target detection mainly includes methods such as clustering and segmentation and deep learning, but it requires a large amount of data to be collected, labeled and trained, which is time-consuming and labor-intensive. At the same time, due to the influence of sensor errors, the accuracy of the target position calculation also needs to be improved.

[0057] This invention employs a probabilistic model to characterize the probabilistic features of the target object, thereby quickly filtering out the target to be detected, saving time and increasing efficiency. It extracts the set of center points of the target point cloud through clustering methods, and filters the set of center points based on the motion relationship of the target object relative to the previous moment, enhancing the robustness of the position calculation, thereby determining the center point of the target object and achieving high-precision positioning of collaborative operation vehicles.

[0058] According to an embodiment of the present invention, a method for positioning a collaborative work vehicle is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0059] This embodiment provides a positioning method for collaborative work vehicles, which can be used in vehicle-mounted computer equipment or electronic devices, such as vehicle controllers, vehicle-mounted computers, etc. Figure 1 This is a flowchart of a positioning method for collaborative work vehicles according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps:

[0060] Step S101: Acquire point cloud data containing target objects on the collaborative operation vehicle in real time.

[0061] Specifically, a lidar device is installed at an appropriate location on the vehicle, and a centering target (i.e., the object) is placed on the collaborative work vehicle. The centering target can be an object that can reflect laser light, such as a reflective strip. The installation positions of the lidar device and the centering target must ensure that the lidar device can capture the reflection data of the centering target (i.e., the object) throughout the entire collaborative work process.

[0062] It should be noted that before acquiring point cloud data using a LiDAR device, the extrinsic parameter calibration of the LiDAR device must be completed first to obtain the transformation matrix from the LiDAR coordinate system to the GPS coordinate system, thereby achieving spatial synchronization between LiDAR and GPS data. Specifically, a set of GPS and LiDAR data can be collected when a vehicle is traveling in a figure-eight pattern. Based on the vehicle's trajectory in the LiDAR coordinate system obtained by the LiDAR odometer and the trajectory in the GPS coordinate system obtained by the GPS device, the extrinsic parameter calibration of the LiDAR to GPS can be achieved through hand-eye calibration.

[0063] The point cloud data in this embodiment of the invention is a dataset of spatial points obtained by scanning with a lidar device. Each point cloud contains coordinates and reflection intensity information. The reflection intensity information of the point cloud is related to the laser incident angle, laser wavelength, lidar energy density, and the surface material and roughness of the target (i.e., the object).

[0064] Step S102: Divide the point cloud data into multiple sliding window regions based on a preset sliding window, and determine the probability density of the point cloud data in each sliding window region.

[0065] like Figure 2 As shown, after the lidar device captures the point cloud data emitted by the target object, the point cloud data is projected onto the yz plane, and the resulting projection plane is divided into N×M grids. Multiple sliding window regions are obtained by moving a preset sliding window for subsequent probability density calculation. Specifically, the size of the sliding window can be determined according to the positioning accuracy requirements and the grid division method, see again... Figure 2 As shown, the preset sliding window size can be 3×3.

[0066] Step S103: Determine the region of interest based on the probability density of the point cloud data within each sliding window region.

[0067] Specifically, by using the probability density of point cloud data within a divided sliding window region, the region of interest (ROI) where the target (i.e. the object) is located is determined. Compared with the method of detecting the target through deep learning, this method can quickly filter out the ROI without large-scale training data, which is more efficient.

[0068] Step S104: Cluster the point cloud data within the region of interest to determine the set of center points of the target object.

[0069] Specifically, point cloud data within a region of interest can be clustered using a density-based clustering algorithm to filter out point cloud data reflected by the target (i.e., the object), thereby determining the set of center points of the target (i.e., the object) in geometric space.

[0070] Step S105: Based on the set of center points of the target object, determine the positioning result of the collaborative operation vehicle.

[0071] Specifically, after obtaining the set of center points of the target (i.e. the object) in geometric space, the center points in the set are further filtered to determine the target center point of the target (i.e. the object). Then, the collaborative operation vehicle is located based on the coordinates of the target center point to obtain the position information of the collaborative operation vehicle.

[0072] The method for locating collaborative work vehicles provided in this embodiment acquires point cloud data containing target objects on the collaborative work vehicle in real time. Based on a preset sliding window, the point cloud data is divided into multiple sliding window regions. The probability density of the point cloud data within each sliding window region is determined, thereby quickly filtering out the region of interest (ROI) where the target object is located. Then, the point cloud data within the ROI region is clustered to determine the set of center points of the target object. Based on this set of center points, the target center point is determined, and the collaborative work vehicle is located. This embodiment of the invention does not require training on large amounts of data, saving time and increasing efficiency. It is also less affected by data transmission errors between devices and exhibits strong robustness.

[0073] This embodiment provides a positioning method for collaborative work vehicles, which can be used in onboard computer equipment or electronic devices, such as vehicle controllers, onboard computers, etc. Figure 3 This is a flowchart of a positioning method for collaborative work vehicles according to an embodiment of the present invention, such as... Figure 3 As shown, the process includes the following steps:

[0074] Step S301: Acquire point cloud data containing target objects on the collaborative work vehicle in real time. For details, please refer to [link to relevant documentation]. Figure 1 Step S101 of the illustrated embodiment will not be described again here.

[0075] Step S302: Divide the point cloud data into multiple sliding window regions based on a preset sliding window, and determine the probability density of the point cloud data in each sliding window region.

[0076] For detailed steps in step S302 above, which involves dividing the point cloud data into multiple sliding window regions based on a preset sliding window, please refer to [link to relevant documentation]. Figure 1 Step S102 of the illustrated embodiment will not be described again here.

[0077] Specifically, the step of determining the probability density of the point cloud data within each sliding window region in step S302 above includes:

[0078] Step S3021: For each sliding window region, calculate the probability distribution parameters corresponding to the point cloud data within the sliding window region based on the pre-stored target object probability model.

[0079] Specifically, point cloud data of the target object can be pre-collected, and the spatial distribution characteristics of the point cloud can be characterized using a three-dimensional normal distribution model to obtain a probability model of the target object. Based on the pre-stored probability model of the target object, the probability distribution parameters (q, Σ) corresponding to the point cloud data within the sliding window area can be obtained. The formula for calculating the probability distribution parameters (q, Σ) is as follows:

[0080]

[0081]

[0082] Where q represents the mean of the point cloud data within the sliding window region, ∑ represents the covariance of the point cloud data within the sliding window region, k represents the number of point cloud data within the sliding window region, and p i This represents the coordinates of the i-th point cloud within the sliding window area.

[0083] Step S3022: For each sliding window region, calculate the probability density of the point cloud data within the sliding window region based on the probability distribution parameters corresponding to the point cloud data within the sliding window region.

[0084] In some optional implementations, the probability density of the point cloud data within the sliding window region can be calculated using the following formula:

[0085]

[0086] Among them, P ij X′ represents the probability density of the point cloud data within the sliding window region. k This represents the point cloud data within the sliding window area.

[0087] Therefore, by using a pre-stored target probability model to determine the probability distribution parameters of point cloud data within the sliding window area, the probability density of the point cloud data within the sliding window area can be obtained. This eliminates the need for real-time analysis of the spatial distribution characteristics of the point cloud data collected by the lidar, resulting in higher efficiency and shorter processing time.

[0088] Step S303: Determine the region of interest based on the probability density of the point cloud data within each sliding window region.

[0089] Specifically, step S303 includes:

[0090] Step S3031: Determine the maximum probability density based on the probability density corresponding to the point cloud data in each sliding window area.

[0091] Generally, the higher the probability density of the sliding window area, the greater the probability that the collaborative work vehicle will hit the target (i.e., the object) in that sliding window area.

[0092] Step S3032: Expand the sliding window region corresponding to the maximum probability density to obtain the region of interest.

[0093] It should be noted that the target (i.e., the object) may happen to be near the boundary of the sliding window region. In order to prevent the omission of the point cloud data of the target and affect the positioning accuracy, the sliding window region corresponding to the maximum probability density is appropriately expanded. For example, the sliding window region with a size of 3×3 is expanded to 6×6, and the expanded region is taken as the region of interest where the target (i.e., the object) is located.

[0094] Specifically, the embodiments of the present invention analyze the probabilistic features of point cloud data to quickly determine the region of interest where the target object is located, without the need to collect a large amount of data for annotation and training, thus achieving higher efficiency.

[0095] Step 304: Cluster the point cloud data within the region of interest to determine the set of center points of the target object. For details, please refer to [link to relevant documentation]. Figure 1 Step S104 of the illustrated embodiment will not be described again here.

[0096] Step 305: Based on the set of center points of the target object, determine the positioning results of the collaborative operation vehicle. For details, please refer to [link to details]. Figure 1 Step S105 of the illustrated embodiment will not be described again here.

[0097] The method for locating collaborative work vehicles provided in this embodiment acquires point cloud data containing target objects on the collaborative work vehicle in real time, calculates the probability density corresponding to the point cloud data within each sliding window region, and determines the maximum probability density, thereby quickly filtering out the region of interest where the target object is located. This eliminates the need for collecting large amounts of data for annotation and training, resulting in higher efficiency. Then, the point cloud data within the region of interest is clustered to locate the collaborative work vehicle. This embodiment of the invention eliminates the need for training on large amounts of data, saving time and increasing efficiency in locating collaborative work vehicles.

[0098] This embodiment provides a positioning method for collaborative work vehicles, which can be used in vehicle-mounted computer equipment or electronic devices, such as vehicle controllers, vehicle-mounted computers, etc. Figure 4 This is a flowchart of a positioning method for collaborative work vehicles according to an embodiment of the present invention, such as... Figure 4 As shown, the process includes the following steps:

[0099] Step S401: Acquire point cloud data containing target objects on the collaborative work vehicle in real time. For details, please refer to [link to relevant documentation]. Figure 3 Step S301 of the illustrated embodiment will not be described again here.

[0100] Step S402: Divide the point cloud data into multiple sliding window regions based on a preset sliding window, and determine the probability density of the point cloud data within each sliding window region. For details, please refer to [link to relevant documentation]. Figure 3 Step S302 of the illustrated embodiment will not be described again here.

[0101] Step S403: Determine the region of interest based on the probability density of the point cloud data within each sliding window region. For details, please refer to [link to relevant documentation]. Figure 3 Step S303 of the illustrated embodiment will not be described again here.

[0102] Step S404: Cluster the point cloud data within the region of interest to determine the set of center points of the target object.

[0103] Specifically, the point cloud data includes the reflection intensity of the point cloud, and step S404 above includes:

[0104] Step S4041: Based on the reflection intensity of the point cloud within the region of interest, filter the point cloud data within the region of interest to obtain the first point cloud set.

[0105] Specifically, point cloud data with reflection intensity lower than a preset reflection prior threshold can be filtered out based on the reflection intensity of the point cloud within the region of interest, thereby filtering out non-target object point clouds within the region of interest and obtaining the first point cloud set.

[0106] Step S4042: Cluster the first point cloud set according to the preset density clustering algorithm to obtain the second point cloud set.

[0107] Specifically, the preset density clustering algorithm can be the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm. The DBSCAN algorithm is a density-based spatial clustering algorithm that is robust to noise. It can find all dense regions of sample points and treat these dense regions as clusters.

[0108] In some optional implementations, step S4042 above includes:

[0109] Step a1, for point cloud p in the first point cloud set i = (x, y, z), which determines the minimum number of points ρ and radius r for the density clustering algorithm, where the radius r can be calculated using the Manhattan distance metric.

[0110] Step a2: Starting from an unvisited point in the first point cloud set, take that point as the center and count the number of points n contained within a circle of radius r. If n ≥ ρ, then the point is marked as the center point; otherwise, it is marked as a noise point.

[0111] Step a3: Repeat step a2 until all points in the first point cloud set have been visited; wherein, if a noise point exists inside a circle with radius r in a point cloud, the noise point is marked as an edge point, otherwise it is still a noise point.

[0112] Therefore, the points in the first point cloud set are clustered by a preset density clustering algorithm, and the centroids of the multiple clusters are used as the points in the second point cloud set to further filter out the point clouds that reflect the target (i.e. the object).

[0113] Step S4043: Filter the second point cloud set according to the pre-stored target object geometric distribution model to determine the center point set of the target object.

[0114] Specifically, the geometric distribution model of the target object can be obtained in advance based on its shape and spatial structure, thereby further filtering the point cloud in the second point cloud set to obtain the set of center points of the target object.

[0115] Figure 5 This is a schematic diagram of the clustering process according to an embodiment of the present invention, such as... Figure 5 As shown, firstly, the point cloud within the region of interest is filtered based on reflection intensity to obtain high reflectivity points (i.e., the first point cloud set). Then, the high reflectivity points (i.e., the first point cloud set) are clustered to obtain cluster points (i.e., the second point cloud set). Finally, the cluster points are further filtered based on the spatial geometric distribution relationship of the target object to obtain the target points (i.e., the set of center points of the target object). This allows for filtering and selection of point cloud data within the region of interest, further determining which points might be the center points of the target object, thus improving the positioning accuracy of the positioning method.

[0116] Step S405: Based on the set of center points of the target object, determine the positioning result of the collaborative operation vehicle.

[0117] Specifically, step S405 includes:

[0118] Step S4051: Filter the set of center points to obtain the target center point.

[0119] To determine the unique center point of the target object and improve positioning accuracy, it is necessary to further filter the points in the set of center points to obtain the target center point.

[0120] In some optional implementations, step S4051 above includes:

[0121] Step b1: Based on the point cloud coordinates of each center point in the center point set, calculate the Euclidean distance between the coordinates of each center point at the current time and the coordinates at the previous time.

[0122] Step b2: Determine the minimum Euclidean distance based on the Euclidean distance of each center point, and take the center point corresponding to the minimum Euclidean distance as the target center point.

[0123] Figure 6 This is a schematic diagram of the motion relationship of the target object according to an embodiment of the present invention, such as... Figure 6 As shown, the change in the coordinates of the center point at the current moment relative to its coordinates at the previous moment is minimal. Therefore, the target center point can be determined by combining the motion information of the target object. Compared with directly using GPS data and radar / laser point cloud data, this eliminates errors that may be caused by sensors or the external environment, further improving the positioning accuracy of the positioning method.

[0124] Specifically, the target center point can be obtained using the following formula:

[0125]

[0126] Where, q sl q represents the GPS coordinates of the center point of the target object. k This represents the GPS coordinates of the center point at time k. Let represent the pose transformation matrix at time k.

[0127] It should be noted that the current vehicle attitude information in the GPS coordinate system can be obtained based on the timestamp information of the GPS and LiDAR devices, thus obtaining the pose transformation matrix of the target center point from the GPS coordinate system to the world coordinate system.

[0128] Step S4052: Perform coordinate transformation on the target center point to determine the world coordinates of the target center point and obtain the positioning result of the collaborative operation vehicle.

[0129] Specifically, the world coordinates of the target center point can be obtained by performing a coordinate transformation using the following formula:

[0130]

[0131] Where, q w The world coordinates of the center point of the target object. This represents the pose transformation matrix.

[0132] Therefore, by calculating the world coordinates of the target center point, the position information of the collaborative operation vehicle can be obtained, and the collaborative operation vehicle can be positioned with high precision, thereby determining the relative positional relationship between the vehicle and the collaborative operation vehicle and ensuring the smooth progress of collaborative operation.

[0133] The positioning method for collaborative work vehicles provided in this embodiment acquires point cloud data containing target objects on the collaborative work vehicle in real time, calculates the probability density corresponding to the point cloud data in each sliding window area, and determines the maximum probability density, thereby quickly filtering out the region of interest where the target object is located. This eliminates the need to collect a large amount of data for annotation and training, resulting in higher efficiency. Then, the point cloud data in the region of interest is clustered, and the set of center points of the target object is filtered out based on the point cloud reflection intensity. Finally, the target center point is determined based on the motion information of the target object, avoiding errors caused by sensors or the external environment, so as to achieve high-precision positioning of collaborative work vehicles.

[0134] The positioning method for collaborative work vehicles in this invention will be described in detail below with reference to a specific embodiment. This embodiment specifically includes the following steps:

[0135] Step c1: Collect point cloud data of the target object using a lidar device and construct a three-dimensional normal distribution model of the target object to obtain the probability distribution parameters (q, Σ) of the target object.

[0136] Step c2: Divide the point cloud data of the target object into N×M grids in the parallel projection direction (yz plane), and calculate the probability density of the point cloud data in each sliding window region based on a 3×3 sliding window.

[0137] Step c3: Select the sliding window region with the highest probability density as the region of interest, and perform density-based clustering on the point cloud within the region of interest to obtain the set of center points of the target object.

[0138] Step c4: Based on the relative motion of the target object relative to the previous moment, filter the set of center points to obtain the target center point, and locate the collaborative operation vehicle.

[0139] Therefore, by calculating the probability density of the point cloud data within each sliding window region, the maximum probability density is determined, thereby quickly filtering out the region of interest where the target object is located. This eliminates the need to collect a large amount of data for annotation and training, resulting in higher efficiency. Then, density-based clustering is performed on the point cloud data within the region of interest. Based on the point cloud reflection intensity and the geometric features of the target object, the set of center points of the target object is filtered out. Finally, the target center point is determined based on the motion information of the target object, eliminating errors caused by sensors or the external environment, and enabling high-precision positioning of collaborative operation vehicles.

[0140] This embodiment also provides a positioning device for collaborative work vehicles, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, hardware implementations, or a combination of software and hardware, are also possible and contemplated.

[0141] This embodiment provides a positioning device for collaborative work vehicles, such as... Figure 7 As shown, it includes:

[0142] The acquisition module 701 is used to acquire point cloud data containing target objects on the collaborative operation vehicle in real time;

[0143] The first processing module 702 is used to divide point cloud data into multiple sliding window regions based on a preset sliding window, and determine the probability density of point cloud data in each sliding window region.

[0144] The second processing module 703 is used to determine the region of interest based on the probability density of the point cloud data within each sliding window region.

[0145] The third processing module 704 is used to cluster the point cloud data within the region of interest to determine the set of center points of the target object;

[0146] The fourth processing module 705 is used to determine the positioning result of the collaborative operation vehicle based on the set of center points of the target object.

[0147] In some alternative implementations, the first processing module 702 includes:

[0148] The first processing unit is used to calculate the probability distribution parameters of the point cloud data within each sliding window area based on the pre-stored target object probability model.

[0149] The second processing unit is used to calculate the probability density of the point cloud data in each sliding window region based on the probability distribution parameters of the point cloud data in the sliding window region.

[0150] In some optional implementations, the second processing unit is used to calculate the probability density of the point cloud data within the sliding window region according to the following formula:

[0151]

[0152] Among them, P ij X′ represents the probability density of the point cloud data within the sliding window region. kLet represent the point cloud data within the sliding window region, k represent the number of point cloud data within the sliding window region, q represent the mean of the point cloud data within the sliding window region, and ∑ represent the covariance of the point cloud data within the sliding window region.

[0153] In some alternative implementations, the second processing module 703 includes:

[0154] The third processing unit is used to determine the maximum probability density based on the probability density corresponding to the point cloud data in each sliding window area;

[0155] The fourth processing unit is used to amplify the sliding window region corresponding to the maximum probability density to obtain the region of interest.

[0156] In some alternative implementations, the third processing module 704 includes:

[0157] The fifth processing unit is used to filter the point cloud data within the region of interest based on the reflection intensity of the point cloud within the region of interest, and obtain the first point cloud set.

[0158] The sixth processing unit is used to cluster the first point cloud set according to a preset density clustering algorithm to obtain the second point cloud set;

[0159] The seventh processing unit is used to filter the second point cloud set according to the pre-stored geometric distribution model of the target object, and determine the set of center points of the target object.

[0160] In some alternative implementations, the fourth processing module 705 includes:

[0161] The eighth processing unit is used to filter the set of center points to obtain the target center point;

[0162] The ninth processing unit is used to perform coordinate transformation on the target center point, determine the world coordinates of the target center point, and obtain the positioning result of the collaborative operation vehicle.

[0163] In some optional implementations, the eighth processing unit includes:

[0164] The first processing subunit is used to calculate the Euclidean distance between the coordinates of each center point at the current time and the coordinates at the previous time, based on the point cloud coordinates of each center point in the center point set.

[0165] The second processing subunit is used to determine the minimum Euclidean distance based on the Euclidean distance between each center point, and to take the center point corresponding to the minimum Euclidean distance as the target center point.

[0166] Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0167] In this embodiment, the positioning device for the collaborative work vehicle is presented in the form of a functional unit. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0168] This invention also provides a computer device having the above-described features. Figure 7 The positioning device for the collaborative work vehicle shown.

[0169] Please see Figure 8 , Figure 8 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 8 As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 8 Take a processor 10 as an example.

[0170] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0171] The memory 20 stores instructions executable by at least one processor 10 to cause the at least one processor 10 to perform the method shown in the above embodiments.

[0172] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0173] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0174] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0175] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0176] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A positioning method for collaborative work vehicles, characterized in that, The method includes: Real-time acquisition of point cloud data containing target objects on collaborative work vehicles; The point cloud data is divided into multiple sliding window regions based on a preset sliding window, and the probability density of the point cloud data in each sliding window region is determined. The region of interest is determined based on the probability density of the point cloud data within each sliding window region; Cluster the point cloud data within the region of interest to determine the set of center points of the target object; Based on the set of center points of the target object, the positioning result of the collaborative operation vehicle is determined; Determining the probability density of point cloud data within each sliding window region includes: For each sliding window region, the probability distribution parameters corresponding to the point cloud data within the sliding window region are calculated based on the pre-stored target object probability model; For each sliding window region, the probability density of the point cloud data within the sliding window region is calculated based on the probability distribution parameters of the point cloud data within the sliding window region. The point cloud data includes the reflection intensity of the point cloud. The step of clustering the point cloud data within the region of interest to determine the set of center points of the target object includes: Based on the reflection intensity of the point cloud within the region of interest, the point cloud data within the region of interest is filtered to obtain a first set of point clouds; The first point cloud set is clustered according to a preset density clustering algorithm to obtain the second point cloud set; The second point cloud set is filtered based on the pre-stored geometric distribution model of the target object to determine the set of center points of the target object.

2. The positioning method for collaborative work vehicles according to claim 1, characterized in that, The formula for calculating the probability density of the point cloud data within the sliding window area is as follows: in, This represents the probability density of the point cloud data within the sliding window area. This represents the point cloud data within the sliding window area. k This indicates the number of point cloud data points within the sliding window area. q This represents the mean value of the point cloud data within the sliding window area. This represents the covariance of the point cloud data within the sliding window area.

3. The positioning method for collaborative work vehicles according to claim 1, characterized in that, The determination of the positioning result of the collaborative operation vehicle based on the set of center points of the target object includes: The target center point is obtained by filtering the set of center points; The target center point is transformed to determine its world coordinates, thus obtaining the positioning result of the collaborative operation vehicle.

4. The positioning method for collaborative operation vehicles according to claim 3, characterized in that, The step of filtering the set of center points to obtain the target center point includes: Based on the point cloud coordinates of each center point in the set of center points, calculate the Euclidean distance of each center point's coordinates at the current time relative to its coordinates at the previous time. Based on the Euclidean distances of each center point, determine the minimum Euclidean distance and take the center point corresponding to the minimum Euclidean distance as the target center point.

5. The positioning method for collaborative work vehicles according to claim 1, characterized in that, The step of determining the region of interest based on the probability density of the point cloud data within each sliding window region includes: The maximum probability density is determined based on the probability density of the point cloud data within each sliding window region; The region of interest is obtained by amplifying the sliding window region corresponding to the maximum probability density.

6. A positioning device for a collaborative work vehicle, used to implement the positioning method for a collaborative work vehicle according to any one of claims 1-5, characterized in that, The device includes: The acquisition module is used to acquire point cloud data containing target objects on collaborative operation vehicles in real time; The first processing module is used to divide the point cloud data into multiple sliding window regions based on a preset sliding window, and determine the probability density of the point cloud data in each sliding window region. The second processing module is used to determine the region of interest based on the probability density of the point cloud data within each sliding window area; The third processing module is used to cluster the point cloud data within the region of interest to determine the set of center points of the target object; The fourth processing module is used to determine the positioning result of the collaborative operation vehicle based on the set of center points of the target object.

7. A computer device, characterized in that, include: A memory and a processor are communicatively connected, the memory storing computer instructions, and the processor executing the computer instructions to perform the positioning method for the cooperative working vehicle as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the positioning method of the cooperative work vehicle according to any one of claims 1 to 5.

Images