A point cloud data denoising method and system based on intelligent driving

Abstract

The present invention relates to the field of unmanned driving technology and provides a point cloud data denoising method and system based on intelligent driving. The present invention proposes to use the gray wolf algorithm to eliminate the noise of the scan factor α in the walking tour algorithm. i,t During the optimization process, different α values are used in the process of local optimization and jumping out of the local optimal point. i,t The optimization was carried out to form a walking combination algorithm, and then the walking combination algorithm was used to optimize the DBSCAN clustering algorithm of the three-dimensional point cloud to obtain the optimal neighborhood radius Eps parameter and the optimal minimum number of points MinPts parameter. The three-dimensional point cloud data of all marked as noise points were corrected through the DBSCAN algorithm, thereby eliminating the noise and improving the recognition accuracy of the three-dimensional point cloud.

Description

Technical Field

[0001] The present invention relates to the field of unmanned driving technology, and in particular to a point cloud data denoising method and system based on intelligent driving. Background Art

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

[0003] Intelligent driving technology has demonstrated widespread application value in transportation, construction machinery, and agriculture. Through unmanned operation, intelligent path planning, and precise operations, it significantly improves operational efficiency, safety, and accuracy. In transportation, intelligent driving improves driving efficiency, ensures safety, and reduces traffic accidents. For construction machinery, unmanned operation reduces personnel risks and improves operational capabilities. For tractors, intelligent agricultural machinery enables precision farming and improves crop efficiency.

[0004] In the field of intelligent driving, LiDAR sensors are one of the core devices for acquiring three-dimensional data of the surrounding environment. LiDAR generates three-dimensional point cloud data by emitting laser beams and receiving return signals. This data accurately reflects information such as the distance, shape, and position of objects, providing critical support for the vehicle's environmental perception. LiDAR-based three-dimensional point cloud data has important applications in object detection, road recognition, and obstacle avoidance. It enables intelligent driving systems to more accurately identify and locate surrounding objects, enabling real-time decision-making. However, in practical applications, 3D point cloud data often contains a large amount of noise. This noise is typically caused by factors such as environmental complexity (such as rain, snow, and fog), sensor accuracy limitations, and physical interference. The presence of noise can reduce the accuracy of intelligent driving systems in recognizing environmental features, leading to problems such as misidentification and missed identification, seriously affecting system stability and safety. Currently, the DBSCAN algorithm, a popular clustering algorithm, is commonly used for noise reduction. However, without optimization of the DBSCAN algorithm, clustering results are poor and robust. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the present invention provides a point cloud data denoising method and system based on intelligent driving. By employing a pedestrian combination algorithm, this method can both find local optimal solutions and transcend local optimal solutions to obtain a global optimal solution. Furthermore, using the optimal solution obtained by the pedestrian combination algorithm to optimize the parameters of the DBSCAN clustering algorithm can enhance the characteristics of point cloud clustering, remove more noise, and improve intelligent driving recognition performance.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] The first aspect of the present invention provides a point cloud data denoising method based on intelligent driving.

[0008] A point cloud data denoising method based on intelligent driving, comprising:

[0009] Segmenting the acquired three-dimensional point cloud data of the target to be segmented in intelligent driving;

[0010] The walking combination algorithm is used to optimize the parameters of the DBSCAN algorithm to obtain the optimal neighborhood radius Eps parameter and the optimal minimum number of points MinPts parameter, so as to cluster the segmented 3D point cloud data, mark the point cloud clusters and noise points, and obtain the 3D point cloud data after removing the noise points.

[0011] The walking combination algorithm includes: in the process of optimizing the walking tour algorithm, using the gray wolf algorithm to improve the scanning factor in the speed update formula of the hiker in the tour algorithm When the number of consecutive times that the hiker's position change is less than the first set value meets the first threshold, the first optimal solution and the number of optimization times are recorded, and a jump process is performed. The jump scan factor is When the number of optimization attempts is less than or equal to the second threshold, the scan factor is used. Continue the optimization iteration. When the number of consecutive times that the hiker's position change is less than the first set value meets the first threshold, record the second optimal solution, update the optimization number, select the best of the first optimal position and the second optimal position, repeat the above optimization process, and output the best solution when the optimization number is greater than the second threshold.

[0012] Furthermore, the scan factor It is expressed by the following formula:

[0013]

[0014] in, Represent the positions of α wolf, β wolf, and δ wolf respectively. α wolf represents the optimal solution, β wolf and δ wolf represent the second-best solution and the third-best solution respectively. represents the position of the gray wolf after iteration t+1, They represent the corresponding movement distances of α wolf, β wolf, and δ wolf according to the estimated position of the prey, They represent the distances between the current search agent and wolf α, wolf β, and wolf δ respectively. α, β, and δ represent the degree of knowledge of wolf α, wolf β, and wolf δ about the location of the prey, i.e., the weights, α+β+δ=1.

[0015] Furthermore, the position change of the hiker is expressed by the following formula:

[0016]

[0017] Among them, Z(t+1) represents the position change of the hiker, β i,t+1 represents the position of the hiker after t+1 iterations, β i,t represents the position of the hiker after t iterations.

[0018] Furthermore, the scan factor It is expressed by the following formula:

[0019]

[0020] in, The range is between [3,9], and the optimization iteration is performed. i,t+1 represents the position of the hiker after t+1 iterations, β i,t represents the position of the hiker after t iterations, Indicates the optimal position of the wolf each time.

[0021] Furthermore, the location update of the hiker in the hiking tour algorithm is implemented by the following process:

[0022] The initial slope formula of the hiking algorithm is:

[0023]

[0024] Where dh and dx represent the hiker’s height difference and travel distance difference, respectively, θ i,t is the slope angle of the path or terrain, S i,t represents the slope of the path or terrain; the hiker's initial speed is given by the Tobler hiking function:

[0025]

[0026] Where W i,t represents the speed of the initial hiker during the iteration time t, in km / h;

[0027] During the search process, the speed of the hiker is determined by the initial speed, the position of the leading hiker, the actual position of the hiker, and the scanning factor. Therefore, the current speed of the i-th hiker is:

[0028] W i,t =W i,t-1 +γ i,t (β best -α i,t β i,t )

[0029] Among them, γ i,t is a random number uniformly distributed between 0 and 1, α i,t is the scanning factor of hiker i, β i,tis the current position of hiker i, β best It is the position of the lead hiker, W i,t represents the current speed of hiker i, W i,t-1 represents the speed of hiker i at the last moment; according to the hiker's speed, the new position update of hiker i should be:

[0030] β i,t+1 =β i,t +W i,t

[0031] Where, β i,t+1 is the updated position of hiker i. The hiker will continuously update its position to reach the leader's position.

[0032] Furthermore, the method of clustering the segmented three-dimensional point cloud data, marking point cloud clusters and noise points, and obtaining three-dimensional point cloud data after removing noise points includes:

[0033] Step (1): Mark the cluster labels of all point clouds as -1, indicating noise points;

[0034] Step (2): Use the walking combination algorithm to obtain the optimal neighborhood radius Eps parameter and the optimal minimum number of points MinPts parameter;

[0035] Step (3): Check one of the points A to determine whether it is a core point. If the number of points in the neighborhood exceeds the threshold MinPts, point A is the core point, and all points in the neighborhood Eps of point A are added to the cluster.

[0036] Step (4): Mark the points that are neither core points nor in the neighborhood of any core points as noise points;

[0037] Step (5): Repeat steps (2), (3), and (4) for each untraversed point in the cluster until all points in the cluster are processed, and obtain the three-dimensional point cloud data after removing the noise points.

[0038] The second aspect of the present invention provides a point cloud data denoising system based on intelligent driving.

[0039] A point cloud data denoising system based on intelligent driving, comprising:

[0040] A preprocessing module is configured to: segment the acquired three-dimensional point cloud data of the target to be segmented in intelligent driving;

[0041] The point cloud clustering module is configured to: use the walking combination algorithm to optimize the parameters of the DBSCAN algorithm to obtain the optimal neighborhood radius Eps parameter and the optimal minimum number of points MinPts parameter to cluster the segmented 3D point cloud data, mark point cloud clusters and noise points, and obtain 3D point cloud data after removing noise points;

[0042] The walking combination algorithm includes: in the process of optimizing the walking tour algorithm, using the gray wolf algorithm to improve the scanning factor in the speed update formula of the hiker in the tour algorithm When the number of consecutive times that the hiker's position change is less than the first set value meets the first threshold, the first optimal solution and the number of optimization times are recorded, and a jump process is performed. The jump scan factor is When the number of optimization attempts is less than or equal to the second threshold, the scan factor is used. Continue the optimization iteration. When the number of consecutive times that the hiker's position change is less than the first set value meets the first threshold, record the second optimal solution, update the optimization number, select the best of the first optimal position and the second optimal position, repeat the above optimization process, and output the best solution when the optimization number is greater than the second threshold.

[0043] A third aspect of the present invention provides a computer-readable storage medium.

[0044] A computer-readable storage medium stores a program thereon, which, when executed by a processor, implements the steps of the point cloud data denoising method based on intelligent driving described in the first aspect.

[0045] A fourth aspect of the present invention provides an electronic device.

[0046] An electronic device includes a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps of the point cloud data denoising method based on intelligent driving described in the first aspect are completed.

[0047] A fifth aspect of the present invention provides a computer program product or computer program.

[0048] The present invention provides a computer program product or computer program, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps of the method for point cloud data denoising based on intelligent driving as described in the first aspect above.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] 1. The present invention provides a point cloud data denoising method and system based on intelligent driving, and proposes to use the gray wolf algorithm to use the scanning factor α in the hiking algorithm. i,j (t) During the optimization process, α i,j (t) Using two different formulas, a hiking combination algorithm is obtained. This algorithm can find the optimal point locally and can jump out of the optimal point to a certain extent, so that the algorithm has the ability to jump out of the local optimal point when processing complex, multi-peak and other three-dimensional point cloud data problems.

[0051] 2. The present invention provides a point cloud data denoising method and system based on intelligent driving. It uses a walking combination algorithm to optimize the DBSCAN clustering algorithm of the three-dimensional point cloud to obtain the optimal neighborhood radius Eps parameter and the optimal minimum number of points MinPts parameter. The three-dimensional point cloud data of all marked as noise points are corrected through the DBSCAN algorithm, thereby eliminating noise and improving the recognition accuracy of the three-dimensional point cloud.

[0052] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0054] Figure 1 The present invention is a flowchart of a point cloud data denoising method based on intelligent driving according to an embodiment. DETAILED DESCRIPTION

[0055] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0056] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0057] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0058] In the absence of conflict, the embodiments and features in the embodiments of this application can be combined with each other.

[0059] Example 1:

[0060] Embodiment 1 of the present invention provides a method for denoising point cloud data based on intelligent driving. The method in this embodiment includes the following steps:

[0061] Segmenting the acquired three-dimensional point cloud data of the target to be segmented in intelligent driving;

[0062] The walking combination algorithm is used to optimize the parameters of the DBSCAN algorithm to obtain the optimal neighborhood radius Eps parameter and the optimal minimum number of points MinPts parameter, so as to cluster the segmented 3D point cloud data, mark the point cloud clusters and noise points, and obtain the 3D point cloud data after removing the noise points.

[0063] The walking combination algorithm includes: in the process of optimizing the walking tour algorithm, using the gray wolf algorithm to improve the scanning factor in the speed update formula of the hiker in the tour algorithm When the number of consecutive times that the hiker's position change is less than the first set value meets the first threshold, the first optimal solution and the number of optimization times are recorded, and a jump process is performed. The jump scan factor is When the number of optimization attempts is less than or equal to the second threshold, the scan factor is used. Continue the optimization iteration. When the number of consecutive times that the hiker's position change is less than the first set value meets the first threshold, record the second optimal solution, update the optimization number, select the best of the first optimal position and the second optimal position, repeat the above optimization process, and output the best solution when the optimization number is greater than the second threshold.

[0064] Among them, the three-dimensional point cloud data is mapped into the three-dimensional point cloud data after the laser radar point cloud data collected by intelligent driving is segmented.

[0065] After the 3D point cloud data is segmented, the segmented data is converted into additional information of the point cloud to form point cloud clusters of all target instances.

[0066] The present embodiment is described in detail below with reference to the accompanying drawings:

[0067] like Figure 1As shown in the figure, the distance-based Euclidean clustering algorithm and the density-based DBSCAN algorithm are commonly used when clustering point cloud data. The density-based DBSCAN algorithm divides areas with sufficient density into clusters and can find clusters of any shape. The DBSCAN algorithm is controlled by the neighborhood radius Eps and the minimum number of points MinPts. Data points are divided into three categories in the DBSCAN algorithm: core points, which are points with more than MinPts neighbors within the radius Eps; boundary points, which are boundary points if they are within the radius Eps of the core point and are not core points themselves; and noise points, which are noise points if they are outside the radius Eps of the core point. The algorithm flow is as follows: extract a point A from the point set. If the number of points in the neighborhood exceeds the threshold MinPts, point A is the core point, and all points in the neighborhood Eps of point A are added to the cluster; repeat the first step for each untraversed point in the cluster until all points in the cluster have been processed; find the points in the point cloud data that are not assigned to the cluster and perform the first two steps until all points are traversed.

[0068] In order to make the point cloud clustering feature of the DBSCAN algorithm more obvious and remove more noise, the walking combination algorithm is used to optimize the neighborhood radius Eps and the minimum number of points MinPts in the DBSCAN algorithm. Specifically:

[0069] Use random numbers to randomly generate the initial positions of multiple hikers in space. For the i-th hiker, its initial position is expressed as:

[0070]

[0071] Where N represents the number of individuals, is the lower bound of the j-th dimension, is the upper bound of the j-th dimension, and λ is a randomly initialized number between 0 and 1.

[0072] The slope angle is initialized using the probability density function of the Weibull distribution:

[0073]

[0074] Here, x is a random variable, which is the initial slope angle faced by the hiker and is a random number between 0° and 50°. m>0 is the scale parameter, and η>0 is the shape parameter, which determines the shape of the distribution.

[0075] The initial slope formula of the hiking algorithm is:

[0076]

[0077] Where dh and dx represent the hiker’s height difference and travel distance difference, respectively, θ i,t is the slope angle of the path or terrain, S i,t represents the slope of the path or terrain. The initial speed of the hiker is given by the Tobler hiking function:

[0078]

[0079] Where W i,t It represents the speed of the initial hiker in iteration number t, in km / h.

[0080] During the search process, the speed of the hiker is determined by the initial speed, the position of the leading hiker, the actual position of the hiker, and the scanning factor. Therefore, the current speed of the i-th hiker is:

[0081] W i,t =W i,t-1 +γ i,t (β best -α i,t β i,t )

[0082] Among them, γ i,t is a random number uniformly distributed between 0 and 1, α i,t is the scan factor of hiker i, which ranges from [1,3]. The scan factor α i,t It can ensure that hiker i does not stray too far from the lead hiker so that he can see the direction of the lead hiker and receive the signal from the lead hiker, β i,t is the current position of hiker i, β best It is the position of the lead hiker, W i,t represents the current speed of hiker i, W i,t-1 represents the speed of hiker i at the last moment. According to the hiker’s speed, the new position of hiker i is updated as:

[0083] β i,t+1 =β i,t +W i,t

[0084] Where, β i,t+1 is the updated position of hiker i. The hiker will continuously update its position to reach the leader's position.

[0085] When hikers explore and develop, they will be affected by the scanning factor α i,t This factor greatly affects the distance between the leader and other hikers. i,t As the range increases, hikers tend to be more inclined to the development stage, that is, to find the optimal solution in the world; on the contrary, when α i,tWhen the range is reduced, hikers will be more inclined to the exploration phase, which helps to accurately search the current local area. The gray wolf algorithm is used to improve it, specifically:

[0086]

[0087] in, Represent the positions of α wolf, β wolf, and δ wolf respectively. α wolf represents the optimal solution, β wolf and δ wolf represent the second-best solution and the third-best solution respectively. represents the optimal position of the gray wolf after iteration t+1, They represent the corresponding movement distances of α wolf, β wolf, and δ wolf according to the estimated position of the prey, They represent the distances between the current search agent and wolf α, wolf β, and wolf δ respectively. α, β, and δ represent the degree of knowledge of wolf α, wolf β, and wolf δ about the location of the prey, i.e., the weights, α+β+δ=1.

[0088] During the search iteration, will continue to decrease, making hikers tend to explore and search locally. Z(t+1) represents the position change of the hiker, β i,t+1 represents the position of the hiker after t+1 iterations, β i,t It represents the position of the hiker after t iterations. n The first set value may be 1% when Z(t+1) < the first set value. n To represent the first threshold, it can be 25 to 40 times, indicating that the hiker's position changes very little. Jump out of this stage and record the position at this time as The scan factor at this time is jumped out, and the scan factor at this time is

[0089]

[0090] Indicates the optimal position of the wolf each time. The range is between [3,9], and the optimization iteration is performed. During the iteration process, Perform optimization, when continuous T n The first set value may be 1% when Z(t+1) < the first set value. n Represents the first threshold, which can be 25 to 40 times. Compare the optimal solution of the second optimization with the optimal solution of the first optimization, select the best one, and record the number of optimization times M. n If M n >T nn , T nnIndicates the second threshold, which can be 5 to 10 times or when the number of iterations reaches the maximum. When the second threshold condition is met, the iteration stops and the optimal position is obtained. That is, the optimal parameters of the DBSCAN algorithm are obtained.

[0091] If M n ≤T nn , then use Continue to perform optimization iteration and repeat the above process.

[0092] Specifically, the DBSCAN algorithm is used to cluster the segmented point cloud data. The process of removing noise points includes:

[0093] (1) The cluster labels of all points are marked as -1, indicating that they are noise points;

[0094] (2) Use the walking combination algorithm to obtain the optimal neighborhood radius Eps parameter and the optimal minimum number of points MinPts parameter;

[0095] (3) Check one of the points A to determine whether it is a core point. If the number of points in the neighborhood exceeds the threshold MinPts, point A is the core point, and all points in the neighborhood Eps of point A are added to the cluster;

[0096] (4) For those points that are neither core points nor in the neighborhood of any core point, they are marked as noise points;

[0097] (5) Repeat steps (2), (3), and (4) for each untraversed point in the cluster until all points in the cluster have been processed and the marked noise points are removed.

[0098] Finally, a three-dimensional point cloud is obtained after removing noise points.

[0099] Example 2:

[0100] A second embodiment of the present invention provides a point cloud data denoising system based on intelligent driving, comprising:

[0101] A preprocessing module is configured to: segment the acquired three-dimensional point cloud data of the target to be segmented in intelligent driving;

[0102] The point cloud clustering module is configured to: use the walking combination algorithm to optimize the parameters of the DBSCAN algorithm to obtain the optimal neighborhood radius Eps parameter and the optimal minimum number of points MinPts parameter to cluster the segmented 3D point cloud data, mark point cloud clusters and noise points, and obtain 3D point cloud data after removing noise points;

[0103] The walking combination algorithm includes: in the process of optimizing the walking tour algorithm, using the gray wolf algorithm to improve the scanning factor in the speed update formula of the hiker in the tour algorithm When the number of consecutive times that the hiker's position change is less than the first set value meets the first threshold, the first optimal solution and the number of optimization times are recorded, and a jump process is performed. The jump scan factor is When the number of optimization attempts is less than or equal to the second threshold, the scan factor is used. Continue the optimization iteration. When the number of consecutive times that the hiker's position change is less than the first set value meets the first threshold, record the second optimal solution, update the optimization number, select the best of the first optimal position and the second optimal position, repeat the above optimization process, and output the best solution when the optimization number is greater than the second threshold.

[0104] In some embodiments, the scan factor It is expressed by the following formula:

[0105]

[0106]

[0107] in, Represent the positions of α wolf, β wolf, and δ wolf respectively. α wolf represents the optimal solution, β wolf and δ wolf represent the second-best solution and the third-best solution respectively. represents the optimal position of the gray wolf after iteration t+1, They represent the corresponding movement distances of α wolf, β wolf, and δ wolf according to the estimated position of the prey, They represent the distances between the current search agent and wolf α, wolf β, and wolf δ respectively. α, β, and δ represent the degree of knowledge of wolf α, wolf β, and wolf δ about the location of the prey, i.e., the weights, α+β+δ=1.

[0108] In some embodiments, the change in the hiker's position is expressed using the following formula:

[0109]

[0110] Among them, Z(t+1) represents the position change of the hiker, β i,t+1 represents the position of the hiker after t+1 iterations, β i,t represents the position of the hiker after t iterations.

[0111] In some embodiments, the scan factor It is expressed by the following formula:

[0112]

[0113] in, The range is between [3,9], and the optimization iteration is performed. i,t+1 represents the position of the hiker after t+1 iterations, β i,t represents the position of the hiker after t iterations, Indicates the optimal position of the wolf each time.

[0114] In some embodiments, the point cloud clustering module is further configured to: update the location of the hiker in the hiking algorithm by using the following process:

[0115] The initial slope formula of the hiking algorithm is:

[0116]

[0117] Where dh and dx represent the hiker’s height difference and travel distance difference, respectively, θ i,t is the slope angle of the path or terrain, S i,t represents the slope of the path or terrain; the hiker's initial speed is given by the Tobler hiking function:

[0118]

[0119] Where W i,t represents the speed of the initial hiker during the iteration time t, in km / h;

[0120] During the search process, the speed of the hiker is determined by the initial speed, the position of the leading hiker, the actual position of the hiker, and the scanning factor. Therefore, the current speed of the i-th hiker is:

[0121] W i,t =W i,t-1 +γ i,t (β best -α i,t β i,t )

[0122] Among them, γ i,t is a random number uniformly distributed between 0 and 1, α i,t is the scanning factor of hiker i, β i,t is the current position of hiker i, β best It is the position of the lead hiker, W i,t represents the current speed of hiker i, W i,t-1 represents the speed of hiker i at the last moment; according to the hiker's speed, the new position update of hiker i should be:

[0123] β i,t+1 =β i,t +W i,t

[0124] Where, β i,t+1 is the updated position of hiker i. The hiker will continuously update its position to reach the leader's position.

[0125] In some embodiments, the point cloud clustering module is further configured to:

[0126] Step (1): Mark the cluster labels of all point clouds as -1, indicating noise points;

[0127] Step (2): Use the walking combination algorithm to obtain the optimal neighborhood radius Eps parameter and the optimal minimum number of points MinPts parameter;

[0128] Step (3): Check one of the points A to determine whether it is a core point. If the number of points in the neighborhood exceeds the threshold MinPts, point A is the core point, and all points in the neighborhood Eps of point A are added to the cluster.

[0129] Step (4): Mark the points that are neither core points nor in the neighborhood of any core points as noise points;

[0130] Step (5): Repeat steps (2), (3), and (4) for each untraversed point in the cluster until all points in the cluster are processed, and obtain the three-dimensional point cloud data after removing the noise points.

[0131] Example 3

[0132] Embodiment 3 of the present invention provides a computer-readable storage medium having a program stored thereon. When the program is executed by a processor, the steps in the point cloud data denoising method system based on intelligent driving as described in embodiment 1 of the present invention are implemented.

[0133] Example 4

[0134] Embodiment 4 of the present invention provides an electronic device, including a memory and a processor, and computer instructions stored in the memory and running on the processor. When the computer instructions are run by the processor, the steps in the point cloud data denoising method system based on intelligent driving described in embodiment 1 of the present invention are implemented.

[0135] Example 5

[0136] This embodiment provides a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps of the point cloud data denoising method system based on intelligent driving described in Example 1.

[0137] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0138] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0139] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0140] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0141] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made by those skilled in the art that fall within the spirit and principles of the present invention and do not require creative effort are intended to be within the scope of protection of the present invention.

Claims

1. A point cloud data denoising method based on intelligent driving, characterized in that: include: Segmenting the acquired three-dimensional point cloud data of the target to be segmented in intelligent driving; The walking combination algorithm is used to optimize the parameters of the DBSCAN algorithm to obtain the optimal neighborhood radius Eps parameter and the optimal minimum number of points MinPts parameter, so as to cluster the segmented 3D point cloud data, mark the point cloud clusters and noise points, and obtain the 3D point cloud data after removing the noise points. The walking combination algorithm includes: in the process of optimizing the walking tour algorithm, using the gray wolf algorithm to improve the scanning factor in the speed update formula of the hiker in the tour algorithm When the number of consecutive times that the hiker's position change is less than the first set value meets the first threshold, the first optimal solution and the number of optimization times are recorded, and a jump process is performed. The jump scan factor is When the number of optimization attempts is less than or equal to the second threshold, the scan factor is used. Continue the optimization iteration. When the number of consecutive times that the hiker's position change is less than the first set value meets the first threshold, record the second optimal solution, update the optimization number, select the best of the first optimal position and the second optimal position, repeat the above optimization process, and output the best solution when the optimization number is greater than the second threshold.

2. The point cloud data denoising method based on intelligent driving according to claim 1, characterized in that: Scan Factor It is expressed by the following formula: in, Represent the positions of α wolf, β wolf, and δ wolf respectively. α wolf represents the optimal solution, β wolf and δ wolf represent the second-best solution and the third-best solution respectively. represents the optimal position of the gray wolf after iteration t+1, They represent the corresponding movement distances of α wolf, β wolf, and δ wolf according to the estimated position of the prey, They represent the distances between the current search agent and wolf α, wolf β, and wolf δ respectively. α, β, and δ represent the degree of knowledge of wolf α, wolf β, and wolf δ about the location of the prey, i.e., the weights, α+β+δ=1.

3. The point cloud data denoising method based on intelligent driving according to claim 1, characterized in that: The change in the position of the hiker is expressed by the following formula: Among them, Z(t+1) represents the position change of the hiker, β i,t+1 represents the position of the hiker after t+1 iterations, β i,t represents the position of the hiker after t iterations.

4. The point cloud data denoising method based on intelligent driving according to claim 1, characterized in that: Scan Factor It is expressed by the following formula: in, The range is between [3,9], and the optimization iteration is performed. i,t+1 represents the position of the hiker after t+1 iterations, β i,t represents the position of the hiker after t iterations, Indicates the optimal position of the wolf each time.

5. The point cloud data denoising method based on intelligent driving according to claim 1, characterized in that: The location update of the hiker in the hiking tour algorithm is achieved by the following process: The initial slope formula of the hiking algorithm is: Where dh and dx represent the hiker’s height difference and travel distance difference, respectively, θ i,t is the slope angle of the path or terrain, S i,t represents the slope of the path or terrain; the hiker's initial speed is given by the Tobler hiking function: Where W i,t represents the speed of the initial hiker during the iteration time t, in km / h; During the search process, the speed of the hiker is determined by the initial speed, the position of the leading hiker, the actual position of the hiker, and the scanning factor. Therefore, the current speed of the i-th hiker is: W i,t =W i,t-1 +g i,t (b best -a i,t b i,t ) Among them, γ i,t is a random number uniformly distributed between 0 and 1, α i,t is the scanning factor of hiker i, β i,t is the current position of hiker i, β best It is the position of the lead hiker, W i,t represents the current speed of hiker i, W i,t-1 represents the speed of hiker i at the last moment; according to the hiker's speed, the new position update of hiker i should be: β i,t+1 ＝β i,t +W i,t Where, β i,t+1 is the updated position of hiker i. The hiker will continuously update its position to reach the leader's position.

6. The point cloud data denoising method based on intelligent driving according to claim 1, characterized in that: The method of clustering the segmented three-dimensional point cloud data, marking point cloud clusters and noise points, and obtaining three-dimensional point cloud data after removing noise points includes: Step (1): Mark the cluster labels of all point clouds as -1, indicating noise points; Step (2): Use the walking combination algorithm to obtain the optimal neighborhood radius Eps parameter and the optimal minimum number of points MinPts parameter; Step (3): Check one of the points A to determine whether it is a core point. If the number of points in the neighborhood exceeds the threshold MinPts, point A is the core point, and all points in the neighborhood Eps of point A are added to the cluster. Step (4): Mark the points that are neither core points nor in the neighborhood of any core points as noise points; Step (5): Repeat steps (2), (3), and (4) for each untraversed point in the cluster until all points in the cluster are processed, and obtain the three-dimensional point cloud data after removing the noise points.

7. A point cloud data denoising system based on intelligent driving, characterized in that: include: A preprocessing module is configured to: segment the acquired three-dimensional point cloud data of the target to be segmented in intelligent driving; The point cloud clustering module is configured to: use the walking combination algorithm to optimize the parameters of the DBSCAN algorithm to obtain the optimal neighborhood radius Eps parameter and the optimal minimum number of points MinPts parameter to cluster the segmented 3D point cloud data, mark point cloud clusters and noise points, and obtain 3D point cloud data after removing noise points; The walking combination algorithm includes: in the process of optimizing the walking tour algorithm, using the gray wolf algorithm to improve the scanning factor in the speed update formula of the hiker in the tour algorithm When the number of consecutive times that the hiker's position change is less than the first set value meets the first threshold, the first optimal solution and the number of optimization times are recorded, and a jump process is performed. The jump scan factor is When the number of optimization attempts is less than or equal to the second threshold, the scan factor is used. Continue the optimization iteration. When the number of consecutive times that the hiker's position change is less than the first set value meets the first threshold, record the second optimal solution, update the optimization number, select the best of the first optimal position and the second optimal position, repeat the above optimization process, and output the best solution when the optimization number is greater than the second threshold.

8. A computer-readable storage medium having a program stored thereon, characterized in that: When the program is executed by a processor, the steps of the point cloud data denoising method based on intelligent driving are implemented.

9. An electronic device, characterized in that: The method comprises a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps of the point cloud data denoising method based on intelligent driving are completed.

10. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor, the steps in the point cloud data denoising method based on intelligent driving are implemented as described in any one of claims 1 to 6.

Images