Point cloud data processing method, electronic equipment and readable medium

Abstract

The embodiment of the invention provides a point cloud data processing method, electronic equipment and a readable medium. The method comprises the following steps: acquiring point cloud data acquired by a laser radar; searching false expansion points based on the reflection intensity; and deleting the false expansion points in the point cloud data to obtain target point cloud data. Therefore, the false expansion point corresponding to the high-reflection expansion phenomenon can be efficiently searched, and the processing efficiency of the high-reflection expansion phenomenon is improved.

Description

Technical Field

[0001] The present invention relates to the field of laser radar technology, and in particular to a point cloud data processing method, an electronic device and a computer-readable medium. Background Art

[0002] In the prior art, when laser radar collects point cloud data, it is easy to have a high-reflection expansion phenomenon. The high-reflection expansion phenomenon refers to the strong reflection and scattering effect of high-reflection objects on lasers. When the laser radar scans highly reflective objects, its echo signal will be stronger than when scanning low-reflection objects, causing the point cloud data to diffuse or expand when it is generated, making the original high-reflection objects appear larger in the point cloud image, and it looks like the high-reflection objects have "expanded".

[0003] The high-reflection expansion phenomenon can affect the reliability of the point cloud data collected by the LiDAR, and may cause the target recognition function of the LiDAR to be abnormal during the use of the LiDAR, resulting in false alarms. When the LiDAR is used for vehicles, it may also cause the vehicle to take unnecessary obstacle avoidance measures, increasing the risk of users driving the vehicle. Summary of the invention

[0004] The embodiment of the present invention provides a method, device, electronic device and computer-readable storage medium for processing point cloud data to solve the high reflection expansion phenomenon occurring in point cloud data collected by a laser radar.

[0005] The embodiment of the present invention discloses a method for processing point cloud data, comprising:

[0006] Acquire point cloud data collected by a laser radar; the point cloud data includes a plurality of data points, and the data points are used to represent the reflection intensity of objects in the surrounding environment of the laser radar;

[0007] Finding a false expansion point based on the reflection intensity;

[0008] The false expansion points are deleted from the point cloud data to obtain target point cloud data.

[0009] Optionally, searching for a false expansion point based on the reflection intensity includes:

[0010] Based on the reflection intensity, searching the data points associated with the high-reflection object as the high-reflection object point;

[0011] Determining a target high-reflection object point among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point;

[0012] Based on the reflection intensity, a false expansion point is searched in a preset false expansion identification area around the target high-reflection object point.

[0013] Optionally, the step of searching, based on the reflection intensity, data points associated with the high-reflection object in the data points as the high-reflection object points comprises:

[0014] Search the data points for data points whose reflection intensity is greater than a preset high-reflection object reflection intensity threshold value as high-reflection object points associated with the high-reflection object.

[0015] Optionally, the step of searching, based on the reflection intensity, data points associated with the high-reflection object in the data points as the high-reflection object points comprises:

[0016] Based on the three-dimensional coordinates of the data point, obtaining the longitudinal position of the data point in the longitudinal direction;

[0017] When the longitudinal position of the data point is higher than the longitudinal position of the laser radar, taking the data point as a target data point;

[0018] Based on the reflection intensity, a data point associated with a high-reflection object is searched in the target data points as a high-reflection object point.

[0019] Optionally, the step of determining a target high-reflection object point among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point comprises:

[0020] If the volume of the high-reflective object corresponding to the high-reflective object point meets the preset volume recognition condition, the high-reflective object point is used as the target high-reflective object point.

[0021] Optionally, the method further comprises:

[0022] For any of the highly reflective object points, taking the highly reflective object point as the center, searching for other highly reflective object points in a preset volume recognition area around the highly reflective object point to obtain a set of highly reflective object points; the set of highly reflective object points includes at least one highly reflective object point;

[0023] Determine the percentage of the number of the high-reflective object points in the high-reflective object point set to the total number of data points in the volume recognition area;

[0024] If the percentage of the number of the high-reflective object points in the high-reflective object point set to the total number of data points in the volume recognition area is greater than the preset large high-reflective object percentage, it is determined that the volume of the high-reflective object corresponding to the high-reflective object point meets the preset volume recognition condition

[0025] If the percentage of the number of highly reflective object points in the highly reflective object point set to the total number of data points in the volume identification area is not greater than the preset percentage of large highly reflective objects, it is determined that the volume of the highly reflective object corresponding to the highly reflective object point does not meet the preset volume identification conditions.

[0026] Optionally, the step of determining a target high-reflection object point among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point comprises:

[0027] Based on the three-dimensional coordinates of the high-reflective object point, obtaining the depth distance of the high-reflective object point relative to the laser radar in the depth direction;

[0028] When the depth distance of the high-reflection object point is less than the preset volume recognition distance, a target high-reflection object point is determined among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point.

[0029] Optionally, the method further comprises:

[0030] In a case where the depth distance of the high-reflection object point is greater than a preset volume recognition distance, the high-reflection object point is used as a target high-reflection object point.

[0031] Optionally, the step of searching for a false expansion point in a preset false expansion identification area around the target high-reflection object point based on the reflection intensity comprises:

[0032] Based on the reflection intensity, searching for undetermined data points in a preset false expansion identification area around the target high-reflection object point;

[0033] Based on the number of high-reflection object points in a preset false expansion identification area around the pending data point, a false expansion point is determined in the pending data point.

[0034] Optionally, the step of searching for undetermined data points in a preset false expansion identification area around the target high-reflection object point based on the reflection intensity includes:

[0035] In a preset false expansion identification area around the target highly reflective object point, searching for at least one data point located on the same object plane as the target highly reflective object point as a coplanar data point;

[0036] Based on the reflection intensity of the target high-reflection object point and the reflection intensity of the coplanar data points, a pending data point is determined among the coplanar data points.

[0037] Optionally, the data points in the point cloud data are arranged in a matrix form to form a point cloud matrix;

[0038] The step of searching at least one data point located on the same object plane as the target highly reflective object point in a preset false expansion identification area around the target highly reflective object point as a coplanar data point comprises:

[0039] In a preset false expansion identification area around the target highly reflective object point, searching the point cloud matrix for data points located in the same column as the target highly reflective object point as longitudinal data points, and / or searching the point cloud matrix for data points located in the same row as the target highly reflective object point as transverse data points;

[0040] Based on the three-dimensional coordinates of the target highly reflective object point and the three-dimensional coordinates of the longitudinal data points, coplanar data points are determined among the longitudinal data points; and / or, based on the three-dimensional coordinates of the target highly reflective object point and the three-dimensional coordinates of the transverse data points, coplanar data points are determined among the transverse data points.

[0041] Optionally, the step of determining a pending data point among the coplanar data points based on the reflection intensity of the target highly reflective object point and the reflection intensity of the coplanar data point comprises:

[0042] If the difference in reflection intensity between the target highly reflective object point and the coplanar data point is greater than a preset reflection intensity difference threshold, the coplanar data point is used as a pending data point.

[0043] Optionally, the step of determining the false expansion point in the pending data point based on the number of high-reflection object points in a preset false expansion identification area around the pending data point includes:

[0044] For any of the pending data points, searching for high-reflection object points in a preset false expansion identification area around the pending data point to obtain the number of high-reflection object points in the false expansion identification area;

[0045] Determine the percentage of the number of highly reflective object points in the false expansion identification area to the total number of data points in the false expansion identification area;

[0046] If the percentage of the number of high-reflection object points in the false expansion identification area to the total number of data points in the false expansion identification area is less than a preset false expansion percentage, the pending data point is taken as a false expansion point.

[0047] Optionally, the step of determining the false expansion point in the pending data point based on the number of high-reflection object points in a preset false expansion identification area around the pending data point includes:

[0048] Based on the three-dimensional coordinates of the pending data point, obtaining the depth distance of the pending data point relative to the laser radar in the depth direction;

[0049] If the depth distance of the pending data point is less than the preset character recognition distance, the pending data point is used as the target pending data point;

[0050] Based on the number of high-reflection object points in a preset false expansion identification area around the target pending data point, a false expansion point is determined in the target pending data point.

[0051] An embodiment of the present invention further provides a point cloud data processing device, comprising:

[0052] A point cloud acquisition module, used to acquire point cloud data collected by the laser radar; the point cloud data includes a number of data points, and the data points are used to represent the three-dimensional coordinates and reflection intensity of objects in the surrounding environment of the laser radar;

[0053] A high-reflection object point search module, used to search for data points associated with high-reflection objects in the data points based on the reflection intensity as high-reflection object points;

[0054] A target high-reflection object point determination module, configured to determine a target high-reflection object point among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point;

[0055] A false expansion point search module, configured to search for a false expansion point in a preset false expansion identification area around the target high-reflection object point based on the reflection intensity;

[0056] The target point cloud data acquisition module is used to delete the false expansion points in the point cloud data to obtain the target point cloud data.

[0057] Optionally, the highly reflective object point search module includes:

[0058] The first high-reflection object point search submodule is used to search the data points for a data point whose reflection intensity is greater than a preset high-reflection object reflection intensity threshold value as a high-reflection object point associated with the high-reflection object.

[0059] Optionally, the highly reflective object point search module includes:

[0060] A longitudinal position acquisition submodule, used for acquiring the longitudinal position of the data point in the longitudinal direction based on the three-dimensional coordinates of the data point;

[0061] a target data point determination submodule, configured to take the data point as a target data point when the longitudinal position of the data point is higher than the longitudinal position of the laser radar;

[0062] The second high-reflection object point search submodule is used to search for data points associated with high-reflection objects in the target data points based on the reflection intensity as high-reflection object points.

[0063] Optionally, the target highly reflective object point determination module includes:

[0064] The first target highly reflective object point determination submodule uses the highly reflective object point as a target highly reflective object point if the volume of the highly reflective object corresponding to the highly reflective object point meets a preset volume recognition condition.

[0065] Optionally, the device further comprises:

[0066] A high-reflection object point set acquisition module is used to, for any of the high-reflection object points, search for other high-reflection object points in a preset volume recognition area around the high-reflection object point with the high-reflection object point as the center, to obtain a high-reflection object point set; the high-reflection object point set includes at least one high-reflection object point;

[0067] A high reflection percentage determination module, used to determine the percentage of the number of the high reflection object points in the high reflection object point set to the total number of data points in the volume recognition area;

[0068] A first recognition condition determination module, configured to determine that the volume of the high-reflective object corresponding to the high-reflective object point satisfies a preset volume recognition condition if the percentage of the number of the high-reflective object points in the high-reflective object point set to the total number of data points in the volume recognition area is greater than a preset large high-reflective object percentage;

[0069] The second recognition condition determination module is used to determine that the volume of the high-reflective object corresponding to the high-reflective object point does not meet the preset volume recognition condition if the percentage of the number of high-reflective object points in the high-reflective object point set to the total number of data points in the volume recognition area is not greater than the preset large-scale high-reflective object percentage.

[0070] Optionally, the target highly reflective object point determination module includes:

[0071] A high-reflection depth distance determination submodule, used to obtain the depth distance of the high-reflection object point relative to the laser radar in the depth direction based on the three-dimensional coordinates of the high-reflection object point;

[0072] The second target high-reflection object point determination submodule is used to determine the target high-reflection object point among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point when the depth distance of the high-reflection object point is less than a preset volume recognition distance.

[0073] Optionally, the device further comprises:

[0074] The volume recognition distance determination module is used to take the high-reflection object point as a target high-reflection object point when the depth distance of the high-reflection object point is greater than a preset volume recognition distance.

[0075] Optionally, the false expansion point search module includes:

[0076] A pending data point search submodule is used to search for pending data points in a preset false expansion recognition area around the target high-reflection object point based on the reflection intensity;

[0077] The false expansion point determination submodule is used to determine the false expansion point in the pending data point based on the number of high-reflection object points in a preset false expansion identification area around the pending data point.

[0078] Optionally, the pending data point search submodule includes:

[0079] A coplanar data point searching unit, configured to search, within a preset false expansion identification area around the target highly reflective object point, at least one data point located on the same object plane as the target highly reflective object point as a coplanar data point;

[0080] The pending data point searching unit is used to determine the pending data point among the coplanar data points based on the reflection intensity of the target high-reflection object point and the reflection intensity of the coplanar data point.

[0081] Optionally, the data points in the point cloud data are arranged in a matrix form to form a point cloud matrix;

[0082] The coplanar data point searching unit comprises:

[0083] A horizontal and vertical data point search subunit is used to search for data points in the point cloud matrix that are located in the same column as the target highly reflective object point as the vertical data point, and / or to search for data points in the point cloud matrix that are located in the same row as the target highly reflective object point as the horizontal data point, within a preset false expansion identification area around the target highly reflective object point;

[0084] A coplanar data point search subunit is used to determine coplanar data points in the longitudinal data points based on the three-dimensional coordinates of the target highly reflective object point and the three-dimensional coordinates of the longitudinal data points; and / or to determine coplanar data points in the transverse data points based on the three-dimensional coordinates of the target highly reflective object point and the three-dimensional coordinates of the transverse data points.

[0085] Optionally, the pending data point searching unit includes:

[0086] The pending data point search subunit is used to take the coplanar data point as a pending data point if the difference in reflection intensity between the target high-reflection object point and the coplanar data point is greater than a preset reflection intensity difference threshold.

[0087] Optionally, the false expansion point determination submodule includes:

[0088] A false expansion region high reflective point search unit is used to search for high reflective object points in a false expansion recognition region preset around any pending data point for any pending data point, and obtain the number of high reflective object points in the false expansion recognition region;

[0089] a false expansion region high reflection point percentage determination unit, used to determine the percentage of the number of high reflection object points in the false expansion identification region to the total number of data points in the false expansion identification region;

[0090] The first false expansion point determination unit is configured to, when a percentage of the number of high-reflection object points in the false expansion identification area to the total number of data points in the false expansion identification area is less than a preset false expansion percentage, take the pending data point as a false expansion point.

[0091] Optionally, the false expansion point determination submodule includes:

[0092] A depth distance determination unit for a pending data point, configured to obtain a depth distance of the pending data point relative to the laser radar in a depth direction based on the three-dimensional coordinates of the pending data point;

[0093] a target pending data point determination unit, configured to use the pending data point as a target pending data point if the depth distance of the pending data point is less than a preset character recognition distance;

[0094] The second false expansion point determination unit is configured to determine a false expansion point in the target data point to be determined based on the number of high-reflection object points in a preset false expansion identification area around the target data point to be determined.

[0095] The embodiment of the present invention further discloses an electronic device, comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus;

[0096] The memory is used to store computer programs;

[0097] The processor is used to implement the method described in the embodiment of the present invention when executing the program stored in the memory.

[0098] The embodiment of the present invention further discloses one or more computer-readable media on which instructions are stored. When executed by one or more processors, the processors are enabled to execute the method described in the embodiment of the present invention.

[0099] The embodiments of the present invention include the following advantages:

[0100] By using the point cloud data processing method provided by the embodiment of the present invention, the point cloud data collected by the laser radar is obtained; based on the reflection intensity, the false expansion points are found; the false expansion points are deleted from the point cloud data to obtain the target point cloud data. In this way, the false expansion points corresponding to the high reflection expansion phenomenon can be efficiently found, thereby improving the processing efficiency of the high reflection expansion phenomenon. BRIEF DESCRIPTION OF THE DRAWINGS

[0101] Figure 1 is a schematic diagram of a laser radar provided in an embodiment of the present invention;

[0102] Figure 2 is a flowchart of a method for processing point cloud data provided in an embodiment of the present invention;

[0103] Figure 3 is a flowchart of a method for processing point cloud data provided in an embodiment of the present invention;

[0104] Figure 4 is a schematic diagram of point cloud data provided in an embodiment of the present invention;

[0105] Figure 5 is a flow chart of a method for processing point cloud data provided in an embodiment of the present invention;

[0106] Figure 6 is a structural block diagram of a point cloud data processing device provided in an embodiment of the present invention;

[0107] Figure 7 is a block diagram of an electronic device provided in an embodiment of the present invention;

[0108] Figure 8 is a schematic diagram of a computer-readable medium provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0109] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0110] Reference Figure 1 , showing a schematic diagram of a laser radar provided in an embodiment of the present invention.

[0111] In an embodiment of the present invention, the laser radar can be set on an object, and the object can be a vehicle, a drone, a robot, a monitoring device, etc., and the present invention does not limit this. As a specific example of the present invention, when the laser radar is set on a vehicle, the laser radar can be set on the top of the vehicle to obtain a larger field of view.

[0112] Generally speaking, a laser radar can have a field of view, which can be the range in which the laser radar can identify objects. The laser radar can emit several laser beams in sequence within the field of view according to a pre-set scanning order, and receive the light reflected back after the laser beam contacts the objects in the surrounding environment. Afterwards, the three-dimensional coordinates of a point on the object can be calculated based on the emission direction of the laser beam, as well as the emission time and reception time of the laser beam, and the reflection intensity of the laser beam can be recorded to obtain the data point. In this way, the laser radar can obtain point cloud data formed by several data points. The point cloud data can be used to describe the location of different objects in the environment around the laser radar, so that the laser radar can perceive the objects in its surrounding environment.

[0113] Specifically, in the three-dimensional coordinate system constructed by the lidar, the X-axis can usually represent the depth direction, the Y-axis can usually represent the horizontal direction, and the Z-axis can usually represent the vertical direction. The plane formed by the X-axis and the Y-axis can be parallel to the ground, and the Z-axis can be perpendicular to the plane formed by the X-axis and the Y-axis.

[0114] like Figure 1 As shown, when the laser radar is set on the top of the vehicle, Figure 1 As can be seen in (a), the Z axis of the laser radar is perpendicular to the ground. Figure 1 As can be seen in (b), the X-axis of the laser radar is parallel to the front direction of the vehicle, and the Y-axis of the laser radar is parallel to the horizontal axis of the vehicle.

[0115] Reference Figure 2 , shows a flowchart of a method for processing point cloud data provided in an embodiment of the present invention, which specifically includes the following steps:

[0116] Acquire point cloud data collected by a laser radar; the point cloud data includes a plurality of data points, and the data points are used to represent the reflection intensity of objects in the surrounding environment of the laser radar;

[0117] Step 201, searching for a false expansion point based on the reflection intensity;

[0118] Step 202: deleting the false expansion points in the point cloud data to obtain target point cloud data.

[0119] Step 203, in the embodiment of the present invention, the laser radar can collect point cloud data according to a preset time period, so as to obtain time frames at different times. For a time frame at a certain time, the point cloud data at that time can be used to analyze whether there is a high reflection expansion phenomenon in it.

[0120] Specifically, the data points collected by the laser radar may record the reflection intensity of the laser reflected by the object. If the object can have a high reflection intensity, the object can be considered as a high-reflection object. However, there may be false expansion points around the high-reflection object due to the high-reflection expansion phenomenon. The false expansion points can be found based on the reflection intensity. Afterwards, the false expansion points can be deleted from the point cloud data to eliminate the high-reflection expansion phenomenon and obtain the target point cloud data. In this way, the false expansion points corresponding to the high-reflection expansion phenomenon can be efficiently found, thereby improving the processing efficiency of the high-reflection expansion phenomenon.

[0121] Reference Figure 3 , shows a flowchart of a method for processing point cloud data provided in an embodiment of the present invention, which may specifically include the following steps:

[0122] Step 301, obtaining point cloud data collected by the laser radar; the point cloud data includes a number of data points, and the data points are used to represent the three-dimensional coordinates and reflection intensity of objects in the surrounding environment of the laser radar;

[0123] In the embodiment of the present invention, the laser radar can collect point cloud data according to a preset time period, so as to obtain time frames at several different times. For a time frame at a certain time, the point cloud data at that time can be used to analyze whether there is a high reflection expansion phenomenon in it.

[0124] Step 302, based on the reflection intensity, searching the data points associated with the high-reflection object as the high-reflection object point;

[0125] Specifically, the data points collected by the laser radar may record the reflection intensity of the laser reflected by the object. If the object can have a higher reflection intensity, the object can be considered a highly reflective object.

[0126] When the reflection intensity recorded in the data point is high, it can be considered that the data point may be located on or around a highly reflective object. Therefore, it can be considered that the data may be associated with a highly reflective object and the data point is regarded as a highly reflective object point.

[0127] Step 303, determining a target high-reflection object point among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point;

[0128] Generally speaking, in point cloud data, large highly reflective objects are more likely to cause serious high-reflection expansion phenomenon, while small highly reflective objects usually cause lighter high-reflection expansion phenomenon. In addition, small highly reflective objects are usually set on larger objects. Suppressing the high-reflection expansion phenomenon of small highly reflective objects may cause the larger object where the small highly reflective object is located to be unable to be normally identified.

[0129] For example, in a vehicle driving scenario, a large, highly reflective object identified by a LiDAR may usually be a traffic sign or other highly reflective sign or label. These signs or labels are usually set independently, and suppressing their high-reflective expansion phenomenon will not have a negative impact on the object's recognition. Smaller, highly reflective objects may usually be reflective strips at the rear of a truck, vehicle license plates, vehicle taillights, etc., which are usually set on the vehicle. Suppressing the high-reflective expansion phenomenon of small, highly reflective objects may have a negative impact on the vehicle's recognition.

[0130] Therefore, before processing a high-reflection object point, the volume of the high-reflection object corresponding to the high-reflection object point can be first determined to determine whether the current high-reflection object point is a data point that needs to be processed to suppress the high-reflection expansion phenomenon. In the case that the volume of the high-reflection object corresponding to the high-reflection object point is large, the high-reflection object point can be determined as a target high-reflection object point that needs further processing.

[0131] Step 304, based on the reflection intensity, searching for a false expansion point in a preset false expansion identification area around the target high-reflection object point;

[0132] After the target high-reflection object point is determined, there may be false expansion points around the high-reflection object where the target high-reflection object point is located. The false expansion points may be located in an area where the high-reflection expansion phenomenon occurs around the high-reflection object.

[0133] In order to reduce possible erroneous deletion and reduce the amount of data processing, a false expansion identification area can be pre-set, and the area where high-reflection expansion occurs around a high-reflection object can usually be located in the false expansion identification area. Therefore, the false expansion points can be found by traversing the data points in the false expansion identification area.

[0134] The false expansion point may generally have a higher reflection intensity, and its reflection intensity may be higher than the reflection intensity of the surrounding environment to a certain extent, and at the same time, its reflection intensity may generally be lower than the reflection intensity of the target high-reflection object point. Therefore, based on the reflection intensity, a data point having a reflection intensity lower than the reflection intensity of the target high-reflection object point and still having a higher reflection intensity may be searched in a preset false expansion identification area around the target high-reflection object point, and the data point may be used as a false expansion point.

[0135] Step 305: Delete the false expansion points in the point cloud data to obtain the target point cloud data.

[0136] After the false expansion points are determined, the false expansion points can be deleted from the point cloud data to obtain the target point cloud data, thereby suppressing the high-reflection expansion phenomenon.

[0137] As a specific example of the present invention, Figure 4 is a schematic diagram of point cloud data provided in an embodiment of the present invention. Figure 4 (a) shows the road sign information in the real scene captured by a camera. Figure 4 (b) is a point cloud data image that has not been processed by the high anti-expansion phenomenon. It can be seen that the road sign information in the point cloud data has obviously expanded compared to the road sign information in the real scene. Figure 4 (c) is a point cloud data image processed by an embodiment of the present invention, from which it can be seen that the high reflection expansion phenomenon is effectively suppressed.

[0138] Through the point cloud data processing method provided by the embodiment of the present invention, the point cloud data collected by the laser radar is obtained; the point cloud data includes a number of data points, and the data points are used to represent the three-dimensional coordinates and reflection intensity of objects in the surrounding environment of the laser radar; based on the reflection intensity, the data points associated with the high-reflection object are searched in the data points as the high-reflection object points; based on the volume of the high-reflection object corresponding to the high-reflection object point, the target high-reflection object point is determined in the high-reflection object point; based on the reflection intensity, the false expansion point is searched in the false expansion identification area preset around the target high-reflection object point; the false expansion point is deleted in the point cloud data to obtain the target point cloud data. In this way, the false expansion point corresponding to the high-reflection expansion phenomenon can be efficiently found and obtained, and the processing efficiency of the high-reflection expansion phenomenon is improved. At the same time, by judging the volume of the high-reflection object corresponding to the high-reflection object point, the high-reflection object with obvious high-reflection expansion phenomenon is processed, the occurrence of the erroneous deletion phenomenon is reduced, and the accuracy of the high-reflection expansion phenomenon processing is improved.

[0139] In one embodiment of the present invention, based on the reflection intensity, the step of searching the data points for data points associated with the highly reflective object as the highly reflective object points includes:

[0140] S11, searching among the data points for data points whose reflection intensity is greater than a preset high-reflection object reflection intensity threshold, as a high-reflection object point associated with the high-reflection object.

[0141] Specifically, a high-reflection object reflection intensity threshold may be preset, and when the reflection intensity of a data point is greater than the high-reflection object reflection intensity threshold, the data point may be considered to have a high reflection intensity and serve as a high-reflection object point associated with the high-reflection object.

[0142] In one embodiment of the present invention, based on the reflection intensity, the step of searching the data points for data points associated with the highly reflective object as the highly reflective object points includes:

[0143] S21, obtaining a longitudinal position of the data point in a longitudinal direction based on the three-dimensional coordinates of the data point;

[0144] S22, when the longitudinal position of the data point is higher than the longitudinal position of the laser radar, taking the data point as a target data point;

[0145] S23, based on the reflection intensity, searching for data points associated with the high-reflection object in the target data points as high-reflection object points.

[0146] Generally speaking, large and highly reflective objects can usually be placed at relatively high positions. Therefore, in order to process highly reflective objects with obvious high-reflection expansion, reduce the occurrence of false deletion, and improve data processing efficiency, data points at higher positions can be identified first.

[0147] Thus, the coordinate data of the data point in the Z-axis direction, that is, the longitudinal direction, can be obtained from the three-dimensional coordinates of the data point, so that the longitudinal position of the data point in the longitudinal direction can be known.

[0148] In a specific implementation, when the longitudinal position of a data point is higher than the longitudinal position of the laser radar, it can be considered that the longitudinal position of the data point is higher, and it may be a data point that needs to be preferentially identified as to whether it is associated with a highly reflective object, and the data point can be used as a target data point.

[0149] Specifically, in the three-dimensional coordinate system constructed by the laser radar, the coordinate value of the data point whose longitudinal position is higher than the longitudinal position of the laser radar can be a positive number in the Z-axis direction, while the coordinate value of the data point whose longitudinal position is lower than the longitudinal position of the laser radar can be a negative number in the Z-axis direction. Therefore, the data point whose coordinate value in the Z-axis direction is a positive number can be used as the target data point.

[0150] Afterwards, based on the reflection intensity, data points associated with the highly reflective object can be searched in the target data points as highly reflective object points, thereby further improving the efficiency of highly reflective object points.

[0151] In one embodiment of the present invention, the step of determining a target high-reflective object point among the high-reflective object points based on the volume of the high-reflective object corresponding to the high-reflective object point includes:

[0152] S31: If the volume of the highly reflective object corresponding to the highly reflective object point meets a preset volume recognition condition, the highly reflective object point is used as a target highly reflective object point.

[0153] In an embodiment of the present invention, a volume recognition condition can be set in advance. If the volume of the high-reflective object corresponding to the high-reflective object point meets the preset volume recognition condition, it can be considered that the high-reflective object point is located on a larger high-reflective object that needs attention, and the high-reflective object point is used as the target high-reflective object point.

[0154] The volume recognition condition can be set according to actual needs, for example, target recognition is performed on the object where the high-reflective object point is located to determine the volume of the object. For another example, it is determined whether the object where the high-reflective object point is located can cover an area of ​​a preset size, etc. The present invention does not limit this.

[0155] In one embodiment of the present invention, the method further comprises:

[0156] S41, for any high-reflection object point, with the high-reflection object point as the center, searching for other high-reflection object points in a preset volume recognition area around the high-reflection object point to obtain a high-reflection object point set; the high-reflection object point set includes at least one high-reflection object point;

[0157] S42, determining the percentage of the number of high-reflection object points in the high-reflection object point set to the total number of data points in the volume recognition area;

[0158] S43, if the percentage of the number of high-reflective object points in the high-reflective object point set to the total number of data points in the volume recognition area is greater than the preset large high-reflective object percentage, then it is determined that the volume of the high-reflective object corresponding to the high-reflective object point meets the preset volume recognition condition

[0159] S44, if the percentage of the number of high-reflective object points in the high-reflective object point set to the total number of data points in the volume recognition area is not greater than the preset large high-reflective object percentage, it is determined that the volume of the high-reflective object corresponding to the high-reflective object point does not meet the preset volume recognition condition.

[0160] In a specific implementation, a volume recognition area may be preset. If the object where the high-reflective object point is located can cover or almost cover the volume recognition area, it can be considered that the object where the high-reflective object point is located has a larger volume.

[0161] The size of the volume recognition area is preset according to rows and columns. For example, the volume recognition area can be set to a row*b column.

[0162] Specifically, the laser radar can emit laser beams in a matrix form, so that the data points can be arranged in the order of rows and columns according to the emitted lasers. An area with a certain row and column size can be selected as the volume recognition area.

[0163] Thereafter, for any high-reflection object point, other high-reflection object points can be searched in a preset volume recognition area around the high-reflection object point with the high-reflection object point as the center, so as to obtain a high-reflection object point set including at least one high-reflection object point. Thereafter, the total number of data points in the volume recognition area can be calculated based on the number of rows and the number of columns in the volume recognition area.

[0164] The percentage of the number of high-reflection object points in the high-reflection object point set to the total number of data points in the volume recognition area can be calculated, and it can be determined whether the percentage of the number of high-reflection object points in the high-reflection object point set to the total number of data points in the volume recognition area is greater than a preset large high-reflection object percentage.

[0165] If the percentage of the number of high-reflective object points in the high-reflective object point set to the total number of data points in the volume recognition area is greater than the preset large high-reflective object percentage, it can be considered that the high-reflective object associated with the high-reflective object point has covered or almost covered the volume recognition area. At this time, it can be determined that the volume of the high-reflective object corresponding to the high-reflective object point meets the preset volume recognition conditions, and the high-reflective object corresponding to the high-reflective object point belongs to a larger volume object.

[0166] If the percentage of high-reflective object points in the high-reflective object point set to the total number of data points in the volume recognition area is not greater than the preset percentage of large high-reflective objects, it can be determined that the volume of the high-reflective object corresponding to the high-reflective object point does not meet the preset volume recognition conditions, and the high-reflective object corresponding to the high-reflective object point does not belong to an object of larger volume, and there is no need to process it to suppress the high-reflective expansion phenomenon.

[0167] In one embodiment of the present invention, the step of determining a target high-reflective object point among the high-reflective object points based on the volume of the high-reflective object corresponding to the high-reflective object point includes:

[0168] S51, based on the three-dimensional coordinates of the high-reflection object point, obtaining a depth distance of the high-reflection object point relative to the laser radar in the depth direction;

[0169] S52, when the depth distance of the high-reflection object point is less than the preset volume recognition distance, determining a target high-reflection object point among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point.

[0170] Specifically, the depth distance of the high-reflective object point relative to the laser radar in the depth direction can be obtained from the three-dimensional coordinates of the high-reflective object point, and the distance of the object relative to the laser radar can be judged based on the depth distance. If the high-reflective object is close to the laser radar, that is, the depth distance of the high-reflective object point is less than the preset volume recognition distance, at this time, since the high-reflective object identified by the laser radar at a closer distance may be small in volume, it is necessary to judge whether the high-reflective object point needs to be used as the target high-reflective object point for processing the expansion phenomenon based on the volume of the high-reflective object corresponding to the high-reflective object point.

[0171] In one embodiment of the present invention, the method further comprises:

[0172] S61, when the depth distance of the high-reflection object point is greater than a preset volume recognition distance, taking the high-reflection object point as a target high-reflection object point.

[0173] Specifically, if the highly reflective object is far away from the LiDAR, the LiDAR will usually be unable to identify a small highly reflective object. That is, at a long distance, the highly reflective object identified by the LiDAR must be a large highly reflective object. In this case, there is no need to determine the volume of the highly reflective object, but the highly reflective object point can be directly used as the target highly reflective object point, further improving processing efficiency.

[0174] In one embodiment of the present invention, the step of searching for a false expansion point in a preset false expansion identification area around a target high-reflection object point based on the reflection intensity includes:

[0175] S71, searching for undetermined data points in a preset false expansion identification area around the target high-reflection object point based on the reflection intensity;

[0176] S72, determining a false expansion point among the pending data points based on the number of highly reflective object points in a preset false expansion identification area around the pending data point.

[0177] In a specific implementation, after determining the target high-reflection object point, the false expansion point can be searched in the false expansion identification area preset around the target high-reflection object point. Since the reflection intensity of the false expansion point is usually lower than that of the high-reflection object point to a certain extent, the undetermined data point can be searched in the false expansion identification area preset around the target high-reflection object point based on the reflection intensity.

[0178] Afterwards, it is possible to further analyze whether there are many high-reflection object points in the preset false expansion identification area around the pending data point. If there are many high-reflection object points, the pending data point may be text information on the sign. In order to avoid accidental deletion of text information, it can be excluded as a false expansion point. If the number of high-reflection object points in the preset false expansion identification area around the pending data point is small, it can be determined that the pending data point is indeed a false expansion point that needs to be adjusted.

[0179] In one embodiment of the present invention, the step of searching for undetermined data points in a preset false expansion identification area around a target high-reflective object point based on the reflection intensity includes:

[0180] S81, searching, within a preset false expansion identification area around a target highly reflective object point, at least one data point located on the same object plane as the target highly reflective object point as a coplanar data point;

[0181] S82, determining a pending data point among the coplanar data points based on the reflection intensity of the target high-reflection object point and the reflection intensity of the coplanar data points.

[0182] Specifically, the preset high-reflection object points are usually located on the same object plane, such as the plane of a traffic sign. Therefore, in the process of determining the false expansion point, at least one data point located on the same object plane as the target high-reflection object point can be firstly found in the preset false expansion identification area around the target high-reflection object point as a coplanar data point.

[0183] Afterwards, the analysis reflection intensity of the coplanar data point can be determined to be lower than that of the highly reflective object point to determine whether it is a pending data point that may belong to false expansion data.

[0184] In one embodiment of the present invention, data points in the point cloud data are arranged in a matrix form to form a point cloud matrix;

[0185] The step of searching, within a preset false expansion identification area around a target highly reflective object point, at least one data point located on the same object plane as the target highly reflective object point as a coplanar data point comprises:

[0186] S91, in a preset false expansion identification area around the target highly reflective object point, searching for data points in the point cloud matrix that are located in the same column as the target highly reflective object point as longitudinal data points, and / or searching for data points in the point cloud matrix that are located in the same row as the target highly reflective object point as transverse data points;

[0187] S92, determining coplanar data points in the longitudinal data points based on the three-dimensional coordinates of the target highly reflective object point and the three-dimensional coordinates of the longitudinal data points; and / or determining coplanar data points in the transverse data points based on the three-dimensional coordinates of the target highly reflective object point and the three-dimensional coordinates of the transverse data points.

[0188] In a specific implementation, we can first search for data points in the point cloud matrix that are located in the same column as the target highly reflective object point in the preset false expansion identification area around the target highly reflective object point as longitudinal data points, and search for data points in the same row as the target highly reflective object point as lateral data points, so as to further analyze whether they are coplanar with the highly reflective object point.

[0189] Specifically, the M points above and N points below the column where the high-reflective object point is located can be traversed to obtain the longitudinal data points, and the U points on the left and V points on the right of the high-reflective object point row can be traversed to obtain the transverse data points. The values ​​of M, N, U, and V can be determined according to actual needs, and the present invention does not limit this.

[0190] Afterwards, the distance differences between the highly reflective object points and the longitudinal data points in different directions can be traversed. If the distance differences between the highly reflective object points and the longitudinal data points in the depth direction and the transverse direction are small, and the distance difference in the longitudinal direction is within a certain range, the longitudinal data points can be considered to be coplanar data points. If the distance differences between the highly reflective object points and the longitudinal data points in the depth direction and the longitudinal direction are small, and the distance difference in the transverse direction is within a certain range, the transverse data points can be considered to be coplanar data points.

[0191] In one embodiment of the present invention, the step of determining the pending data point among the coplanar data points based on the reflection intensity of the target highly reflective object point and the reflection intensity of the coplanar data point includes:

[0192] S101: If the difference in reflection intensity between the target highly reflective object point and the coplanar data point is greater than a preset reflection intensity difference threshold, the coplanar data point is used as a pending data point.

[0193] Afterwards, the difference in reflection intensity between the target highly reflective object point and the coplanar data point can be determined to determine whether the coplanar data point is a pending data point that may need to be processed. When the difference in reflection intensity between the target highly reflective object point and the coplanar data point is greater than a preset reflection intensity difference threshold, it can be considered that the two have an obvious reflection intensity difference, the coplanar data point may be a false expansion point, and the coplanar data point is used as a pending data point.

[0194] In one embodiment of the present invention, the step of determining the false expansion point in the pending data point based on the number of high-reflection object points in a preset false expansion identification area around the pending data point includes:

[0195] S111, for any pending data point, searching for high-reflection object points in a preset false expansion identification area around the pending data point, and obtaining the number of high-reflection object points in the false expansion identification area;

[0196] S112, determining the percentage of the number of high-reflection object points in the false expansion identification area to the total number of data points in the false expansion identification area;

[0197] S113, if the percentage of the number of high-reflection object points in the false expansion identification area to the total number of data points in the false expansion identification area is less than a preset false expansion percentage, the pending data point is regarded as a false expansion point.

[0198] Thereafter, for any pending data point, it can be determined whether it is surrounded by high-reflection object points to determine whether it is text information with low reflection intensity in the high-reflection object.

[0199] Therefore, a false expansion identification area can be preset around the pending data point, and high-reflection object points can be searched in the false expansion identification area to obtain the number of high-reflection object points in the false expansion identification area.

[0200] Thereafter, the percentage of the number of high-reflection object points in the false expansion identification area to the total number of data points in the false expansion identification area may be calculated to determine whether the high-reflection object points account for a high proportion in the false expansion identification area.

[0201] When the percentage of the number of high-reflection object points in the false expansion identification area to the total number of data points in the false expansion identification area is less than the preset false expansion percentage, since there are fewer high-reflection object points around the pending data point, the pending data point can be considered to be a false data point.

[0202] When the percentage of the number of high-reflection object points in the false expansion identification area to the total number of data points in the false expansion identification area is greater than the preset false expansion percentage, since there are many high-reflection object points around the pending data point, it can be considered that the pending data point is not a false data point, but may be text information in a high-reflection object, thereby avoiding the text information in the high-reflection object from being accidentally deleted.

[0203] In one embodiment of the present invention, the step of determining the false expansion point in the pending data point based on the number of high-reflection object points in a preset false expansion identification area around the pending data point includes:

[0204] S121, based on the three-dimensional coordinates of the pending data point, obtaining a depth distance of the pending data point relative to the laser radar in the depth direction;

[0205] S122, if the depth distance of the pending data point is less than the preset character recognition distance, the pending data point is used as a target pending data point;

[0206] S123, determining false expansion points in the target data points to be determined based on the number of high-reflection object points in a preset false expansion identification area around the target data points to be determined.

[0207] In a specific implementation, since the depth distance of the pending data point is far relative to the depth of the laser radar in the depth direction, that is, the depth distance of the pending data point is greater than the preset character recognition distance, it is difficult for the laser radar to recognize the text information in the highly reflective object. In this case, there is no need to judge the false expansion point, and the pending data point can be directly used as a false expansion point.

[0208] However, if the depth distance of the pending data point is less than the preset character recognition distance, since the laser radar can normally recognize the text information in the highly reflective object at this time, it is still necessary to determine the false expansion point in the target pending data point through the number of highly reflective object points in the preset false expansion recognition area around the target pending data point.

[0209] As a specific example of the present invention, Figure 5 is a flow chart of a method for processing point cloud data provided in an embodiment of the present invention. The method for processing point cloud data may include the following steps:

[0210] 501: Traverse each point of the point cloud data of the current frame in the order of laser radar scanning.

[0211] 502: Determine whether each point satisfies the following three conditions at the same time:

[0212] A. The Z-axis distance value (z) of the point is greater than 0, that is, the point must be above the laser radar field of view.

[0213] B. The reflection intensity (intensity) of this point is greater than or equal to the reflection intensity threshold (high_ref) of a highly reflective object.

[0214] C.

[0215] a) The X-axis distance value (x) of the point is greater than or equal to the distance threshold of the highly reflective object (far_distance).

[0216] b) If the X-axis distance value (x) of the point is less than the high-reflective object distance threshold (far_distance), it is necessary to traverse each point in the a row and b column centered on the point, count the total number (sum) of points in the area that satisfy the reflection intensity (intensity) greater than or equal to the high-reflective object reflection intensity threshold (high_ref), and determine whether the percentage (sum / (a*b)) of the total number of points in the area that satisfy the reflection intensity (intensity) greater than or equal to the high-reflective object reflection intensity threshold (high_ref) is greater than or equal to the large high-reflective object space percentage threshold (p).

[0217] 503: If condition 502 is met, perform the following operations:

[0218] A. Longitudinal high-reflection expansion suppression algorithm: traverse the M points above and N points below the column where the high-reflection object point is located.

[0219] B. Lateral high-reflection expansion suppression algorithm: traverse the U points on the left and V points on the right of the row where the high-reflection object point is located.

[0220] 504, 506: Longitudinal high reflection expansion suppression: If the traversed point is found to meet both of the following conditions during the traversal process, the traversed point flag is set to false:

[0221] A. The traversed point and the high-reflection object point share the same plane vertically. High-reflection object (mostly high-reflection signs and traffic signs):

[0222] a) The absolute value of the distance difference between the traversed point and the highly reflective object point on the X-axis (Δx) is less than the coplanar distance difference threshold (F_dis_v) in the longitudinal X-axis direction.

[0223] b) The absolute value of the distance difference between the traversed point and the highly reflective object point on the Y axis (Δy) is less than the coplanar distance difference threshold (H_dis_v) in the longitudinal Y axis direction.

[0224] c) The absolute value of the distance difference between the traversed point and the highly reflective object point on the Z axis (Δz) is less than the coplanar distance difference threshold (V_dis_v) in the longitudinal Z axis direction.

[0225] B. The traversed points meet the reflection intensity characteristics of high-reflection expansion false points: the reflection intensity difference between the traversed point and the high-reflection object point (the reflection intensity of the high-reflection object point minus the reflection intensity of the traversed point, i.e. Δintensity) is greater than or equal to the reflection intensity difference threshold (intensity_diff) between the high-reflection object and the false expansion point.

[0226] 505,507: Lateral high reflection expansion suppression: If a point is found to meet both of the following conditions during the traversal process, the traversed point flag is set to false:

[0227] A. The traversed points and the highly reflective object point share the same plane horizontally with the highly reflective object (mostly highly reflective signs and traffic signs):

[0228] a) The absolute value of the distance difference between the traversed point and the highly reflective object point on the X-axis (Δx) is less than the coplanar distance difference threshold (F_dis_h) in the transverse X-axis direction.

[0229] b) The absolute value of the distance difference between the traversed point and the highly reflective object point on the Y axis (Δy) is less than the coplanar distance difference threshold (H_dis_h) in the lateral Y axis direction.

[0230] c) The absolute value of the distance difference between the traversed point and the highly reflective object point on the Z axis (Δz) is less than the coplanar distance difference threshold (V_dis_h) in the transverse Z axis direction.

[0231] B. The traversed points meet the reflection intensity characteristics of high-reflection expansion false points: the reflection intensity difference between the traversed point and the high-reflection object point (the reflection intensity of the high-reflection object point minus the reflection intensity of the traversed point, i.e. Δintensity) is greater than or equal to the reflection intensity difference threshold (intensity_diff) between the high-reflection object and its false expansion point.

[0232] 508: Traverse each point of the current frame point cloud data whose identification bit is false and whose X-axis distance value (x) is less than or equal to the low anti-information distance threshold (close_distance).

[0233] 509: Traverse each point in the c rows and d columns centered on the point where the flag bit is false.

[0234] 510: If the traversed point meets the following conditions: the reflection intensity (intensity) of the point is greater than or equal to the reflection intensity threshold (high_ref) of the high-reflective object.

[0235] 511: If each point in the c rows and d columns centered on the point with the flag bit set to false satisfies the above conditions, the high reflection point count (highref_num) in the area is increased by one, and the counting is stopped after each point in the area is traversed.

[0236] 512: Determine whether the percentage (highref_num / (c*d)) of the high reflection point count (highref_num) in the region to the total number of points (c*d) in the region is greater than or equal to the percentage threshold (q) in the high reflection object plane.

[0237] 513: If the condition 112 is met, update the point whose flag bit is false to true.

[0238] 514: If condition 112 is not met, the flag bit of the point will not be changed and will remain false.

[0239] 515: Output all points whose identifier is true in the current frame.

[0240] It should be noted that, for the sake of simplicity, the method embodiments are described as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present invention are not limited by the order of the actions described, because according to the embodiments of the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present invention.

[0241] Reference Figure 6 , shows a structural block diagram of a point cloud data processing device provided in an embodiment of the present invention, which may specifically include the following modules:

[0242] The point cloud acquisition module 601 is used to acquire point cloud data collected by the laser radar; the point cloud data includes a number of data points, and the data points are used to represent the three-dimensional coordinates and reflection intensity of objects in the surrounding environment of the laser radar;

[0243] A high-reflection object point search module 602 is used to search for a data point associated with a high-reflection object in the data points based on the reflection intensity as a high-reflection object point;

[0244] A target high-reflection object point determination module 603 is used to determine a target high-reflection object point among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point;

[0245] A false expansion point search module 604 is used to search for false expansion points in a preset false expansion identification area around the target high-reflection object point based on the reflection intensity;

[0246] The target point cloud data acquisition module 605 is used to delete the false expansion points in the point cloud data to obtain the target point cloud data.

[0247] Optionally, the highly reflective object point search module includes:

[0248] The first high-reflection object point search submodule is used to search the data points for a data point whose reflection intensity is greater than a preset high-reflection object reflection intensity threshold value as a high-reflection object point associated with the high-reflection object.

[0249] Optionally, the highly reflective object point search module includes:

[0250] A longitudinal position acquisition submodule, used for acquiring the longitudinal position of the data point in the longitudinal direction based on the three-dimensional coordinates of the data point;

[0251] a target data point determination submodule, configured to take the data point as a target data point when the longitudinal position of the data point is higher than the longitudinal position of the laser radar;

[0252] The second high-reflection object point search submodule is used to search for data points associated with high-reflection objects in the target data points based on the reflection intensity as high-reflection object points.

[0253] Optionally, the target highly reflective object point determination module includes:

[0254] The first target highly reflective object point determination submodule uses the highly reflective object point as a target highly reflective object point if the volume of the highly reflective object corresponding to the highly reflective object point meets a preset volume recognition condition.

[0255] Optionally, the device further comprises:

[0256] A high-reflection object point set acquisition module is used to, for any of the high-reflection object points, search for other high-reflection object points in a preset volume recognition area around the high-reflection object point with the high-reflection object point as the center, to obtain a high-reflection object point set; the high-reflection object point set includes at least one high-reflection object point;

[0257] A high reflection percentage determination module, used to determine the percentage of the number of the high reflection object points in the high reflection object point set to the total number of data points in the volume recognition area;

[0258] A first recognition condition determination module, configured to determine that the volume of the high-reflective object corresponding to the high-reflective object point satisfies a preset volume recognition condition if the percentage of the number of the high-reflective object points in the high-reflective object point set to the total number of data points in the volume recognition area is greater than a preset large high-reflective object percentage;

[0259] The second recognition condition determination module is used to determine that the volume of the high-reflective object corresponding to the high-reflective object point does not meet the preset volume recognition condition if the percentage of the number of high-reflective object points in the high-reflective object point set to the total number of data points in the volume recognition area is not greater than the preset large-scale high-reflective object percentage.

[0260] Optionally, the target highly reflective object point determination module includes:

[0261] A high-reflection depth distance determination submodule, used to obtain the depth distance of the high-reflection object point relative to the laser radar in the depth direction based on the three-dimensional coordinates of the high-reflection object point;

[0262] The second target high-reflection object point determination submodule is used to determine the target high-reflection object point among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point when the depth distance of the high-reflection object point is less than a preset volume recognition distance.

[0263] Optionally, the device further comprises:

[0264] The volume recognition distance determination module is used to take the high-reflection object point as a target high-reflection object point when the depth distance of the high-reflection object point is greater than a preset volume recognition distance.

[0265] Optionally, the base false expansion point search module includes:

[0266] A pending data point search submodule is used to search for pending data points in a preset false expansion recognition area around the target high-reflection object point based on the reflection intensity;

[0267] The false expansion point determination submodule is used to determine the false expansion point in the pending data point based on the number of high-reflection object points in a preset false expansion identification area around the pending data point.

[0268] Optionally, the pending data point search submodule includes:

[0269] A coplanar data point searching unit, configured to search, within a preset false expansion identification area around the target highly reflective object point, at least one data point located on the same object plane as the target highly reflective object point as a coplanar data point;

[0270] The pending data point searching unit is used to determine the pending data point among the coplanar data points based on the reflection intensity of the target high-reflection object point and the reflection intensity of the coplanar data point.

[0271] Optionally, the data points in the point cloud data are arranged in a matrix form to form a point cloud matrix;

[0272] The coplanar data point searching unit comprises:

[0273] A horizontal and vertical data point search subunit is used to search for data points in the point cloud matrix that are located in the same column as the target highly reflective object point as the vertical data point, and / or to search for data points in the point cloud matrix that are located in the same row as the target highly reflective object point as the horizontal data point, within a preset false expansion identification area around the target highly reflective object point;

[0274] A coplanar data point search subunit is used to determine coplanar data points in the longitudinal data points based on the three-dimensional coordinates of the target highly reflective object point and the three-dimensional coordinates of the longitudinal data points; and / or to determine coplanar data points in the transverse data points based on the three-dimensional coordinates of the target highly reflective object point and the three-dimensional coordinates of the transverse data points.

[0275] Optionally, the pending data point searching unit includes:

[0276] The pending data point search subunit is used to take the coplanar data point as a pending data point if the difference in reflection intensity between the target high-reflection object point and the coplanar data point is greater than a preset reflection intensity difference threshold.

[0277] Optionally, the false expansion point determination submodule includes:

[0278] A false expansion region high reflective point search unit is used to search for high reflective object points in a false expansion recognition region preset around any pending data point for any pending data point, and obtain the number of high reflective object points in the false expansion recognition region;

[0279] a false expansion region high reflection point percentage determination unit, used to determine the percentage of the number of high reflection object points in the false expansion identification region to the total number of data points in the false expansion identification region;

[0280] The first false expansion point determination unit is configured to, when a percentage of the number of high-reflection object points in the false expansion identification area to the total number of data points in the false expansion identification area is less than a preset false expansion percentage, take the pending data point as a false expansion point.

[0281] Optionally, the false expansion point determination submodule includes:

[0282] A depth distance determination unit for a pending data point, configured to obtain a depth distance of the pending data point relative to the laser radar in a depth direction based on the three-dimensional coordinates of the pending data point;

[0283] a target pending data point determination unit, configured to use the pending data point as a target pending data point if the depth distance of the pending data point is less than a preset character recognition distance;

[0284] The second false expansion point determination unit is configured to determine a false expansion point in the target data point to be determined based on the number of high-reflection object points in a preset false expansion identification area around the target data point to be determined.

[0285] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0286] In addition, an embodiment of the present invention further provides an electronic device, such as Figure 7 As shown, it includes a processor 701, a communication interface 702, a memory 703 and a communication bus 704, wherein the processor 701, the communication interface 702, and the memory 703 communicate with each other through the communication bus 704.

[0287] Memory 703, used for storing computer programs;

[0288] The processor 701 is used to execute the program stored in the memory 703 to implement the following steps:

[0289] Acquire point cloud data collected by a laser radar; the point cloud data includes a plurality of data points, and the data points are used to represent the three-dimensional coordinates and reflection intensity of objects in the surrounding environment of the laser radar;

[0290] Based on the reflection intensity, searching the data points associated with the high-reflection object as the high-reflection object point;

[0291] Determining a target high-reflection object point among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point;

[0292] Based on the reflection intensity, searching for a false expansion point in a preset false expansion identification area around the target high-reflection object point;

[0293] The false expansion points are deleted from the point cloud data to obtain target point cloud data.

[0294] Optionally, the step of searching, based on the reflection intensity, data points associated with the high-reflection object in the data points as the high-reflection object points comprises:

[0295] Search the data points for data points whose reflection intensity is greater than a preset high-reflection object reflection intensity threshold value as high-reflection object points associated with the high-reflection object.

[0296] Optionally, the step of searching, based on the reflection intensity, data points associated with the high-reflection object in the data points as the high-reflection object points comprises:

[0297] Based on the three-dimensional coordinates of the data point, obtaining the longitudinal position of the data point in the longitudinal direction;

[0298] When the longitudinal position of the data point is higher than the longitudinal position of the laser radar, taking the data point as a target data point;

[0299] Based on the reflection intensity, a data point associated with a high-reflection object is searched in the target data points as a high-reflection object point.

[0300] Optionally, the step of determining a target high-reflection object point among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point comprises:

[0301] If the volume of the high-reflective object corresponding to the high-reflective object point meets the preset volume recognition condition, the high-reflective object point is used as the target high-reflective object point.

[0302] Optionally, the method further comprises:

[0303] For any of the highly reflective object points, taking the highly reflective object point as the center, searching for other highly reflective object points in a preset volume recognition area around the highly reflective object point to obtain a set of highly reflective object points; the set of highly reflective object points includes at least one highly reflective object point;

[0304] Determine the percentage of the number of the high-reflective object points in the high-reflective object point set to the total number of data points in the volume recognition area;

[0305] If the percentage of the number of the high-reflective object points in the high-reflective object point set to the total number of data points in the volume recognition area is greater than the preset large high-reflective object percentage, it is determined that the volume of the high-reflective object corresponding to the high-reflective object point meets the preset volume recognition condition

[0306] If the percentage of the number of highly reflective object points in the highly reflective object point set to the total number of data points in the volume identification area is not greater than the preset percentage of large highly reflective objects, it is determined that the volume of the highly reflective object corresponding to the highly reflective object point does not meet the preset volume identification conditions.

[0307] Optionally, the step of determining a target high-reflection object point among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point comprises:

[0308] Based on the three-dimensional coordinates of the high-reflective object point, obtaining the depth distance of the high-reflective object point relative to the laser radar in the depth direction;

[0309] When the depth distance of the high-reflection object point is less than the preset volume recognition distance, a target high-reflection object point is determined among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point.

[0310] Optionally, the method further comprises:

[0311] In a case where the depth distance of the high-reflection object point is greater than a preset volume recognition distance, the high-reflection object point is used as a target high-reflection object point.

[0312] Optionally, the step of searching for a false expansion point in a preset false expansion identification area around the target high-reflection object point based on the reflection intensity comprises:

[0313] Based on the reflection intensity, searching for undetermined data points in a preset false expansion identification area around the target high-reflection object point;

[0314] Based on the number of high-reflection object points in a preset false expansion identification area around the pending data point, a false expansion point is determined in the pending data point.

[0315] Optionally, the step of searching for undetermined data points in a preset false expansion identification area around the target high-reflection object point based on the reflection intensity includes:

[0316] In a preset false expansion identification area around the target highly reflective object point, searching for at least one data point located on the same object plane as the target highly reflective object point as a coplanar data point;

[0317] Based on the reflection intensity of the target high-reflection object point and the reflection intensity of the coplanar data points, a pending data point is determined among the coplanar data points.

[0318] Optionally, the data points in the point cloud data are arranged in a matrix form to form a point cloud matrix;

[0319] The step of searching at least one data point located on the same object plane as the target highly reflective object point in a preset false expansion identification area around the target highly reflective object point as a coplanar data point comprises:

[0320] In a preset false expansion identification area around the target highly reflective object point, searching the point cloud matrix for data points located in the same column as the target highly reflective object point as longitudinal data points, and / or searching the point cloud matrix for data points located in the same row as the target highly reflective object point as transverse data points;

[0321] Based on the three-dimensional coordinates of the target highly reflective object point and the three-dimensional coordinates of the longitudinal data points, coplanar data points are determined among the longitudinal data points; and / or, based on the three-dimensional coordinates of the target highly reflective object point and the three-dimensional coordinates of the transverse data points, coplanar data points are determined among the transverse data points.

[0322] Optionally, the step of determining a pending data point among the coplanar data points based on the reflection intensity of the target highly reflective object point and the reflection intensity of the coplanar data point comprises:

[0323] If the difference in reflection intensity between the target highly reflective object point and the coplanar data point is greater than a preset reflection intensity difference threshold, the coplanar data point is used as a pending data point.

[0324] Optionally, the step of determining the false expansion point in the pending data point based on the number of high-reflection object points in a preset false expansion identification area around the pending data point comprises:

[0325] For any of the pending data points, searching for high-reflection object points in a preset false expansion identification area around the pending data point to obtain the number of high-reflection object points in the false expansion identification area;

[0326] Determine the percentage of the number of highly reflective object points in the false expansion identification area to the total number of data points in the false expansion identification area;

[0327] If the percentage of the number of high-reflection object points in the false expansion identification area to the total number of data points in the false expansion identification area is less than a preset false expansion percentage, the pending data point is taken as a false expansion point.

[0328] Optionally, the step of determining the false expansion point in the pending data point based on the number of high-reflection object points in a preset false expansion identification area around the pending data point includes:

[0329] Based on the three-dimensional coordinates of the pending data point, obtaining the depth distance of the pending data point relative to the laser radar in the depth direction;

[0330] If the depth distance of the pending data point is less than the preset character recognition distance, the pending data point is used as the target pending data point;

[0331] Based on the number of high-reflection object points in a preset false expansion identification area around the target pending data point, a false expansion point is determined in the target pending data point.

[0332] The communication bus mentioned in the above terminal can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0333] The communication interface is used for communication between the above terminal and other devices.

[0334] The memory may include a random access memory (RAM) or a non-volatile memory, such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0335] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0336] like Figure 8 As shown, in another embodiment provided by the present invention, a computer-readable storage medium 801 is also provided, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer executes the point cloud data processing method described in the above embodiment.

[0337] In another embodiment provided by the present invention, a computer program product including instructions is also provided, and when the computer is run on the computer, the computer executes the method for processing point cloud data described in the above embodiment.

[0338] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state hard disk Solid State Disk (SSD)), etc.

[0339] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0340] Each embodiment in this specification is described in a related manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0341] The above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention are included in the protection scope of the present invention.

Claims

1. A method for processing point cloud data, characterized in that: include: Acquire point cloud data collected by a laser radar; the point cloud data includes a plurality of data points, and the data points are used to represent the reflection intensity of objects in the surrounding environment of the laser radar; Finding a false expansion point based on the reflection intensity; The false expansion points are deleted from the point cloud data to obtain target point cloud data.

2. The method according to claim 1, characterized in that The step of searching for a false expansion point based on the reflection intensity includes: Based on the reflection intensity, searching the data points associated with the high-reflection object as the high-reflection object point; Determining a target high-reflection object point among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point; Based on the reflection intensity, a false expansion point is searched in a preset false expansion identification area around the target high-reflection object point.

3. The method according to claim 2, characterized in that The step of searching, based on the reflection intensity, data points associated with the high-reflection object in the data points as the high-reflection object points comprises: Search the data points for data points whose reflection intensity is greater than a preset high-reflection object reflection intensity threshold value as high-reflection object points associated with the high-reflection object.

4. The method according to claim 2 or 3, characterized in that: The step of searching, based on the reflection intensity, data points associated with the high-reflection object in the data points as the high-reflection object points comprises: Based on the three-dimensional coordinates of the data point, obtaining the longitudinal position of the data point in the longitudinal direction; When the longitudinal position of the data point is higher than the longitudinal position of the laser radar, taking the data point as a target data point; Based on the reflection intensity, a data point associated with a high-reflection object is searched in the target data points as a high-reflection object point.

5. The method according to claim 2, characterized in that: The step of determining a target high-reflection object point among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point comprises: If the volume of the high-reflective object corresponding to the high-reflective object point meets the preset volume recognition condition, the high-reflective object point is used as the target high-reflective object point.

6. The method according to claim 5, characterized in that The method further comprises: For any of the highly reflective object points, taking the highly reflective object point as the center, searching for other highly reflective object points in a preset volume recognition area around the highly reflective object point to obtain a set of highly reflective object points; the set of highly reflective object points includes at least one highly reflective object point; Determine the percentage of the number of the high-reflective object points in the high-reflective object point set to the total number of data points in the volume recognition area; If the percentage of the number of the high-reflective object points in the high-reflective object point set to the total number of data points in the volume recognition area is greater than a preset large high-reflective object percentage, it is determined that the volume of the high-reflective object corresponding to the high-reflective object point meets the preset volume recognition condition; If the percentage of the number of highly reflective object points in the highly reflective object point set to the total number of data points in the volume identification area is not greater than the preset percentage of large highly reflective objects, it is determined that the volume of the highly reflective object corresponding to the highly reflective object point does not meet the preset volume identification conditions.

7. The method according to claim 2, characterized in that The step of determining a target high-reflection object point among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point comprises: Based on the three-dimensional coordinates of the high-reflective object point, obtaining the depth distance of the high-reflective object point relative to the laser radar in the depth direction; When the depth distance of the high-reflection object point is less than the preset volume recognition distance, a target high-reflection object point is determined among the high-reflection object points based on the volume of the high-reflection object corresponding to the high-reflection object point.

8. The method according to claim 7, characterized in that The method further comprises: In a case where the depth distance of the high-reflection object point is greater than a preset volume recognition distance, the high-reflection object point is used as a target high-reflection object point.

9. The method according to claim 2, characterized in that: The step of searching for a false expansion point in a preset false expansion identification area around the target high-reflection object point based on the reflection intensity comprises: Based on the reflection intensity, searching for undetermined data points in a preset false expansion identification area around the target high-reflection object point; Based on the number of high-reflection object points in a preset false expansion identification area around the pending data point, a false expansion point is determined in the pending data point.

10. The method according to claim 9, characterized in that The step of searching for undetermined data points in a preset false expansion identification area around the target high-reflection object point based on the reflection intensity comprises: In a preset false expansion identification area around the target highly reflective object point, searching for at least one data point located on the same object plane as the target highly reflective object point as a coplanar data point; Based on the reflection intensity of the target high-reflection object point and the reflection intensity of the coplanar data points, a pending data point is determined among the coplanar data points.

11. The method according to claim 10, characterized in that The data points in the point cloud data are arranged in a matrix form to form a point cloud matrix; The step of searching at least one data point located on the same object plane as the target highly reflective object point in a preset false expansion identification area around the target highly reflective object point as a coplanar data point comprises: In a preset false expansion identification area around the target highly reflective object point, searching the point cloud matrix for data points located in the same column as the target highly reflective object point as longitudinal data points, and / or searching the point cloud matrix for data points located in the same row as the target highly reflective object point as transverse data points; Based on the three-dimensional coordinates of the target highly reflective object point and the three-dimensional coordinates of the longitudinal data points, coplanar data points are determined among the longitudinal data points; and / or, based on the three-dimensional coordinates of the target highly reflective object point and the three-dimensional coordinates of the transverse data points, coplanar data points are determined among the transverse data points.

12. The method according to claim 10 or 11, characterized in that: The step of determining the pending data points among the coplanar data points based on the reflection intensity of the target highly reflective object point and the reflection intensity of the coplanar data points comprises: If the difference in reflection intensity between the target highly reflective object point and the coplanar data point is greater than a preset reflection intensity difference threshold, the coplanar data point is used as a pending data point.

13. The method according to claim 9, characterized in that The step of determining the false expansion point in the pending data point based on the number of high-reflection object points in a preset false expansion identification area around the pending data point comprises: For any of the pending data points, searching for high-reflection object points in a preset false expansion identification area around the pending data point to obtain the number of high-reflection object points in the false expansion identification area; Determine the percentage of the number of highly reflective object points in the false expansion identification area to the total number of data points in the false expansion identification area; If the percentage of the number of high-reflection object points in the false expansion identification area to the total number of data points in the false expansion identification area is less than a preset false expansion percentage, the pending data point is taken as a false expansion point.

14. The method according to claim 9 or 13, characterized in that: The step of determining the false expansion point in the pending data point based on the number of high-reflection object points in a preset false expansion identification area around the pending data point comprises: Based on the three-dimensional coordinates of the pending data point, obtaining the depth distance of the pending data point relative to the laser radar in the depth direction; If the depth distance of the pending data point is less than the preset character recognition distance, the pending data point is used as the target pending data point; Based on the number of high-reflection object points in a preset false expansion identification area around the target pending data point, a false expansion point is determined in the target pending data point.

15. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; The memory is used to store computer programs; The processor is used to implement the method according to any one of claims 1 to 14 when executing the program stored in the memory.

16. A computer-readable medium having instructions stored thereon, which, when executed by one or more processors, cause the processors to perform the method according to any one of claims 1 to 14.

Images