Point cloud data detection method and device, equipment and storage medium

Abstract

The present disclosure provides a point cloud data detection method and device, equipment and a storage medium, relates to the technical field of computers, and in particular to the technical field of automatic driving, high-precision maps and navigation. The specific implementation scheme is that a detection device acquires a plurality of point cloud objects, the plurality of point cloud objects are point cloud objects in target point cloud data composed of a plurality of groups of point cloud data, and each group of point cloud data includes at least one point cloud object. Then, the detection device determines at least one feature pair from the plurality of point cloud objects, and the point cloud objects in one feature pair are the same. For each feature pair, the detection device acquires a first distance corresponding to the feature pair to acquire at least one first distance, and the first distance is the distance between the point cloud objects in the feature pair. The detection device determines first information based on the at least one first distance, and the first information is used to indicate whether the target point cloud data is abnormal in splicing.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and in particular to the fields of autonomous driving, high-precision maps, and navigation technology, specifically to a method, apparatus, device, and storage medium for detecting point cloud data. Background Technology

[0002] With the development of transportation, roads are becoming increasingly complex, leading to a greater demand for high-definition maps. High-definition maps, also known as high-precision maps, possess accurate vehicle location information and rich road element data. In the creation of high-definition maps, electronic devices typically need to collect point cloud data of a region multiple times and then stitch these collected point cloud data together to obtain a stitched point cloud dataset. The electronic device can then use this stitched point cloud dataset to create a high-definition map. Summary of the Invention

[0003] This disclosure provides a method, apparatus, device, and storage medium for detecting point cloud data.

[0004] Firstly, this disclosure provides a method for detecting point cloud data, including:

[0005] A point cloud data detection device (hereinafter referred to as the "detection device") acquires multiple point cloud objects. These multiple point cloud objects are point cloud objects in a target point cloud dataset composed of multiple sets of point cloud data, with each set of point cloud data including at least one point cloud object. The detection device then determines at least one feature pair from the multiple point cloud objects, where the point cloud objects in a feature pair are identical. For each feature pair, the detection device acquires a first distance corresponding to the feature pair, thus obtaining at least one first distance, which is the distance between the point cloud objects in the feature pair. Based on at least one first distance, the detection device determines first information, which is used to indicate whether the target point cloud data has an abnormal stitching pattern.

[0006] Secondly, this disclosure provides a point cloud data detection device, comprising:

[0007] The acquisition unit is used to acquire multiple point cloud objects, which are point cloud objects in a target point cloud data composed of multiple sets of point cloud data stitched together, with each set of point cloud data including at least one point cloud object. The processing unit is used to determine at least one feature pair from the multiple point cloud objects, where the point cloud objects in a feature pair are identical. The processing unit is also used to acquire a first distance corresponding to each feature pair, thereby acquiring at least one first distance, where the first distance is the distance between the point cloud objects in the feature pair. The processing unit is further used to determine first information based on at least one first distance, where the first information is used to indicate whether the target point cloud data has a stitching anomaly.

[0008] Thirdly, this disclosure provides an electronic device, including:

[0009] At least one processor; and

[0010] A memory communicatively connected to at least one processor; wherein the memory stores instructions executable by at least one processor, the instructions being executed by at least one processor to enable at least one processor to perform any of the methods in the first aspect.

[0011] Fourthly, this disclosure provides a non-transitory computer-readable storage medium storing computer instructions, comprising:

[0012] Computer instructions are used to cause the computer to perform any of the methods in the first aspect.

[0013] Fifthly, this disclosure provides a computer program product, including:

[0014] A computer program, any one of the methods of a computer program in the first aspect of being executed by a processor.

[0015] The technology disclosed herein solves the problem that the detection of point cloud data relies on manual quality inspection, and improves the accuracy of detecting whether point cloud data is abnormal.

[0016] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0017] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0018] Figure 1 This is an example diagram of stitched point cloud data provided in an embodiment of this disclosure;

[0019] Figure 2 This is a flowchart illustrating a point cloud data detection method provided in an embodiment of this disclosure;

[0020] Figure 3 This is a flowchart illustrating another point cloud data detection method provided in this embodiment of the disclosure;

[0021] Figure 4 This is an example diagram of a point cloud object provided in an embodiment of this disclosure;

[0022] Figure 5 This is an example diagram of a distance provided in an embodiment of this disclosure;

[0023] Figure 6This is a flowchart illustrating another point cloud data detection method provided in this embodiment of the disclosure;

[0024] Figure 7 This is a schematic diagram of the structure of a point cloud data detection device provided in an embodiment of this disclosure;

[0025] Figure 8 This is a block diagram of an electronic device for a point cloud data detection method provided in an embodiment of this disclosure. Detailed Implementation

[0026] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0027] The collection, storage, use, processing, transmission, provision, and disclosure of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0028] Before providing a detailed description of the point cloud data detection method of this disclosure, the application scenarios of this disclosure will be introduced first.

[0029] First, the application scenarios of the embodiments of this disclosure will be introduced.

[0030] With the development of transportation, roads are becoming increasingly complex, leading to a growing demand for high-precision maps. High-precision maps, also known as high-resolution maps, are used by autonomous vehicles. They possess accurate vehicle location information and rich road element data, helping cars anticipate complex road conditions such as slope, curvature, and heading, thereby better mitigating potential risks.

[0031] Currently, the production of high-precision maps typically begins with multiple point cloud data acquisitions for a specific road segment. Simultaneously, data is collected using the Global Navigation Satellite System (GNSS) and Inertial Measurement Unit (IMU). Based on this data, global pose information is estimated, and the global pose of the point cloud data acquisition device is obtained from the estimation results. This is then used to stitch together the multiple point cloud data sets. Finally, electronic equipment can use the stitched point cloud data to create a high-precision map.

[0032] For example, such as Figure 1 As shown, it illustrates the stitched point cloud data 101, which consists of point cloud data collected in two separate datasets (e.g., point cloud data A and point cloud data B). Point cloud data A is represented by dashed lines (e.g., arrow 103, sign 105, manhole cover 107, and lane line 109), while point cloud data B is represented by solid lines (e.g., arrow 102, sign 104, manhole cover 106, and lane line 108). The stitched point cloud data 101 can include multiple point cloud objects, such as directional arrows (e.g., arrows 102 and 103), signs (e.g., signs 104 and 105), and manhole covers (e.g., manhole covers 106 and 107).

[0033] However, due to interference from external factors (such as dim lighting), the point cloud data collected by electronic devices may be inaccurate, which in turn may cause anomalies in the stitched point cloud data (such as the same point cloud objects in different point cloud data not overlapping), ultimately resulting in low accuracy of the high-precision map.

[0034] For example, in combination Figure 1 It can be seen that arrows 102 and 103 are the same arrow, but they do not completely overlap; sign 104 and sign 105 are the same sign, but they also do not completely overlap.

[0035] Currently, staff can perform quality checks on the stitched point cloud data to determine whether identical point cloud objects overlap. However, manually inspecting the stitched point cloud data is not only less accurate but also less efficient.

[0036] To address the aforementioned problems, this disclosure provides a method for detecting point cloud data, applicable to scenarios involving the detection of point cloud data. In this method, a detection device acquires multiple point cloud objects, which are point cloud objects within a target point cloud dataset composed of different point cloud data. The detection device then identifies at least one feature pair from the multiple point cloud objects, where the point cloud objects in the feature pair are identical. For each feature pair, the detection device acquires a first distance corresponding to the feature pair, obtaining at least one first distance, which is the distance between the point cloud objects in the feature pair. Based on this first distance, the detection device then determines whether the target point cloud data exhibits splicing anomalies.

[0037] Understandably, after acquiring multiple point cloud objects, the detection device identifies at least one feature pair, resulting in point cloud objects identical to each of these feature pairs. Then, the detection device acquires at least one first distance. Since this first distance represents the distance between point cloud objects within a feature pair, it reflects the degree of overlap between identical point cloud objects in different point cloud datasets. Thus, based on at least one first distance—the degree of overlap between multiple sets of identical point cloud objects—the detection device can determine whether the target point cloud data exhibits stitching anomalies. This achieves the goal of detecting stitched point cloud data, and by calculating the distance between identical point cloud objects, the accuracy of detecting stitching anomalies in the stitched point cloud data can be improved.

[0038] It should be noted that the target point cloud data (i.e., the stitched point cloud data) is not limited in the embodiments disclosed herein. For example, the target point cloud data can be point cloud data generated in real time. As another example, the target point cloud data can be updated point cloud data.

[0039] For example, when the target point cloud data is real-time generated point cloud data, it can be point cloud data stitched together after the acquisition device collects multiple sets of point cloud data at a high frequency. For instance, while a car is driving, it can collect point cloud data frames every 100 milliseconds and stitch together multiple point cloud data frames (such as point cloud keyframes) to generate the target point cloud data. Alternatively, when the target point cloud data is updated point cloud data, it can be point cloud data stitched together after the acquisition device collects multiple sets of point cloud data at a low frequency. For instance, the acquisition device can collect one set of point cloud data per day and stitch together the point cloud data frames collected each day to generate the target point cloud data.

[0040] It should be noted that the embodiments of this disclosure do not limit the detection equipment. The detection equipment in the embodiments of this disclosure can be an electronic device, such as a tablet computer, mobile phone, desktop computer, laptop computer, handheld computer, notebook computer, ultra-mobile personal computer (UMPC), netbook, as well as cellular phone, personal digital assistant (PDA), augmented reality (AR) / virtual reality (VR) device, in-vehicle device, etc. The embodiments of this disclosure do not impose special limitations on the specific form of the electronic device.

[0041] Optionally, the electronic device can also be a server, which can be a physical server or a cloud server. Alternatively, the server can be a server cluster.

[0042] Optionally, the testing equipment can also be used for automobiles.

[0043] Optionally, the detection device (such as a vehicle) can collect point cloud data. The vehicle may include sensors for sensing its surroundings. These sensors may include one or more of the following: visual cameras, infrared cameras, ultrasonic sensors, millimeter-wave radar, and lidar (LiDAR). Different sensors can provide different detection accuracy and range. Cameras can be mounted in front of, behind, or at other locations on the vehicle. Visual cameras can capture real-time information about the interior and exterior of the vehicle and present it to the driver and / or passengers. Furthermore, by analyzing the images captured by the visual cameras, information such as traffic light signals, intersection conditions, and the operating status of other vehicles can be obtained. Infrared cameras can capture objects in night vision conditions. Ultrasonic sensors can be mounted around the vehicle to measure the distance to objects outside the vehicle using the strong directionality of ultrasound. Millimeter-wave radar can be mounted in front of, behind, or at other locations on the vehicle to measure the distance to objects outside the vehicle using the characteristics of electromagnetic waves. LiDAR can be mounted in front of, behind, or at other locations on the vehicle to detect object edges and shape information for object recognition and tracking. Due to the Doppler effect, the radar device can also measure changes in the speed of the vehicle and moving objects.

[0044] The execution entity of the point cloud data detection method provided in this disclosure can be a detection device (such as a car, electronic device, etc.). Furthermore, the device can also be the central processing unit (CPU) of the detection device, or a detection module within the detection device used for detecting point cloud data. This disclosure uses an example of a detection device executing the point cloud data detection method to illustrate the point cloud data detection method provided in this disclosure.

[0045] like Figure 2 As shown, this is a method for detecting point cloud data provided in an embodiment of the present disclosure. The method includes:

[0046] S201, The detection equipment acquires multiple point cloud objects.

[0047] Among them, multiple point cloud objects are point cloud objects in target point cloud data composed of multiple sets of point cloud data spliced ​​together, and each set of point cloud data includes at least one point cloud object.

[0048] For example, the target point cloud data is composed of point cloud data A and point cloud data B. The target point cloud data includes: point cloud object a, point cloud object b, point cloud object c, point cloud object d, and point cloud object e. Among them, point cloud object a, point cloud object b, and point cloud object c are point cloud objects in point cloud data A, and point cloud object d and point cloud object e are point cloud objects in point cloud data B.

[0049] In one possible implementation, the detection device can acquire the stitched point cloud data. Then, the detection device can perform semantic segmentation on the stitched point cloud data to obtain multiple point cloud objects.

[0050] It should be noted that the point cloud object in this embodiment can also be referred to as a semantic feature.

[0051] In another possible implementation, the detection device stores point cloud objects from each set of point cloud data.

[0052] Optionally, the region corresponding to each set of point cloud data is smaller than a first preset region range threshold. Alternatively, the region corresponding to each set of point cloud data is larger than a second preset region range threshold, where the first preset region range threshold is greater than the second preset region range threshold.

[0053] It should be noted that the preset region range threshold is not limited in the embodiments disclosed herein. For example, the preset region range threshold can be 10×10. Another example is that the preset region range threshold can be 5×10. Yet another example is that the preset region range threshold can be 6×8.

[0054] S202, The detection device determines at least one feature pair from multiple point cloud objects.

[0055] In one feature pair, the point cloud objects are identical.

[0056] For example, in combination Figure 1 Arrow 102 and arrow 103 can be a feature pair, sign 104 and sign 105 can be a feature pair, and manhole cover 106 and manhole cover 107 can be a feature pair.

[0057] In one possible implementation, the detection device can divide multiple point cloud objects into at least one feature pair according to preset conditions. For each point cloud object, the detection device can acquire the object type of the point cloud object, the identifier of the point cloud data group to which the point cloud object belongs, and the location information of the point cloud data. Then, the detection device can divide the multiple point cloud objects into at least one feature pair according to the object type of the point cloud object, the point cloud data identifier, and the location information of the point cloud data, based on the preset conditions.

[0058] The feature pair includes a first object and a second object. Preset conditions include: the object type of the first object and the object type of the second object are the same; the point cloud data group where the first object is located is different from the point cloud data group where the second object is located; the second distance is less than a first preset distance threshold; and the second distance is the minimum distance in the distance set, which includes: the distance between the center point of the first object and the center points of all point cloud objects except the first object; the distance between the center point of the second object and the center points of all point cloud objects except the second object; the second distance is the distance between the center points of the first object and the center points of the second object.

[0059] In this embodiment of the disclosure, the object type may include a line type and a surface type. The object type of the first object is the same as the object type of the second object.

[0060] For example, linear point cloud objects may include lane lines, curbs, guardrails, poles, etc., while area point cloud objects may include signs, ground arrows, speed limit signs, ground, etc.

[0061] It is understandable that if the first object and the second object are the same object, then the object types of the first object and the second object are the same.

[0062] In one possible design, a set of point cloud data corresponds to a single point cloud data identifier. The detection device can determine whether two point cloud objects belong to the same point cloud data set based on their corresponding point cloud data identifiers. If the two point cloud objects have the same identifier, the detection device determines they belong to the same point cloud data set. If the two point cloud objects have different identifiers, the detection device also determines they belong to the same point cloud data set.

[0063] It should be noted that the point cloud data identifier in this disclosure is not limited. For example, the point cloud data identifier can be the ID of the task that collects the point cloud data (such as taskid). Another example is that the point cloud data identifier can be the timestamp of the point cloud data collection (such as timestamp). Yet another example is that the point cloud data identifier can be a combination of ID and timestamp.

[0064] In this embodiment of the disclosure, the detection device can acquire a second distance. Then, the detection device can compare the second distance with a first preset distance threshold to determine whether the second distance is less than the first preset distance threshold.

[0065] In one possible design, the second distance can be determined using Equation 1.

[0066] S(f i ,f j )=|C i -Cj Formula 1.

[0067] Wherein, S(f i ,f j f is used to represent the distance between the center point of the i-th point cloud object and the center point of the j-th point cloud object. i f is used to represent the i-th point cloud object. i C is used to represent the j-th point cloud object. i C is used to represent the center point of the i-th point cloud object. j Used to represent the center point of the j-th point cloud object.

[0068] It should be noted that the embodiments disclosed herein do not limit the first preset distance threshold. For example, the first preset distance threshold can be 5 centimeters. Another example is that the first preset distance threshold can be 10 centimeters. Yet another example is that the first preset distance threshold can be 1 meter.

[0069] It is understandable that if the distance between two point cloud objects is less than the first preset distance threshold, it means that the two point cloud objects are close to each other and have a high degree of overlap, indicating that the two point cloud objects may be a feature pair.

[0070] In this embodiment of the disclosure, the detection device can acquire a set of distances and determine the minimum distance in the set of distances.

[0071] For example, suppose the point cloud objects in point cloud data A include point cloud object a and point cloud object b, and the point cloud objects in point cloud data B include point cloud object d and point cloud object e. If the first object is point cloud object a and the second object is point cloud object d, then the distance set A includes the distance between point cloud object a and point cloud object d, the distance between point cloud object a and point cloud object e, and the distance between point cloud object e and point cloud object b. Among these, the distance between point cloud object a and point cloud object d is the minimum distance in the distance set A.

[0072] Understandably, the detection equipment divides feature pairs according to preset conditions. If two point cloud objects have the same object type, belong to different point cloud data groups, and the distance between them is less than a first preset distance threshold, then the two point cloud objects are determined to be the same point cloud object. In this way, feature pairs can be divided based on object type, their respective point cloud data groups, and the distance between them, thus improving the accuracy of feature pair determination.

[0073] It should be noted that the target point cloud data may be composed of multiple sets of point cloud data. When the target point cloud data is composed of three or more sets of point cloud data, there may be three or more identical point cloud objects in the target point cloud data.

[0074] In some embodiments, the detection device can divide multiple sets of point cloud data into multiple point cloud datasets, each point cloud dataset including two sets of point cloud data. Then, the detection device can acquire multiple point cloud objects from each point cloud dataset and determine at least one feature pair from the multiple point cloud objects.

[0075] S203. For each feature pair, the detection device acquires the first distance corresponding to the feature pair to acquire at least one first distance.

[0076] The first distance is the distance between point cloud objects in the feature pair.

[0077] In one possible implementation, the detection device can acquire the position information of two point cloud objects in a feature pair, and determine the first distance between the feature pairs based on the position information of the two point cloud objects.

[0078] S204. The detection device determines first information based on at least one first distance.

[0079] The first piece of information is used to indicate whether the target point cloud data is spliced ​​abnormally.

[0080] It should be noted that, in this embodiment of the disclosure, point cloud data stitching anomaly refers to the situation where, in the stitched point cloud data, the same object from different point cloud data sets is not located in the same position after stitching. That is, the same object from different point cloud data sets does not overlap after stitching.

[0081] In one possible design, the first information may include either the second or the third information. The second information is used to indicate an anomaly in the target point cloud data stitching, while the third information is used to indicate that the target point cloud data stitching is normal.

[0082] Optionally, the second information may include the object identifier of the point cloud object with the stitching anomaly.

[0083] It should be noted that the object identifier of the point cloud object is not limited in the embodiments disclosed herein. For example, the object identifier can be the object name of the point cloud object. Another example is that the object identifier can be the object ID of the point cloud object. Yet another example is that the object identifier can be the location information of the point cloud object.

[0084] In one possible implementation, the detection device can compare at least one first distance with a third preset distance threshold to determine the number of first target distances, where the first target distance is less than the third preset distance threshold. If the number of first target distances is less than the first preset number threshold, the detection device determines the first information as second information. If the number of first target distances is greater than or equal to the first preset number threshold, the detection device determines the first information as third information.

[0085] For example, if the third preset distance threshold is 1 meter, the first preset quantity threshold is 2, and at least one first distance includes 10 meters, 2 meters, and 0.1 meters, then the detection device can generate second information. For instance, the second information may include the position information of the point cloud objects corresponding to the feature pairs at 10 meters and 2 meters, respectively.

[0086] Optionally, the detection device can determine the number of second target distances based on at least one first distance, where the second target distance is a first distance greater than or equal to a third preset distance threshold. Then, the detection device can determine first information based on the number of second target distances.

[0087] Optionally, the detection device can determine the first information based on the ratio of the distance to the first target to the first distance. Alternatively, the detection device can determine the first information based on the ratio of the distance to the second target to the first distance.

[0088] Understandably, after acquiring multiple point cloud objects, the detection device identifies at least one feature pair. Since these multiple point cloud objects are point cloud objects within a target point cloud dataset composed of multiple sets of point cloud data stitched together, and each set of point cloud data includes at least one point cloud object, identical point cloud objects can be obtained from different sets of point cloud data. Then, the detection device acquires a first distance corresponding to each feature pair. This first distance, representing the distance between point cloud objects within the feature pair, reflects the degree of overlap between identical point cloud objects in different sets of point cloud data. Thus, the detection device can determine whether the target point cloud data exhibits stitching anomalies based on at least one first distance, i.e., the degree of overlap between multiple sets of identical point cloud objects. This achieves the goal of detecting stitched point cloud data, and by calculating the distance between identical point cloud objects, the accuracy of detecting stitching anomalies in the stitched point cloud data can be improved.

[0089] It should be noted that multiple point cloud objects may contain various object types. If the same method is used to obtain the first distance for feature pairs when different point cloud objects have different object types, the first distance may be inaccurate, thus affecting the accuracy of the detected and stitched point cloud data.

[0090] In some embodiments, the detection device can obtain a first distance corresponding to a feature pair according to the type of point cloud object. The following describes an embodiment of this disclosure using the example of obtaining a first distance corresponding to a feature pair, where the feature pair includes a first object and a second object.

[0091] like Figure 3 As shown, another point cloud data detection method provided in this embodiment of the present disclosure is provided. In this method, step S202 may include:

[0092] S301, The detection device acquires the attribute information of the first object and the attribute information of the second object.

[0093] The attribute information may include: object type and location information.

[0094] In one possible implementation, the detection device stores the relationship between semantic features (i.e., point cloud objects) and object types. The detection device can determine the object type of a first object based on a first object, and determine the object type of a second object based on a second object.

[0095] For example, if the first and second objects are lane lines, then the first and second objects are line types. If the first and second objects are guardrails, then the first and second objects are line types. If the first and second objects are ground arrows, then the first and second objects are area types.

[0096] In one possible design, the position information of a point cloud object may include: the position information of the center point of the point cloud object and the method information of the point cloud object.

[0097] It should be noted that the representation of the position information of point cloud objects in this embodiment is not limited. For example, the position information of point cloud objects can be represented based on the world coordinate system. As another example, the position information of point cloud objects can be represented by bounding boxes.

[0098] For example, the location information of a point cloud object can be represented by a bounding box. (e.g.) Figure 4 As shown, the point cloud object 401 can be represented by its length, width, height, 3D direction (i.e., direction information), and center point A to represent the 3D geometric range and orientation of the point cloud object a. The direction information can include direction a (e.g., direct(x)), direction b (e.g., direct(y)), and direction c (e.g., direct(z)), and the position of the center point A is (a1, a2, a3).

[0099] It should be noted that the specific methods for obtaining the location information of point cloud objects can refer to the methods for obtaining the location of point cloud data in conventional technologies, which will not be elaborated here.

[0100] S302. The detection equipment determines the target algorithm based on the target type.

[0101] The target types are the object type of the first object and the object type of the second object. The target algorithm is used to determine the distance between point cloud objects of the two target types.

[0102] In some embodiments, the distance between two point cloud objects of the same target type may include two types of distances: a first type of distance and a second type of distance. The first type of distance may be used to indicate the length distance, and the second type of distance may be used to indicate the angle between the two point cloud objects.

[0103] In one possible design, the first distance may include: a first type of distance. If the target type is linear, the target algorithm includes: a first type of algorithm, which determines the first type of distance between the first object and the second object based on the position information of the center point of the first object, the position information of the center point of the second object, and the orientation information of the first object.

[0104] Optionally, the first type of algorithm can be represented by Formula 2.

[0105]

[0106]

[0107] Wherein, Δ1(f i ,f j This is used to represent the first-class distance between the i-th point cloud object (i.e., the first object) and the j-th point cloud object (i.e., the second object) when the target type is linear. Used to represent the orientation information of the i-th point cloud object.

[0108] If the target type is a planar type, the target algorithm includes a second type of algorithm, which is: based on the position information of the center point of the first object, the position information of the center point of the second object, and the orientation information of the first object, the first type of distance between the first object and the second object is determined.

[0109] Optionally, the second type of algorithm can be represented by Equation 4.

[0110]

[0111] Wherein, Δ2(f i ,f j ) is used to represent the first type distance between the i-th point cloud object (i.e., the first object) and the j-th point cloud object (i.e., the second object) when the target type is a face.

[0112] In other words, the first type of distance between the first object and the second object is the distance between the center point of the first object and the plane of the second object.

[0113] Understandably, for point cloud objects of linear and planar types, using different algorithms to determine the length distance between two point cloud objects can improve the accuracy of determining the first distance, thereby improving the accuracy of detecting whether the stitched point cloud data is abnormal.

[0114] In another possible design, the first distance may include a second type of distance. The target algorithm also includes a third type of algorithm. The third type of algorithm is: determining the second type of distance between the first object and the second object based on the orientation information of the first object and the orientation information of the second object.

[0115] In other words, the second type of distance is the angular distance between the first object and the second object.

[0116] In this embodiment of the disclosure, for any second-class distance corresponding to a feature pair, the detection device can use a third-class algorithm to determine the second-class distance corresponding to the feature pair.

[0117] Optionally, the second type of algorithm can be represented by Formula 5.

[0118]

[0119] Among them, Δ3(f i ,f j This is used to represent the second type of distance between the i-th point cloud object and the j-th point cloud object. Used to represent the orientation information of the j-th point cloud object.

[0120] For example, such as Figure 5 As shown, the feature pair includes point cloud object 501 and point cloud object 502, and the second type distance between point cloud object 501 and point cloud object 502 is an included angle 503 (e.g., 50 degrees).

[0121] Understandably, the first distance can include a second type of distance, which indicates the angle between two point cloud objects. The target algorithm also includes a third type of algorithm, which determines the second type of distance between the first and second objects based on the orientation information of the first and second objects. In this way, the angle between the two point cloud objects can be determined, and thus the degree of overlap between the two point cloud objects can be determined.

[0122] S303. The detection device determines the first distance based on the position information of the first object and the position information of the second object, according to the target algorithm.

[0123] In one possible implementation, the detection device can determine a first type of distance between the first object and the second object based on the position information of the center point of the first object, the position information of the center point of the second object, and the orientation information of the first object.

[0124] In other words, the detection equipment can acquire only the first type of distance.

[0125] In another possible implementation, the detection device can determine a second type of distance between the first object and the second object based on the orientation information of the first object and the orientation information of the second object.

[0126] In other words, the detection equipment can acquire only the second type of distance.

[0127] In another possible implementation, the detection device can determine a first type of distance between the first object and the second object based on the position information of the center point of the first object, the position information of the center point of the second object, and the orientation information of the first object. Furthermore, the detection device can determine a second type of distance between the first object and the second object based on the orientation information of the first object and the orientation information of the second object.

[0128] In other words, the detection equipment can acquire both Type I and Type II distances.

[0129] Based on the above technical solution, the detection device can acquire attribute information of a first object and a second object. The attribute information includes object type and location information. Then, the detection device can determine a target algorithm based on the target type. Thus, based on the location information of the first and second objects, the detection device can determine a first distance according to the target algorithm. Therefore, this solution improves the accuracy of the determined distance—specifically, the degree of overlap between two point cloud objects—by selecting an appropriate algorithm based on the type of the point cloud object, thereby enabling the detection of anomalies in the stitched point cloud data.

[0130] like Figure 6 As shown, another point cloud data detection method provided in this embodiment of the present disclosure is provided, in which S204 may include:

[0131] S601. The detection device determines a third distance based on at least one first distance.

[0132] In one possible implementation, the third distance is used to reflect the overall situation of at least one feature in the target point cloud data relative to the corresponding first distance. The detection device can determine the third distance based on at least one first distance according to a preset processing algorithm.

[0133] It should be noted that the preset processing algorithm is not limited in the embodiments of this disclosure. For example, the preset processing algorithm can be an algorithm for determining the mean. Another example is that the preset processing algorithm can be an algorithm for determining the variance. Yet another example is that the preset processing algorithm can be an algorithm for determining the standard deviation. The following describes the embodiments of this disclosure using an algorithm for determining the mean as an example.

[0134] In one possible design, the third distance can be the average of at least one first distance.

[0135] Optionally, the third distance can be represented by Equation 6.

[0136]

[0137] Where β represents the average of the first distances corresponding to the N feature pairs (i.e., at least one feature pair), Δ k The distance is used to represent the first distance of the k-th feature pair, where N is a positive integer.

[0138] In some embodiments, when at least one first distance is a first type of distance or at least one first distance is a second type of distance, the detection device may determine a third distance based on at least one first distance.

[0139] In other embodiments, where at least one first distance includes a first type of distance and a second type of distance, the detection device can determine a fourth distance and a fifth distance, and a third distance includes the fourth distance and the fifth distance. The fourth distance is determined by the first type of distance among at least one distance, and the fifth distance is determined by the second type of distance among at least one distance.

[0140] Optionally, the third distance can also be the maximum or minimum distance among at least one of the first distances.

[0141] S602. The detection equipment determines whether the third distance is greater than the second preset distance threshold.

[0142] In some embodiments, if the third distance is greater than the second preset distance threshold, the detection device executes S603.

[0143] In some embodiments, if the third distance is less than or equal to the second preset distance threshold, the detection device executes S604.

[0144] In some implementations, where the third distance includes the fourth and fifth distances, the second preset distance threshold includes the third distance threshold and the fourth distance threshold.

[0145] In one possible design, a third distance greater than a second preset distance threshold may include: a fourth distance greater than a third distance threshold, and a fifth distance greater than a fourth distance threshold. A third distance less than or equal to a second preset distance threshold may include: a fourth distance less than or equal to a third distance threshold, and / or, a fifth distance less than or equal to a fourth distance threshold.

[0146] S603, The detection equipment determines that the first information is the second information.

[0147] S604, The detection equipment determines that the first information is the third information.

[0148] Understandably, the detection device determines a third distance based on at least one first distance, and this third distance reflects the overall situation of at least one first distance. Subsequently, if the third distance is greater than a second preset distance threshold, it indicates that at least one first distance is generally too large, meaning the overlap of feature pairs is low, indicating abnormal stitching of the target point cloud data. If the third distance is less than or equal to the second preset distance threshold, it indicates that at least one first distance is generally too small, meaning the overlap of feature pairs is high, indicating normal stitching of the target point cloud data.

[0149] The foregoing primarily describes the solutions provided by the embodiments of this disclosure from the perspective of computer devices. It is understood that, in order to achieve the above functions, the computer device includes corresponding hardware structures and / or software modules for executing each function. Those skilled in the art should readily recognize that, based on the point cloud data detection method steps described in conjunction with the embodiments disclosed herein, this disclosure can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this disclosure.

[0150] This disclosure embodiment can divide the point cloud data detection method into functional modules or functional units according to the above method example. For example, each function can be divided into a separate functional module or functional unit, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware or in software functional modules or functional units. The module or unit division in this disclosure embodiment is illustrative and only represents one logical functional division; other division methods may be used in actual implementation.

[0151] like Figure 7 The diagram shown is a structural schematic of a point cloud data detection device provided in an embodiment of this disclosure. The point cloud data detection device may include an acquisition unit 701 and a processing unit 702.

[0152] Acquisition unit 701 is used to acquire multiple point cloud objects, which are point cloud objects in target point cloud data composed of multiple sets of point cloud data, each set of point cloud data including at least one point cloud object. Processing unit 702 is used to determine at least one feature pair from the multiple point cloud objects, where the point cloud objects in a feature pair are identical. Processing unit 702 is further used to acquire a first distance corresponding to each feature pair, thereby acquiring at least one first distance, where the first distance is the distance between point cloud objects in the feature pair. Processing unit 702 is further used to determine first information based on at least one first distance, the first information indicating whether the target point cloud data has a splicing anomaly.

[0153] Optionally, the processing unit 702 is further configured to divide multiple point cloud objects into at least one feature pair according to preset conditions. The feature pair includes: a first object and a second object; the preset conditions include: the object type of the first object and the object type of the second object are the same; the point cloud data group where the first object is located is different from the point cloud data group where the second object is located; a second distance is less than a first preset distance threshold; and the second distance is the minimum distance in a distance set, which includes: the distance between the center point of the first object and the center points of all point cloud objects except the first object; the distance between the center point of the second object and the center points of all point cloud objects except the second object; and the second distance is the distance between the center points of the first object and the center points of the second object.

[0154] Optionally, the acquisition unit 701 is further configured to acquire attribute information of the first object and the second object, the attribute information including object type and location information. The processing unit 702 is further configured to determine a target algorithm based on the target type, where the target type is the object type of the first object and the object type of the second object, and the target algorithm is used to determine the distance between point cloud objects of the two target types. The processing unit 702 is further configured to determine a first distance based on the location information of the first object and the location information of the second object, according to the target algorithm, the first distance being the distance between the first object and the second object.

[0155] Optionally, the object type includes linear and area types. The position information of the point cloud object includes the position information of the center point of the point cloud object and the orientation information of the point cloud object. The first distance includes a first type of distance, which is used to indicate length distance. If the target type is linear, the target algorithm includes a first type of algorithm, which determines the first type of distance between the first object and the second object based on the position information of the center point of the first object, the position information of the center point of the second object, and the orientation information of the first object. If the target type is area type, the target algorithm includes a second type of algorithm, which determines the first type of distance between the first object and the second object based on the position information of the center point of the first object, the position information of the center point of the second object, and the orientation information of the first object.

[0156] Optionally, the first distance also includes a second type of distance, which indicates the angle between two point cloud objects. The target algorithm also includes a third type of algorithm, which determines the second type of distance between the first object and the second object based on the orientation information of the first object and the orientation information of the second object.

[0157] Optionally, the first information includes either second information or third information, wherein the second information is used to indicate an anomaly in the target point cloud data stitching, and the third information is used to indicate that the target point cloud data stitching is normal. The processing unit 702 is further configured to determine a third distance based on at least one first distance. The processing unit 702 is further configured to determine the first information as the second information if the third distance is greater than a second preset distance threshold. The processing unit 702 is further configured to determine the first information as the third information if the third distance is less than or equal to the second preset distance threshold.

[0158] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0159] Figure 8 A schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0160] like Figure 8As shown, device 800 includes a computing unit 801, which can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) 802 or a computer program loaded from storage unit 808 into random access memory (RAM) 803. RAM 803 may also store various programs and data required for the operation of device 800. The computing unit 801, ROM 802, and RAM 803 are interconnected via bus 804. Input / output (I / O) interface 805 is also connected to bus 804.

[0161] Multiple components in device 800 are connected to I / O interface 805, including: input unit 806, such as keyboard, mouse, etc.; output unit 807, such as various types of monitors, speakers, etc.; storage unit 808, such as disk, optical disk, etc.; and communication unit 809, such as network card, modem, wireless transceiver, etc. Communication unit 809 allows device 800 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0162] The computing unit 801 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as the point cloud data detection method. For example, in some embodiments, the point cloud data detection method can be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed on device 800 via ROM 802 and / or communication unit 809. When the computer program is loaded into RAM 803 and executed by the computing unit 801, one or more steps of the point cloud data detection method described above can be performed. Alternatively, in other embodiments, the computing unit 801 can be configured to perform the point cloud data detection method by any other suitable means (e.g., by means of firmware).

[0163] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0164] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0165] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0166] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0167] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0168] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0169] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0170] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for detecting point cloud data, comprising: Multiple point cloud objects are obtained. The multiple point cloud objects are point cloud objects in the target point cloud data composed of multiple sets of point cloud data. Each set of point cloud data includes at least one point cloud object. At least one feature pair is determined from the plurality of point cloud objects, wherein the point cloud objects in a feature pair are the same, the object types of the point cloud objects in the feature pair are the same, and the point cloud objects in the feature pair come from different point cloud data groups. For each feature pair, obtain the first distance corresponding to the feature pair to obtain at least one first distance, wherein the first distance is the distance between point cloud objects in the feature pair; Based on the at least one first distance, first information is determined, which is used to indicate whether the target point cloud data has an abnormal stitching.

2. The method according to claim 1, wherein, Determining at least one feature pair from the plurality of point cloud objects includes: The plurality of point cloud objects are divided into at least one feature pair according to preset conditions; The feature pair includes a first object and a second object; the preset conditions include: the object type of the first object and the object type of the second object are the same; the point cloud data group where the first object is located is different from the point cloud data group where the second object is located; the second distance is less than a first preset distance threshold; the second distance is the minimum distance in the distance set, the distance set includes: the distance between the center point of the first object and the center point of the point cloud objects other than the first object among the plurality of point cloud objects; the distance between the center point of the second object and the center point of the point cloud objects other than the second object among the plurality of point cloud objects; the second distance is the distance between the center point of the first object and the center point of the second object.

3. The method according to claim 1 or 2, wherein, The feature pair includes a first object and a second object. Obtaining the first distance corresponding to the feature pair includes: Obtain the attribute information of the first object and the attribute information of the second object, wherein the attribute information includes: object type and location information; Based on the target type, a target algorithm is determined, wherein the target type is the object type of the first object and the object type of the second object, and the target algorithm is used to determine the distance between two point cloud objects of the target type; Based on the location information of the first object and the location information of the second object, the first distance is determined according to the target algorithm. The first distance is the distance between the first object and the second object.

4. The method according to claim 3, wherein the object type includes: Linear and planar types, the position information of the point cloud object includes: the position information of the center point of the point cloud object, the orientation information of the point cloud object, and the first distance includes: a first type of distance, which is used to indicate length distance; If the target type is the linear type, then the target algorithm includes: a first type of algorithm, which is: determining a first type of distance between the first object and the second object based on the position information of the center point of the first object, the position information of the center point of the second object, and the direction information of the first object; If the target type is the planar type, then the target algorithm includes a second type of algorithm, which is: determining a first type of distance between the first object and the second object based on the position information of the center point of the first object, the position information of the center point of the second object, and the orientation information of the first object.

5. The method according to claim 3, wherein the first distance further comprises: The second type of distance is used to indicate the angle between two point cloud objects; The target algorithm also includes a third type of algorithm, which is: determining a second type of distance between the first object and the second object based on the orientation information of the first object and the orientation information of the second object.

6. The method according to claim 1 or 2, wherein, The first information includes: a second information or a third information, wherein the second information is used to indicate that the target point cloud data splicing is abnormal, and the third information is used to indicate that the target point cloud data splicing is normal; Determining the first information based on the at least one first distance includes: Based on the at least one first distance, determine the third distance; If the third distance is greater than the second preset distance threshold, then the first information is determined to be the second information; If the third distance is less than or equal to the second preset distance threshold, then the first information is determined to be the third information.

7. A point cloud data detection device, comprising: The acquisition unit is used to acquire multiple point cloud objects, wherein the multiple point cloud objects are point cloud objects in target point cloud data composed of multiple sets of point cloud data spliced ​​together, and each set of point cloud data includes at least one point cloud object. A processing unit is configured to determine at least one feature pair from the plurality of point cloud objects, wherein the point cloud objects in a feature pair are the same, the object types of the point cloud objects in the feature pair are the same, and the point cloud objects in the feature pair come from different point cloud data groups. The processing unit is further configured to obtain a first distance corresponding to each feature pair, so as to obtain at least one first distance, wherein the first distance is the distance between point cloud objects in the feature pair; The processing unit is further configured to determine first information based on the at least one first distance, wherein the first information is used to indicate whether the target point cloud data has an abnormal stitching.

8. The apparatus according to claim 7, The processing unit is further configured to divide the plurality of point cloud objects into at least one feature pair according to preset conditions; in, The feature pair includes: a first object and a second object; the preset conditions include: the object type of the first object and the object type of the second object are the same, the point cloud data group where the first object is located is different from the point cloud data group where the second object is located, the second distance is less than the first preset distance threshold, and the second distance is the minimum distance in the distance set, wherein the distance set includes: the distance between the center point of the first object and the center point of the point cloud objects other than the first object among the plurality of point cloud objects, and the distance between the center point of the second object and the center point of the point cloud objects other than the second object among the plurality of point cloud objects; the second distance is the distance between the center point of the first object and the center point of the second object.

9. The apparatus according to claim 7 or 8, wherein, The feature pair includes a first object and a second object; The acquisition unit is further configured to acquire attribute information of the first object and attribute information of the second object, wherein the attribute information includes: object type and location information; The processing unit is further configured to determine a target algorithm based on a target type, wherein the target type is the object type of the first object and the object type of the second object, and the target algorithm is configured to determine the distance between two point cloud objects of the target type; The processing unit is further configured to determine the first distance based on the location information of the first object and the location information of the second object, according to the target algorithm, wherein the first distance is the distance between the first object and the second object.

10. The apparatus of claim 9, wherein the object type includes: Linear and planar types, the position information of the point cloud object includes: the position information of the center point of the point cloud object, the orientation information of the point cloud object, and the first distance includes: a first type of distance, which is used to indicate length distance; If the target type is the linear type, then the target algorithm includes: a first type of algorithm, which is: determining a first type of distance between the first object and the second object based on the position information of the center point of the first object, the position information of the center point of the second object, and the direction information of the first object; If the target type is the planar type, then the target algorithm includes a second type of algorithm, which is: determining a first type of distance between the first object and the second object based on the position information of the center point of the first object, the position information of the center point of the second object, and the orientation information of the first object.

11. The apparatus according to claim 9, wherein the first distance further comprises: The second type of distance is used to indicate the angle between two point cloud objects; The target algorithm also includes a third type of algorithm, which is: determining a second type of distance between the first object and the second object based on the orientation information of the first object and the orientation information of the second object.

12. The apparatus according to claim 7 or 8, wherein the first information includes: The second information or the third information, wherein the second information is used to indicate that the target point cloud data splicing is abnormal, and the third information is used to indicate that the target point cloud data splicing is normal; The processing unit is further configured to determine a third distance based on the at least one first distance; The processing unit is further configured to determine the first information as the second information if the third distance is greater than the second preset distance threshold; The processing unit is further configured to determine the first information as the third information if the third distance is less than or equal to the second preset distance threshold.

13. An electronic device, comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-6.

14. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-6.

15. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1-6.

Images