Method, device, storage medium and program product for testing point cloud registration effect

Abstract

The embodiment of the application provides a kind of point cloud registration effect test method, equipment, storage medium and program product, the method includes obtaining the point cloud data in the preset time period, point cloud data includes multiple overlapping point clouds, multiple overlapping point clouds are consistent, determine the target offset corresponding to the continuous change of the preset time period, and based on target offset, offset processing is carried out to point cloud data, obtain the point cloud data after offset, according to the point cloud data after offset and target offset, the processing effect of the point cloud registration model to be evaluated is tested, and test result is obtained.The method provided in the embodiment of the application can not only realize automatic evaluation, improve efficiency and reduce cost, but also can better simulate the deviation characteristics of point cloud data, improve the accuracy of the processing effect of the point cloud registration model.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the technical field of high-precision map, and particularly relate to a point cloud registration effect testing method and device, a storage medium and a program product. BACKGROUND

[0002] Laser point cloud registration processing is one of the key processes of high-precision map production. By performing registration processing on the collected laser point cloud, the "ghosting" of the laser point cloud is eliminated, and the consistency of the data is ensured. Whether the registration processing effect of the laser point cloud registration model on the laser point cloud with deviation meets the requirements is crucial, and therefore, the processing effect of the laser point cloud registration model needs to be accurately evaluated.

[0003] In related technologies, the point cloud registration processing data evaluation method usually directly uses real deviation data, and manually measures the offset vector T of the homonymic point on the data as the true value to evaluate the processing effect of the point cloud registration model on the real deviation data.

[0004] However, in the process of implementing the present application, the inventors found that at least the following problems exist in the prior art: In the above-mentioned method, the deviation of the homonymic point is manually measured on the laser point cloud, which requires a large amount of manpower, is low in efficiency, and is high in cost. SUMMARY

[0005] Embodiments of the present application provide a point cloud registration effect testing method, device, storage medium and program product to improve efficiency and reduce cost.

[0006] In a first aspect, embodiments of the present application provide a point cloud registration effect testing method, comprising:

[0007] Obtaining point cloud data in a preset time period; the point cloud data comprises a plurality of overlapping point clouds; the plurality of overlapping point clouds are consistent;

[0008] Selecting a target time point from the preset time period, configuring a maximum offset value for the target time point, and determining a target offset value according to the maximum offset value, the target time point and the preset time period;

[0009] Based on the target offset value, performing offset processing on the point cloud data to obtain offset point cloud data;

[0010] According to the offset point cloud data and the target offset value, the processing effect of the to-be-evaluated point cloud registration model is tested to obtain a test result.

[0011] In a possible design, the target offset value is determined according to the maximum offset value, the target time point and the preset time period, comprising:

[0012] a first offset configured for a time period between a starting time point of the preset time period and the target time point, the first offset continuously changing from a first value to the maximum offset value over time;

[0013] a second offset configured for a time period between the target time point and an ending time point of the preset time period, the second offset continuously changing from the maximum offset value to a second value over time, the first value and the second value being less than the maximum offset value.

[0014] In a possible design, the first offset is incrementally changed, and the second offset is decrementally changed.

[0015] In a possible design, the offset processing of the point cloud data based on the target offset comprises:

[0016] obtaining a POS track corresponding to the point cloud data;

[0017] offset processing the POS track based on the target offset to obtain an offset POS track;

[0018] offset processing the point cloud data based on an offset between the offset POS track and the POS track to obtain offset point cloud data.

[0019] In a possible design, the testing of the processing effect of the point cloud registration model to be evaluated based on the offset point cloud data and the target offset to obtain a test result comprises:

[0020] inputting the offset point cloud data into the point cloud registration model to be evaluated to obtain registered point cloud data and a corresponding offset estimate value, the offset point cloud data comprising a plurality of offset overlapping point clouds, and the offset estimate value comprising at least one actual relative offset of at least one pair of offset overlapping point clouds in the plurality of offset overlapping point clouds;

[0021] determining, according to the target offset, at least one theoretical relative offset of at least one pair of offset overlapping point clouds in the plurality of offset overlapping point clouds;

[0022] determining a test result according to the at least one actual relative offset and the at least one theoretical relative offset.

[0023] In a possible design, the determining of the test result according to the at least one actual relative offset and the at least one theoretical relative offset comprises:

[0024] For each of the at least one actual relative offset, a difference between the actual relative offset and a corresponding theoretical relative offset is calculated, and if the difference is less than or equal to a preset threshold, it is determined that the overlapping point cloud pair corresponding to the actual relative offset is a successfully registered point cloud pair;

[0025] According to a ratio between a number of successfully registered point cloud pairs and a total number of overlapping point cloud pairs, a registration success rate is determined.

[0026] According to the registration success rate, the test result is determined.

[0027] In a second aspect, embodiments of the present application provide a point cloud registration effect testing device, comprising:

[0028] An acquisition module is configured to acquire point cloud data in a preset time period; the point cloud data comprises a plurality of overlapping point clouds; the plurality of overlapping point clouds are consistent;

[0029] A determination module is configured to select a target time point from the preset time period, configure a maximum offset value for the target time point, and determine a target offset value according to the maximum offset value, the target time point and the preset time period.

[0030] An offset module is configured to perform offset processing on the point cloud data based on the target offset value, and obtain offset point cloud data.

[0031] A test module is configured to test a processing effect of a point cloud registration model to be tested according to the offset point cloud data and the target offset value, and obtain a test result.

[0032] In a third aspect, embodiments of the present application provide a point cloud registration effect testing device, comprising at least one processor and a memory.

[0033] The memory stores computer execution instructions.

[0034] The at least one processor executes the computer execution instructions stored in the memory, so that the at least one processor executes the method as described in the first aspect and various possible designs of the first aspect.

[0035] In a fourth aspect, embodiments of the present application provide a computer readable storage medium, which stores computer execution instructions, and when a processor executes the computer execution instructions, the method as described in the first aspect and various possible designs of the first aspect is implemented.

[0036] In a fifth aspect, an embodiment of the present application provides a computer program product, comprising a computer program, which, when executed by a processor, implements the method in the first aspect above and various possible designs of the first aspect.

[0037] The point cloud registration effect test method, device, storage medium and program product provided by the embodiment of the present application, the method comprises: acquiring point cloud data in a preset time period, the point cloud data comprising a plurality of overlapping point clouds, the plurality of overlapping point clouds being consistent; determining a continuously changing target offset corresponding to the preset time period; performing offset processing on the point cloud data based on the target offset to obtain offset point cloud data; and testing the processing effect of a point cloud registration model to be evaluated based on the offset point cloud data and the target offset to obtain a test result. The method provided by the embodiment of the present application can increase the offset of the point cloud data that is overlapping but has no ghosting, and the offset is a continuously changing offset that changes with the collection time, and the offset point cloud data is obtained, and then the processing effect of the point cloud registration model is tested based on the offset point cloud data, so that not only the automatic evaluation, the efficiency and the cost reduction can be realized, but also the deviation characteristics of the point cloud data can be simulated well, and the accuracy of the test of the processing effect of the point cloud registration model is improved. BRIEF DESCRIPTION OF DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0039] Figure 1 The schematic diagram of the point cloud data with ghosting provided by the embodiment of the present application;

[0040] Figure 2 The application scenario schematic diagram of the point cloud registration effect test method provided by the embodiment of the present application;

[0041] Figure 3 The flowchart of the point cloud registration effect test method provided by the embodiment of the present application Figure 1 ;

[0042] Figure 4 The flowchart of the point cloud registration effect test method provided by the embodiment of the present application Figure 2 ;

[0043] Figure 5 The structural schematic diagram of the point cloud registration effect test device provided by the embodiment of the present application;

[0044] Figure 6 A hardware structure schematic diagram of a test device for point cloud registration effect provided by an embodiment of the present application is shown. DETAILED DESCRIPTION

[0045] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below in connection with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some, but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0046] Data collection of a high-definition map is usually completed by a professional mobile measurement system (containing sensors such as a Global Navigation Satellite System (GNSS), an Inertial Navigation System (INS), a Lidar, and a Camera) carried. Due to the precision of the collection device and the influence of the collection environment, the laser point cloud data collected at a time usually has an absolute position deviation. When multiple laser point cloud data are superimposed, the same element appears in multiple positions, resulting in "ghosting" of the laser point cloud (as shown in Figure 1 The figure includes elements such as rectangular signs and linear traffic markings. Taking the sign as an example, two signs partially overlap, like "ghosting"), which interferes with subsequent mapping and does not meet the precision requirements of a high-definition map. Therefore, the collected laser point cloud needs to be registered to eliminate "ghosting" of the laser point cloud and ensure the consistency of the data. Laser point cloud registration processing is one of the key processes of high-definition map production. In addition, due to the complex and diverse collection environment, the deviation of the repeatedly collected laser point cloud is different in size. In addition, the data scene is complex and diverse. Therefore, it is crucial to determine whether the registration processing effect of the laser point cloud registration model on the laser point cloud with a deviation meets the requirements. Therefore, the processing effect of the laser point cloud registration model needs to be accurately evaluated. If the laser point cloud registration processing effect cannot be tested and evaluated sufficiently, it is difficult to promote the iterative research and development of the laser point cloud registration model, which is not conducive to the efficient and accurate registration of the laser point cloud.

[0047] In related technologies, the point cloud registration processing data evaluation method usually directly uses real deviation data and manually measures the offset vector T of the homonym as the ground truth.

[0048] However, in the above method, the homonym deviation is measured manually on the laser point cloud, which requires a large amount of manpower, is low in efficiency, and is high in cost.

[0049] In order to solve the above technical problems, the present inventors have found that the offset of point cloud data with good consistency (with overlap but without ghosting) can be automatically added by a computer to obtain simulation data, and the processing effect of a point cloud registration model is evaluated based on the simulation data and the automatically added offset. In addition, for point cloud data collection of a high-precision map, the line-shaped feature data is usually collected for a long time, and the point cloud data is dynamically collected by a collection vehicle driving along a road. Therefore, the collection process is likely to be affected by signal quality and the like, resulting in the accumulation of the deviation of the POS trajectory. In turn, the deviation of the POS trajectory gradually decreases as the signal quality recovers. Based on this, the offset of the point cloud data also changes accordingly during the process. Therefore, the inventors believe that the deviation of the real point cloud data should be continuously changed. In order to obtain more realistic simulation data, a continuously changing offset can be automatically added. Thus, not only can the automatic evaluation be realized to improve the efficiency and reduce the cost, but also the deviation characteristics of the point cloud data can be better simulated to improve the accuracy of the test of the processing effect of the point cloud registration model. Based on this, the embodiments of the present application provide a quality test method for point cloud registration processing.

[0050] Figure 2 The application scenario of the test method for point cloud registration effect provided by the embodiments of the present application is shown in FIG. 1. Figure 2 As shown in FIG. 1, a high-precision collection device is in communication connection with a point cloud registration effect test device. The high-precision collection device is used to collect point cloud data, the point cloud data includes a plurality of overlapping point clouds, and the plurality of overlapping point clouds have consistency, and the point cloud data is sent to the test device, and the test device is used to test the point cloud registration effect based on the point cloud data. The high-precision collection device can be a collection vehicle provided with a high-precision laser radar sensor.

[0051] In the implementation process, a high-precision acquisition device, such as an acquisition vehicle, obtains point cloud data by a high-precision laser radar sensor during driving along a road, the point cloud data includes a plurality of overlapping point clouds, the plurality of overlapping point clouds are consistent, and the point cloud data is sent to a test device. The test device obtains point cloud data in a preset time period, determines a continuously changing target offset corresponding to the preset time period, and performs offset processing on the point cloud data based on the target offset to obtain offset point cloud data. The processing effect of the point cloud registration model to be tested is tested based on the offset point cloud data and the target offset, and a test result is obtained. The point cloud registration effect testing method provided in the embodiment of the application increases the offset of the point cloud data that overlaps without ghosting, and the offset is a continuously changing offset that changes with the acquisition time. The offset point cloud data is obtained, and then the processing effect of the point cloud registration model is tested based on the offset point cloud data. Therefore, not only can the automatic evaluation be realized, the efficiency be improved, and the cost be reduced, but also the deviation characteristics of the point cloud data can be simulated well, and the accuracy of the test of the processing effect of the point cloud registration model can be improved.

[0052] It should be noted that Figure 2 The scene diagram shown is only an example, and the point cloud registration effect testing method and the scene described in the embodiment of the application are used to more clearly illustrate the technical solutions of the embodiment of the application, and do not constitute a limitation on the technical solutions provided by the embodiment of the application. Those skilled in the art can know that the technical solutions provided by the embodiment of the application are also applicable to similar technical problems as the system evolves and new business scenarios appear.

[0053] The technical solutions of the application will be described in detail in the following specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes can not be described in some embodiments.

[0054] Figure 3 Flowchart of the point cloud registration effect testing method provided by the embodiment of the application Figure 1 .

[0055] As Figure 3 shown, the method comprises:

[0056] 301, obtaining point cloud data in a preset time period; the point cloud data includes a plurality of overlapping point clouds; the plurality of overlapping point clouds are consistent.

[0057] The execution subject of the embodiment can be a computer, a tablet computer or other terminal device, and can also be a server.

[0058] In this embodiment, there are various ways to obtain point cloud data. Laser point cloud data with overlap and no ghosting can be selected from high-precision map laser point cloud data collected in the area. High-precision map laser point cloud data with good consistency can also be collected by a high-precision collection device. Consistent point cloud data with good consistency can also be obtained after eliminating ghosting through point cloud registration processing. This embodiment does not limit this.

[0059] 302. From the selected target time point in the preset time period, a maximum offset value is configured for the target time point, and a target offset value is determined according to the maximum offset value, the target time point, and the preset time period.

[0060] Specifically, considering that the high-precision map laser point cloud data is linear ground object data collected for a long time, and due to the accumulation of POS trajectory deviation, the corresponding point cloud data also continuously changes with the change of collection time. Therefore, in order to accurately simulate the deviation characteristics of the real point cloud data, the determined target offset value is a target offset value that continuously changes with time corresponding to the preset time period. Further, the point cloud data can be offset processed based on the determined target offset value.

[0061] In some embodiments, determining the target offset value according to the maximum offset value, the target time point, and the preset time period can include: configuring a first offset value that continuously changes from a first value to the maximum offset value with time increasing for a time period between a starting time point of the preset time period and the target time point; configuring a second offset value that continuously changes from the maximum offset value to a second value with time increasing for a time period between the target time point and an ending time point of the preset time period; and the first value and the second value are both less than the maximum offset value.

[0062] In some embodiments, the first offset value is a cumulative incremental change with time, and the second offset value is a cumulative decremental change with time.

[0063] For example, the target time point can be selected as a middle time point of the preset time period, or a one-third time point. The target time point can be set according to the actual deviation change characteristics in the actual acquisition process. For example, during the acquisition process of the acquisition vehicle, a tunnel is encountered, the global navigation satellite system (GNSS) signal becomes poor, and after the trajectory is predicted by the inertial navigation system (INS), the deviation starts to accumulate. After the acquisition vehicle exits the tunnel, the GNSS signal is restored, and the deviation starts to decrease. Therefore, the target time point can be selected as the time when the acquisition vehicle exits the tunnel. This embodiment is not limited in this regard. During the period before the target time point in the preset time period, the offset can be linearly increased from zero to a preset maximum offset value. During the period after the target time point, the offset can be linearly decreased from the preset maximum offset value to zero. Of course, the offset can also be nonlinear. The actual INS deviation characteristics in the actual acquisition process can be used to determine the offset. This embodiment is not limited in this regard.

[0064] 303. Offset the point cloud data based on the target offset to obtain offset point cloud data.

[0065] In some embodiments, after the offset is determined, the point cloud data can be offset in various ways.

[0066] In one implementation, to improve data processing efficiency, the target offset can be directly superimposed on the point cloud data to obtain offset point cloud data.

[0067] In another implementation, to obtain offset POS trajectory corresponding to the offset point cloud, so as to facilitate other processing operations. Offset the point cloud data based on the target offset to obtain offset point cloud data can include: obtaining a POS trajectory corresponding to the point cloud data; offsetting the POS trajectory based on the target offset to obtain an offset POS trajectory; and offsetting the point cloud data based on the offset between the offset POS trajectory and the POS trajectory to obtain offset point cloud data.

[0068] For example, as shown in Figure 4 During the POS trajectory disturbance process, the vehicle-mounted laser point cloud deviation has the characteristics of increasing from small to large and then decreasing from large to small. Therefore, the POS trajectory is disturbed according to this principle. The trajectory in the time period [t start , t end ] is disturbed, and the disturbance formula is as follows:

[0069]

[0070] wherein POS i , POD' i are the trajectory coordinates before and after disturbance at time i, Δ = (Δ x , Δ y , Δ z ) is the maximum trajectory disturbance deviation amount in the time period.

[0071] During the disturbance of the laser point cloud, the conversion equation of the laser point cloud from the INS coordinate to the world coordinate system is:

[0072]

[0073] wherein,

[0074] is the coordinate of the point cloud in the INS coordinate system.

[0075] is the coordinate of the point cloud in the world coordinate system.

[0076] (B, L, H) is the longitude, latitude and geodetic height of the origin of the INS coordinate.

[0077] (r, p, y) is the roll angle, pitch angle and heading angle measured by the INS.

[0078] is the rotation matrix of the local horizontal coordinate system to the world coordinate system.

[0079] is the rotation matrix of the INS coordinate system to the local horizontal coordinate system.

[0080] is the translation vector of the local horizontal coordinate system to the world coordinate system.

[0081] After the disturbance of the trajectory, the laser point cloud needs to be disturbed accordingly. According to the above laser point cloud coordinate conversion equation and the error propagation principle of the vehicle-mounted laser point cloud, the laser point disturbance formula is:

[0082] X i ' = X i + POS' i - POS i (3)

[0083] wherein POS i , POS' i are the trajectory coordinates before and after disturbance at time i, x i , x i ' are the coordinates of the point cloud before and after disturbance at time i, respectively.

[0084] 304. Based on the offset point cloud data and the target offset, test the processing effect of the point cloud registration model to be evaluated, and obtain the test results.

[0085] Specifically, after obtaining the offset point cloud data, the target offset can be used as the ground truth, i.e., the theoretical value, and the offset point cloud data can be used as the input point cloud for the point cloud registration model to be evaluated. After registering the offset point cloud data through the point cloud registration model to be evaluated, the registered point cloud and the corresponding offset estimate are output. Then, the test result can be determined based on the comparison between the offset estimate and the target offset.

[0086] In some embodiments, testing the processing effect of the point cloud registration model to be evaluated based on the offset point cloud data and the target offset to obtain test results may include: inputting the offset point cloud data into the point cloud registration model to be evaluated to obtain registered point cloud data and corresponding offset estimates; the offset point cloud data includes multiple offset overlapping point clouds; the offset estimates include the actual relative offsets corresponding to at least one offset overlapping point cloud pair among the multiple offset overlapping point clouds; determining the theoretical relative offsets corresponding to at least one offset overlapping point cloud pair among the multiple offset overlapping point clouds based on the target offset; and determining the test results based on the at least one actual relative offset and the at least one theoretical relative offset.

[0087] In some embodiments, determining a test result based on the at least one actual relative offset and the at least one theoretical relative offset includes: for each of the at least one actual relative offset, calculating the difference between the actual relative offset and the corresponding theoretical relative offset; if the difference is less than or equal to a preset threshold, determining the overlapping point cloud pair corresponding to the actual relative offset as a successfully registered point cloud pair; determining the registration success rate based on the ratio between the number of successfully registered point cloud pairs and the total number of overlapping point cloud pairs; and determining the test result based on the registration success rate.

[0088] For example, such as Figure 4 As shown, after obtaining the offset point cloud data, i.e., after generating the offset laser point cloud data, this offset point cloud data is used as the input to the laser point cloud registration model. During the registration process, the laser point cloud registration model selects two offset overlapping point clouds from multiple offset overlapping point clouds in the offset point cloud data as an overlapping point cloud pair. One of the offset overlapping point clouds in the pair is designated as the target point cloud, and the other as the source point cloud. The time intervals for the source and target point clouds are t and t', respectively. s , t t, the output of the laser point cloud registration model is an offset estimation value T = (t x , t y , t z ) corresponding to the offset point cloud pair after the offset, where t x , t y , t z respectively represent the offset values in each dimension.

[0089] After obtaining the offset estimation value, the offset true value, that is, the theoretical relative offset between the two offset point clouds in the overlapping point cloud pair, can be calculated.

[0090] The time points are t s , t t respectively.

[0091]

[0092] Given a registration success threshold Thresh, if the distance between T ground truth and T is ‖T ground truth -T‖<Thresh, it indicates that the overlapping point cloud pair registration is successful, otherwise it fails.

[0093] The test result can be characterized by the registration success rate. The number of overlapping point cloud pairs can be multiple. For example, if the number of overlapping point clouds in the offset point cloud data is 3, that is, the point cloud data includes 3 overlapping point clouds a, b and c, then the overlapping point cloud pair can include 3, that is, ab, bc and ac, that is, the total number of overlapping point cloud pairs is 3.

[0094] The calculation formula of the registration success rate is:

[0095]

[0096] Where Count all is the total number of overlapping point cloud pairs, and Count sucess is the number of successfully registered point cloud pairs.

[0097] The point cloud registration effect test method provided in the embodiment can increase the offset of the overlapping point cloud data without ghosting, and the offset is a continuously changing offset with the change of the collection time, to obtain the offset point cloud data, and then test the processing effect of the point cloud registration model based on the offset point cloud data, so as to not only realize automatic evaluation, improve efficiency and reduce cost, but also better simulate the deviation characteristics of the point cloud data and improve the accuracy of the test of the processing effect of the point cloud registration model.

[0098] Figure 5 A structural schematic diagram of a test device for point cloud registration effect is provided in the embodiments of the present application. As shown in the figure, the test device 50 for point cloud registration effect comprises an acquisition module 501, a determination module 502, an offset module 503 and a test module 504. Figure 5

[0099] The acquisition module 501 is configured to acquire point cloud data in a preset time period; the point cloud data comprises a plurality of overlapping point clouds; the plurality of overlapping point clouds are consistent;

[0100] The determination module 502 is configured to select a target time point from the preset time period, configure a maximum offset value for the target time point, and determine a target offset value according to the maximum offset value, the target time point and the preset time period.

[0101] The offset module 503 is configured to perform offset processing on the point cloud data based on the target offset value to obtain offset point cloud data.

[0102] The test module 504 is configured to test the processing effect of a point cloud registration model to be evaluated according to the offset point cloud data and the target offset value, and obtain a test result.

[0103] The test device for point cloud registration effect provided in the embodiments of the present application can increase the offset value for the overlapping point cloud data without ghosting, and the offset value is a continuously changing offset value varying with the collection time, so as to obtain offset point cloud data, and then test the processing effect of the point cloud registration model based on the offset point cloud data, thereby not only realizing automatic evaluation, improving efficiency and reducing cost, but also better simulating the deviation characteristics of the point cloud data and improving the accuracy of the test of the processing effect of the point cloud registration model.

[0104] In some embodiments, the determination module 502 is specifically configured to:

[0105] configure a first offset value continuously changing from a first value to the maximum offset value with the increase of time for a time period between a starting time point of the preset time period and the target time point;

[0106] configure a second offset value continuously changing from the maximum offset value to a second value with the increase of time for a time period between the target time point and an ending time point of the preset time period; the first value and the second value are both less than the maximum offset value.

[0107] In some embodiments, the first offset value is an increasing value, and the second offset value is a decreasing value.

[0108] ​In some embodiments, the offset module 503 is specifically configured to:

[0109] obtain a POS trajectory corresponding to the point cloud data;

[0110] perform offset processing on the POS trajectory based on the target offset amount, to obtain an offset POS trajectory;

[0111] perform offset processing on the point cloud data based on the offset amount between the offset POS trajectory and the POS trajectory, to obtain offset point cloud data.

[0112] In some embodiments, the test module 504 is specifically configured to:

[0113] input the offset point cloud data into a point cloud registration model to be tested, to obtain registered point cloud data and a corresponding offset estimate; the offset point cloud data includes a plurality of offset overlapping point clouds; the offset estimate includes at least one actual relative offset amount corresponding to each of the plurality of offset overlapping point clouds;

[0114] determine, according to the target offset amount, at least one theoretical relative offset amount corresponding to each of the plurality of offset overlapping point clouds;

[0115] determine a test result according to the at least one actual relative offset amount and the at least one theoretical relative offset amount.

[0116] In some embodiments, the test module 504 is specifically configured to:

[0117] for each actual relative offset amount in the at least one actual relative offset amount, calculate a difference value between the actual relative offset amount and a corresponding theoretical relative offset amount, and if the difference value is less than or equal to a preset threshold, determine that an overlapping point cloud pair corresponding to the actual relative offset amount is a successfully registered point cloud pair;

[0118] determine a registration success rate according to a ratio between a number of successfully registered point cloud pairs and a total number of overlapping point cloud pairs;

[0119] determine the test result according to the registration success rate.

[0120] The point cloud registration effect test device provided by the embodiments of the present application can be used to execute the method embodiments described above, and has similar implementation principles and technical effects, which will not be described here.

[0121] Figure 6 The hardware structure diagram of the point cloud registration effect test device provided by the embodiments of the present application. The device can be a computer, a messaging device, a tablet device, a medical device, etc.

[0122] The device 60 can include one or more of the following components: a processing component 601, a memory 602, a power supply component 603, a multimedia component 604, an audio component 605, an input / output (I / O) interface 606, a sensor component 607, and a communication component 608.

[0123] The processing component 601 usually controls overall operations of the device 60, such as operations associated with displaying, making phone calls, data communications, camera operations and recording operations. The processing component 601 can include one or more processors 609 to execute instructions to complete all or part of steps of the above methods. In addition, the processing component 601 can include one or more modules to facilitate the interaction between the processing component 601 and other components. For example, the processing component 601 can include a multimedia module to facilitate the interaction between the multimedia component 604 and the processing component 601.

[0124] The memory 602 is configured to store various types of data to support operations of the device 60. Examples of these data include instructions for any application or method operating on the device 60, contact data, phonebook data, messages, pictures, videos, and the like. The memory 602 can be implemented by any type of volatile or non-volatile storage devices or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic or optical disk.

[0125] The power supply component 603 provides power for the various components of the device 60. The power supply component 603 can include a power supply management system, one or more power supplies, and other components associated with generating, managing and distributing power for the device 60.

[0126] The multimedia component 604 includes a screen providing an output interface between the device 600 and a user. In some embodiments, the screen includes a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen can be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touch, swiping, and gestures on the touch panel. The touch sensors can not only sense a boundary of a touching or swiping action, but also detect duration and pressure associated with the touching or swiping action. In some embodiments, the multimedia component 604 includes a front camera and / or a rear camera. When the device 600 is in an operation mode, such as a shooting mode or a video mode, the front camera and / or the rear camera can receive external multimedia data. Each of the front and rear camera can be a fixed optical lens system or have a focal length and optical zooming capability.

[0127] The audio component 605 is configured to output and / or input audio signals. For example, the audio component 605 includes a microphone (MIC) configured to receive external audio signals when the device 600 is in an operation mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals can be further stored in the memory 602 or transmitted via the communication component 608. In some embodiments, the audio component 605 also includes a speaker for outputting audio signals.

[0128] The I / O interface 606 provides an interface between the processing component 601 and peripheral interface modules, such as a keypad, a click wheel, buttons, and so on. The buttons can include, but are not limited to, a home button, a volume button, a start button, and a lock button.

[0129] The sensor component 607 includes one or more sensors to provide various state assessments for the device 600. For example, the sensor component 607 can assess an on / off state of the device 600, relative positioning of components, such as a display and a keypad of the device 600, a position change of the device 600 or a component of the device 600, presence or absence of user contact with the device 600, an orientation or acceleration / deceleration of the device 600, and a temperature change of the device 600. The sensor component 607 can include a proximity sensor configured to assess presence of a nearby object without any physical touch. The sensor component 607 can further include a light sensor, such as a CMOS or CCD image sensor, for use in an imaging application. In some embodiments, the sensor component 607 can further include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0130] The communication component 608 is configured to facilitate wired or wireless communication between the device 60 and other devices. The device 60 can access a wireless network based on a communication standard, such as WiFi, 2G, or 3G, or a combination thereof. In an exemplary embodiment, the communication component 608 receives a broadcast signal or broadcast related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 608 further includes a Near Field Communication (NFC) module to facilitate close proximity communication. For example, the NFC module can be implemented based on Radio Frequency Identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

[0131] In an exemplary embodiment, the device 60 can be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, micro-controllers, microprocessors or other electronic components, for performing the above-described methods.

[0132] In an exemplary embodiment, a non-transitory computer readable storage medium including instructions, such as the memory 602 including instructions, is also provided, which can be executed by the processor 609 of the device 60 to complete the above-described methods. For example, the non-transitory computer readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disc, and an optical data storage device, etc.

[0133] The above-described computer readable storage medium can be implemented by any type of volatile or non-volatile storage devices or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disc or optical disc. The readable storage medium can be any available medium which can be accessed by a general or special purpose computer.

[0134] An exemplary readable storage medium is coupled to the processor, so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be an integral part of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (ASIC). Of course, the processor and the readable storage medium can also exist as discrete components in the device.

[0135] Those skilled in the art can understand that all or part of the steps of the above-mentioned method embodiments can be completed by program instruction related hardware. The foregoing program can be stored in a computer readable storage medium. The program executes to perform the steps of the above-mentioned method embodiments; and the foregoing storage medium includes various storage media that can store program codes, such as ROM, RAM, magnetic disk or optical disk.

[0136] The embodiments of the present application also provide a computer program product, comprising a computer program, which, when executed by a processor, implements the point cloud registration effect testing method executed by the point cloud registration effect testing device.

[0137] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for testing point cloud registration effectiveness, characterized in that, include: Acquire point cloud data within a preset time period; the point cloud data includes multiple overlapping point clouds; the multiple overlapping point clouds are consistent; Select a target time point within the preset time period, configure a maximum offset for the target time point, and determine the target offset based on the maximum offset, the target time point, and the preset time period; Based on the target offset, the point cloud data is offset to obtain offset point cloud data; Based on the offset point cloud data and the target offset, the processing effect of the point cloud registration model to be evaluated is tested, and the test results are obtained.

2. The method according to claim 1, characterized in that, Determining the target offset based on the maximum offset value, the target time point, and the preset time period includes: Configure a first offset for the time period between the start time of the preset time period and the target time point, which continuously changes from a first value to the maximum value of the offset as time increases; Configure a second offset for the time period between the target time point and the end time point of the preset time period, which continuously changes from the maximum offset value to a second value as time increases; both the first value and the second value are less than the maximum offset value.

3. The method according to claim 2, characterized in that, The first offset changes incrementally, while the second offset changes incrementally.

4. The method according to claim 1, characterized in that, The step of offsetting the point cloud data based on the target offset to obtain offset point cloud data includes: Obtain the POS trajectory corresponding to the point cloud data; The POS trajectory is offset based on the target offset to obtain the offset POS trajectory; The point cloud data is offset based on the offset POS trajectory and the offset between the POS trajectories to obtain offset point cloud data.

5. The method according to any one of claims 1-4, characterized in that, The process involves testing the processing effect of the point cloud registration model to be evaluated based on the offset point cloud data and the target offset, and obtaining test results, including: The offset point cloud data is input into the point cloud registration model to be evaluated to obtain the registered point cloud data and the corresponding offset estimate; the offset point cloud data includes multiple offset overlapping point clouds; the offset estimate includes the actual relative offset of at least one offset overlapping point cloud pair among the multiple offset overlapping point clouds respectively; Based on the target offset, determine the theoretical relative offset of at least one pair of offset overlapping point clouds among the multiple offset overlapping point clouds; The test results are determined based on the at least one actual relative offset and the at least one theoretical relative offset.

6. The method according to claim 5, characterized in that, Determining the test result based on the at least one actual relative offset and the at least one theoretical relative offset includes: For each of the at least one actual relative offsets, calculate the difference between the actual relative offset and the corresponding theoretical relative offset. If the difference is less than or equal to a preset threshold, then determine that the overlapping point cloud pair corresponding to the actual relative offset is a successfully registered point cloud pair. The registration success rate is determined by the ratio between the number of successfully registered point cloud pairs and the total number of overlapping point cloud pairs. The test result is determined based on the registration success rate.

7. A testing device for point cloud registration effect, characterized in that, include: The acquisition module is used to acquire point cloud data within a preset time period; the point cloud data includes multiple overlapping point clouds; the multiple overlapping point clouds are consistent. The determination module is used to select a target time point from the preset time period, configure a maximum offset value for the target time point, and determine the target offset based on the maximum offset value, the target time point, and the preset time period; The offset module is used to offset the point cloud data based on the target offset amount to obtain offset point cloud data; The testing module is used to test the processing effect of the point cloud registration model to be evaluated based on the offset point cloud data and the target offset, and obtain the test results.

8. A testing device for point cloud registration effect, characterized in that, include: At least one processor and memory; The memory stores computer-executed instructions; The at least one processor executes computer execution instructions stored in the memory, causing the at least one processor to perform the test method for point cloud registration effect as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by the processor, implement the point cloud registration effect testing method as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the test method for point cloud registration effect as described in any one of claims 1 to 6.

Images