CN117994764A - Data processing method, device, equipment and storage medium - Google Patents
Data processing method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN117994764A CN117994764A CN202410250228.XA CN202410250228A CN117994764A CN 117994764 A CN117994764 A CN 117994764A CN 202410250228 A CN202410250228 A CN 202410250228A CN 117994764 A CN117994764 A CN 117994764A
- Authority
- CN
- China
- Prior art keywords
- point cloud
- cloud data
- environmental information
- simulation algorithm
- processed
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000003672 processing method Methods 0.000 title abstract description 11
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 124
- 230000007613 environmental effect Effects 0.000 claims abstract description 111
- 238000012545 processing Methods 0.000 claims abstract description 95
- 238000000034 method Methods 0.000 claims abstract description 54
- 238000011156 evaluation Methods 0.000 claims abstract description 42
- 230000002159 abnormal effect Effects 0.000 claims abstract description 31
- 238000001914 filtration Methods 0.000 claims abstract description 26
- 230000033001 locomotion Effects 0.000 claims description 54
- 230000003287 optical effect Effects 0.000 claims description 22
- 238000004088 simulation Methods 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 5
- 230000008569 process Effects 0.000 description 12
- 238000005491 wire drawing Methods 0.000 description 12
- 238000012854 evaluation process Methods 0.000 description 8
- 230000008447 perception Effects 0.000 description 7
- 238000011157 data evaluation Methods 0.000 description 4
- 230000033228 biological regulation Effects 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 230000002238 attenuated effect Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/30—Noise filtering
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Traffic Control Systems (AREA)
Abstract
The application discloses a data processing method, a data processing device, data processing equipment and a storage medium, and belongs to the technical field of computers. The method comprises the following steps: acquiring environment information of an automatic driving vehicle and point cloud data to be processed; based on the environmental information, acquiring simulated point cloud data corresponding to the environmental information by using a preset ray tracing simulation algorithm; based on the simulated point cloud data, carrying out evaluation processing on the point cloud data to be processed to obtain an evaluation processing result; based on the evaluation processing result, carrying out anomaly identification processing on the point cloud data to be processed so as to obtain abnormal point cloud data; and filtering the abnormal point cloud data in the point cloud data to be processed to obtain target point cloud data. The method and the device improve the efficiency and reliability of filtering processing of the abnormal point cloud data.
Description
Technical Field
The application relates to the technical field of computers, in particular to the technical fields of intelligent transportation, automatic driving and the like, and particularly relates to a data processing method, device, equipment and storage medium.
Background
The point cloud data collected by the lidar of an autonomous vehicle is often affected by stringing noise. The drawing noise refers to a phenomenon that a laser beam is elongated when being received due to relative motion between a laser radar and a target in a perceived environment, so that false target point cloud points are generated or the target point cloud points are shifted in position. Such noise not only affects the accurate identification of the target location by the autopilot system, but may also mislead the trajectory planning and decision making module, thereby reducing the performance of the overall autopilot system.
Therefore, there is a need for an effective method for filtering the wire drawing noise to improve the environmental perception performance of the automatic driving system.
Disclosure of Invention
The application provides a data processing method, a device, equipment and a storage medium, which solve the problem that wiredrawing noise data in point cloud data cannot be effectively filtered, and the technical scheme is as follows:
In a first aspect, there is provided a method of data processing, the method comprising:
Acquiring environment information of an automatic driving vehicle and point cloud data to be processed;
Based on the environmental information, acquiring simulated point cloud data corresponding to the environmental information by using a preset ray tracing simulation algorithm;
Based on the simulated point cloud data, carrying out evaluation processing on the point cloud data to be processed to obtain an evaluation processing result;
Based on the evaluation processing result, carrying out anomaly identification processing on the point cloud data to be processed so as to obtain abnormal point cloud data;
And filtering the abnormal point cloud data in the point cloud data to be processed to obtain target point cloud data.
In one possible implementation manner, the preset ray tracing simulation algorithm includes a path simulation algorithm, based on the environmental information, the obtaining simulated point cloud data corresponding to the environmental information by using the preset ray tracing simulation algorithm includes:
Simulating a beam propagation path of a lidar of the autonomous vehicle using the path simulation algorithm based on the environmental information;
And obtaining simulated point cloud data corresponding to the environment information based on the environment information and the beam propagation path.
In one possible implementation manner, the preset ray tracing simulation algorithm further includes a relative motion simulation algorithm, based on the environmental information, the obtaining simulated point cloud data corresponding to the environmental information by using the preset ray tracing simulation algorithm includes:
simulating the relative motion of the autonomous vehicle and the target object by using the relative motion simulation algorithm based on the environment information so as to obtain motion data of the autonomous vehicle and motion data of the target object in the environment information;
And obtaining simulated point cloud data corresponding to the environmental information based on the motion data of the automatic driving vehicle and the motion data of the target object in the environmental information.
In one possible implementation manner, the preset light ray tracing simulation algorithm includes a variable light path simulation algorithm, based on the environmental information, the obtaining simulated point cloud data corresponding to the environmental information by using the preset light ray tracing simulation algorithm includes:
Simulating a variable optical path propagation path of a beam of a laser radar of the autonomous vehicle using the variable optical path simulation algorithm based on the environmental information;
And obtaining simulated point cloud data corresponding to the environmental information based on the environmental information and the variable optical path propagation path.
In one possible implementation manner, the preset ray tracing simulation algorithm further includes an energy simulation algorithm, based on the environmental information, the obtaining simulated point cloud data corresponding to the environmental information by using the preset ray tracing simulation algorithm includes:
Simulating energy data of a beam of a lidar of the autonomous vehicle using the energy simulation algorithm based on the environmental information;
and obtaining simulated point cloud data corresponding to the environmental information based on the environmental information and the energy data.
In one possible implementation manner, the performing, based on a result of the evaluation process, anomaly identification processing on the point cloud data to be processed to obtain abnormal point cloud data includes:
Based on the result of the evaluation processing, performing anomaly identification processing on the point cloud data to be processed by using a preset anomaly identification algorithm;
and acquiring abnormal point cloud data based on the result of the abnormal identification processing.
In a second aspect, there is provided an apparatus for data processing, the apparatus comprising:
The acquisition unit is used for acquiring the environment information of the automatic driving vehicle and the point cloud data to be processed;
The simulation unit is used for obtaining simulated point cloud data corresponding to the environment information by utilizing a preset ray tracing simulation algorithm based on the environment information;
The evaluation unit is used for performing evaluation processing on the point cloud data to be processed based on the simulated point cloud data so as to obtain an evaluation processing result;
The identification unit is used for carrying out abnormal identification processing on the point cloud data to be processed based on the evaluation processing result so as to obtain abnormal point cloud data;
And the filtering unit is used for filtering the abnormal point cloud data in the point cloud data to be processed so as to obtain target point cloud data.
In a third aspect, there is provided a computer readable storage medium having stored therein at least one instruction that is loaded and executed by a processor to implement the aspects and methods of any one of the possible implementations as described above.
In a fourth aspect, there is provided an electronic device comprising:
at least one processor; and
A memory communicatively coupled to the at least one processor; wherein,
The memory stores instructions executable by the at least one processor to enable the at least one processor to perform the aspects and methods of any one of the possible implementations described above.
In a fifth aspect, there is provided a computer program product comprising a computer program which, when executed by a processor, implements the method of the aspects and any one of the possible implementations described above.
In a sixth aspect, there is provided an autonomous vehicle comprising an electronic device as described above.
The technical scheme provided by the application has the beneficial effects that at least:
as can be seen from the above technical solution, on the one hand, according to the embodiment of the present application, the environmental information of the automatic driving vehicle and the point cloud data to be processed may be obtained, and further, based on the environmental information, the simulated point cloud data corresponding to the environmental information may be obtained by using a preset ray tracing simulation algorithm, and based on the simulated point cloud data, the point cloud data to be processed may be evaluated to obtain an evaluation result, and based on the evaluation result, the point cloud data to be processed may be subjected to an anomaly identification process to obtain the anomaly point cloud data, so that the anomaly point cloud data in the point cloud data to be processed may be filtered to obtain the target point cloud data.
It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the application or to delineate the scope of the application. Other features of the present application will become apparent from the description that follows.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method of data processing provided by one embodiment of the present application;
FIG. 2 is a schematic diagram of a method of data processing provided by another embodiment of the present application;
fig. 3 is a block diagram of an apparatus for data processing according to another embodiment of the present application.
Detailed Description
Exemplary embodiments of the present application will now be described with reference to the accompanying drawings, in which various details of the embodiments of the present application are included to facilitate understanding, and are to be considered merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the application. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
It will be apparent that the described embodiments are some, but not all, embodiments of the application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
It should be noted that, the terminal device in the embodiment of the present application may include, but is not limited to, smart devices such as a mobile phone, a Personal digital assistant (Personal DIGITAL ASSISTANT, PDA), a wireless handheld device, and a Tablet Computer (Tablet Computer); the display device may include, but is not limited to, a personal computer, a television, or the like having a display function.
In addition, the term "and/or" herein is merely an association relationship describing an association object, and means that three relationships may exist, for example, a and/or B may mean: a exists alone, A and B exist together, and B exists alone. In addition, the character "/" herein generally indicates that the front and rear associated objects are an "or" relationship.
The rapid development of autopilot technology has enabled lidar to play a key role in sensing the environment. The lidar acquires three-dimensional point cloud data of the surrounding environment by emitting a laser beam and measuring its return time. However, due to the wide variety of disturbances and noise present in real scenes, point cloud data collected by lidar is often affected by wire drawing noise, which can lead to inaccuracy and instability of the environmental perception.
At present, a filtering method for laser radar point cloud wiredrawing noise is mainly concentrated on two layers of software and hardware. At the software level, filtering algorithms, such as gaussian filtering, median filtering, etc., are typically employed to post-process the lidar point cloud data. However, these conventional filtering methods often have difficulty in sufficiently retaining detailed information of the target point cloud, and may cause erroneous judgment of the real target. At the hardware level, some laser radar devices adopt a multi-line laser radar technology, and the influence of wiredrawing noise is reduced by increasing the number of laser beams. However, this method increases the cost and complexity of the system, and is not conducive to popularization in large-scale applications.
Therefore, it is desirable to provide a data processing method, which can effectively filter the wire drawing noise in the point cloud data, so as to ensure inaccuracy and instability of environmental perception.
Referring to fig. 1, a flow chart of a method for data processing according to an embodiment of the application is shown. The data processing method specifically comprises the following steps:
And 101, acquiring environmental information of the automatic driving vehicle and point cloud data to be processed.
Step 102, based on the environmental information, obtaining simulated point cloud data corresponding to the environmental information by using a preset ray tracing simulation algorithm.
And 103, based on the simulated point cloud data, performing evaluation processing on the point cloud data to be processed to obtain an evaluation processing result.
And 104, performing anomaly identification processing on the point cloud data to be processed based on the result of the evaluation processing so as to obtain abnormal point cloud data.
Step 105, filtering the abnormal point cloud data in the point cloud data to be processed to obtain target point cloud data.
Thus, after the target point cloud data is obtained, the automatic driving system can utilize the target point cloud data to identify the target object in the vehicle driving environment, so that the identification processing result is used for track planning and other decision planning.
It should be noted that the environmental information of the autonomous vehicle may include, but is not limited to, objects, obstacles, and other objects that may be sensed by the sensor. The environmental information of the autonomous vehicle may include image information acquired by a camera sensor.
It should be noted that, a thread execution based on the graphics processor (graphics processing unit, GPU) may use a ray tracing simulation algorithm to simulate the perception of environmental information by a laser beam of one path, so as to obtain corresponding simulated point cloud data.
Note that the outlier cloud data may include wire drawing noise data.
It should be noted that, part or all of the execution bodies of steps 101 to 105 may be an application located at the local terminal, or may be a functional unit such as a plug-in unit or a software development kit (Software Development Kit, SDK) provided in the application located at the local terminal, or may be a processing engine located in a server on the network side, or may be a distributed system located on the network side, for example, a processing engine or a distributed system in a data processing platform on the network side, which is not limited in this embodiment.
It will be appreciated that the application may be a local program (NATIVEAPP) installed on the local terminal, or may also be a web page program (webApp) of a browser on the local terminal, which is not limited in this embodiment.
In this way, the simulated point cloud data corresponding to the environmental information can be obtained by utilizing the ray tracing simulation algorithm, and then the point cloud data to be processed is evaluated according to the simulated point cloud data, so that the wire drawing noise data in the point cloud data can be accurately identified based on the evaluation result, the abnormal point cloud data can be effectively filtered, the negative influence of the wire drawing noise data on the performance of the automatic driving system can be reduced or eliminated, and the reliability and the stability of the automatic driving system on the environmental perception are improved.
Alternatively, in one possible implementation of the present embodiment, the preset ray tracing simulation algorithm may include a path simulation algorithm. In step 102, the path simulation algorithm may be specifically used to simulate the beam propagation path of the lidar of the autopilot vehicle based on the environmental information, so that simulated point cloud data corresponding to the environmental information may be obtained based on the environmental information and the beam propagation path.
In this implementation, the path simulation algorithm may be a simulation algorithm based on the physical principles of the propagation path of the light beam.
In another specific implementation of this implementation, first, propagation path information of a beam of the laser radar of the autonomous vehicle from the target object that is transmitted to the received environmental information may be calculated based on the environmental information to simulate a beam propagation path of the laser radar of the autonomous vehicle. And secondly, obtaining the simulated point cloud data corresponding to the environment information based on the target object and the light beam propagation path in the environment information.
In this volumetric implementation, the simulated point cloud data may include, but is not limited to, location data.
In this way, the beam propagation path of the laser radar of the automatic driving vehicle can be simulated by utilizing a path simulation algorithm to obtain simulated point cloud data corresponding to environmental information, and simulated point cloud data of symbol laser beam propagation path rules, namely correct point cloud data meeting the light propagation principle, can be obtained, so that the point cloud data to be processed can be evaluated by utilizing the simulated point cloud data in the follow-up process, and the effectiveness of point cloud data evaluation is improved.
Optionally, in a possible implementation manner of this embodiment, the preset ray tracing simulation algorithm may further include a relative motion simulation algorithm. In step 102, the relative motion between the autopilot and the target object may be simulated by using the relative motion simulation algorithm based on the environmental information, so as to obtain motion data of the autopilot and motion data of the target object in the environmental information, and further, simulated point cloud data corresponding to the environmental information may be obtained based on the motion data of the autopilot and the motion data of the target object in the environmental information.
In this implementation, the movement data of the autonomous vehicle may include, but is not limited to, speed, position, etc. of the vehicle. The motion data of the target object in the environmental information may include, but is not limited to, a speed, a position, etc. of the target object. Here, the motion data of the target object in the environmental information may be acquired by other sensors. The relative motion simulation algorithm may be a doppler effect based algorithm.
In one specific implementation of this implementation, based on the environmental information, the relative motion of the autonomous vehicle and the target object is simulated using an algorithm based on the doppler effect to obtain motion data of the autonomous vehicle and motion data of the target object in the environmental information.
In this volumetric implementation, the simulated point cloud data may include, but is not limited to, location data.
In this way, the relative motion of the automatic driving vehicle and the target object can be simulated by utilizing a relative motion simulation algorithm, so that simulated point cloud data corresponding to the environment information can be obtained based on the obtained motion data of the automatic driving vehicle and the motion data of the target object in the environment information, and the simulated point cloud data conforming to the relative motion rule of the target object and the laser radar can be obtained, so that the point cloud data to be processed can be evaluated by utilizing the simulated point cloud data in the follow-up process, and the effectiveness of point cloud data evaluation is improved.
It should be noted that, the specific implementation procedure provided in the present implementation manner may be combined with the various specific implementation procedures provided in the foregoing implementation manner to implement the data processing method of the present embodiment. The detailed description may refer to the relevant content in the foregoing implementation, and will not be repeated here.
Optionally, in a possible implementation manner of this embodiment, the preset light ray tracing simulation algorithm may further include a variable light path simulation algorithm. In step 102, the variable optical path propagation path of the beam of the laser radar of the autopilot vehicle may be simulated by using the variable optical path simulation algorithm based on the environmental information, and further simulated point cloud data corresponding to the environmental information may be obtained based on the environmental information and the variable optical path propagation path.
In this implementation, the variable optical path simulation algorithm may include a simulation algorithm based on multiple reflections and refractions. The variable optical path propagation path may include a new light propagation direction.
In one specific implementation of this implementation, first, a new light propagation direction of a beam of a lidar of an autonomous vehicle may be simulated using a simulation algorithm based on multiple reflections and refractions based on environmental information. Secondly, the position of the cloud point of the simulation point is determined based on the new light propagation direction. Again, simulated point cloud data corresponding to the environmental information may be obtained based on the environmental information and the location of the simulated point cloud points.
In this way, the variable light path simulation algorithm is utilized to simulate the variable light path propagation path of the light beam of the laser radar of the automatic driving vehicle so as to obtain simulated point cloud data corresponding to the environmental information, and the simulated point cloud data conforming to the rule of the variable light path propagation path can be obtained, namely, the correct point cloud data meeting the light propagation principle can be obtained, so that the point cloud data to be processed can be evaluated by utilizing the simulated point cloud data in the follow-up process, and the effectiveness of point cloud data evaluation is improved.
It should be noted that, the specific implementation procedure provided in the present implementation manner may be combined with the various specific implementation procedures provided in the foregoing implementation manner to implement the data processing method of the present embodiment. The detailed description may refer to the relevant content in the foregoing implementation, and will not be repeated here.
Optionally, in a possible implementation manner of this embodiment, the preset ray tracing simulation algorithm may further include an energy simulation algorithm. In step 102, the energy data of the beam of the lidar of the autopilot vehicle may be simulated by using the energy simulation algorithm based on the environmental information, and further simulated point cloud data corresponding to the environmental information may be obtained based on the environmental information and the energy data.
In this implementation, the point cloud data may include point cloud point energy. An energy simulation algorithm may be used to simulate the light energy of a point cloud.
Here, the simulated point cloud data may include, but is not limited to, location data and energy data.
In this way, the energy data of the laser radar beam of the automatic driving vehicle can be simulated by utilizing an energy simulation algorithm to obtain simulated point cloud data corresponding to environmental information, so that the simulated point cloud data conforming to the beam energy attenuation law of the laser radar can be obtained, namely, the correct point cloud data meeting the light energy attenuation principle can be obtained, the point cloud data to be processed can be conveniently evaluated by utilizing the simulated point cloud data, and the effectiveness of point cloud data evaluation is improved.
It should be noted that, the specific implementation procedure provided in the present implementation manner may be combined with the various specific implementation procedures provided in the foregoing implementation manner to implement the data processing method of the present embodiment. The detailed description may refer to the relevant content in the foregoing implementation, and will not be repeated here.
Alternatively, in one possible implementation of the present embodiment, the preset ray tracing simulation algorithm may include a plurality of simulation algorithms based on different ray tracing principles. In step 103, specifically, the confidence evaluation process may be performed on the point cloud data to be processed based on the simulated point cloud data obtained by using at least one simulation algorithm, so as to obtain at least one confidence value corresponding to each point cloud data, so that the at least one confidence value corresponding to each point cloud data can be used as a result of the evaluation process corresponding to each point cloud data, and a result of the evaluation process is obtained based on a result of the evaluation process corresponding to each point cloud data.
In this implementation, the preset ray tracing simulation algorithm may include a path simulation algorithm, a relative motion simulation algorithm, a variable optical path simulation algorithm, and an energy simulation algorithm.
In a specific implementation process of the implementation manner, based on the simulated point cloud data obtained by using a path simulation algorithm, the simulated point cloud data obtained by using a relative motion simulation algorithm, the simulated point cloud data obtained by using a variable optical path simulation algorithm and the simulated point cloud data obtained by using an energy simulation algorithm, confidence evaluation processing is performed on the point cloud data to be processed, and thus four confidence values corresponding to each point cloud data can be obtained, so that the confidence values of four dimensions corresponding to each point cloud data can be used as the result of the evaluation processing corresponding to each point cloud data, and the result of the evaluation processing is obtained based on the result of the evaluation processing corresponding to each point cloud data.
Optionally, in one possible implementation manner of this embodiment, in step 104, specifically, a preset anomaly identification algorithm may be used to perform anomaly identification processing on the point cloud data to be processed based on a result of the evaluation processing, and further, abnormal point cloud data may be obtained based on a result of the anomaly identification processing.
In this implementation, the preset anomaly identification algorithm may include, but is not limited to, a 3 Sigma (Sigma) criterion algorithm, a random sample consensus algorithm (Random Sample Consensus, RANSAC), a machine learning algorithm, and the like.
In a specific implementation process of the implementation manner, the 3Sigma criterion algorithm is utilized to perform anomaly identification processing on the result of evaluation processing of the point cloud data to be processed, and further abnormal point cloud data can be obtained based on the result of anomaly identification processing.
In this way, the abnormal point cloud data can be obtained by performing the abnormal recognition processing on the result of the evaluation processing of the point cloud data to be processed by using the preset abnormal recognition algorithm, so that the accuracy of recognizing the abnormal point cloud data can be improved.
It should be noted that, the specific implementation procedure provided in the present implementation manner may be combined with the various specific implementation procedures provided in the foregoing implementation manner to implement the data processing method of the present embodiment. The detailed description may refer to the relevant content in the foregoing implementation, and will not be repeated here.
For better understanding of the method according to the embodiment of the present application, the method according to the embodiment of the present application is described below with reference to the accompanying drawings and specific application scenarios.
Fig. 2 is a flow chart of a method for data processing according to another embodiment of the present application, as shown in fig. 2. The application scene of the method can be a scene for identifying and filtering wiredrawing noise data in the point cloud data.
Step 201, acquiring environment information of an automatic driving vehicle and point cloud data to be processed.
In the present embodiment, here, the environmental information may be environmental information obtained by performing preprocessing based on time synchronization. The point cloud data to be processed may also be point cloud data obtained by performing preprocessing based on time synchronization.
Step 202, based on the environmental information, simulating a beam propagation path of a laser radar of an automatic driving vehicle by using a path simulation algorithm, and obtaining first simulation point cloud data corresponding to the environmental information.
And 203, based on the environmental information, simulating the relative motion of the automatic driving vehicle and the target object by using a relative motion simulation algorithm so as to obtain second simulation point cloud data corresponding to the environmental information.
Step 204, based on the environmental information, simulating a variable optical path propagation path of the beam of the laser radar of the autonomous vehicle by using a variable optical path simulation algorithm to obtain third simulated point cloud data corresponding to the environmental information.
Step 205, based on the environmental information, simulating energy data of a beam of the laser radar of the autonomous vehicle by using an energy simulation algorithm to obtain fourth simulated point cloud data corresponding to the environmental information.
In this embodiment, a thread of the GPU may perform a ray tracing simulation algorithm to simulate the perception of environmental information by a beam of a laser beam, so as to obtain corresponding simulated point cloud data.
It can be understood that based on parallel computing of the GPU, tasks such as different paths of ray tracing of a laser beam or intersection detection of a ray and different targets are allocated to multiple cores of the GPU to perform parallel computing by utilizing the SIMD instruction set, so that large-scale parallel processing is realized.
In this embodiment, the path simulation algorithm, the relative motion simulation algorithm, the variable optical path simulation algorithm, and the energy simulation algorithm are all algorithms based on the ray tracing principle. Here, based on the environmental information, the path simulation algorithm, the relative motion simulation algorithm, the variable optical path simulation algorithm, and the energy simulation algorithm may be used to perform ray tracing processing on the laser beam of the laser radar, so as to obtain the simulated point cloud data.
It will be appreciated that here, the path simulation principle may be to accurately simulate the propagation path of each laser beam. The entire propagation of the laser beam from transmission to reception is calculated, including interactions with the target object in the environmental information. The principle of relative motion simulation may be to obtain from a sensor the relative motion between a lidar sensor and a target object in the surrounding environment. The relative motion includes, but is not limited to, motion of a vehicle, motion of a target object, and other potential sources of motion. The principle of the variable optical path simulation algorithm, i.e. the multiple reflection and refraction simulation, may be to calculate the multiple reflections and refraction that may occur when the laser beam interacts with the target object surface. Multiple reflection and refraction paths from the emission to the target object surface, the target object surface to the laser radar sensor. For each reflection and refraction, a new ray propagation direction is calculated to determine the position of the point cloud. The energy simulation, i.e. the principle of simulating the energy attenuation of a laser beam, may refer to the fact that the laser beam encounters various object surfaces during propagation, including target objects, ground and other obstacles. When the laser beam interacts with these surfaces, a portion of the light energy may be reflected, refracted, scattered, or absorbed, resulting in a gradual decrease in the energy of the laser beam. The energy attenuation may cause the energy of the target object to be attenuated differently at different distances along the propagation path of the laser beam, so that the energy information of the point cloud can be simulated.
Step 206, performing confidence evaluation processing on the point cloud data to be processed based on the first simulated point cloud data, the second simulated point cloud data, the third simulated point cloud data and the fourth simulated point cloud data to obtain a first confidence value, a second confidence value, a third confidence value and a fourth confidence value corresponding to each point cloud point in the point cloud data to be processed.
In this embodiment, matching processing may be performed on the first simulated point cloud data, the second simulated point cloud data, the third simulated point cloud data, and the fourth simulated point cloud data and the point cloud data to be processed, so as to obtain a confidence level corresponding to each point cloud data.
Here, the result of the evaluation process of each point cloud point may include a first confidence value, a second confidence value, a third confidence value, and a fourth confidence value corresponding to each point cloud point.
For example, if there are four kinds of analog point cloud points corresponding to one point cloud point to be processed, the confidence values of the four dimensions corresponding to the point cloud point to be processed are all 1, that is, the confidence values of the four dimensions are (1, 1). If two corresponding simulated point clouds exist, the confidence values of the corresponding four dimensions are (1, 0).
Step 207, performing anomaly identification processing on the point cloud data to be processed by using a preset anomaly identification algorithm based on the first confidence value, the second confidence value, the third confidence value and the fourth confidence value to obtain anomaly point cloud data.
In this embodiment, the preset anomaly identification algorithm may include, but is not limited to, a 3 Sigma (Sigma) criterion algorithm, a random sample consensus algorithm (Random Sample Consensus, RANSAC), a machine learning algorithm, and the like.
Illustratively, based on the first confidence value, the second confidence value, the third confidence value and the fourth confidence value, performing anomaly identification processing on the point cloud data to be processed by using a 3Sigma criterion to obtain abnormal point cloud data.
It will be appreciated that in conjunction with the 3Sigma criterion, point cloud data that does not conform to the ray tracing principles may be identified and filtered. The stringing noise data typically appears as some outlier points in the point cloud that may deviate from the normal motion pattern. These potential stringing noise data are effectively captured by marking points that do not meet the 3Sigma criterion as outliers.
And step 208, filtering the abnormal point cloud data to obtain target point cloud data.
Thus, after the target point cloud data after the filtering processing is obtained, the target point cloud data after the filtering processing can be output to a downstream module so as to be used for processing such as track planning, target object detection and other behavior decisions, and the safe running of the automatic driving vehicle is controlled.
In this embodiment, based on the parallel computing capability of the GPU, tasks such as different paths of ray tracing of the laser radar beam or intersection detection of the ray and different targets are allocated to multiple cores of the GPU to perform parallel computing by using the SIMD instruction set, so that large-scale parallel processing is realized. In the GPU-based local memory (L1) and the shared memory (L2), the access delay to the global memory is reduced. Aiming at the hardware characteristics of the GPU, the hardware thread bundles are optimized through stream processing and asynchronous calculation, and each thread bundle can be guaranteed to be executed efficiently.
Based on the technical scheme of the embodiment, the radar wiredrawing noise filtering algorithm based on GPU optical tracking combines the high parallel computing capability of the GPU by utilizing the optical tracking principle, and can process laser radar point cloud data more efficiently and with low cost when processing automatic driving large-scale point cloud real-time data, thereby effectively reducing or eliminating the negative influence of wiredrawing noise on the performance of an automatic driving system, improving the environment perception performance of the automatic driving system, reducing the position offset of a target point and the appearance of false target points in the point cloud data, and further providing more accurate and reliable input data for track planning and decision of vehicles. By realizing the efficient filtering of the laser radar point cloud data, the safety, stability and adaptability of the automatic driving system can be improved, and the application of the automatic driving technology in various traffic scenes is promoted.
It should be noted that, for simplicity of description, the foregoing method embodiments are all described as a series of acts, but it should be understood and appreciated by those skilled in the art that the present application is not limited by the order of acts, as some steps may be performed in other orders or concurrently in accordance with the present application. Further, those skilled in the art will also appreciate that the embodiments described in the specification are all preferred embodiments, and that the acts and modules referred to are not necessarily required for the present application.
In the foregoing embodiments, the descriptions of the embodiments are emphasized, and for parts of one embodiment that are not described in detail, reference may be made to related descriptions of other embodiments.
Fig. 3 is a block diagram showing a structure of an apparatus for data processing according to an embodiment of the present application, as shown in fig. 3. The apparatus 300 for data processing of the present embodiment may include an acquisition unit 301, an analog unit 302, an evaluation unit 303, an identification unit 304, and a filtering unit 305. The acquiring unit 301 is configured to acquire environmental information of an autonomous vehicle and point cloud data to be processed; the simulation unit 302 is configured to obtain simulated point cloud data corresponding to the environmental information by using a preset ray tracing simulation algorithm based on the environmental information; an evaluation unit 303, configured to perform evaluation processing on the point cloud data to be processed based on the simulated point cloud data, so as to obtain a result of the evaluation processing; an identifying unit 304, configured to perform an anomaly identification process on the point cloud data to be processed based on a result of the evaluation process, so as to obtain abnormal point cloud data; and a filtering unit 305, configured to perform filtering processing on the abnormal point cloud data in the point cloud data to be processed, so as to obtain target point cloud data.
It should be noted that, part or all of the data processing apparatus in this embodiment may be an application located at the local terminal, or may also be a functional unit such as a plug-in unit or a software development kit (Software Development Kit, SDK) disposed in the application located at the local terminal, or may also be a processing engine located in a server on the network side, or may also be a distributed system located on the network side, for example, a processing engine or a distributed system in a data processing platform on the network side, which is not limited in this embodiment.
It will be appreciated that the application may be a local program (NATIVEAPP) installed on the local terminal, or may also be a web page program (webApp) of a browser on the local terminal, which is not limited in this embodiment.
Optionally, in one possible implementation manner of this embodiment, the preset light ray tracing simulation algorithm includes a path simulation algorithm, and the simulation unit 302 may be specifically configured to simulate, based on the environmental information, a beam propagation path of the lidar of the autopilot vehicle by using the path simulation algorithm, and obtain, based on the environmental information and the beam propagation path, simulated point cloud data corresponding to the environmental information.
Optionally, in one possible implementation manner of this embodiment, the preset light ray tracing simulation algorithm further includes a relative motion simulation algorithm, and the simulation unit 302 may specifically be configured to simulate, based on the environmental information, relative motion of the autonomous vehicle and the target object by using the relative motion simulation algorithm, so as to obtain motion data of the autonomous vehicle and motion data of the target object in the environmental information, and obtain simulated point cloud data corresponding to the environmental information based on the motion data of the autonomous vehicle and the motion data of the target object in the environmental information.
Optionally, in one possible implementation manner of this embodiment, the preset light trace simulation algorithm includes a variable light path simulation algorithm, and the simulation unit 302 may be specifically configured to simulate, based on the environmental information, a variable light path propagation path of a light beam of the laser radar of the autopilot vehicle by using the variable light path simulation algorithm, and obtain, based on the environmental information and the variable light path propagation path, simulated point cloud data corresponding to the environmental information.
Optionally, in one possible implementation manner of this embodiment, the preset light ray tracing simulation algorithm further includes an energy simulation algorithm, and the simulation unit 302 may be specifically configured to simulate, based on the environmental information, energy data of a beam of the laser radar of the autopilot vehicle by using the energy simulation algorithm, and obtain, based on the environmental information and the energy data, simulated point cloud data corresponding to the environmental information.
Optionally, in one possible implementation manner of this embodiment, the identifying unit 304 may be specifically configured to perform, based on a result of the evaluation process, an anomaly identification process on the point cloud data to be processed by using a preset anomaly identification algorithm, and obtain the anomaly point cloud data based on the result of the anomaly identification process.
In this embodiment, the environmental information of the automatic driving vehicle and the point cloud data to be processed may be obtained by the obtaining unit, and then the simulated point cloud data corresponding to the environmental information may be obtained by the simulation unit based on the environmental information and using a preset ray tracing simulation algorithm, the evaluation unit performs the evaluation processing on the point cloud data to be processed based on the simulated point cloud data to obtain the result of the evaluation processing, and the recognition unit performs the anomaly recognition processing on the point cloud data to be processed based on the result of the evaluation processing to obtain the anomaly point cloud data, so that the filtering unit may perform the filtering processing on the anomaly point cloud data in the point cloud data to be processed to obtain the target point cloud data.
One embodiment of the present application provides a computer-readable storage medium having stored therein at least one instruction that is loaded and executed by a processor to implement a method of data processing as described above.
According to embodiments of the present application, the present application also provides an electronic device, a readable storage medium and a computer program product.
One embodiment of the present application provides an autonomous vehicle including an electronic device as described above. Specifically, the autonomous vehicle may be a vehicle of the L2 class and above. Such as unmanned delivery vehicles, unmanned logistics vehicles, etc.
In the technical scheme of the application, related personal information of the user, such as collection, storage, use, processing, transmission, provision, disclosure and other processes of images, attribute data and the like of the user, accords with the regulations of related laws and regulations and does not violate the popular regulations.
It should be noted that: in the data processing apparatus provided in the foregoing embodiment, only the division of the functional modules is used for illustration, and in practical application, the above-mentioned functional allocation may be performed by different functional modules according to needs, that is, the internal structure of the data processing apparatus is divided into different functional modules, so as to perform all or part of the functions described above. In addition, the apparatus for data processing and the method embodiment for data processing provided in the foregoing embodiments belong to the same concept, and specific implementation processes of the apparatus for data processing and the method embodiment are detailed in the method embodiment, and are not repeated herein.
It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program for instructing relevant hardware, where the program may be stored in a computer readable storage medium, and the storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.
The above description should not be taken as limiting the embodiments of the application, but rather should be construed to cover all modifications, equivalents, improvements, etc. that may fall within the spirit and principles of the embodiments of the application.
Claims (11)
1. A method of data processing, the method comprising:
Acquiring environment information of an automatic driving vehicle and point cloud data to be processed;
Based on the environmental information, acquiring simulated point cloud data corresponding to the environmental information by using a preset ray tracing simulation algorithm;
Based on the simulated point cloud data, carrying out evaluation processing on the point cloud data to be processed to obtain an evaluation processing result;
Based on the evaluation processing result, carrying out anomaly identification processing on the point cloud data to be processed so as to obtain abnormal point cloud data;
And filtering the abnormal point cloud data in the point cloud data to be processed to obtain target point cloud data.
2. The method of claim 1, wherein the predetermined ray tracing simulation algorithm comprises a path simulation algorithm, wherein obtaining simulated point cloud data corresponding to the environmental information using the predetermined ray tracing simulation algorithm based on the environmental information comprises:
Simulating a beam propagation path of a lidar of the autonomous vehicle using the path simulation algorithm based on the environmental information;
And obtaining simulated point cloud data corresponding to the environment information based on the environment information and the beam propagation path.
3. The method according to claim 1 or 2, wherein the predetermined ray tracing simulation algorithm further includes a relative motion simulation algorithm, and based on the environmental information, obtaining simulated point cloud data corresponding to the environmental information using the predetermined ray tracing simulation algorithm includes:
simulating the relative motion of the autonomous vehicle and the target object by using the relative motion simulation algorithm based on the environment information so as to obtain motion data of the autonomous vehicle and motion data of the target object in the environment information;
And obtaining simulated point cloud data corresponding to the environmental information based on the motion data of the automatic driving vehicle and the motion data of the target object in the environmental information.
4. The method of claim 1, wherein the predetermined ray tracing simulation algorithm comprises a variable optical path simulation algorithm, and wherein obtaining simulated point cloud data corresponding to the environmental information using the predetermined ray tracing simulation algorithm based on the environmental information comprises:
Simulating a variable optical path propagation path of a beam of a laser radar of the autonomous vehicle using the variable optical path simulation algorithm based on the environmental information;
And obtaining simulated point cloud data corresponding to the environmental information based on the environmental information and the variable optical path propagation path.
5. The method according to any one of claims 1 to 4, wherein the predetermined ray tracing simulation algorithm further includes an energy simulation algorithm, and based on the environmental information, obtaining simulated point cloud data corresponding to the environmental information using the predetermined ray tracing simulation algorithm includes:
Simulating energy data of a beam of a lidar of the autonomous vehicle using the energy simulation algorithm based on the environmental information;
and obtaining simulated point cloud data corresponding to the environmental information based on the environmental information and the energy data.
6. The method according to claim 1, wherein the performing anomaly identification processing on the point cloud data to be processed based on the result of the evaluation processing to obtain anomaly point cloud data includes:
Based on the result of the evaluation processing, performing anomaly identification processing on the point cloud data to be processed by using a preset anomaly identification algorithm;
and acquiring abnormal point cloud data based on the result of the abnormal identification processing.
7. An apparatus for data processing, the apparatus comprising:
The acquisition unit is used for acquiring the environment information of the automatic driving vehicle and the point cloud data to be processed;
The simulation unit is used for obtaining simulated point cloud data corresponding to the environment information by utilizing a preset ray tracing simulation algorithm based on the environment information;
The evaluation unit is used for performing evaluation processing on the point cloud data to be processed based on the simulated point cloud data so as to obtain an evaluation processing result;
The identification unit is used for carrying out abnormal identification processing on the point cloud data to be processed based on the evaluation processing result so as to obtain abnormal point cloud data;
And the filtering unit is used for filtering the abnormal point cloud data in the point cloud data to be processed so as to obtain target point cloud data.
8. An electronic device, comprising:
at least one processor; and
A memory communicatively coupled to the at least one processor; wherein,
The memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-6.
9. A non-transitory computer readable storage medium storing computer instructions for causing the computer to perform the method of any one of claims 1-6.
10. A computer program product comprising a computer program which, when executed by a processor, implements the method according to any of claims 1-6.
11. An autonomous vehicle comprising the electronic device of claim 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410250228.XA CN117994764A (en) | 2024-03-05 | 2024-03-05 | Data processing method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410250228.XA CN117994764A (en) | 2024-03-05 | 2024-03-05 | Data processing method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117994764A true CN117994764A (en) | 2024-05-07 |
Family
ID=90897518
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410250228.XA Pending CN117994764A (en) | 2024-03-05 | 2024-03-05 | Data processing method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117994764A (en) |
-
2024
- 2024-03-05 CN CN202410250228.XA patent/CN117994764A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113902897B (en) | Training of target detection model, target detection method, device, equipment and medium | |
| CN108629231B (en) | Obstacle detection method, apparatus, device and storage medium | |
| CN106772435B (en) | A kind of unmanned plane barrier-avoiding method and device | |
| CN109087510B (en) | Traffic monitoring method and device | |
| US20130016896A1 (en) | 3D Visualization of Light Detection and Ranging Data | |
| CN113264066A (en) | Obstacle trajectory prediction method and device, automatic driving vehicle and road side equipment | |
| CN113177968A (en) | Target tracking method and device, electronic equipment and storage medium | |
| CN113780064A (en) | Target tracking method and device | |
| CN113392794B (en) | Vehicle line crossing identification method and device, electronic equipment and storage medium | |
| US20240203031A1 (en) | System and method for accelerated ray tracing with asynchronous operation and ray transformation | |
| CN112686951A (en) | Method, device, terminal and storage medium for determining robot position | |
| CN114091515A (en) | Obstacle detection method, obstacle detection device, electronic apparatus, and storage medium | |
| CN115728734B (en) | Laser radar shielding detection method and device, storage medium and laser radar | |
| CN115147831A (en) | Training method and device of three-dimensional target detection model | |
| CN114241448A (en) | Method and device for obtaining heading angle of obstacle, electronic equipment and vehicle | |
| CN114119465A (en) | Point cloud data processing method and device | |
| CN111736167B (en) | Method and device for obtaining laser point cloud density | |
| CN117994764A (en) | Data processing method, device, equipment and storage medium | |
| CN112540616B (en) | Laser point cloud generation method and device based on unmanned driving | |
| CN113516013B (en) | Target detection method, target detection device, electronic equipment, road side equipment and cloud control platform | |
| CN114674328A (en) | Map generation method, map generation device, electronic device, storage medium, and vehicle | |
| CN114140497A (en) | Target vehicle 3D real-time tracking method and system | |
| EP4124888B1 (en) | A method for processing a lidar point cloud to simulate the presence of an aerosol such as fog | |
| CN113640764B (en) | Radar one-dimensional range profile identification method and device based on multi-dimension one-dimensional convolution | |
| CN118762235A (en) | Label determining method, device, equipment and medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |