CN115294564A - Method, device, medium and electronic equipment for detecting identification effectiveness of point cloud data - Google Patents
Method, device, medium and electronic equipment for detecting identification effectiveness of point cloud data Download PDFInfo
- Publication number
- CN115294564A CN115294564A CN202210941964.0A CN202210941964A CN115294564A CN 115294564 A CN115294564 A CN 115294564A CN 202210941964 A CN202210941964 A CN 202210941964A CN 115294564 A CN115294564 A CN 115294564A
- Authority
- CN
- China
- Prior art keywords
- area
- point cloud
- cloud data
- corrected
- perception result
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/64—Three-dimensional objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/761—Proximity, similarity or dissimilarity measures
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The application relates to the technical field of automatic driving, and provides a method, a device, a medium and electronic equipment for detecting identification effectiveness of point cloud data. The method comprises the steps of obtaining preset point cloud data; the point cloud data comprises a to-be-corrected area, wherein the to-be-corrected area is an inaccurately identified point cloud area, the point cloud data is identified to obtain a perception result area, a target perception result area overlapped with the to-be-corrected area is determined according to the to-be-corrected area and the perception result area, and then a similarity parameter between the target perception result area and the to-be-corrected area is determined; and judging the effectiveness of the point cloud data identification based on the similarity parameter. The method can save the time consumption for detecting the validity of point cloud data identification and improve the efficiency for detecting the validity of point cloud data identification.
Description
Technical Field
The embodiment of the application relates to the technical field of automatic driving, in particular to a method, a device, a medium and electronic equipment for detecting identification effectiveness of point cloud data.
Background
Currently, point cloud data is typically identified using perceptual algorithms to enable object identification, such as laser detection and ranging (LIDAR) algorithms. If data in partial areas in the point cloud data are not enough to support a perception algorithm for judgment, the situation that accurate identification cannot be achieved can be caused, and missing detection and false detection can occur. Therefore, the perception algorithm often needs to be updated on a targeted basis.
In the related technology, the upgraded sensing algorithm needs to be tested in a real scene to detect whether the problem is solved, and the detection of the effectiveness of point cloud data identification usually involves many detection links and consumes a long time. Therefore, when the effectiveness of point cloud data identification is detected, the detection efficiency is low.
Disclosure of Invention
The embodiment of the application provides a method, a device, a medium and an electronic device for detecting the identification effectiveness of point cloud data, and the detection efficiency of the identification effectiveness of the point cloud data can be improved.
In a first aspect, an embodiment of the present application provides a method for detecting validity of point cloud data identification, including:
acquiring preset point cloud data, wherein the point cloud data comprises a to-be-corrected area, and the to-be-corrected area is an inaccurately identified point cloud area;
identifying the point cloud data to obtain a perception result area;
determining a target perception result area overlapped with the area to be corrected according to the area to be corrected and the perception result area;
determining a similarity parameter between the target perception result area and the area to be corrected;
and judging the effectiveness of the point cloud data identification based on the similarity parameter.
The method for detecting the identification effectiveness of the point cloud data comprises the steps of firstly obtaining preset point cloud data; the point cloud data comprises a to-be-corrected area, wherein the to-be-corrected area is an inaccurately identified point cloud area, the point cloud data is identified to obtain a perception result area, a target perception result area overlapped with the to-be-corrected area is determined according to the to-be-corrected area and the perception result area, and then a similarity parameter between the target perception result area and the to-be-corrected area is determined; and judging the effectiveness of the point cloud data identification based on the similarity parameter. According to the method, the preset point cloud data are identified, the similarity calculation is carried out based on the identification result, and the effectiveness of the point cloud data identification is judged according to the similarity parameter obtained by the similarity calculation.
In an optional embodiment, the point cloud data further includes a regular time range corresponding to the region to be corrected;
the identifying the point cloud data to obtain a perception result area comprises the following steps:
determining target point cloud data corresponding to the area to be corrected according to the regular time range and the point cloud data;
and identifying the target point cloud data to obtain a perception result area.
In this embodiment, the point cloud data further includes a regular time range corresponding to the area to be corrected; determining target point cloud data corresponding to the area to be corrected according to the regular time range and the point cloud data; the target point cloud data is identified to obtain a perception result area, and an acquisition mechanism of the target point cloud data is provided, so that the perception result area can be generated more efficiently, the time consumption for detecting the effectiveness of point cloud data identification is further saved, and the efficiency for detecting the effectiveness of point cloud data identification is improved.
In an optional embodiment, after the point cloud data is identified to obtain the sensing result area, the method further includes:
and if the to-be-corrected area and the perception result area are determined not to be overlapped according to the to-be-corrected area and the perception result area, performing point cloud data identification validity detection abnormal alarm.
In this embodiment, if it is determined that the area to be corrected and the sensing result area are not overlapped according to the area to be corrected and the sensing result area, an abnormal alarm for the identification and validity detection of the point cloud data is performed. According to the method, monitoring aiming at the condition that the area to be corrected and the sensing result area are not overlapped is set, and an abnormal warning mechanism is provided, so that the effectiveness of point cloud data identification can be determined more efficiently, the time consumption for detecting the effectiveness of point cloud data identification is saved, and the efficiency for detecting the effectiveness of point cloud data identification is improved.
In an alternative embodiment, the perception result region has a corresponding perception result time; the perception result time is used for determining a starting point cloud data frame and an ending point cloud data frame corresponding to a perception result area when the point cloud data are identified;
the determining a target perception result region overlapping with the region to be corrected according to the region to be corrected and the perception result region includes:
and if the current perception result area and the area to be corrected are overlapped in the perception result time range corresponding to the current perception result area, taking the current perception result area as a target perception result area.
In this embodiment, the sensing result region has a corresponding sensing result time, and in the process of determining the target sensing result region overlapping with the to-be-corrected region, if the current sensing result region overlaps with the to-be-corrected region within the sensing result time range corresponding to the current sensing result region, the current sensing result region is taken as the target sensing result region. According to the method, the target perception result area overlapped with the area to be corrected is determined based on the perception result time corresponding to the perception result area, so that the calculation amount during the determination of the target perception result area can be reduced, the time consumption for detecting the effectiveness of point cloud data identification is further saved, and the efficiency for detecting the effectiveness of point cloud data identification is improved.
In an optional embodiment, the determining a similarity parameter between the target perception result region and the region to be corrected includes:
determining a characteristic perception result area according to the target perception result area and the corresponding target perception result time range; the characteristic perception result area is obtained by splicing the tracks of each frame of the target perception result area in the corresponding target perception result time range;
determining a target feature perception result area overlapped with the area to be corrected according to the feature perception result area and the area to be corrected;
and obtaining the similarity parameter according to the target feature perception result region and the region to be corrected.
In this embodiment, the tracks of each frame of the target perception result region in the corresponding target perception result time range are spliced to obtain a feature perception result region; and determining a target feature perception result region overlapped with the region to be corrected according to the feature perception result region and the region to be corrected, and obtaining the similarity parameter according to the target feature perception result region and the region to be corrected. The method provides a mechanism for carrying out association evaluation on the target perception result area and the area to be corrected, can efficiently determine the similarity parameter of the target perception result area and the area to be corrected based on the target perception result time range and the target perception result area, and further improves the detection efficiency of the effectiveness of point cloud data identification.
In an optional embodiment, the obtaining the similarity parameter according to the target feature perception result region and the region to be corrected includes:
determining a first area of the target feature perception result region and a second area of the to-be-corrected region according to the target feature perception result region and the to-be-corrected region;
and taking the ratio of the first area to the second area as the similarity parameter.
In this embodiment, according to the target feature perception result region and the region to be corrected, a first area of the target feature perception result region and a second area of the region to be corrected are determined; and taking the ratio of the first area to the second area as the similarity parameter. The similarity parameter determined by the method is high in evaluation scale identification degree, the similarity between the target feature perception result area and the area to be corrected can be accurately reflected, the calculation amount required for determining the similarity parameter is small, the accuracy of evaluating the upgrade result of the upgraded perception algorithm can be improved, and the detection efficiency of the effectiveness of point cloud data identification is further improved.
In an optional embodiment, the determining the validity of the point cloud data identification based on the similarity parameter includes:
and under the condition that the similarity parameter is greater than a preset similarity threshold value, judging that the point cloud data identification is effective.
In this embodiment, if the similarity parameter is greater than a preset similarity threshold, it is determined that the point cloud data identification is valid. According to the method, the point cloud data can be judged to be effective on the basis of the similarity parameter conveniently by setting the preset similarity threshold, and the detection efficiency of the point cloud data identification effectiveness is further improved.
In a second aspect, an embodiment of the present application further provides an apparatus for detecting validity of point cloud data identification, including:
the device comprises a data acquisition unit, a data processing unit and a data processing unit, wherein the data acquisition unit is used for acquiring preset point cloud data, and the point cloud data comprises a to-be-corrected area, wherein the to-be-corrected area is a point cloud area which is not accurately identified;
the correction identification unit is used for identifying the point cloud data to obtain a perception result area;
an overlap determining unit, configured to determine, according to the to-be-corrected region and the sensing result region, a target sensing result region that overlaps with the to-be-corrected region;
a similarity determining unit, configured to determine a similarity parameter between the target sensing result region and the region to be corrected;
and the upgrading identification unit is used for judging the effectiveness of the point cloud data identification based on the similarity parameter.
In an optional embodiment, the point cloud data further includes a regular time range corresponding to the area to be corrected; the correction identification unit is specifically configured to:
determining target point cloud data corresponding to the area to be corrected according to the regular time range and the point cloud data;
and identifying the target point cloud data to obtain a perception result area.
In an alternative embodiment, the apparatus further comprises:
and the abnormal alarm unit is used for carrying out point cloud data identification validity detection abnormal alarm if the to-be-corrected area and the perception result area are determined not to be overlapped according to the to-be-corrected area and the perception result area.
In an alternative embodiment, the perception result region has a corresponding perception result time; the perception result time is used for determining a starting point cloud data frame and an ending point cloud data frame corresponding to a perception result area when the point cloud data are identified; the overlap determining unit is specifically configured to:
and if the current perception result area and the area to be corrected are overlapped in the perception result time range corresponding to the current perception result area, taking the current perception result area as a target perception result area.
In an optional embodiment, the similarity determining unit is specifically configured to:
determining a characteristic perception result area according to the target perception result area and the corresponding target perception result time range; the characteristic perception result area is an area obtained by splicing the tracks of each frame of the target perception result area in the corresponding target perception result time range;
determining a target feature perception result area overlapped with the area to be corrected according to the feature perception result area and the area to be corrected;
and obtaining the similarity parameter according to the target feature perception result region and the region to be corrected.
In an optional embodiment, the similarity determining unit is specifically configured to:
determining a first area of the target feature perception result region and a second area of the to-be-corrected region according to the target feature perception result region and the to-be-corrected region;
and taking the ratio of the first area to the second area as the similarity parameter.
In an optional embodiment, the upgrade identification unit is specifically configured to:
and under the condition that the similarity parameter is greater than a preset similarity threshold value, judging that the point cloud data identification is effective.
In a third aspect, an embodiment of the present application further provides a computer-readable storage medium, where a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the method for detecting validity of point cloud data identification in the first aspect is implemented.
In a fourth aspect, the present application further provides an electronic device, including a memory and a processor, where the memory stores a computer program executable on the processor, and when the computer program is executed by the processor, the method for detecting validity of point cloud data identification of the first aspect is implemented.
In a fifth aspect, the present application further provides a computer program product, which includes computer instructions, when the computer instructions are executed by a computing device, the computing device may perform the method according to any one of the first aspect.
For technical effects brought by any one implementation manner of the second aspect to the fifth aspect, reference may be made to technical effects brought by a corresponding implementation manner of the first aspect, and details are not described here.
Drawings
Fig. 1 is a schematic flow chart of a method for detecting identification validity of point cloud data according to an embodiment of the present disclosure;
fig. 2 is a schematic flow chart of a region for obtaining a sensing result of the detection method for detecting the identification effectiveness of point cloud data according to the embodiment of the present application;
fig. 3 is a schematic flow chart illustrating a similarity parameter determination process of a method for detecting point cloud data identification validity according to an embodiment of the present disclosure;
fig. 4 is a schematic diagram of a feature perception result region obtained by stitching in the detection method for point cloud data identification validity provided in the embodiment of the present application;
fig. 5 is a schematic flowchart of another method for detecting validity of point cloud data identification according to an embodiment of the present disclosure;
fig. 6 is a schematic structural diagram of an apparatus for detecting validity of point cloud data identification according to an embodiment of the present disclosure;
fig. 7 is a schematic structural diagram of another apparatus for detecting validity of point cloud data identification according to an embodiment of the present disclosure;
fig. 8 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be described in further detail with reference to the accompanying drawings, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It should be noted that references to the terms "comprising" and "having," and variations thereof, in the context of this application are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
Some of the words that appear in the text are explained below:
(1) Point cloud data: the point cloud data refers to a set of sampling points with spatial coordinates obtained by a laser radar.
(2) LIDAR (Light Detection and Ranging) algorithm: the LIDAR algorithm is a programmed algorithm for object recognition of the point cloud data of the LIDAR. The laser radar uses a laser as a transmitting light source, obtains point cloud data by adopting a photoelectric detection technical means, and then carries out data processing on the point cloud data through a LIDAR algorithm so as to identify an object.
Currently, point cloud data is typically identified using perceptual algorithms to enable object identification, such as laser detection and ranging (LIDAR) algorithms. If data in partial areas in the point cloud data are not enough to support a perception algorithm for judgment, the situation that accurate identification cannot be achieved can be caused, and missing detection and false detection can occur. Therefore, the perception algorithm often needs to be updated on a targeted basis.
In the related art, the upgraded perception algorithm needs to be tested in a real scene to detect whether the problem is solved, and the detection of the effectiveness of point cloud data identification usually involves many detection links and consumes a long time. Therefore, when the effectiveness of point cloud data identification is detected, the detection efficiency is low.
In order to solve the above problems, an embodiment of the present application provides a method for detecting validity of point cloud data identification, which includes first obtaining preset point cloud data; the point cloud data comprise a point cloud area to be corrected, wherein the point cloud area to be corrected is an inaccurately identified point cloud area, the point cloud data are identified to obtain a perception result area, a target perception result area overlapped with the area to be corrected is determined according to the area to be corrected and the perception result area, and then a similarity parameter of the target perception result area and the area to be corrected is determined; and judging the effectiveness of point cloud data identification based on the similarity parameters. According to the method, the preset point cloud data are identified, the similarity calculation is carried out based on the identification result, and the effectiveness of the point cloud data identification is judged according to the similarity parameter obtained by the similarity calculation.
The technical solutions provided by the embodiments of the present application will be described in detail below with reference to the accompanying drawings.
Fig. 1 illustrates a schematic flow chart of detection of validity of point cloud data identification according to an embodiment of the present application. As shown in fig. 1, the method may include the steps of:
step S101, acquiring preset point cloud data; the point cloud data includes an area to be corrected.
Wherein, the area to be corrected is an area which is not accurately identified.
In specific implementation, the point cloud data comprises a to-be-corrected area and a normal area. The normal area is an accurately identified area.
In some embodiments of the present application, the preset point cloud data may be obtained by configuring through the following processes: determining point cloud data which are not accurately identified by a perception algorithm; determining an inaccurate identification time period and an object activity area of an inaccurate identification object by combining the point cloud data and the video image corresponding to the point cloud data; when the point cloud data is read by a player and a point cloud image corresponding to the point cloud data is displayed on a display frame by frame, an inaccurate identification area frame is drawn in an object moving area of an inaccurate identification object. And taking the interior of the inaccurate identification area frame as an area to be corrected, and taking the inaccurate identification area frame as a normal area.
In specific implementation, the point cloud data which is not accurately identified by the perception algorithm can be determined by comparing the video image with the object identification result of the perception algorithm. The inaccurately identified region box is typically a rectangular box.
Exemplarily, it is assumed that preset point cloud Data _1 includes a region to be corrected Scope _ Tar and a normal region Scope _ fine. The region to be modified Scope _ Tar may be an inner region of a rectangular frame Rec _ Tar, and the normal region Scope _ fine may be an outer region of the rectangular frame Rec _ Tar.
And S102, identifying point cloud data to obtain a perception result area.
In specific implementation, the point cloud data is identified through the upgraded sensing algorithm, and a sensing result area is obtained. The sensing result area obtained by identifying the point cloud data can be multiple.
Exemplarily, the point cloud Data _1 is identified through an upgraded perception algorithm, and a plurality of perception result areas (Scope _ Test _1, scope _ Test _2, …, scope _ Test _ I, …, scope _ Test _ n) are obtained.
In some embodiments of the present application, the point cloud data further comprises a regular time range corresponding to the area to be corrected.
In some embodiments of the present application, the preset point cloud data may be obtained by configuring through the following processes: determining point cloud data which are not accurately identified by a perception algorithm; determining an inaccurate identification time period and an object activity area of an inaccurate identification object by combining the point cloud data and the video image corresponding to the point cloud data; when the point cloud data is read by a player and the point cloud image corresponding to the point cloud data is displayed on a display frame by frame, an inaccurate identification area frame is drawn in an object moving area of an inaccurate identification object, and a certain time range is selected as a regular time range corresponding to the inaccurate identification area frame in the omission time period. And taking the interior of the inaccurate identification area frame as an area to be corrected, and taking the inaccurate identification area frame as a normal area.
It will be appreciated that the regular time range may be part or all of the inaccurately identified time period.
In some embodiments of the present application, the preset point cloud data may further include a plurality of regions to be corrected, where each region to be corrected is an internal region of a corresponding region frame that is not accurately identified. The regular time ranges corresponding to the inaccurate identification area frames of the areas to be corrected are not overlapped.
Fig. 2 is a schematic flow chart of a method for detecting validity of point cloud data identification according to an embodiment of the present application to obtain a perception result region. As shown in fig. 2, identifying point cloud data to obtain a sensing result region may be implemented by the following steps:
step S201, determining target point cloud data corresponding to the area to be corrected according to the regular time range and the point cloud data.
In specific implementation, the regular time range includes a starting time and an ending time for determining a point cloud data frame in the point cloud data, for example, assuming that the starting time of the point cloud data frame is a 2s time point of the point cloud data, and the ending time of the point cloud data frame is a 4s time point of the point cloud data, the regular time range is a 2 s-4 s time range, and the point cloud data frame between the 2s time point and the 4s time point of the point cloud data can be determined through the regular time range, so as to obtain the target point cloud data corresponding to the area to be corrected.
Step S202, identifying the target point cloud data to obtain a perception result area.
In specific implementation, when the upgraded perception algorithm identifies point cloud data, each point cloud data frame in a regular time range is identified.
The method of the embodiment provides a mechanism for acquiring the target point cloud data, and only the sensing result area corresponding to the target point cloud data is determined, so that the efficiency of the process of generating the sensing result area is higher, the time consumption for detecting the effectiveness of point cloud data identification is further saved, and the efficiency for detecting the effectiveness of point cloud data identification is improved.
Step S103, determining a target perception result area overlapped with the area to be corrected according to the area to be corrected and the perception result area.
In some embodiments of the present application, the perception result region has a corresponding perception result time; the perception result time is used for determining a starting point cloud data frame and an ending point cloud data frame corresponding to the perception result area when the point cloud data are identified; determining a target perception result area overlapped with the area to be corrected according to the area to be corrected and the perception result area, wherein the method can be realized by the following processes: acquiring the perception result areas one by one, and executing the following operations when acquiring one perception result area: and if the current perception result area and the area to be corrected are overlapped in the perception result time range corresponding to the current perception result area, taking the current perception result area as a target perception result area.
Exemplarily, it is assumed that the sensing result time corresponding to the sensing result region Scope _ Test _ I is (2.5 s to 3 s), and the point cloud Data _1 has 2 point cloud Data frames in total within a time range of 2.5s to 3s, including a starting point cloud Data frame PC _ I and an ending point cloud Data frame PC _ j. Determining a target perception result region overlapped with the region to be corrected according to the region to be corrected and the perception result region, specifically obtaining the perception result regions (Scope _ Test _1, scope _ Test _2, …, scope _ Test _ I, … and Scope _ Test _ n) one by one, and executing the following repeated operations when obtaining one perception result region: only taking the current sensing result region as Scope _ Test _ I as an example for explanation, if the current sensing result region Scope _ Test _ I and the region to be corrected Scope _ Tar are overlapped within the sensing result time range (2.5 s-3 s) corresponding to the current sensing result region Scope _ Test _ I, the current sensing result region Scope _ Test _ I is taken as the target sensing result region.
And step S104, determining a similarity parameter between the target perception result area and the area to be corrected.
In some embodiments of the present application, determining a similarity parameter between a target sensing result region and a region to be corrected, as shown in fig. 3, includes the following steps:
step S301, determining a characteristic perception result area according to the target perception result area and the corresponding target perception result time range.
The characteristic perception result area is obtained by splicing the tracks of each frame of the target perception result area in the corresponding target perception result time range.
Exemplarily, taking a target perception result area as Scope _ Test _ I, a target perception result time range corresponding to the target perception result area is (2.5 s-3 s), and a point cloud Data _1 has 2 point cloud Data frames in total in the time range of 2.5 s-3 s, as shown in fig. 4, tracks 401 and 402 of each frame of the target perception result area Scope _ Test _ I in the corresponding target perception result time range (2.5 s-3 s) are spliced to obtain an area Scope _ Test _ Joint, and the area Scope _ Test _ Joint is used as a feature perception result area.
Step S302, according to the feature perception result area and the area to be corrected, a target feature perception result area which is overlapped with the area to be corrected is determined.
Illustratively, referring to fig. 4, a target feature perception result region 404 overlapping with the region to be corrected 400 is determined according to the feature perception result region Scope _ Test _ Joint and the region to be corrected 400.
Step S303, a similarity parameter is obtained according to the target feature perception result area and the area to be corrected.
Illustratively, the similarity parameter Similar is obtained according to the target feature perception result area 404 and the area to be corrected 400.
In the method of this embodiment, the tracks of each frame of the target sensing result region in the corresponding target sensing result time range are spliced to obtain a feature sensing result region; and then according to the characteristic perception result area and the area to be corrected, determining a target characteristic perception result area overlapped with the area to be corrected, and according to the target characteristic perception result area and the area to be corrected, obtaining a similarity parameter.
In some embodiments of the present application, a similarity parameter is obtained according to the target feature sensing result region and the region to be corrected, and specifically: determining a first area of the target feature perception result region and a second area of the to-be-corrected region according to the target feature perception result region and the to-be-corrected region; and taking the ratio of the first area to the second area as a similarity parameter.
Exemplarily, taking the target feature sensing result area 404 and the area to be corrected 400 of fig. 4 as an example, according to the target feature sensing result area 404 and the area to be corrected 400, a first area S1 of the target feature sensing result area 404 and a second area S2 of the area to be corrected 400 are determined, and a ratio of the first area S1 to the second area S2 is used as a similarity parameter Similar.
And step S105, judging the effectiveness of point cloud data identification based on the similarity parameter.
In some embodiments of the present application, the determining the validity of the point cloud data identification based on the similarity parameter may specifically be: and under the condition that the similarity parameter is greater than a preset similarity threshold value, judging that the point cloud data identification is effective.
In the embodiment of the present application, the similarity threshold may be set to different values according to the specific requirement of the perceptual algorithm upgrade. For example, the similarity threshold may be set to 85%, and if the similarity parameter similarity is greater than 85%, the point cloud data identification is determined to be valid.
The method for detecting the identification effectiveness of the point cloud data comprises the steps of firstly obtaining preset point cloud data; the point cloud data comprises a to-be-corrected area, wherein the to-be-corrected area is an inaccurately identified point cloud area, then the point cloud data is identified to obtain a perception result area, a target perception result area overlapped with the to-be-corrected area is determined according to the to-be-corrected area and the perception result area, and then a similarity parameter of the target perception result area and the to-be-corrected area is determined; and judging the effectiveness of point cloud data identification based on the similarity parameters. According to the method, the preset point cloud data are identified, and the similarity calculation is carried out based on the identification result, so that the effectiveness of point cloud data identification is judged according to the similarity parameter obtained by the similarity calculation, a live-action test is not needed in the detection process, the effectiveness of point cloud data identification can be determined more conveniently, the time consumption for detecting the effectiveness of point cloud data identification is reduced, and the detection efficiency of the effectiveness of point cloud data identification is improved.
In some embodiments of the present application, the method further comprises the step of performing a point cloud data identification validity detection anomaly alarm. And identifying the point cloud data to obtain a perception result area, and if the area to be corrected and the perception result area are determined not to be overlapped according to the area to be corrected and the perception result area, performing point cloud data identification validity detection abnormal alarm.
In specific implementation, in the process of configuring and obtaining preset point cloud data, the to-be-corrected area avoids selecting a point cloud data frame with two objects with overlapped positions. When the upgraded sensing algorithm identifies point cloud data, a group of sensing result areas are output according to point cloud data frames in a regular time range, at most one target sensing result area is arranged in each group of sensing result areas, and the target sensing result area and the area to be corrected have an overlapping part in the corresponding sensing result time. And if the to-be-corrected area and the perception result area are determined not to be overlapped according to the to-be-corrected area and the perception result area, performing point cloud data identification validity detection abnormal alarm.
In the method of the embodiment, if the to-be-corrected area and the sensing result area are determined not to be overlapped according to the to-be-corrected area and the sensing result area, the point cloud data identification validity detection abnormal alarm is performed. According to the method, monitoring aiming at the condition that the area to be corrected and the perception result area are not overlapped is set, and an abnormal alarm mechanism is provided, so that the effectiveness of point cloud data identification can be determined more efficiently, the time consumption for detecting the effectiveness of point cloud data identification is saved, and the efficiency for detecting the effectiveness of point cloud data identification is improved.
Although the embodiments of the present application provide the operational steps of the methods as shown in the above embodiments or figures, more or less operational steps may be included in the above methods based on conventional or non-inventive efforts. In steps where no necessary causal relationship exists logically, the order of execution of these steps is not limited to the order of execution provided by the embodiments of the present application. The above methods may be performed in the order of the embodiments or in parallel in the method shown in the drawings when the method is executed in an actual process or apparatus.
Fig. 5 is a schematic flow chart of another method for detecting validity of point cloud data identification according to an embodiment of the present disclosure. As shown in fig. 5, the following steps may be included:
step S501, acquiring preset point cloud data; the point cloud data comprises a region to be corrected and a regular time range corresponding to the region to be corrected.
Wherein, the area to be corrected is an area which is not accurately identified.
In the point cloud data, the point cloud area except the area to be corrected is a normal area. The normal area is an accurately identified area.
Step S502, determining target point cloud data corresponding to the area to be corrected according to the regular time range and the point cloud data.
And S503, identifying the target point cloud data to obtain a perception result area.
Wherein, the perception result area has corresponding perception result time; and the perception result time is used for determining a starting point cloud data frame and an ending point cloud data frame corresponding to the perception result area when the point cloud data are identified. And identifying the target point cloud data, wherein a plurality of perception result areas can be obtained.
Step S504, according to the area to be corrected and the perception result area, determining a target perception result area overlapped with the area to be corrected.
In specific implementation, the process of determining the target sensing result area overlapped with the area to be corrected according to the area to be corrected and the sensing result area may specifically be that, if the current sensing result area and the area to be corrected are overlapped within the sensing result time range corresponding to the current sensing result area, the current sensing result area is taken as the target sensing result area.
In the embodiment of the application, the sensing result areas are acquired one by one, and each sensing result area is acquired, and if the acquired sensing result area and the area to be corrected are judged to be overlapped in the sensing result time range corresponding to the current sensing result area, the acquired sensing result area is used as the target sensing result area.
And step S505, determining a characteristic perception result area according to the target perception result area and the corresponding target perception result time range.
The characteristic perception result area is obtained by splicing the tracks of each frame of the target perception result area in the corresponding target perception result time range.
Step S506, according to the feature perception result area and the area to be corrected, a target feature perception result area which is overlapped with the area to be corrected is determined.
And step S507, obtaining a similarity parameter according to the target feature perception result area and the area to be corrected.
In some embodiments, the similarity parameter is obtained according to the target feature sensing result region and the region to be corrected, and may be a first area of the target feature sensing result region and a second area of the region to be corrected, which are determined according to the target feature sensing result region and the region to be corrected; and taking the ratio of the first area to the second area as a similarity parameter.
Step S508, if the similarity parameter is greater than the preset similarity threshold, it is determined that the point cloud data identification is valid.
According to the method, live-action test is not needed after the perception algorithm is upgraded, a new target feature perception result area is obtained through recognition of preset point cloud data after the algorithm is upgraded, the target feature perception result area is compared with a problem area of the perception algorithm before the algorithm is upgraded, effectiveness of point cloud data recognition is judged, whether the effectiveness of point cloud data recognition has a problem or not can be found quickly, effective obtaining of the effectiveness of point cloud data recognition is achieved, time and material resource consumption is reduced, efficiency of detecting the effectiveness of point cloud data recognition can be improved, and monitoring cost is reduced.
Based on the same inventive concept, the embodiment of the present application further provides a device for detecting validity of point cloud data identification, and as the device is a device corresponding to the method for detecting validity of point cloud data identification in the embodiment of the present application, and the principle of the device for solving the problem is similar to that of the method, the implementation of the device may refer to the implementation process of the method embodiment, and repeated details are omitted.
Fig. 6 shows a schematic structural diagram of a device for detecting validity of point cloud data identification according to an embodiment of the present application. As shown in fig. 6, the apparatus for detecting the identification effectiveness of point cloud data includes: a data acquisition unit 601, a correction identification unit 602, an overlap determination unit 603, a similarity determination unit 604, and an upgrade identification unit 605; wherein, the first and the second end of the pipe are connected with each other,
a data obtaining unit 601, configured to obtain preset point cloud data, where the point cloud data includes a to-be-corrected area, and the to-be-corrected area is an area of a point cloud that is not accurately identified;
a correction identification unit 602, configured to identify point cloud data to obtain a sensing result area;
an overlap determining unit 603, configured to determine, according to the region to be corrected and the sensing result region, a target sensing result region that overlaps with the region to be corrected;
a similarity determining unit 604, configured to determine a similarity parameter between the target sensing result region and the region to be corrected;
and an upgrade identification unit 605 for determining validity of the point cloud data identification based on the similarity parameter.
In an optional embodiment, the point cloud data further comprises a regular time range corresponding to the area to be corrected; the modification identifying unit 602 is specifically configured to:
determining target point cloud data corresponding to the area to be corrected according to the regular time range and the point cloud data;
and identifying the target point cloud data to obtain a perception result area.
In an alternative embodiment, the perception result area has a corresponding perception result time; the perception result time is used for determining a starting point cloud data frame and an ending point cloud data frame corresponding to the perception result area when the point cloud data are identified; the overlap determining unit 603 is specifically configured to:
and if the current perception result area and the area to be corrected are overlapped in the perception result time range corresponding to the current perception result area, taking the current perception result area as a target perception result area.
In an alternative embodiment, the similarity determining unit 604 is specifically configured to:
determining a characteristic perception result area according to the target perception result area and the corresponding target perception result time range; the characteristic perception result area is obtained by splicing the tracks of each frame of the target perception result area in the corresponding target perception result time range;
determining a target feature perception result area overlapped with the area to be corrected according to the feature perception result area and the area to be corrected;
and obtaining a similarity parameter according to the target feature perception result area and the area to be corrected.
In an alternative embodiment, the similarity determining unit 604 is specifically configured to:
determining a first area of the target feature perception result region and a second area of the to-be-corrected region according to the target feature perception result region and the to-be-corrected region;
and taking the ratio of the first area to the second area as a similarity parameter.
In an alternative embodiment, the upgrade identification unit 605 is specifically configured to:
and under the condition that the similarity parameter is greater than a preset similarity threshold value, judging that the point cloud data identification is effective.
In an alternative embodiment, as shown in fig. 7, the apparatus further comprises:
and the abnormal warning unit 701 is configured to identify the point cloud data to obtain a sensing result region, and if it is determined that the region to be corrected and the sensing result region are not overlapped according to the region to be corrected and the sensing result region, perform abnormal warning for point cloud data identification validity detection.
The electronic equipment is based on the same inventive concept as the method embodiment, and the embodiment of the application also provides the electronic equipment. The electronic device can be used for detecting the effectiveness of point cloud data identification. In one embodiment, the electronic device may be a server, a terminal device, or other electronic devices. In this embodiment, the electronic device may be configured as shown in fig. 8, and include a memory 801, a communication module 803, and one or more processors 802.
A memory 801 for storing computer programs executed by the processor 802. The memory 801 may mainly include a program storage area and a data storage area, where the program storage area may store an operating system, programs required for running an instant messaging function, and the like; the storage data area can store various instant messaging information, operation instruction sets and the like.
The memory 801 may be a volatile memory (volatile memory), such as a random-access memory (RAM); the memory 801 may also be a non-volatile memory (non-volatile memory) such as, but not limited to, a read-only memory (rom), a flash memory (flash memory), a Hard Disk Drive (HDD) or a solid-state drive (SSD), or the memory 801 may be any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. The memory 801 may be a combination of the above memories.
The processor 802 may include one or more Central Processing Units (CPUs), or be a digital processing unit, etc. The processor 802 is configured to implement the above-described method for detecting the identification validity of the point cloud data when calling the computer program stored in the memory 801.
The communication module 803 is used for communicating with the terminal device and other servers.
The embodiment of the present application does not limit the specific connection medium among the memory 801, the communication module 803 and the processor 802. In fig. 8, the memory 801 and the processor 802 are connected by a bus 804, the bus 804 is represented by a thick line in fig. 8, and the connection manner between other components is merely illustrative and not limited. The bus 804 may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one thick line is shown in FIG. 8, but this is not intended to represent only one bus or type of bus.
The embodiment of the application also provides a computer storage medium, wherein computer executable instructions are stored in the computer storage medium and used for realizing the method for detecting the identification effectiveness of the point cloud data in any embodiment of the application.
According to an aspect of the application, a computer program product or computer program is provided, comprising computer instructions, the computer instructions being stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions, so that the computer device executes the detection method for the identification effectiveness of the point cloud data in the above embodiments. The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present application, and shall be covered by the scope of the present application.
Claims (10)
1. A method for detecting the identification effectiveness of point cloud data is characterized by comprising the following steps:
acquiring preset point cloud data, wherein the point cloud data comprise a to-be-corrected area, and the to-be-corrected area is an inaccurately identified point cloud area;
identifying the point cloud data to obtain a perception result area;
determining a target perception result area overlapped with the area to be corrected according to the area to be corrected and the perception result area;
determining a similarity parameter between the target perception result area and the area to be corrected;
and judging the effectiveness of the point cloud data identification based on the similarity parameter.
2. The method of claim 1, wherein the point cloud data further comprises a regular time range corresponding to the area to be corrected;
the identifying the point cloud data to obtain a perception result area comprises the following steps:
determining target point cloud data corresponding to the area to be corrected according to the regular time range and the point cloud data;
and identifying the target point cloud data to obtain a perception result area.
3. The method of claim 1, wherein after identifying the point cloud data to obtain a perception result region, the method further comprises:
and if the to-be-corrected area and the perception result area are determined not to be overlapped according to the to-be-corrected area and the perception result area, performing point cloud data identification validity detection abnormal alarm.
4. The method of claim 1, wherein the perception result region has a corresponding perception result time; the perception result time is used for determining a starting point cloud data frame and an ending point cloud data frame corresponding to a perception result area when the point cloud data are identified;
the determining, according to the region to be corrected and the sensing result region, a target sensing result region overlapping with the region to be corrected includes:
and if the current perception result area and the area to be corrected are overlapped in the perception result time range corresponding to the current perception result area, taking the current perception result area as a target perception result area.
5. The method according to claim 4, wherein the determining the similarity parameter between the target perception result area and the area to be corrected comprises:
determining a characteristic perception result area according to the target perception result area and the corresponding target perception result time range; the characteristic perception result area is obtained by splicing the tracks of each frame of the target perception result area in the corresponding target perception result time range;
determining a target feature perception result area overlapped with the area to be corrected according to the feature perception result area and the area to be corrected;
and obtaining the similarity parameter according to the target feature perception result region and the region to be corrected.
6. The method according to claim 5, wherein obtaining the similarity parameter according to the target feature perception result region and the region to be corrected comprises:
determining a first area of the target feature perception result region and a second area of the to-be-corrected region according to the target feature perception result region and the to-be-corrected region;
and taking the ratio of the first area to the second area as the similarity parameter.
7. The method of claim 6, wherein the determining the validity of the point cloud data identification based on the similarity parameter comprises:
and under the condition that the similarity parameter is greater than a preset similarity threshold value, judging that the point cloud data identification is effective.
8. A detection device for point cloud data identification effectiveness is characterized by comprising:
the data acquisition unit is used for acquiring preset point cloud data; the point cloud data comprises a to-be-corrected area, wherein the to-be-corrected area is a point cloud area which is not accurately identified;
the correction identification unit is used for identifying the point cloud data to obtain a perception result area;
an overlap determining unit, configured to determine, according to the to-be-corrected region and the sensing result region, a target sensing result region that overlaps with the to-be-corrected region;
a similarity determining unit, configured to determine a similarity parameter between the target sensing result region and the region to be corrected;
and the upgrading identification unit is used for judging the effectiveness of the point cloud data identification based on the similarity parameter.
9. A computer-readable storage medium having a computer program stored therein, the computer program characterized by: the computer program, when executed by a processor, implements the method of any one of claims 1 to 7.
10. An electronic device comprising a memory and a processor, the memory having stored thereon a computer program operable on the processor, the computer program, when executed by the processor, implementing the method of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210941964.0A CN115294564A (en) | 2022-08-08 | 2022-08-08 | Method, device, medium and electronic equipment for detecting identification effectiveness of point cloud data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210941964.0A CN115294564A (en) | 2022-08-08 | 2022-08-08 | Method, device, medium and electronic equipment for detecting identification effectiveness of point cloud data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115294564A true CN115294564A (en) | 2022-11-04 |
Family
ID=83828122
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210941964.0A Pending CN115294564A (en) | 2022-08-08 | 2022-08-08 | Method, device, medium and electronic equipment for detecting identification effectiveness of point cloud data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115294564A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117129979A (en) * | 2023-10-25 | 2023-11-28 | 深圳市迅龙软件有限公司 | Laser radar calibration method and system based on machine learning model |
-
2022
- 2022-08-08 CN CN202210941964.0A patent/CN115294564A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117129979A (en) * | 2023-10-25 | 2023-11-28 | 深圳市迅龙软件有限公司 | Laser radar calibration method and system based on machine learning model |
| CN117129979B (en) * | 2023-10-25 | 2024-02-13 | 深圳市迅龙软件有限公司 | Laser radar calibration method and system based on machine learning model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109165162B (en) | Test method and related equipment for vehicle-mounted system | |
| CN107230329B (en) | Fire detector calibration method, device and equipment | |
| EP3674750A1 (en) | Obstacle positioning method, device and terminal | |
| CN115294564A (en) | Method, device, medium and electronic equipment for detecting identification effectiveness of point cloud data | |
| CN105763505A (en) | Operation method and device based on user account | |
| US11094049B2 (en) | Computing device and non-transitory storage medium implementing target object identification method | |
| JP6158967B1 (en) | Environmental pollution prediction system and method | |
| CN110580220A (en) | method for measuring execution time of code segment and terminal equipment | |
| CN112182579B (en) | Process list generation method and device and abnormal process detection method and device | |
| CN116311144A (en) | Method and device for predicting vehicle steering and computer readable storage medium | |
| CN116303320A (en) | Real-time task management method, device, equipment and medium based on log file | |
| CN111143424A (en) | Characteristic scene data mining method and device and terminal | |
| CN112231194B (en) | Index abnormity root analysis method and device and computer readable storage medium | |
| EP3936977A1 (en) | Application program management method and apparatus, and storage medium | |
| CN113839912B (en) | Method, device, medium and equipment for analyzing abnormal host by active and passive combination | |
| CN114399669A (en) | Target detection method and device | |
| CN114034972A (en) | Intelligent cable fault determination method and device based on image data | |
| CN111143423B (en) | Dynamic scene labeling data mining method and device and terminal | |
| CN107908516B (en) | Data display method and device | |
| CN110704216A (en) | Application no-response detection method and device for terminal and electronic equipment | |
| CN111124847B (en) | Website user behavior analysis method and device | |
| CN112863096A (en) | Monitoring method and device | |
| US20230288394A1 (en) | Crop condition monitoring system and crop condition monitoring method using the same | |
| CN114152349B (en) | Temperature measurement method and device, storage medium and electronic equipment | |
| CN116597292A (en) | Image processing method, device, electronic equipment and storage 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 |