CN115293980A - Small-size dynamic noise filtering method and device based on historical information - Google Patents
Small-size dynamic noise filtering method and device based on historical information Download PDFInfo
- Publication number
- CN115293980A CN115293980A CN202210916049.6A CN202210916049A CN115293980A CN 115293980 A CN115293980 A CN 115293980A CN 202210916049 A CN202210916049 A CN 202210916049A CN 115293980 A CN115293980 A CN 115293980A
- Authority
- CN
- China
- Prior art keywords
- point cloud
- cloud data
- frame
- small
- current
- 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
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000001914 filtration Methods 0.000 title claims abstract description 28
- 238000005070 sampling Methods 0.000 claims description 8
- 230000008569 process Effects 0.000 description 10
- 239000000428 dust Substances 0.000 description 7
- 238000010586 diagram Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 238000001514 detection method Methods 0.000 description 4
- 230000002829 reductive effect Effects 0.000 description 4
- 230000003068 static effect Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000002592 echocardiography Methods 0.000 description 3
- 230000000670 limiting effect Effects 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 229910021389 graphene Inorganic materials 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000036961 partial effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
Images
Classifications
-
- G06T5/70—
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/48—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S17/00
- G01S7/4802—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S17/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/48—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S17/00
- G01S7/495—Counter-measures or counter-counter-measures using electronic or electro-optical means
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/53—Querying
- G06F16/532—Query formulation, e.g. graphical querying
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Optical Radar Systems And Details Thereof (AREA)
Abstract
The application provides a small-size dynamic noise filtering method and device based on historical information, electronic equipment and a machine readable storage medium, wherein the method comprises the following steps: acquiring a point cloud data set corresponding to a current point cloud frame; constructing a set to be queried based on point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame; constructing a neighborhood corresponding to target point cloud data in a point cloud data set corresponding to the current point cloud frame, and counting the number N of the point cloud data in the neighborhood corresponding to the target point cloud data in the set to be inquired; and comparing the number N of the point cloud data with a threshold T for determining whether the target point cloud data is small-size dynamic noise, if the number N of the point cloud data is less than the threshold T, determining the target point cloud data as small-size dynamic noise, and filtering the small-size dynamic noise.
Description
Technical Field
The application relates to the technical field of computer vision, in particular to a small-size dynamic noise filtering method and device based on historical information, electronic equipment and a machine readable storage medium.
Background
In the field of computer vision technology, a laser radar is commonly used for acquiring point cloud data. In the process of acquiring the point cloud data of the measuring environment by the laser radar, the point cloud data of small-size dynamic objects such as dust, tail gas, plastic bags, floating leaves, flying birds and the like in the measuring environment are small-size dynamic noise points for the measuring environment. Therefore, how to filter out small-size dynamic noise is a technical problem to be solved urgently in the field.
Disclosure of Invention
The application provides a small-size dynamic noise point filtering method based on historical information, which is characterized by comprising the following steps:
acquiring a point cloud data set corresponding to a current point cloud frame;
constructing a set to be queried based on the point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame;
constructing a neighborhood corresponding to target point cloud data in a point cloud data set corresponding to the current point cloud frame, and counting the number N of the point cloud data in the neighborhood corresponding to the target point cloud data in the set to be inquired;
and comparing the number N of the point cloud data with a threshold T for determining whether the target point cloud data is small-size dynamic noise, if the number N of the point cloud data is less than the threshold T, determining the target point cloud data as small-size dynamic noise, and filtering the small-size dynamic noise.
Optionally, the constructing a set to be queried based on the point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame includes:
and acquiring point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame, overlapping the point cloud data, and constructing a set to be queried.
Optionally, the obtaining of the point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame includes:
and acquiring point cloud data corresponding to the current point cloud frame, and selecting point cloud data corresponding to at least one frame of point cloud frame before the current point cloud frame at preset intervals.
Optionally, the constructing a neighborhood corresponding to target point cloud data in the point cloud data set corresponding to the current point cloud frame includes:
and determining target point cloud data in a point cloud data set corresponding to the current point cloud frame, and determining a flat column space taking the target point cloud data as a center as a neighborhood corresponding to the target point cloud data, wherein the bottom surface of the flat column space is parallel to a horizontal plane.
Optionally, the method further includes:
the radius of the flat cylinder space is determined based on the angular resolution of the lidar, and/or the degree of voxel sampling.
The application provides a small-size dynamic noise filtering device based on historical information, the device includes:
the point cloud data acquisition module is used for acquiring a point cloud data set corresponding to the current point cloud frame;
the query set building module is used for building a query set based on the point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame;
the point cloud data counting module is used for constructing a neighborhood corresponding to target point cloud data in a point cloud data set corresponding to the current point cloud frame, and counting the number N of the point cloud data in the neighborhood corresponding to the target point cloud data in the set to be inquired;
and the noise filtering module is used for comparing the number N of the point cloud data with a threshold T for determining whether the target point cloud data is small-size dynamic noise, and if the number N of the point cloud data is smaller than the threshold T, determining the target point cloud data as the small-size dynamic noise and filtering the small-size dynamic noise.
Optionally, the constructing a set to be queried based on the point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame includes:
and acquiring point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame, overlapping the point cloud data, and constructing a set to be queried.
Optionally, the acquiring point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame includes:
and acquiring point cloud data corresponding to the current point cloud frame, and selecting point cloud data corresponding to at least one frame of point cloud frame before the current point cloud frame at preset intervals.
Optionally, the constructing a neighborhood corresponding to target point cloud data in the point cloud data set corresponding to the current point cloud frame includes:
determining target point cloud data in a point cloud data set corresponding to the current point cloud frame, and determining a flat column space with the target point cloud data as a center as a neighborhood corresponding to the target point cloud data, wherein the bottom surface of the flat column space is parallel to a horizontal plane.
Optionally, the apparatus further comprises:
the radius of the flat cylinder space is determined based on the angular resolution of the lidar, and/or the degree of voxel sampling.
The present application further provides an electronic device, comprising:
a processor;
a memory for storing processor-executable instructions;
wherein the processor implements the steps of the above method by executing the executable instructions.
The present application also provides a machine-readable storage medium having stored thereon computer instructions which, when executed by a processor, implement the steps of the above-described method.
In summary, the method can construct a set to be queried through the history information, query the number of point cloud data in the neighborhood of the target point cloud data in the set to be queried, and judge whether the target point cloud data is small-size dynamic noise based on the number of the point cloud data, so that the small-size dynamic noise can be filtered.
Drawings
FIG. 1 is a flow diagram illustrating a method for small-scale dynamic noise filtering based on historical information in accordance with an exemplary embodiment;
FIG. 2 is a block diagram of a small-scale dynamic noise filtering apparatus based on historical information according to an exemplary embodiment;
fig. 3 is a hardware block diagram of an electronic device in which a small-sized dynamic noise filtering apparatus based on historical information is located according to an exemplary embodiment.
Detailed Description
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.
It should be noted that: in other embodiments, the steps of the corresponding methods are not necessarily performed in the order shown and described in this specification. In some other embodiments, the method may include more or fewer steps than those described herein. Moreover, a single step described in this specification may be broken down into multiple steps for description in other embodiments; multiple steps described in this specification may be combined into a single step in other embodiments.
In order to make those skilled in the art better understand the technical solutions in the embodiments of the present disclosure, the related technologies related to the embodiments of the present disclosure will be briefly described below.
Point cloud data: laser pulses are emitted outwards by the laser radar, and are reflected from the ground or the surface of an object to form a plurality of echoes, the echoes return to the laser radar sensor, and the processed reflected data is called point cloud data.
Application scenario overview
In the field of computer vision technology, a laser radar is commonly used for acquiring point cloud data. In the process of acquiring the point cloud data of the measuring environment by the laser radar, the point cloud data of small-size dynamic objects such as dust, tail gas, plastic bags, floating leaves, flying birds and the like in the measuring environment are small-size dynamic noise points for the measuring environment.
For example, in the field of autonomous driving, autonomous vehicles typically scan a scene using a laser radar to obtain point cloud data. However, due to the high-precision characteristic of the laser radar, dust, tail gas, fallen leaves and other small-size dynamic objects in a scene do not cause interference to vehicle driving, point cloud data can be generated in the process that the laser radar scans the scene, and the point cloud data can be regarded as small-size dynamic noise points and cause interference to path planning of an automatic driving vehicle, so that mistaken parking is caused.
In practical application, in order to avoid the small-size dynamic noise generated by the small-size dynamic object from interfering with the path planning of the autonomous vehicle, which results in a mis-stop, the following two schemes are commonly used:
one is to add acquisition equipment, such as a millimeter wave radar in addition to the laser radar. The spatial region without obstacle information in the data acquired by the millimeter wave radar can be used as the region where the small-size dynamic object exists, and the point cloud data in the region can be removed from the point cloud data set acquired by the laser radar.
And secondly, the detectable category is perfected, and the small-size dynamic object is classified and identified. However, since the forms of small-sized dynamic objects such as dust, tail gas and the like are varied, a large number of samples need to be marked to realize effective identification; and a large number of samples are labeled, and the samples are only removed when the path planning is used after the types are accurately identified, so that the meaning of type detection is not great, and the waste of computing resources is caused.
Inventive concept
As described above, the small-sized dynamic noise generated by the small-sized dynamic objects interferes with the route planning of the autonomous vehicle, resulting in a mis-stop.
In view of this, the present specification aims to provide a scheme for filtering small-size dynamic noise points based on historical information.
The core concept of the specification is as follows:
and constructing a set to be queried according to the historical information, querying the number of point cloud data in the neighborhood of the target point cloud data in the set to be queried, and judging whether the target point cloud data is small-size dynamic noise points or not based on the number of the point cloud data.
By the method, the small-size dynamic noise points can be filtered based on the historical information without increasing the cost of acquisition equipment, and the waste of computational resources for classifying and identifying the small-size dynamic objects corresponding to the small-size dynamic noise points is avoided.
The present application is described below by using specific embodiments and in conjunction with specific application scenarios.
Referring to fig. 1, fig. 1 is a flow chart illustrating small-scale dynamic noise filtering based on historical information according to an exemplary embodiment, where the method performs the following steps:
step 102: and acquiring a point cloud data set corresponding to the current point cloud frame.
Step 104: and constructing a set to be queried based on the point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame.
Step 106: and constructing a neighborhood corresponding to target point cloud data in a point cloud data set corresponding to the current point cloud frame, and counting the number N of the point cloud data in the neighborhood corresponding to the target point cloud data in the set to be inquired.
Step 108: and comparing the number N of the point cloud data with a threshold T for determining whether the target point cloud data is small-size dynamic noise points, if the number N of the point cloud data is smaller than the threshold T, determining the target point cloud data as the small-size dynamic noise points, and filtering the small-size dynamic noise points.
In computer vision technical field, can use laser radar to obtain some cloud data, laser radar can scan the scene that locates at present through many scanning pencil of laser radar, obtains the some cloud data of current scene, and laser radar outwards launches laser pulse, from ground or object surface reflection, forms a plurality of echoes, returns laser radar sensor, and the reflection data after handling forms some cloud data. In the process of acquiring the point cloud data of the measuring environment by the laser radar, the point cloud data of small-size dynamic objects such as dust, tail gas, plastic bags, floating leaves, flying birds and the like in the measuring environment are small-size dynamic noise points for the measuring environment.
For example, in the field of autonomous driving, to meet the positioning requirement, an autonomous vehicle may generally scan a scene using a laser radar to obtain point cloud data. However, due to the high-precision characteristic of the laser radar, dust, tail gas, fallen leaves and other small-size dynamic objects in a scene do not cause interference to vehicle driving, point cloud data can be generated in the process that the laser radar scans the scene, and the point cloud data can be regarded as small-size dynamic noise points and cause interference to path planning of an automatic driving vehicle, so that mistaken parking is caused.
In order to avoid adverse effects caused by small-size dynamic noise, historical information can be used for filtering the small-size dynamic noise. The scheme described in the specification can acquire a point cloud data set corresponding to a current point cloud frame, and can construct a set to be queried based on the current point cloud frame and point cloud data corresponding to at least one previous point cloud frame.
After the point cloud data is acquired, the point cloud data corresponding to at least one frame of point cloud frame before the point cloud data can be constructed into a set to be queried, so that the current point cloud frame and the set to be queried can be compared conveniently, and small-size dynamic noise points can be identified. And point cloud data can be screened in combination with class detection, so that the comparison times are reduced, and the efficiency of removing small-size dynamic noise points is improved.
In an embodiment shown, the laser radar may obtain a point cloud data set corresponding to a current scene, may perform category detection on the point cloud data set, determine point cloud data corresponding to a detectable object in the set, and remove the point cloud data.
For example, in the field of automatic driving, point cloud data of a current scene acquired by a laser radar carried by a vehicle in a coordinate system corresponding to the laser radar may be converted into point cloud data of the coordinate system corresponding to the vehicle according to preset parameters of the laser radar, and voxel filtering and sampling may be performed on the point cloud data to obtain a current scene data set. The method comprises the steps that a current scene data set can be subjected to class detection, and point cloud data corresponding to a detectable object are determined; the method can be used for fitting the ground points in the current scene data set, determining point cloud data corresponding to the ground points, removing the point cloud data corresponding to the detectable objects and the point cloud data corresponding to the ground points, and reducing the comparison times.
After the query range of the small-size dynamic noise point is reduced by the method, a set to be queried can be constructed, and the target point cloud data in the target point cloud set with the reduced range is compared with the set to be queried, so that the small-size dynamic noise point is identified.
In an embodiment shown, point cloud data corresponding to a current point cloud frame and at least one previous point cloud frame may be obtained, and the point cloud data of the current point cloud frame and the point cloud data of the previous point cloud frame are overlapped to construct a set to be queried.
Due to the fact that the number of the point clouds after being superposed is large, the time is consumed in the neighborhood searching process, and part of most relevant data can be selected for processing. For example, in the field of automatic driving, part of target point cloud data in a target point cloud set in a coordinate system corresponding to a vehicle may be mapped to a global coordinate system, where the global coordinate system may have three coordinate axes of x, y, and z, and a xoy plane formed by an origin o, an x axis, and a y axis is a horizontal plane, and the z axis is perpendicular to the horizontal plane, and since a dynamic small-sized dynamic object with a longer distance has a smaller influence on the vehicle traveling, the part of the target point cloud data may be selected in a certain range around the vehicle, for example, point cloud data located in a range of 40 meters from the front, 10 meters from the left and right, and 20 meters from the back may be selected. The range can be reasonably selected according to the vehicle running speed and the frame rate of the laser radar.
In an embodiment shown, in order to further enlarge the interval of the dynamic object in the set to be queried and to have more obvious distinction degree with the static object, the point cloud data corresponding to the current point cloud frame may be acquired, and the point cloud data corresponding to at least one frame of point cloud frame before the current point cloud frame is selected at preset intervals.
For example, a current point cloud frame and 8 previous point cloud frames can be marked as point cloud frame 0, point cloud frame 1 \8230, 8230, point cloud frame 8, where point cloud frame 0 is the current frame, point cloud frame 1 is the point cloud frame one frame before the current frame, and so on, point cloud frame 8 is the point cloud frame eight frames before the current frame. The point cloud frames can be selected at intervals of one frame, and a set to be queried is constructed.
After the set to be queried is constructed, the target point cloud data can be subjected to radius neighborhood search in the set to be queried, and the number of point cloud data in a neighborhood corresponding to the target point cloud data is determined so as to be used for judging whether the target point cloud data is a small-size dynamic noise point.
In one embodiment shown, a neighborhood corresponding to the target point cloud data in the point cloud data set corresponding to the current point cloud frame may be constructed, and the number N of point cloud data in the neighborhood corresponding to the target point cloud data in the set to be queried may be counted.
It should be noted that the object corresponding to the small-size dynamic noise point may be an object in a moving state in the air, for example, the object may be dust, tail gas, fallen leaves, a bird, a plastic bag floating in the air, and the like.
Compared with an object in a static state, the object corresponding to the small-size dynamic noise point has position change in the direction vertical to the horizontal plane, so that the search range in the direction vertical to the horizontal plane can be reduced when the neighborhood is established for the target point cloud data, the counted number of the point cloud data corresponding to the small-size dynamic noise point is less when the target point cloud data is compared with the set to be inquired, and the accuracy of identifying the small-size dynamic noise point is improved.
For example, because the point cloud data acquired by the laser radar has errors, it cannot be completely accurate, and when the point cloud data in the coordinate system corresponding to the vehicle is converted into the global coordinate system, the corresponding position of the object in the point cloud frame will also deviate to some extent. Because the coordinate system corresponding to the vehicle is parallel to the global coordinate system, the deviation is mainly horizontal deviation, and meanwhile, because the position change of the small-size dynamic object in the z-axis direction is large, the neighborhood can be constructed into a flat cylinder space, so that under the condition of deviation, the static object still has more point cloud data in the flat cylinder space, and the small-size dynamic object has more obvious distinction degree.
In one illustrated embodiment, target point cloud data in a point cloud data set corresponding to a current point cloud frame may be determined, and a flat column space with the target point cloud data as a center is determined as a neighborhood corresponding to the target point cloud data, where a bottom surface of the flat column space is parallel to a horizontal plane.
In one embodiment shown, the radius of the flat cylinder space is determined based on the angular resolution of the lidar, and/or the degree of voxel sampling.
Because the laser radar sends out S laser points by 1 degree of rotation, the farther an object is from the laser radar, the wider the region of the S point distribution is, the larger the interval between the points is, the more reasonable the neighbor needs to be searched in a larger range, and therefore the radius R of the flat column space can be determined according to the angular resolution of the used laser radar and the voxel sampling degree. The radius of the flat cylinder space can also be determined by the distance between the laser radar and the detected object, for example, the effective radius R of the flat cylinder space can also be R = L + (L-20) × 0.02, where L is the distance between the laser radar and the detected object.
In one embodiment shown, the number N of point cloud data is compared with a threshold T for determining whether the target point cloud data is a small-size dynamic noise, and if the number N of point cloud data is smaller than the threshold T, the target point cloud data is determined as the small-size dynamic noise, and the small-size dynamic noise is filtered out.
In an embodiment shown, because the point cloud data of the current point cloud frame is superimposed in the set to be queried, if the proportion of the point cloud data corresponding to the current point cloud frame of the small-size dynamic object in the neighborhood is too high, misjudgment can be caused, so that the number M of the point cloud data belonging to the current frame in the number N of the point cloud data can be obtained, and whether the target point cloud data is the small-size dynamic noise point or not can be judged through the two.
For example, for a moving non-scattered plastic bag, the number of neighbors of which may exceed the value of T, N <2T and M >0.5N may be used as conditions for filtering out small-size dynamic noise.
Referring to fig. 2, fig. 2 is a block diagram of a small-sized dynamic noise filtering apparatus based on historical information according to an exemplary embodiment. The above-mentioned device includes:
a point cloud data acquiring module 210, configured to acquire a point cloud data set corresponding to a current point cloud frame;
a to-be-queried set constructing module 220, configured to construct a to-be-queried set based on point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame;
a point cloud data statistics module 230, configured to construct a neighborhood corresponding to target point cloud data in a point cloud data set corresponding to the current point cloud frame, and count, in the set to be queried, the number N of point cloud data in the neighborhood corresponding to the target point cloud data;
and a noise filtering module 240 for comparing the number N of the point cloud data with a threshold T for determining whether the target point cloud data is a small-size dynamic noise, and if the number N of the point cloud data is smaller than the threshold T, determining the target point cloud data as the small-size dynamic noise and filtering the small-size dynamic noise.
Optionally, the constructing a set to be queried based on the point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame includes:
and acquiring point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame, overlapping the point cloud data, and constructing a set to be queried.
Optionally, the obtaining of the point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame includes:
and acquiring point cloud data corresponding to the current point cloud frame, and selecting point cloud data corresponding to at least one frame of point cloud frame before the current point cloud frame at preset intervals.
Optionally, the constructing a neighborhood corresponding to target point cloud data in the point cloud data set corresponding to the current point cloud frame includes:
and determining target point cloud data in a point cloud data set corresponding to the current point cloud frame, and determining a flat column space taking the target point cloud data as a center as a neighborhood corresponding to the target point cloud data, wherein the bottom surface of the flat column space is parallel to a horizontal plane.
Optionally, the apparatus further comprises:
the radius of the flat cylinder space is determined based on the angular resolution of the lidar, and/or the degree of voxel sampling.
Referring to fig. 3, fig. 3 is a hardware structure diagram of an electronic device where a small-sized dynamic noise filtering apparatus based on historical information is located according to an exemplary embodiment. On the hardware level, the device includes a processor 302, an internal bus 304, a network interface 306, a memory 408, and a non-volatile memory 310, although it may include hardware required for other services. One or more embodiments of the present description may be implemented in software, such as by processor 302 reading a corresponding computer program from non-volatile storage 310 into memory 408 and then executing. Of course, besides software implementation, the one or more embodiments in this specification do not exclude other implementations, such as logic devices or combinations of software and hardware, and so on, that is, the execution subject of the following processing flow is not limited to each logic unit, and may also be hardware or logic devices.
For the device embodiments, since they substantially correspond to the method embodiments, reference may be made to the partial description of the method embodiments for relevant points. The above-described embodiments of the apparatus are only illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution in the specification. One of ordinary skill in the art can understand and implement without inventive effort.
The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. A typical implementation device is a computer, which may be in the form of a personal computer, laptop, cellular telephone, camera phone, smart phone, personal digital assistant, media player, navigation device, email messaging device, game console, tablet computer, wearable device, or a combination of any of these devices.
In a typical configuration, a computer includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include forms of volatile memory in a computer readable medium, random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.
Computer-readable media, including both permanent and non-permanent, removable and non-removable media, may implement the information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Disks (DVD) or other optical storage, magnetic cassettes, magnetic disk storage, quantum memory, graphene-based storage media or other magnetic storage devices, or any other non-transmission medium, that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising one of 8230; \8230;" 8230; "does not exclude the presence of additional like elements in a process, method, article, or apparatus that comprises that element.
The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
The terminology used in the description of the one or more embodiments is for the purpose of describing the particular embodiments only and is not intended to be limiting of the description of the one or more embodiments. As used in this specification and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.
It should be understood that although the terms first, second, third, etc. may be used in one or more embodiments of the present description to describe various information, such information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of one or more embodiments herein. The word "if" as used herein may be interpreted as "at" \8230; "or" when 8230; \8230; "or" in response to a determination ", depending on the context.
The above description is intended only to be exemplary of the one or more embodiments of the present disclosure, and should not be taken as limiting the one or more embodiments of the present disclosure, as any modifications, equivalents, improvements, etc. that come within the spirit and scope of the one or more embodiments of the present disclosure are intended to be included within the scope of the one or more embodiments of the present disclosure.
Claims (12)
1. A small-size dynamic noise filtering method based on historical information is characterized by comprising the following steps:
acquiring a point cloud data set corresponding to a current point cloud frame;
constructing a set to be queried based on point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame;
constructing a neighborhood corresponding to target point cloud data in a point cloud data set corresponding to the current point cloud frame, and counting the number N of the point cloud data in the neighborhood corresponding to the target point cloud data in the set to be inquired;
and comparing the number N of the point cloud data with a threshold T for determining whether the target point cloud data is small-size dynamic noise points, if the number N of the point cloud data is smaller than the threshold T, determining the target point cloud data as the small-size dynamic noise points, and filtering the small-size dynamic noise points.
2. The method of claim 1, wherein constructing the set to be queried based on the point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame comprises:
and acquiring point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame, and overlapping the point cloud data to construct a set to be queried.
3. The method of claim 2, wherein obtaining point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame comprises:
and acquiring point cloud data corresponding to the current point cloud frame, and selecting point cloud data corresponding to at least one frame of point cloud frame before the current point cloud frame at preset intervals.
4. The method of claim 1, wherein constructing a neighborhood corresponding to target point cloud data in the point cloud data set corresponding to the current point cloud frame comprises:
determining target point cloud data in a point cloud data set corresponding to the current point cloud frame, and determining a flat column space with the target point cloud data as a center as a neighborhood corresponding to the target point cloud data, wherein the bottom surface of the flat column space is parallel to a horizontal plane.
5. The method of claim 4, further comprising:
the radius of the flat cylinder space is determined based on the angular resolution of the lidar, and/or the degree of voxel sampling.
6. A small-size dynamic noise filtering device based on historical information, the device comprising:
the point cloud data acquisition module is used for acquiring a point cloud data set corresponding to the current point cloud frame;
the device comprises a to-be-queried set building module, a to-be-queried set building module and a query module, wherein the to-be-queried set building module is used for building a to-be-queried set based on point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame;
the point cloud data counting module is used for constructing a neighborhood corresponding to target point cloud data in a point cloud data set corresponding to the current point cloud frame, and counting the number N of the point cloud data in the neighborhood corresponding to the target point cloud data in the set to be inquired;
and the noise filtering module is used for comparing the number N of the point cloud data with a threshold T for determining whether the target point cloud data is small-size dynamic noise, and if the number N of the point cloud data is smaller than the threshold T, determining the target point cloud data as the small-size dynamic noise and filtering the small-size dynamic noise.
7. The apparatus of claim 6, wherein the constructing a set to be queried based on point cloud data corresponding to the current point cloud frame and at least one frame of point cloud frames before the current point cloud frame comprises:
and acquiring point cloud data corresponding to the current point cloud frame and at least one previous point cloud frame, and overlapping the point cloud data to construct a set to be queried.
8. The apparatus of claim 7, wherein obtaining point cloud data corresponding to the current point cloud frame and at least one frame of point cloud frame before the current point cloud frame comprises:
and acquiring point cloud data corresponding to the current point cloud frame, and selecting point cloud data corresponding to at least one frame of point cloud frame before the current point cloud frame at preset intervals.
9. The apparatus of claim 6, wherein the constructing a neighborhood corresponding to the target point cloud data in the point cloud data set corresponding to the current point cloud frame comprises:
and determining target point cloud data in a point cloud data set corresponding to the current point cloud frame, and determining a flat column space taking the target point cloud data as a center as a neighborhood corresponding to the target point cloud data, wherein the bottom surface of the flat column space is parallel to a horizontal plane.
10. The apparatus of claim 9, further comprising:
the radius of the flat cylinder space is determined based on the angular resolution of the lidar, and/or the degree of voxel sampling.
11. A machine-readable storage medium having stored thereon computer instructions which, when executed by a processor, implement the steps of the method of any one of claims 1-5.
12. An electronic device, comprising:
a processor;
a memory for storing processor-executable instructions;
wherein the processor implements the steps of the method of any one of claims 1-5 by executing the executable instructions.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210916049.6A CN115293980A (en) | 2022-08-01 | 2022-08-01 | Small-size dynamic noise filtering method and device based on historical information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210916049.6A CN115293980A (en) | 2022-08-01 | 2022-08-01 | Small-size dynamic noise filtering method and device based on historical information |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115293980A true CN115293980A (en) | 2022-11-04 |
Family
ID=83825478
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210916049.6A Pending CN115293980A (en) | 2022-08-01 | 2022-08-01 | Small-size dynamic noise filtering method and device based on historical information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115293980A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110109142A (en) * | 2019-04-04 | 2019-08-09 | 深圳市速腾聚创科技有限公司 | Point cloud filtering method, device, computer equipment and storage medium |
| CN111199555A (en) * | 2019-12-13 | 2020-05-26 | 意诺科技有限公司 | Millimeter wave radar target identification method |
| CN113192206A (en) * | 2021-04-28 | 2021-07-30 | 华南理工大学 | Three-dimensional model real-time reconstruction method and device based on target detection and background removal |
| CN113947552A (en) * | 2021-11-25 | 2022-01-18 | 中山大学 | Laser radar snow removal method and system integrating intensity and space-time geometric characteristics |
| WO2022133770A1 (en) * | 2020-12-23 | 2022-06-30 | 深圳元戎启行科技有限公司 | Method for generating point cloud normal vector, apparatus, computer device, and storage medium |
-
2022
- 2022-08-01 CN CN202210916049.6A patent/CN115293980A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110109142A (en) * | 2019-04-04 | 2019-08-09 | 深圳市速腾聚创科技有限公司 | Point cloud filtering method, device, computer equipment and storage medium |
| CN111199555A (en) * | 2019-12-13 | 2020-05-26 | 意诺科技有限公司 | Millimeter wave radar target identification method |
| WO2022133770A1 (en) * | 2020-12-23 | 2022-06-30 | 深圳元戎启行科技有限公司 | Method for generating point cloud normal vector, apparatus, computer device, and storage medium |
| CN113192206A (en) * | 2021-04-28 | 2021-07-30 | 华南理工大学 | Three-dimensional model real-time reconstruction method and device based on target detection and background removal |
| CN113947552A (en) * | 2021-11-25 | 2022-01-18 | 中山大学 | Laser radar snow removal method and system integrating intensity and space-time geometric characteristics |
Non-Patent Citations (1)
| Title |
|---|
| 林时雨 等: ""基于时间序列及邻域分析的管道点云障碍物检测"", 《中国知网》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107817496B (en) | Object detection system suitable for automatic vehicle | |
| CN112526513B (en) | Millimeter wave radar environment map construction method and device based on clustering algorithm | |
| CN113487759B (en) | Parking patrol method and device, mobile patrol equipment and patrol system | |
| US20240005674A1 (en) | Road edge recognition based on laser point cloud | |
| CN112513679A (en) | Target identification method and device | |
| CN115265519A (en) | Online point cloud map construction method and device | |
| CN115205803A (en) | Automatic driving environment sensing method, medium and vehicle | |
| CN112612975A (en) | Method, device, equipment and storage medium for identifying type of commonly-transported goods of vehicle | |
| CN109102026A (en) | A kind of vehicle image detection method, apparatus and system | |
| CN114882701A (en) | Parking space detection method and device, electronic equipment and machine readable storage medium | |
| CN111483464A (en) | Dynamic automatic driving lane changing method, equipment and storage medium based on road side unit | |
| CN114241448A (en) | Method and device for obtaining heading angle of obstacle, electronic equipment and vehicle | |
| CN113189610A (en) | Map-enhanced autonomous driving multi-target tracking method and related equipment | |
| CN115293980A (en) | Small-size dynamic noise filtering method and device based on historical information | |
| CN114631124A (en) | Three-dimensional point cloud segmentation method and device and movable platform | |
| US20230025981A1 (en) | Method and apparatus with grid map generation | |
| CN116385999A (en) | Parking space identification method, device and equipment | |
| CN113203424B (en) | Multi-sensor data fusion method and device and related equipment | |
| CN113465615B (en) | Lane line generation method and related device | |
| CN115656982A (en) | Water surface clutter removal method and device, computer equipment and storage medium | |
| CN112835063B (en) | Method, device, equipment and storage medium for determining dynamic and static properties of object | |
| CN112700387A (en) | Laser data processing method, device and equipment and storage medium | |
| CN111488771B (en) | OCR hooking method, device and equipment | |
| CN114509774A (en) | Positioning method, positioning system, vehicle, and computer-readable storage medium | |
| CN111723797B (en) | Method and system for determining bounding box of three-dimensional target |
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 |