CN117789198B - Method for realizing point cloud degradation detection based on 4D millimeter wave imaging radar - Google Patents
Method for realizing point cloud degradation detection based on 4D millimeter wave imaging radar Download PDFInfo
- Publication number
- CN117789198B CN117789198B CN202410221574.5A CN202410221574A CN117789198B CN 117789198 B CN117789198 B CN 117789198B CN 202410221574 A CN202410221574 A CN 202410221574A CN 117789198 B CN117789198 B CN 117789198B
- Authority
- CN
- China
- Prior art keywords
- point cloud
- point
- shape
- millimeter wave
- cloud
- 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.)
- Active
Links
- 230000015556 catabolic process Effects 0.000 title claims abstract description 55
- 238000006731 degradation reaction Methods 0.000 title claims abstract description 55
- 238000003384 imaging method Methods 0.000 title claims abstract description 34
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000001514 detection method Methods 0.000 title claims abstract description 32
- 230000004927 fusion Effects 0.000 claims abstract description 30
- 230000000593 degrading effect Effects 0.000 claims abstract description 11
- 238000012545 processing Methods 0.000 claims abstract description 10
- 239000013598 vector Substances 0.000 claims description 41
- 238000000605 extraction Methods 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 5
- 238000012217 deletion Methods 0.000 description 3
- 230000037430 deletion Effects 0.000 description 3
- 238000003672 processing method Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
Landscapes
- Radar Systems Or Details Thereof (AREA)
- Image Processing (AREA)
Abstract
The invention relates to a point cloud degradation detection method for realizing point cloud information entropy fusion point cloud shape based on a 4D millimeter wave imaging radar, which comprises the following steps: inputting 4D millimeter wave Lei Dadian cloud, performing point cloud pretreatment on the cloud, and removing outlier noise points; calculating the information entropy of the point cloud, judging whether the currently acquired point cloud is degraded, if so, performing further fitting processing on the shape of the current point cloud, and judging whether the shape tends to be regular, if so, degrading the point cloud, otherwise, not degrading the current point cloud. The invention also relates to a corresponding device, processor and computer readable storage medium thereof. The method, the device, the processor and the computer readable storage medium for detecting the point cloud degradation based on the 4D millimeter wave imaging radar realize the point cloud information entropy fusion point cloud shape, calculate the information entropy directly from the self space distribution condition of the point cloud, judge whether the point cloud is degraded according to the information entropy, and further improve the accuracy of degradation detection.
Description
Technical Field
The invention relates to the field of 4D millimeter wave imaging radars, in particular to the technical field of point cloud degradation detection, and specifically relates to a point cloud degradation detection method, a device, a processor and a computer readable storage medium for realizing point cloud information entropy fusion point cloud shape based on a 4D millimeter wave imaging radar.
Background
In the SLAM technical field, point cloud is a very common input, and point cloud matching is an important module for SLAM front-end processing. The point cloud is typically generated by a lidar or a 4D millimeter wave radar. For some environments similar, in a scene with weak textures and no textures, the distribution conditions of point clouds tend to be consistent, and basically, different points cannot be proposed for feature matching, so that an incorrect matching result is caused, and the pose precision of an automatic driving vehicle estimated by the SLAM front end is very low, so that the reconstruction of environment information cannot be completed. The point cloud in this scenario is the degenerate point cloud.
At present, aiming at SLAM mapping under similar scenes, whether the current point cloud is degenerated is generally judged from the result of point cloud matching, and the degeneracy direction of a certain dimension is calculated, for example, based on the score of point cloud matching, the ratio of the maximum eigenvalue to the minimum eigenvalue of a Hessian matrix in the matching solving process and the average error value after matching. Although the method can detect the degenerated point cloud to a certain extent, the method is mainly used for the laser radar point cloud, has poor applicability to the 4D millimeter wave Lei Dadian cloud, is not robust because the 4D millimeter wave radar point cloud is very sparse and has larger noise, and can not detect the non-degenerated point cloud or detect the normal point cloud as the degenerated point cloud because different threshold parameters are required to be set under different degenerated scenes.
Therefore, it is necessary to propose a scheme capable of simultaneously overcoming the defect of poor robustness of the existing laser radar point cloud degradation detection method and sparse 4D millimeter wave imaging radar point cloud, so as to improve the applicability of the 4D millimeter wave radar point cloud degradation detection.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides a point cloud degradation detection method, a device, a processor and a computer readable storage medium thereof for realizing point cloud information entropy fusion point cloud shape based on a 4D millimeter wave imaging radar.
In order to achieve the above object, the method, the device, the processor and the computer readable storage medium for detecting the degradation of the point cloud based on the 4D millimeter wave imaging radar to realize the entropy fusion of the point cloud information and the shape of the point cloud according to the present invention are as follows:
the method for detecting the point cloud degradation based on the 4D millimeter wave imaging radar to realize the point cloud information entropy fusion point cloud shape is mainly characterized by comprising the following steps:
(1) Inputting 4D millimeter wave Lei Dadian cloud, performing point cloud pretreatment on the cloud, and removing outlier noise points;
(2) Calculating point cloud information entropy, judging whether the point cloud acquired at present is degenerated or not, and if so, entering a step (3);
(3) And performing further fitting processing on the current point cloud shape, judging whether the shape tends to be regular, if so, degrading the point cloud, otherwise, not degrading the current point cloud.
Preferably, the step (1) specifically includes the following steps:
(1.1) inputting the 4D millimeter wave Lei Dadian cloud at the current time N is the number of input point clouds;
(1.2) cloud of 4D millimeter waves Lei Dadian Each point/>,/>Searching N adjacent points near the radius R, and describing/>, if N meets the following conditionIs a noise point:
;
Wherein alpha is a set threshold value of the number of the neighboring points;
(1.3) if Point Is the noise point, will/>From 4D millimeter wave Lei Dadian cloud/>The deletion in the cloud is carried out to obtain the preprocessed point cloud/>M is the number of the point clouds after pretreatment.
Preferably, the step (2) specifically includes the following steps:
(2.1) calculating a Point cloud center Point in the following manner :
;
Wherein m is point cloudIn (1)/(number of point clouds)For the dot/>Coordinates of (c);
(2.2) calculating the Point cloud covariance as follows :
;
(2.3) Calculating the Point cloud information entropy in the following manner:
;
Wherein e is natural logarithm;
(2.4) determining the point cloud in the following manner Whether degradation occurs:
;
Wherein, Entropy threshold for point cloud degradation information;
(2.5) if a point cloud And (3) if the degradation occurs, entering a step, otherwise, not degrading the current point cloud, and exiting degradation detection.
Preferably, the step (3) specifically includes the following steps:
(3.1) performing point cloud region of interest (ROI) extraction processing: at the point cloud Selecting x, y, z direction/>Points within the range, get the point cloud/>T is the number of point clouds after extracting the ROI, i.e./>;
(3.2) Point cloudExtracting the outline and obtaining boundary points of the point cloud outline;
(3.3) detecting gradient change of the boundary point cloud based on the obtained point cloud outline boundary points, and finding out the vertex of the current point cloud shape according to the gradient change;
And (3.4) connecting the obtained t vertexes in pairs in sequence to form t-1 vectors, judging whether the t-1 vectors are on the same plane, if not, the current point cloud is not in a regular shape, and if not, the point cloud is not degraded, otherwise, the point cloud is degraded.
Preferably, the step (3.2) specifically includes the following steps:
(3.2.1) computing Point cloud Each point/>Normal/>:
(3.2.2) Calculation pointsNearby N neighbor points/>Normal vector/>;
(3.2.3) Judging whether the included angle theta meets the following conditions:
;
(3.2.4) obtaining boundary points of the point cloud contour through the contour extraction The number of boundary points.
Preferably, the step (3.3) specifically includes the following steps:
(3.3.1) calculating a contour point gradient in the following manner :
;
;
Wherein,Is boundary point/>Coordinate point of/>Is boundary point/>And/>A distance therebetween;
(3.3.2) gradient variation difference was calculated as follows :
;
Wherein,The previous contour point gradient;
(3.3.3) computing Point cloud shape vertices The boundary point/>, is considered to be present if the following condition is satisfiedIs the vertex/>, of the point cloud shape:
;
Wherein,Is a gradient change threshold.
Preferably, the step (3.4) specifically includes the following steps:
(3.4.1) acquiring a point cloud shape vector: vertex point And/>Composition vector/>Vertex/>And/>Composition vector/>Obtain the shape vector/>, of the vertex;
(3.4.2) Calculating vectorsAnd shape vector/>Plane equation/>, determined between other vectors in (a)Wherein/>Parameters of plane equation are determined;
(3.4.3) determining whether all plane equations satisfy the following condition:
;
wherein t represents the number of vertices of the point cloud shape.
(3.4.4) If all planes meet the above condition, indicating that all vectors are in the same plane, the shape of the point cloud tends to be regular and the point cloud is degraded currently, otherwise, the point cloud is not degraded.
The point cloud degradation detection device for realizing the point cloud information entropy fusion point cloud shape based on the 4D millimeter wave imaging radar comprises:
a processor configured to execute computer-executable instructions;
And the memory stores one or more computer executable instructions which, when executed by the processor, realize the step of the point cloud degradation detection method for realizing the point cloud information entropy fusion point cloud shape based on the 4D millimeter wave imaging radar.
The point cloud degradation detection processor based on the 4D millimeter wave imaging radar realizes the point cloud information entropy fusion point cloud shape, wherein the processor is configured to execute computer executable instructions, and when the computer executable instructions are executed by the processor, the step of the point cloud degradation detection method based on the 4D millimeter wave imaging radar realizes the point cloud information entropy fusion point cloud shape is realized.
The computer readable storage medium has a computer program stored thereon, the computer program being executable by a processor to implement the above-described method for detecting the degradation of a point cloud based on the 4D millimeter wave imaging radar to realize the entropy fusion point cloud shape of the point cloud information.
The point cloud degradation detection method, the device, the processor and the computer readable storage medium thereof for realizing the point cloud information entropy fusion point cloud shape based on the 4D millimeter wave imaging radar solve the problems of limitation and false detection that laser radar point cloud degradation detection can only be judged by a front-back frame matching result, and judge whether the point cloud really degrades or not directly from the spatial position distribution inside the point cloud, the relativity between the points and the shape rule of the point cloud.
Drawings
Fig. 1 is a flowchart of a point cloud degradation detection method for realizing a point cloud information entropy fusion point cloud shape based on a 4D millimeter wave imaging radar.
Detailed Description
In order to more clearly describe the technical contents of the present invention, a further description will be made below in connection with specific embodiments.
Before describing in detail embodiments that are in accordance with the present invention, it should be observed 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.
Before explaining the technical scheme in detail, the following explanation is made on some technical features:
SLAM: building and positioning in real time;
pose: the position and posture comprises position information, x, y, z, posture information and roll, pitch, yaw;
ROI: region of Interesting, a region of interest.
Referring to fig. 1, the method for detecting the point cloud degradation based on the 4D millimeter wave imaging radar to realize the point cloud information entropy fusion point cloud shape includes the following steps:
(1) Inputting 4D millimeter wave Lei Dadian cloud, performing point cloud pretreatment on the cloud, and removing outlier noise points;
(2) Calculating point cloud information entropy, judging whether the point cloud acquired at present is degenerated or not, and if so, entering a step (3);
(3) And performing further fitting processing on the current point cloud shape, judging whether the shape tends to be regular, if so, degrading the point cloud, otherwise, not degrading the current point cloud.
As a preferred embodiment of the present invention, the step (1) specifically includes the steps of:
(1.1) inputting the 4D millimeter wave Lei Dadian cloud at the current time N is the number of input point clouds;
(1.2) cloud of 4D millimeter waves Lei Dadian Each point/>,/>N neighbor points are searched near their radius R, explaining/>, if the following condition is satisfiedIs a noise point:
;
Wherein alpha is a set threshold value of the number of the neighboring points;
(1.3) if Point Is the noise point, will/>From 4D millimeter wave Lei Dadian cloud/>The deletion in the cloud is carried out to obtain the preprocessed point cloud/>M is the number of the point clouds after pretreatment.
As a preferred embodiment of the present invention, the step (2) specifically includes the following steps:
(2.1) calculating a Point cloud center Point in the following manner :
;
Wherein m is point cloudIn (1)/(number of point clouds)For the dot/>Coordinates of (c);
(2.2) calculating the Point cloud covariance as follows :
;
(2.3) Calculating the Point cloud information entropy in the following manner:
;
Wherein e is natural logarithm;
(2.4) determining the point cloud in the following manner Whether degradation occurs:
;
Wherein, Entropy threshold for point cloud degradation information;
(2.5) if a point cloud And (3) if the degradation occurs, entering a step, otherwise, not degrading the current point cloud, and exiting degradation detection.
As a preferred embodiment of the present invention, the step (3) specifically includes the following steps:
(3.1) performing point cloud region of interest (ROI) extraction processing: at the point cloud Selecting x, y, z direction/>Points within the range, get the point cloud/>T is the number of point clouds after extracting the ROI, i.e
;
(3.2) Point cloudExtracting the outline and obtaining boundary points of the point cloud outline;
(3.3) detecting gradient change of the boundary point cloud based on the obtained point cloud outline boundary points, and finding out the vertex of the current point cloud shape according to the gradient change;
And (3.4) connecting the obtained t vertexes in pairs in sequence to form t-1 vectors, judging whether the t-1 vectors are on the same plane, if not, the current point cloud is not in a regular shape, and if not, the point cloud is not degraded, otherwise, the point cloud is degraded.
As a preferred embodiment of the present invention, the step (3.2) specifically includes the steps of:
(3.2.1) computing Point cloud Each point/>Normal/>:
(3.2.2) Calculation pointsNearby N neighbor points/>Normal vector/>;
(3.2.3) Judging whether the included angle theta meets the following conditions:
;
(3.2.4) obtaining boundary points of the point cloud contour through the contour extraction The number of boundary points.
As a preferred embodiment of the present invention, the step (3.3) specifically includes the steps of:
(3.3.1) calculating a contour point gradient in the following manner :
;
;
Wherein,Is boundary point/>Coordinate point of/>Is boundary point/>And/>A distance therebetween;
(3.3.2) gradient variation difference was calculated as follows :
;
Wherein,The previous contour point gradient;
(3.3.3) computing Point cloud shape vertices The boundary point/>, is considered to be present if the following condition is satisfiedIs the vertex/>, of the point cloud shape:
;
Wherein,Is a gradient change threshold.
As a preferred embodiment of the present invention, the step (3.4) specifically includes the steps of:
(3.4.1) acquiring a point cloud shape vector: vertex point And/>Composition vector/>Vertex/>And/>Composition vector/>Obtain the shape vector/>, of the vertex;
(3.4.2) Calculating vectorsAnd shape vector/>Plane equation/>, determined between other vectors in (a)Wherein/>Parameters of plane equation are determined;
(3.4.3) determining whether all plane equations satisfy the following condition:
;
wherein t represents the number of vertices of the point cloud shape.
(3.4.4) If all planes meet the above condition, indicating that all vectors are in the same plane, the shape of the point cloud tends to be regular and the point cloud is degraded currently, otherwise, the point cloud is not degraded.
Referring to fig. 1, the method for detecting the point cloud degradation based on the 4D millimeter wave imaging radar to realize the point cloud information entropy fusion point cloud shape specifically includes three processing modules of point cloud preprocessing, point cloud information entropy and point cloud shape, and the specific processing flow among the modules is as follows:
(1) Point cloud preprocessing
Because the 4D millimeter wave Lei Dadian cloud has high sparsity and noise error, the Lei Dadian cloud is subjected to filtering and denoising pretreatment before the point cloud degradation detection, and some outlier noise points are removed. The specific method comprises the following steps:
inputting 4D millimeter wave Lei Dadian cloud at current moment N is the number of input point clouds;
For a pair of Each point/>,/>Searching N adjacent points near the radius R;
Description is made if N satisfies the following condition Is a noise point:
;
Alpha is the threshold value of the number of the neighbor points, if Is the noise point, will/>From/>The deletion in the cloud is carried out to obtain the preprocessed point cloud/>M is the number of the point clouds after pretreatment.
(2) Point cloud information entropy
The point cloud information entropy is used for explaining the distribution correlation of the internal space of the point cloud. The method comprises the steps of calculating a point cloud center, calculating a point cloud covariance, calculating a point cloud information entropy, and judging whether 4 processes are degraded or not:
① Computing a point cloud center point The following formula:
;
m is point cloud Point cloud number,/>For the dot/>Is defined by the coordinates of (a).
② Computing point cloud covariance:
;
③ Calculating point cloud information entropy:
;
E is the natural logarithm.
④ Point cloud degradation determination
If it isThe following conditions are satisfied to describe the point cloud/>And if not, entering the point cloud shape module to further judge whether the point cloud really degenerates.
;
Entropy threshold for point cloud degradation information.
(3) Point cloud shape
Point cloud information entropyThe point cloud is larger in degradation probability, and whether the point cloud really degrades is further judged by fitting out whether the shape of the point cloud is regular or not. The method comprises 4 processes of point cloud ROI extraction, point cloud contour extraction, point cloud shape fitting and point cloud shape trend rule judgment.
(3.1) Point cloud ROI extraction
To efficiently fit the approximate shape of the point cloud, the point cloud is formed byIs selected from x, y and z directions based on (a)Points within the range, get the point cloud/>T is the number of point clouds after extracting the ROI, i.e
;
(3.2) Point cloud contour extraction
At the point cloudExtract the rough outline/>S is the number of contour points, and the specific flow is as follows:
a) Calculate each point Normal/>The normal is calculated as the most basic processing method in space geometry;
b) Calculation of Nearby N neighbor points/>Normal vectorThe normal vector is calculated as a space geometric basic processing method;
c) If the angle theta satisfies the following condition, Is a contour point
;
Obtaining boundary points of the point cloud contour through contour extractionThe number of boundary points.
(3.3) Point cloud shape fitting
Utilizing the contour points extracted in (3.2) to fit the shape of the point cloud, namely detecting gradient change of the contour point cloud, and finding out the vertex of the shape according to the gradient changeThe specific process flow is as follows:
a) Calculating contour points Is a gradient of (2);
;
;
For/> Coordinate point of/>For/>And/>Distance between them.
B) Calculating gradient change difference:
;
C) Computing shape verticesThe following condition is considered/>Is the shape vertex/>:
;
Is a gradient change threshold.
(3.4) Point cloud shape trending rule judgment
In (3.3), the vertices of the point cloud shape are obtainedAnd (3) connecting the t vertexes in pairs in sequence to form t-1 vectors, checking whether the t-1 vectors are on the same plane, if not, carrying out regular shape, and if not, carrying out degradation on the point cloud, otherwise, carrying out degradation. The specific process is as follows:
a) Vertex point And/>Composition vector/>Vertex/>And/>Composition vectorObtain the shape vector/>, of the vertex;
B) Calculation ofAnd/>Plane equation/>, determined between other vectors in (a),/>Is a determined parameter of a plane equation, and the plane equation is a basic processing method in geometry.
C) If all plane equations are satisfied:
;
The description is the same plane, the shape of the point cloud tends to be regular, the point cloud is degraded, and otherwise, the point cloud is not degraded.
The point cloud degradation detection device for realizing the point cloud information entropy fusion point cloud shape based on the 4D millimeter wave imaging radar comprises:
a processor configured to execute computer-executable instructions;
And the memory stores one or more computer executable instructions which, when executed by the processor, realize the step of the point cloud degradation detection method for realizing the point cloud information entropy fusion point cloud shape based on the 4D millimeter wave imaging radar.
The point cloud degradation detection processor based on the 4D millimeter wave imaging radar realizes the point cloud information entropy fusion point cloud shape, wherein the processor is configured to execute computer executable instructions, and when the computer executable instructions are executed by the processor, the step of the point cloud degradation detection method based on the 4D millimeter wave imaging radar realizes the point cloud information entropy fusion point cloud shape is realized.
The computer readable storage medium has a computer program stored thereon, the computer program being executable by a processor to implement the above-described method for detecting the degradation of a point cloud based on the 4D millimeter wave imaging radar to realize the entropy fusion point cloud shape of the point cloud information.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and further implementations are included within the scope of the preferred embodiment of the present invention in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present invention.
It is to be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution device.
Those of ordinary skill in the art will appreciate that all or a portion of the steps carried out in the method of the above-described embodiments may be implemented by a program to instruct related hardware, and the program may be stored in a computer readable storage medium, where the program when executed includes one or a combination of the steps of the method embodiments.
The above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, or the like.
In the description of the present specification, reference to the terms "one embodiment," "some embodiments," "examples," "specific examples," or "embodiments," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
While embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and not to be construed as limiting the invention, and that variations, modifications, alternatives and variations may be made to the above embodiments by one of ordinary skill in the art within the scope of the invention.
The point cloud degradation detection method, the device, the processor and the computer readable storage medium thereof for realizing the point cloud information entropy fusion point cloud shape based on the 4D millimeter wave imaging radar solve the problems of limitation and false detection that laser radar point cloud degradation detection can only be judged by a front-back frame matching result, and judge whether the point cloud really degrades or not directly from the spatial position distribution inside the point cloud, the relativity between the points and the shape rule of the point cloud.
In this specification, the invention has been described with reference to specific embodiments thereof. It will be apparent that various modifications and variations can be made without departing from the spirit and scope of the invention. The specification and drawings are, accordingly, to be regarded in an illustrative rather than a restrictive sense.
Claims (7)
1. A point cloud degradation detection method for realizing point cloud information entropy fusion point cloud shape based on a 4D millimeter wave imaging radar is characterized by comprising the following steps:
(1) Inputting 4D millimeter wave Lei Dadian cloud, performing point cloud pretreatment on the cloud, and removing outlier noise points;
(2) Calculating point cloud information entropy, judging whether the point cloud acquired at present is degenerated or not, and if so, entering a step (3);
(3) Performing further fitting processing on the current point cloud shape, judging whether the point cloud shape tends to be regular, if so, degrading the point cloud, otherwise, not degrading the current point cloud;
the step (1) specifically comprises the following steps:
(1.1) inputting the 4D millimeter wave Lei Dadian cloud at the current time N+1 is the number of input point clouds;
(1.2) cloud of 4D millimeter waves Lei Dadian Each point/>,/>Searching N adjacent points near the radius R, and describing the point/>, if N meets the following conditionIs a noise point:
;
Wherein alpha is a set threshold value of the number of the neighboring points;
(1.3) if Point Is the noise point, then the point/>From 4D millimeter wave Lei Dadian cloud/>Obtaining the first point cloud/>, after pretreatmentM+1 is the number of point clouds after pretreatment;
The step (2) specifically comprises the following steps:
(2.1) calculating a Point cloud center Point in the following manner :
;
Wherein m+1 is the first point cloudIn (1)/(number of point clouds)For the dot/>Coordinates of (c);
(2.2) calculating the Point cloud covariance as follows :
;
(2.3) Calculating the Point cloud information entropy in the following manner:
;
Wherein e is natural logarithm;
(2.4) determining the first point cloud in the following manner Whether degradation occurs:
;
Wherein, Entropy threshold for point cloud degradation information;
(2.5) if the first point cloud If so, entering a step (3), otherwise, not degrading the current point cloud, and exiting degradation detection;
The step (3) specifically comprises the following steps:
(3.1) performing point cloud region of interest (ROI) extraction processing: at a first point cloud Selecting x, y, z direction/>Points within the range, a second point cloud/>, is obtainedT+1 is the number of point clouds after extracting the ROI, i.e
;
(3.2) For a second Point cloudExtracting the outline and obtaining boundary points of the point cloud outline;
(3.3) detecting gradient change of the boundary point cloud based on the obtained point cloud outline boundary points, and finding out the vertex of the current point cloud shape according to the gradient change;
And (3.4) connecting the obtained k+1 vertexes in pairs to form k vectors in sequence, judging whether the k vectors are on the same plane, if not, the current point cloud is not in a regular shape, and if not, the point cloud is not degraded, otherwise, the point cloud is degraded.
2. The method for detecting the point cloud degradation based on the 4D millimeter wave imaging radar to realize the point cloud information entropy fusion point cloud shape according to claim 1, wherein the step (3.2) specifically comprises the following steps:
(3.2.1) computing a second Point cloud Each point/>Normal/>;
(3.2.2) Calculation pointsNearby N neighbor points/>,/>Normal vector;
(3.2.3) Judging whether the included angle theta meets the following conditions:
;
(3.2.4) obtaining boundary points of the point cloud contour through the contour extraction S is the number of boundary points.
3. The method for detecting the point cloud degradation based on the 4D millimeter wave imaging radar to realize the entropy fusion of the point cloud information and the point cloud shape according to claim 2, wherein the step (3.3) specifically comprises the following steps:
(3.3.1) calculating a contour point gradient in the following manner :
;
;
Wherein,Is boundary point/>Coordinate point of/>Is boundary point/>And/>A distance therebetween;
(3.3.2) gradient variation difference was calculated as follows :
;
Wherein,The previous contour point gradient;
(3.3.3) computing Point cloud shape vertices The boundary point/>, is considered to be present if the following condition is satisfiedIs the vertex/>, of the point cloud shapeWherein/>:
;
Wherein,Is a gradient change threshold.
4. The method for detecting the point cloud degradation based on the 4D millimeter wave imaging radar to realize the entropy fusion of the point cloud information and the point cloud shape according to claim 3, wherein the step (3.4) specifically comprises the following steps:
(3.4.1) acquiring a point cloud shape vector: vertex point And/>Composition vector/>Vertex/>And/>Composition vector/>,/>Obtain the shape vector/>, of the vertex;
(3.4.2) Calculating vectorsAnd shape vector/>Plane equation/>, determined between other vectors in (a)Wherein/>Parameters of plane equation are determined;
(3.4.3) determining whether all plane equations satisfy the following condition:
;
(3.4.4) if all planes meet the above condition, indicating that all vectors are in the same plane, the shape of the point cloud tends to be regular and the point cloud is degraded currently, otherwise, the point cloud is not degraded.
5. A point cloud degradation detection device for realizing point cloud information entropy fusion point cloud shape based on a 4D millimeter wave imaging radar is characterized by comprising:
a processor configured to execute computer-executable instructions;
A memory storing one or more computer-executable instructions which, when executed by the processor, implement the steps of the method for detecting point cloud degradation based on 4D millimeter wave imaging radar to achieve a point cloud information entropy fusion point cloud shape of any one of claims 1 to 4.
6. A point cloud degradation detection processor for realizing a point cloud information entropy fusion point cloud shape based on a 4D millimeter wave imaging radar, wherein the processor is configured to execute computer executable instructions, which when executed by the processor, realize the steps of the point cloud degradation detection method for realizing the point cloud information entropy fusion point cloud shape based on the 4D millimeter wave imaging radar according to any one of claims 1 to 4.
7. A computer-readable storage medium, having stored thereon a computer program executable by a processor to implement the steps of the method for detecting point cloud degradation based on 4D millimeter wave imaging radar to realize point cloud information entropy fusion point cloud shape according to any one of claims 1 to 4.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410221574.5A CN117789198B (en) | 2024-02-28 | 2024-02-28 | Method for realizing point cloud degradation detection based on 4D millimeter wave imaging radar |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410221574.5A CN117789198B (en) | 2024-02-28 | 2024-02-28 | Method for realizing point cloud degradation detection based on 4D millimeter wave imaging radar |
Publications (2)
Publication Number | Publication Date |
---|---|
CN117789198A CN117789198A (en) | 2024-03-29 |
CN117789198B true CN117789198B (en) | 2024-05-14 |
Family
ID=90400398
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202410221574.5A Active CN117789198B (en) | 2024-02-28 | 2024-02-28 | Method for realizing point cloud degradation detection based on 4D millimeter wave imaging radar |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117789198B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN118068286B (en) * | 2024-04-16 | 2024-07-05 | 上海几何伙伴智能驾驶有限公司 | Method for realizing degradation environment detection aiming at influence of 4D millimeter wave Lei Dadian cloud positioning |
Citations (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110246112A (en) * | 2019-01-21 | 2019-09-17 | 厦门大学 | Three-dimensional point cloud quality evaluating method in the room laser scanning SLAM based on deep learning |
CN111507968A (en) * | 2020-04-20 | 2020-08-07 | 北京英迈琪科技有限公司 | Image fusion quality detection method and device |
WO2020235979A1 (en) * | 2019-05-23 | 2020-11-26 | 삼성전자 주식회사 | Method and device for rendering point cloud-based data |
CN112560972A (en) * | 2020-12-21 | 2021-03-26 | 北京航空航天大学 | Target detection method based on millimeter wave radar prior positioning and visual feature fusion |
CN112581457A (en) * | 2020-12-23 | 2021-03-30 | 武汉理工大学 | Pipeline inner surface detection method and device based on three-dimensional point cloud |
CN114581389A (en) * | 2022-02-24 | 2022-06-03 | 华侨大学 | Point cloud quality analysis method based on three-dimensional edge similarity characteristics |
CN115113170A (en) * | 2022-07-21 | 2022-09-27 | 电子科技大学 | Laser radar edge feature prediction method based on indoor feature degradation environment |
CN115236616A (en) * | 2022-07-28 | 2022-10-25 | 中国人民解放军陆军工程大学 | Radar system quality determination method and system based on combined weighted gray cloud model |
WO2023093824A1 (en) * | 2021-11-26 | 2023-06-01 | 中兴通讯股份有限公司 | Point cloud quality evaluation method, and device and storage medium |
CN116205886A (en) * | 2023-03-07 | 2023-06-02 | 重庆邮电大学 | Point cloud quality assessment method based on relative entropy |
CN116299535A (en) * | 2023-03-06 | 2023-06-23 | 上海交通大学 | Laser radar SLAM degradation detection method and system based on geometric information |
CN116309144A (en) * | 2023-03-08 | 2023-06-23 | 广州大学 | Point cloud shape completion method based on diffusion probability model |
CN116359873A (en) * | 2023-03-29 | 2023-06-30 | 上海几何伙伴智能驾驶有限公司 | Method, device, processor and storage medium for realizing SLAM processing of vehicle-end 4D millimeter wave radar by combining fisheye camera |
WO2023197601A1 (en) * | 2022-04-14 | 2023-10-19 | 北京大学 | Gradient field-based point cloud repair method |
WO2023232165A1 (en) * | 2022-06-01 | 2023-12-07 | 湖南大学无锡智能控制研究院 | Multi-radar data fusion obstacle detection method, and system |
CN117761722A (en) * | 2023-12-25 | 2024-03-26 | 广州高新兴机器人有限公司 | Laser radar SLAM degradation detection method, system, electronic equipment and storage medium |
-
2024
- 2024-02-28 CN CN202410221574.5A patent/CN117789198B/en active Active
Patent Citations (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110246112A (en) * | 2019-01-21 | 2019-09-17 | 厦门大学 | Three-dimensional point cloud quality evaluating method in the room laser scanning SLAM based on deep learning |
WO2020235979A1 (en) * | 2019-05-23 | 2020-11-26 | 삼성전자 주식회사 | Method and device for rendering point cloud-based data |
CN111507968A (en) * | 2020-04-20 | 2020-08-07 | 北京英迈琪科技有限公司 | Image fusion quality detection method and device |
CN112560972A (en) * | 2020-12-21 | 2021-03-26 | 北京航空航天大学 | Target detection method based on millimeter wave radar prior positioning and visual feature fusion |
CN112581457A (en) * | 2020-12-23 | 2021-03-30 | 武汉理工大学 | Pipeline inner surface detection method and device based on three-dimensional point cloud |
WO2023093824A1 (en) * | 2021-11-26 | 2023-06-01 | 中兴通讯股份有限公司 | Point cloud quality evaluation method, and device and storage medium |
CN114581389A (en) * | 2022-02-24 | 2022-06-03 | 华侨大学 | Point cloud quality analysis method based on three-dimensional edge similarity characteristics |
WO2023197601A1 (en) * | 2022-04-14 | 2023-10-19 | 北京大学 | Gradient field-based point cloud repair method |
WO2023232165A1 (en) * | 2022-06-01 | 2023-12-07 | 湖南大学无锡智能控制研究院 | Multi-radar data fusion obstacle detection method, and system |
CN115113170A (en) * | 2022-07-21 | 2022-09-27 | 电子科技大学 | Laser radar edge feature prediction method based on indoor feature degradation environment |
CN115236616A (en) * | 2022-07-28 | 2022-10-25 | 中国人民解放军陆军工程大学 | Radar system quality determination method and system based on combined weighted gray cloud model |
CN116299535A (en) * | 2023-03-06 | 2023-06-23 | 上海交通大学 | Laser radar SLAM degradation detection method and system based on geometric information |
CN116205886A (en) * | 2023-03-07 | 2023-06-02 | 重庆邮电大学 | Point cloud quality assessment method based on relative entropy |
CN116309144A (en) * | 2023-03-08 | 2023-06-23 | 广州大学 | Point cloud shape completion method based on diffusion probability model |
CN116359873A (en) * | 2023-03-29 | 2023-06-30 | 上海几何伙伴智能驾驶有限公司 | Method, device, processor and storage medium for realizing SLAM processing of vehicle-end 4D millimeter wave radar by combining fisheye camera |
CN117761722A (en) * | 2023-12-25 | 2024-03-26 | 广州高新兴机器人有限公司 | Laser radar SLAM degradation detection method, system, electronic equipment and storage medium |
Non-Patent Citations (1)
Title |
---|
降雨条件下车载激光雷达感知局限性;邢星宇 等;同济大学学报(自然科学版);20230531;第第51卷卷(第第5期期);第785-793页 * |
Also Published As
Publication number | Publication date |
---|---|
CN117789198A (en) | 2024-03-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20210056708A1 (en) | Target detection and training for target detection network | |
CN110717489B (en) | Method, device and storage medium for identifying text region of OSD (on Screen display) | |
CN117789198B (en) | Method for realizing point cloud degradation detection based on 4D millimeter wave imaging radar | |
CN111681197B (en) | Remote sensing image unsupervised change detection method based on Siamese network structure | |
US8625903B2 (en) | Pose estimation | |
CN112669318B (en) | Surface defect detection method, device, equipment and storage medium | |
CN111712731A (en) | Target detection method and system and movable platform | |
EP2249311A1 (en) | Systems and methods for extracting planar features, matching the planar features, and estimating motion from the planar features | |
CN106485651B (en) | The image matching method of fast robust Scale invariant | |
CN108225319B (en) | Monocular vision rapid relative pose estimation system and method based on target characteristics | |
US10339396B2 (en) | Vehicle accessibility determination device | |
KR20210026412A (en) | Method for Real-time Object Detection Based on Lidar Sensor and Camera Using CNN | |
US20100074473A1 (en) | System and method of extracting plane features | |
CN110428425B (en) | Sea-land separation method of SAR image based on coastline vector data | |
CN114118252A (en) | Vehicle detection method and detection device based on sensor multivariate information fusion | |
EP3825956B1 (en) | Processing a 3d signal of a shape attribute over a real object | |
CN116051822A (en) | Concave obstacle recognition method and device, processor and electronic equipment | |
Wietrzykowski et al. | Stereo plane R-CNN: Accurate scene geometry reconstruction using planar segments and camera-agnostic representation | |
CN113822361B (en) | SAR image similarity measurement method and system based on Hamming distance | |
KR102468927B1 (en) | Method and system for detecting variation amount of object in sar images | |
Kovacs et al. | Edge detection in discretized range images | |
CN114202684A (en) | Point cloud data projection method, system and device suitable for water surface environment | |
Ganbold et al. | Coarse-to-fine evolutionary method for fast horizon detection in maritime images | |
Quach et al. | Low-level track finding and completion using random fields | |
CN117590371B (en) | Method for realizing global parking space state detection based on 4D millimeter wave imaging radar |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |