CN117670715A - Object point cloud normal determining method, device, equipment, robot and storage medium - Google Patents
Object point cloud normal determining method, device, equipment, robot and storage medium Download PDFInfo
- Publication number
- CN117670715A CN117670715A CN202311698859.XA CN202311698859A CN117670715A CN 117670715 A CN117670715 A CN 117670715A CN 202311698859 A CN202311698859 A CN 202311698859A CN 117670715 A CN117670715 A CN 117670715A
- Authority
- CN
- China
- Prior art keywords
- point
- normal
- determining
- point cloud
- object point
- 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 48
- 238000009499 grossing Methods 0.000 claims abstract description 60
- 238000012937 correction Methods 0.000 claims abstract description 38
- 238000004590 computer program Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 abstract description 21
- 230000003044 adaptive effect Effects 0.000 description 9
- 238000004891 communication Methods 0.000 description 5
- 230000000875 corresponding effect Effects 0.000 description 5
- 238000005452 bending Methods 0.000 description 3
- 230000002596 correlated effect Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000000007 visual effect Effects 0.000 description 3
- 238000005094 computer simulation Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000000513 principal component analysis Methods 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
Abstract
The application relates to the technical field of robot vision, and provides an object point cloud normal determining method, device, equipment, a robot and a storage medium, wherein the method comprises the following steps: acquiring an object point cloud; extracting a target point from the object point cloud; determining a modified smooth radius according to the curvature of the target point; determining a field point set based on the modified smooth radius; the field point set is a point set with the distance between the field point set and the target point in the object point cloud being smaller than or equal to the correction smooth radius; a normal to the target point is determined based on the set of domain points. By means of the method, the normal line of the object point cloud is determined in a self-adaptive object point cloud normal line smoothing mode, the correction smooth radius is determined according to the curvature of the target point, the smoothing requirements of different curvature areas can be met, and therefore accuracy of object point cloud normal line calculation is improved.
Description
Technical Field
The application relates to the technical field of robot vision, in particular to an object point cloud normal determining method, an object point cloud normal determining device, object point cloud normal determining equipment, a robot and a storage medium.
Background
The point cloud is sampled on the object surface, and the normal line of the object surface is the point cloud normal line, so that the geometric estimation of the object surface can be firstly carried out, and the point cloud normal line can be calculated. The normal line calculation method generally estimates the normal line of a point by performing plane fitting or surface fitting from some points in the vicinity of the point.
However, there is often a measurement error when capturing point cloud data by a depth camera. If the irregular data caused by the errors are directly reconstructed by taking the curved surface, the reconstructed curved surface is unsmooth or has a leak, and when the normal is calculated on the object point cloud based on the unsmooth or the leak curved surface, the obtained normal is inaccurate. For example, there may be a problem that the normal angles of adjacent regions are large, that is, a part of the normal is not accurate, and the like.
Disclosure of Invention
The present application aims to solve at least one of the technical problems existing in the related art. Therefore, the method for determining the normal of the object point cloud can meet the smooth requirements of different curvature areas and improve the accuracy of calculating the normal of the object point cloud.
The application also provides a normal determining device, equipment, a robot and a storage medium of the object point cloud.
According to an embodiment of the first aspect of the application, a method for determining a normal line of an object point cloud includes: acquiring an object point cloud; extracting a target point from the object point cloud; determining a modified smooth radius according to the curvature of the target point; determining a field point set based on the modified smooth radius; the field point set is a point set with the distance between the field point set and the target point in the object point cloud being smaller than or equal to the correction smooth radius; a normal to the target point is determined based on the set of domain points.
According to the method for determining the normal of the object point cloud, a self-adaptive object point cloud normal smoothing mode is provided, and the smoothing requirements of different curvature areas are met, so that the accuracy of object point cloud normal calculation is improved.
According to one embodiment of the present application, determining a modified smoothing radius based on a curvature of a target point includes: determining a curvature of the target point; obtaining an initial smooth radius and a smooth coefficient; and determining a modified smooth radius according to the curvature of the target point, the initial smooth radius and the smooth coefficient.
According to one embodiment of the present application, determining a set of domain points based on a modified smooth radius includes: comparing the distances between other points except the target point in the object point cloud and the target point one by one, and taking the point with the distance smaller than or equal to the correction smooth radius as the field point; based on all the domain points, a domain point set is obtained.
According to one embodiment of the present application, determining a normal to a target point based on a set of domain points includes: determining the normal line of each domain point in the domain point set; obtaining an average normal based on the sum of the normals of all the domain points and the number of the domain points in the domain point set; and carrying out unitization treatment on the average normal line to obtain the normal line of the target point.
According to one embodiment of the present application, the modified smoothing radius r is: r=w×d/k; where k is the curvature of the target point, d is the initial smoothing radius, and w is the smoothing coefficient.
According to one embodiment of the present application, obtaining an object point cloud includes: acquiring a semantic grid map based on an indoor environment; and extracting object point clouds based on the semantic grid map to obtain the object point clouds.
According to one embodiment of the present application, after determining the normal of the target point based on the field point set, it includes: judging whether all points in the object point cloud determine normal lines; if at least one point of the normal is not determined by the existence of the object point cloud, the target point is extracted from the at least one point.
According to an embodiment of the second aspect of the present application, a normal determining device for an object point cloud includes: the object point cloud module is used for acquiring object point clouds; the target point module is used for extracting a target point from the object point cloud; the modified smooth radius module is used for determining a modified smooth radius according to the curvature of the target point; the domain point set module is used for determining a domain point set based on the modified smooth radius; the field point set is a point set with the distance between the field point set and the target point in the object point cloud being smaller than or equal to the correction smooth radius; and the normal line module is used for determining the normal line of the target point based on the field point set.
An electronic device according to an embodiment of the third aspect of the present application includes a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements any one of the object point cloud normal determination methods described above when executing the program.
The robot according to the fourth aspect of the present application comprises a control unit; the control unit is used for executing any one of the object point cloud normal determining methods.
A non-transitory computer readable storage medium according to an embodiment of the fifth aspect of the present application has stored thereon a computer program which, when executed by a processor, implements any of the object point cloud normal determination methods described above.
The above technical solutions in the embodiments of the present application have at least one of the following technical effects:
the method has the advantages that the self-adaptive object point cloud normal smoothing mode is adopted, the correction smoothing radius is determined according to the curvature of the target point, the original uniform smoothing radius is corrected according to the curvatures of different target points, the smoothing requirements of different curvature areas can be met, and therefore the accuracy of object point cloud normal calculation is improved.
Additional aspects and advantages of the application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the application.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the related art, the drawings that are required to be used in the embodiments or the related technical descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to the drawings without inventive effort for a person having ordinary skill in the art.
Fig. 1 is one of flow diagrams of a method for determining a normal line of an object point cloud according to an embodiment of the present application;
FIG. 2 is a second flow chart of a method for determining a normal line of an object point cloud according to an embodiment of the present application;
fig. 3 is a schematic structural diagram of a normal determining device of an object point cloud according to an embodiment of the present application;
fig. 4 is a schematic entity structure of an electronic device according to an embodiment of the present application.
Detailed Description
Embodiments of the present application are described in further detail below with reference to the accompanying drawings and examples. The following examples are illustrative of the present application but are not intended to limit the scope of the present application.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," 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 embodiments of the present application. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.
The embodiment of the application provides a normal line determining method of an object point cloud. Referring to fig. 1, fig. 1 is a flow chart of a method for determining a normal line of an object point cloud according to an embodiment of the present application, in this embodiment, the method for determining the normal line of the object point cloud may include steps 110 to 150, where each step is specifically as follows:
step 110: an object point cloud is acquired.
The object point cloud may include at least one object point cloud therein. Each object point cloud includes at least one point therein. The object point cloud is one of the point cloud data, and various acquisition modes of the object point cloud can be realized, for example, the object point cloud can be scanned by a sensor or converted by a computer model.
Sensor scanning: a sensor such as a laser radar or a depth camera can be adopted to acquire point cloud data by scanning the surrounding environment, and object point clouds are screened from the point cloud data. The lidar may calculate the distance by emitting a laser beam and measuring the time it is reflected back, thereby obtaining point cloud data. The depth camera may utilize infrared radiation and an image sensor to measure object distances in the scene, thereby generating point cloud data.
Computer model conversion: two-dimensional images or three-dimensional models can be converted into object point clouds using computer vision and image processing techniques. For example, the object point cloud can be recovered from the image by means of structured light, stereo vision or multi-view image capturing, etc.
Step 120: the target point is extracted from the object point cloud.
The target point is extracted from the object point cloud, wherein the target point is the point to be determined by normal calculation.
In some embodiments, it is of course also possible to perform normal calculation determination on points in the object point cloud one by one, so as to obtain normal information of the whole object point cloud. The normal information may provide information support for normal calculation of the present embodiment.
Step 130: and determining a correction smooth radius according to the curvature of the target point.
The curvature and the original normal of the target point can be obtained by performing plane fitting or surface fitting and normal calculation through the target point and some points near the target point.
For example, the curvature of the target point may be acquired by principal component analysis or the like.
And determining a modified smoothing radius based on the curvature of the target point, thereby meeting the smoothing requirements of the areas with different curvatures.
In some embodiments, the curvature of the target point and the degree of curvature of the object surface are positively correlated, and the curvature of the target point and the angle of normal are negatively correlated. Namely:
the larger the curvature of the target point is, the larger the bending degree of the surface of the object is, and the smaller the normal included angle of the adjacent area is after the object is smoothed; conversely, the smaller the curvature of the target point is, the smaller the bending degree of the object surface is, and the larger the normal included angle of the adjacent area is after smoothing.
The normal angle of the adjacent area can influence the accuracy of normal calculation of the target point, for example, the problem that partial normal is inaccurate when the normal angle of the adjacent area is larger can exist. Based on this, the method for determining the modified smooth radius according to the curvature of the target point provided in this embodiment can adapt to the smooth requirements of different curvature areas, thereby improving the accuracy of object point cloud normal calculation.
Step 140: a set of domain points is determined based on the modified smoothing radius.
And determining a field point set of the target point based on the modified smoothing radius which is obtained in the step 130 and is suitable for the smoothing requirements of the different curvature areas. The field point set is a point set with the distance between the object point cloud and the target point being smaller than or equal to the correction smooth radius.
Therefore, in this step, it is necessary to compare the distances between points other than the target point in the object point cloud and the target point, thereby obtaining a domain point set.
Step 150: a normal to the target point is determined based on the set of domain points.
Finally, a normal to the target point is obtained based on the set of domain points determined by the modified smoothing radius. Specifically, the normal line of the target point may be obtained from the normal line data in the domain point set and the number of all the points in the domain point set.
In the above, the present embodiment provides a method for determining a normal line of an object point cloud, including: acquiring an object point cloud; extracting a target point from the object point cloud; determining a modified smooth radius according to the curvature of the target point; determining a field point set based on the modified smooth radius; the field point set is a point set with the distance between the field point set and the target point in the object point cloud being smaller than or equal to the correction smooth radius; a normal to the target point is determined based on the set of domain points. By adopting the self-adaptive object point cloud normal smoothing mode, the correction smoothing radius can be determined according to the curvature of the target point, the original uniform smoothing radius can be corrected according to the curvature of different target points, and the smoothing requirements of different curvature areas can be met, so that the accuracy of object point cloud normal calculation is improved.
Based on the above embodiment, the step of determining the modified smooth radius according to the curvature of the target point may specifically include:
step 131: the curvature of the target point is determined.
Step 132: an initial smoothing radius and smoothing coefficient are obtained.
Step 133: and determining a modified smooth radius according to the curvature of the target point, the initial smooth radius and the smooth coefficient.
In the present embodiment, the correction smoothing radius may be determined according to the curvature of the target point, the initial smoothing radius, and the smoothing coefficient. The curvature of the target point can be obtained through calculation, the initial smooth radius and the smooth coefficient are preset values, and a specific value can be set by a person skilled in the art according to actual situations.
Based on the above embodiment, the modified smoothing radius r and the curvature k of the target point may be positively correlated, that is: the larger the curvature k of the target point is, the larger the correction smooth radius r is; conversely, the smaller the curvature k of the target point, the smaller the correction smoothing radius r.
Illustratively, the modified smooth radius r may be expressed as: r=w×d/k.
Where k is the curvature of the target point, d is the initial smoothing radius, and w is the smoothing coefficient.
The specific arrangement of the modified smooth radius r is given above. According to the formula and the principle, when the curvature k of the target point is larger, the bending degree of the object surface is larger, so that when normal calculation is performed, in order to ensure the accuracy of the normal, the smoothing process can be performed through larger correction smoothing radius, and the smoothing requirement of the target point when the curvature is larger is met.
Conversely, when the curvature k of the target point is smaller, the curvature k of the object surface is smaller, so that the smoothing treatment can be performed by smaller modified smoothing radius, and the smoothing requirement of the target point at the smaller curvature is satisfied.
Based on the above embodiment, the step of determining the domain point set based on the modified smooth radius may specifically include:
step 141: and comparing the distances between other points except the target point in the object point cloud and the target point one by one, and taking the point with the distance smaller than or equal to the correction smooth radius as the field point.
Step 142: based on all the domain points, a domain point set is obtained.
In this embodiment, it is necessary to compare the distances between points other than the target point in the object point cloud and the target point one by one, and obtain the domain point based on the comparison result.
Specifically, in the distance comparison process, when the distance between some other point and the target point is smaller than or equal to the correction smooth radius, the other point is marked as a domain point of the target point; otherwise, when the distance between a certain object point and the target point is larger than the correction smooth radius, the mark of the domain point is not made.
When object points except the target point in the object point cloud are subjected to distance comparison one by one, a plurality of field points can be obtained, all the field points are summarized, and finally a field point set can be obtained.
Based on the above embodiment, the step of determining the normal line of the target point based on the domain point set may specifically include:
step 151: a normal to each of the domain points in the set of domain points is determined.
Step 152: and obtaining an average normal based on the sum of the normals of all the domain points and the number of the domain points in the domain point set.
Step 153: and carrying out unitization treatment on the average normal line to obtain the normal line of the target point.
In the present embodiment, a specific method of determining the normal line of the target point based on the domain point set is provided. Firstly, determining the normal line of each field point in the field point set in sequence; since the field point set can comprise a plurality of field points, the normal information of the field point set can also comprise normals corresponding to each point one by one, namely a plurality of normals are included; the average normal can be obtained by adding the normals of all the domain points in the domain point set and dividing the normals into the number of the domain points in the domain point set.
Further, since the normal line is a unit vector and the average normal line is subjected to addition processing, it is necessary to unitize the average normal line, and then the normal line to the target point can be obtained.
It should be noted that, the normal line obtained in step 151 may be a normal line obtained based on a uniform smooth radius; the normal line obtained based on the adaptive correction smoothing radius in the above embodiment may be obtained.
Based on the above embodiment, the step of obtaining the object point cloud may specifically include:
step 111: a semantic grid map is acquired based on the indoor environment.
Step 112: and extracting object point clouds based on the semantic grid map to obtain the object point clouds.
In this embodiment, the object point cloud may be obtained from a semantic grid map. The semantic grid map refers to a 3D semantic grid that can represent real-world shape information using a plurality of cube grids. Each grid has corresponding unique grid information, where the grid information includes semantic information and index values.
For example, the semantic grid map may be a robot applied indoors, and thus the acquired environment image data may be image data of an indoor environment within a movable range of the robot. The robot can be provided with a visual sensor, and the semantic grid map can be obtained through shooting of the visual sensor and data processing of the robot.
In some embodiments, the visual sensor may be a depth camera, which may include a binocular RGB camera, a structured light camera, a TOF (Time of Flight) camera, or the like.
Based on the above embodiment, the step after determining the normal of the target point based on the domain point set may further include:
step 161: and judging whether all points in the object point cloud determine normal lines.
Step 162: if at least one point of the normal is not determined by the existence of the object point cloud, the target point is extracted from the at least one point.
In this embodiment, normal determination by means of adaptive object point cloud normal smoothing may be performed for all points in the object point cloud. It is therefore necessary to make normal determination for points in the object point cloud one by one.
Specifically, after determining the normal line of the target point each time, it is required to determine whether all points in the object point cloud determine the normal line by means of smoothing the normal line of the adaptive object point cloud, and if at least one point in the object point cloud exists, which does not pass through the adaptive correction smoothing radius determination normal line in the above embodiment mode, the target point is extracted from at least one point, and the adaptive correction smoothing radius determination normal line process is continued.
If no point of the normal line is determined by the adaptive correction smoothing radius in the object point cloud, the adaptive object point cloud normal line smoothing of all normal lines in the object point cloud is determined, and the normal line of each point in the object point cloud is determined.
Based on the above embodiment, in order to save calculation force, the calculation efficiency is improved. Before the step of determining the modified smooth radius according to the curvature of the target point, whether the corresponding modified smooth radius needs to be determined or not can be judged according to the curvature of the target point, and the subsequent step is executed, namely, the curvature of the target point is compared with a preset threshold value to obtain a comparison result, and whether the subsequent step is executed or not is determined according to the comparison result.
For example, comparing the curvature of the target point with a first preset threshold, and when the curvature of the target point is less than or equal to the first preset threshold, determining a modified smoothing radius according to the curvature of the target point and executing the subsequent steps; otherwise, when the curvature of the target point is greater than the first preset threshold, ending the normal determination of the target point by adopting a mode of correcting the smooth radius to determine the normal, and adopting a uniform and fixed smooth radius to determine the normal.
For example, comparing the curvature of the target point with a second preset threshold, and when the curvature of the target point is greater than or equal to the second preset threshold, determining a modified smoothing radius according to the curvature of the target point and executing the subsequent steps; otherwise, when the curvature of the target point is smaller than the second preset threshold, ending the normal determination of the target point by adopting a mode of correcting the smooth radius to determine the normal, and adopting a uniform and fixed smooth radius to determine the normal.
For example, comparing the curvature of the target point with a first preset threshold value and a second preset threshold value, and when the curvature of the target point is greater than or equal to the second preset threshold value or the curvature of the target point is less than or equal to the first preset threshold value, determining a modified smoothing radius according to the curvature of the target point and executing the subsequent steps; otherwise, when the curvature of the target point is larger than the first preset threshold value and smaller than the second preset threshold value, the normal line of the target point is determined by correcting the smooth radius, and the normal line can be determined by adopting a uniform and fixed smooth radius.
It should be noted that the first preset threshold and the second preset threshold may be set by those skilled in the art according to practical use, and are not limited herein.
Referring to fig. 2, fig. 2 is a second flowchart of a method for determining a normal line of an object point cloud according to an embodiment of the present application, where in the method for determining a normal line of an object point cloud may include steps 210 to 250, each of the steps is specifically as follows:
step 210: one point is sequentially acquired from the object point cloud.
For example, at first, the number of points in the object point cloud is determined, each point is given a sequence number, and one point is sequentially acquired from the object point cloud according to the size of the sequence number, so that the normal line determining method of the above embodiment is performed, that is, the sequentially acquired point is taken as the target point.
Optionally, prior to the examination step 210, the original normals and curvatures in the object point cloud may also be determined by a uniform smooth radius as subsequent calculation parameters.
Step 220: a modified smooth radius is calculated from the point curvature.
Based on the above embodiment, the correction calculation is performed according to the curvature of the target point, and the correction smooth radius is obtained.
Step 230: and searching a field point set in the object point cloud, wherein the distance between the field point set and the object point cloud is smaller than or equal to the correction smooth radius.
Based on the above embodiment, the object point cloud is found, the domain point from which the distance from the object point cloud is smaller than and equal to the correction smoothing radius is determined, and the domain point set of the object point is determined based on all the domain points.
Step 240: and calculating a new normal of the point according to the original normal of the field point set.
And adding and summarizing normals corresponding to each field point in the field point set to obtain a first total value, dividing the first total value by the number of field points in the field point set to obtain an average normal of the object point cloud, and unitizing the average normal to obtain a new normal of the cloud.
Step 250: it is determined whether all points in the object point cloud recalculate the new normal.
Judging whether all points in the object point cloud calculate new normals, if all points in the object point cloud calculate new normals in an adaptive mode, continuing to execute step 260; if there is at least one point in the object point cloud where the new normal is not calculated in an adaptive manner, the process returns to step 210, where the next point in the object point cloud where the new normal is not calculated is selected until all points calculate the new normal.
Step 260: the original normal of the object point cloud is replaced with the new normal.
When all points in the object point cloud calculate new normals, the original normals of the points in the corresponding object point cloud are replaced by the new normal lines in a unified mode.
In the embodiment, after new normal lines of all points in the object point cloud are uniformly calculated, old normal lines are uniformly replaced by normal lines, and the normal line smoothing radius is adaptively adjusted by combining curvature information, so that the accuracy of object normal line calculation is improved, and meanwhile, the calculation efficiency is also improved.
Referring to fig. 3, fig. 3 is a schematic structural diagram of a normal determining device for an object point cloud according to an embodiment of the present application. In this embodiment, the normal determining apparatus of the object point cloud may include an object point cloud module 310, a target point module 320, a modified smooth radius module 330, a field point set module 340, and a normal module 350. Specifically:
the object point cloud module 310 is configured to obtain an object point cloud.
The target point module 320 is configured to extract a target point from the object point cloud.
The modified smooth radius module 330 is configured to determine a modified smooth radius according to the curvature of the target point.
A domain point set module 340 for determining a domain point set based on the modified smooth radius; the field point set is a point set with the distance between the field point set and the target point in the object point cloud being smaller than or equal to the correction smooth radius.
A normal module 350 for determining a normal to the target point based on the set of domain points.
The normal determining device for the object point cloud provided by the embodiment comprises an object point cloud module, a target point module, a correction smooth radius module, a field point set module and a normal module, wherein the correction smooth radius module adopts a self-adaptive object point cloud normal smoothing mode, the correction smooth radius can be determined according to the curvature of the target point, the original uniform smooth radius can be corrected according to the curvature of different target points, the smooth requirements of different curvature areas can be met, and therefore the accuracy of object point cloud normal calculation is improved.
Based on the above embodiment, the modified smooth radius module 330 is specifically configured to:
determining a curvature of the target point; obtaining an initial smooth radius and a smooth coefficient; and determining a modified smooth radius according to the curvature of the target point, the initial smooth radius and the smooth coefficient.
Based on the above embodiment, the domain point set module 340 is specifically configured to:
comparing the distances between other points except the target point in the object point cloud and the target point one by one, and taking the point with the distance smaller than or equal to the correction smooth radius as the field point; based on all the domain points, a domain point set is obtained.
Based on the above embodiment, the normal module 350 is specifically configured to:
determining the normal line of each domain point in the domain point set; obtaining an average normal based on the sum of the normals of all the domain points and the number of the domain points in the domain point set; and carrying out unitization treatment on the average normal line to obtain the normal line of the target point.
Based on the above embodiment, the modified smoothing radius r is: r=w×d/k; where k is the curvature of the target point, d is the initial smoothing radius, and w is the smoothing coefficient.
Based on the above embodiment, the object point cloud module 310 is specifically configured to:
acquiring a semantic grid map based on an indoor environment; and extracting object point clouds based on the semantic grid map to obtain the object point clouds.
Based on the above embodiment, the normal determining device of the object point cloud further includes a judging module, where the judging module is specifically configured to:
judging whether all points in the object point cloud determine normal lines; if at least one point of the normal is not determined by the existence of the object point cloud, the target point is extracted from the at least one point.
In yet another aspect, an embodiment of the present application further provides an electronic device. Referring to fig. 4, fig. 4 is a schematic physical structure of an electronic device according to an embodiment of the present application, where the electronic device may include: processor 410, communication interface (Communications Interface) 420, memory 430 and communication bus 440, wherein processor 410, communication interface 420 and memory 430 communicate with each other via communication bus 440. The processor 410 may invoke logic instructions in the memory 430 to perform the following method of determining the normal to the object point cloud:
acquiring an object point cloud; extracting a target point from the object point cloud; determining a modified smooth radius according to the curvature of the target point; determining a field point set based on the modified smooth radius; the field point set is a point set with the distance between the field point set and the target point in the object point cloud being smaller than or equal to the correction smooth radius; a normal to the target point is determined based on the set of domain points.
Further, the logic instructions in the memory 430 described above may be implemented in the form of software functional units and may be stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the related art or in a part of the technical solution, in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In yet another aspect, an embodiment of the present application discloses a robot, where the robot includes a control unit, and the control unit is configured to execute the method for determining the normal line of the object point cloud provided by the above embodiments of the method, for example, including:
acquiring an object point cloud; extracting a target point from the object point cloud; determining a modified smooth radius according to the curvature of the target point; determining a field point set based on the modified smooth radius; the field point set is a point set with the distance between the field point set and the target point in the object point cloud being smaller than or equal to the correction smooth radius; a normal to the target point is determined based on the set of domain points.
In yet another aspect, embodiments of the present application further provide a non-transitory computer readable storage medium having stored thereon a computer program that, when executed by a processor, is implemented to perform the method for determining a normal of an object point cloud provided in the above embodiments, for example, including:
acquiring an object point cloud; extracting a target point from the object point cloud; determining a modified smooth radius according to the curvature of the target point; determining a field point set based on the modified smooth radius; the field point set is a point set with the distance between the field point set and the target point in the object point cloud being smaller than or equal to the correction smooth radius; a normal to the target point is determined based on the set of domain points.
In summary, the embodiments of the present application provide an object point cloud normal determining method, apparatus, device, robot, and storage medium, where the method includes: acquiring an object point cloud; extracting a target point from the object point cloud; determining a modified smooth radius according to the curvature of the target point; determining a field point set based on the modified smooth radius; the field point set is a point set with the distance between the field point set and the target point in the object point cloud being smaller than or equal to the correction smooth radius; a normal to the target point is determined based on the set of domain points. By means of the method, the normal line of the object point cloud is determined in a self-adaptive object point cloud normal line smoothing mode, the correction smooth radius is determined according to the curvature of the target point, the smoothing requirements of different curvature areas can be met, and therefore accuracy of object point cloud normal line calculation is improved.
The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on such understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the related art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform the method described in the respective embodiments or some parts of the embodiments.
Finally, it should be noted that the above embodiments are only for illustrating the present application, and are not limiting of the present application. While the present application has been described in detail with reference to the embodiments, those skilled in the art will understand that various combinations, modifications, or equivalents of the technical solutions of the present application may be made without departing from the spirit and scope of the technical solutions of the present application, and all such modifications are intended to be covered by the claims of the present application.
Claims (11)
1. A method for determining a normal to an object point cloud, comprising:
acquiring an object point cloud;
extracting a target point from the object point cloud;
determining a modified smooth radius according to the curvature of the target point;
determining a set of domain points based on the modified smooth radius; wherein the field point set is a point set in the object point cloud, and the distance between the field point set and the target point is smaller than or equal to the correction smooth radius;
a normal to the target point is determined based on the set of domain points.
2. The method according to claim 1, wherein the determining a modified smoothing radius from a curvature of the target point includes:
determining a curvature of the target point;
obtaining an initial smooth radius and a smooth coefficient;
and determining a modified smooth radius according to the curvature of the target point, the initial smooth radius and the smooth coefficient.
3. The method of claim 1, wherein determining the set of domain points based on the modified smoothed radius comprises:
comparing the distances between other points except the target point in the object point cloud and the target point one by one, and taking the point with the distance smaller than or equal to the correction smooth radius as a field point;
and obtaining the domain point set based on all the domain points.
4. The method of determining a normal to an object point cloud according to claim 1, wherein the determining a normal to the target point based on the set of domain points includes:
determining a normal line of each domain point in the domain point set;
obtaining an average normal based on the sum of the normals of all the domain points and the number of the domain points in the domain point set;
and carrying out unitization treatment on the average normal line to obtain the normal line of the target point.
5. The method for determining the normal line of the object point cloud according to claim 2, wherein the modified smoothing radius r is:
r＝w*d/k；
where k is the curvature of the target point, d is the initial smoothing radius, and w is the smoothing coefficient.
6. The method for determining a normal to an object point cloud according to any of claims 1 to 5, wherein the acquiring the object point cloud comprises:
acquiring a semantic grid map based on an indoor environment;
and extracting object point clouds based on the semantic grid map to obtain the object point clouds.
7. The method according to any one of claims 1 to 5, wherein after determining the normal of the target point based on the field point set, comprising:
judging whether all points in the object point cloud have determined normals;
and if at least one point with an undetermined normal exists in the object point cloud, extracting a target point from the at least one point.
8. A normal determining apparatus for an object point cloud, comprising:
the object point cloud module is used for acquiring object point clouds;
a target point module for extracting a target point from the object point cloud;
the modified smooth radius module is used for determining a modified smooth radius according to the curvature of the target point;
a domain point set module for determining a domain point set based on the modified smooth radius; wherein the field point set is a point set in the object point cloud, and the distance between the field point set and the target point is smaller than or equal to the correction smooth radius;
and the normal module is used for determining the normal of the target point based on the field point set.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the method of determining the normal of an object point cloud according to any of claims 1 to 7 when executing the program.
10. A robot comprising a control unit; the control unit is configured to perform the normal determination method of the object point cloud according to any one of claims 1 to 7.
11. A non-transitory computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when executed by a processor, implements the method of determining the normals of an object point cloud according to any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311698859.XA CN117670715A (en) | 2023-12-11 | 2023-12-11 | Object point cloud normal determining method, device, equipment, robot and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311698859.XA CN117670715A (en) | 2023-12-11 | 2023-12-11 | Object point cloud normal determining method, device, equipment, robot and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117670715A true CN117670715A (en) | 2024-03-08 |
Family
ID=90065913
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311698859.XA Pending CN117670715A (en) | 2023-12-11 | 2023-12-11 | Object point cloud normal determining method, device, equipment, robot and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117670715A (en) |
-
2023
- 2023-12-11 CN CN202311698859.XA patent/CN117670715A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20190236807A1 (en) | Method and device for determining external parameter of stereoscopic camera | |
| Kang et al. | Automatic targetless camera–lidar calibration by aligning edge with gaussian mixture model | |
| CN109640066B (en) | Method and device for generating high-precision dense depth image | |
| EP3488424A1 (en) | Systems and methods for improved surface normal estimation | |
| KR20170068462A (en) | 3-Dimensional Model Generation Using Edges | |
| CA3114457A1 (en) | Volume measurement method and system, apparatus and computer-readable storage medium | |
| CN110487274B (en) | SLAM method and system for weak texture scene, navigation vehicle and storage medium | |
| CN111105452B (en) | Binocular vision-based high-low resolution fusion stereo matching method | |
| Shivakumar et al. | Real time dense depth estimation by fusing stereo with sparse depth measurements | |
| CN111798507A (en) | Power transmission line safety distance measuring method, computer equipment and storage medium | |
| CN111179333B (en) | Defocus blur kernel estimation method based on binocular stereo vision | |
| CN112381862A (en) | Full-automatic registration method and device for CAD (computer-aided design) model and triangular mesh | |
| CN111739071A (en) | Rapid iterative registration method, medium, terminal and device based on initial value | |
| CN110793441A (en) | High-precision object geometric dimension measuring method and device | |
| Zhu et al. | Triangulation of well-defined points as a constraint for reliable image matching | |
| CN117670715A (en) | Object point cloud normal determining method, device, equipment, robot and storage medium | |
| CN114387353A (en) | Camera calibration method, calibration device and computer readable storage medium | |
| CN112258635B (en) | Three-dimensional reconstruction method and device based on improved binocular matching SAD algorithm | |
| CN111833395B (en) | Direction-finding system single target positioning method and device based on neural network model | |
| CN113421292A (en) | Three-dimensional modeling detail enhancement method and device | |
| CN108230377B (en) | Point cloud data fitting method and system | |
| CN113409376A (en) | Method for filtering laser radar point cloud based on depth estimation of camera | |
| CN113065483A (en) | Positioning method, positioning device, electronic equipment, medium and robot | |
| CN111630569B (en) | Binocular matching method, visual imaging device and device with storage function | |
| CN113496521B (en) | Method and device for generating depth image and camera external parameter by using multiple color pictures |
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 |