Method and device for identifying fracture surface of object fragment

Abstract

The invention discloses an article fragment fracture surface identification method and device, and the method comprises the steps: taking the mass center of an initial fragment point cloud as an original point, recalculating the coordinate information of each data point in the initial fragment point cloud, and obtaining the fragment point cloud; determining adjacent points of the data points by taking each data point in the fragment point cloud as a center and a set distance as a radius; determining spatial distribution characteristics of the data points according to a relative position relationship between the data points and adjacent points of the data points; and according to the spatial distribution characteristics of the data points, determining the data points of the fracture surface as target data points in the data points, and representing the fracture surface in a region formed by the target data points in the fragment point cloud. The method does not depend on manual marking of a user, and realizes automatic identification of the area representing the fracture surface in the fragment point cloud.

Description

Technical Field

[0001] The present application belongs to the field of point cloud data processing, and specifically relates to a method and device for identifying fracture surfaces of object fragments. Background Art

[0002] With the development of science and technology, computer technology and digital technology, the three-dimensional reconstruction of cultural relics has also entered the digital stage. Digital modeling can not only restore the appearance of cultural relics buildings, but also provide important assistance for restoration and protection work. In the process of cultural relic restoration, early cultural relic experts and researchers manually restored cultural relics based on historical materials and experience. This method is not only laborious and time-consuming, but also likely to cause secondary wear and tear on cultural relics during the splicing process. With the development of computer technology, cultural relics can be input into specific software for simulated splicing using laser scanning and other methods with digital information as the carrier, and the first step of splicing is the identification and extraction of fracture surfaces. The determination of fracture surfaces currently still relies on manual recognition by the naked eye.

[0003] Therefore, how to automatically determine the fracture surface area based on the point cloud of object fragments is a technical problem that needs to be solved urgently. Summary of the invention

[0004] The purpose of this application is to automatically determine the area where the fracture surface is located for the fragment point cloud of the object fragment.

[0005] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by the practice of the present application.

[0006] According to one embodiment of the present application, the present application provides a method for identifying a fracture surface of an object fragment, the method comprising:

[0007] Collect point clouds of object fragments to obtain initial fragment point clouds;

[0008] Taking the centroid of the initial fragment point cloud as the origin, recalculating the coordinate information of each data point in the initial fragment point cloud to obtain the fragment point cloud;

[0009] Taking each data point in the fragment point cloud as the center and a set distance as the radius, determine the neighboring points of the data point;

[0010] Determining the spatial distribution characteristics of the data point according to the relative position relationship between the data point and the neighboring points of the data point;

[0011] According to the spatial distribution characteristics of each of the data points, the data points belonging to the fracture surface are determined as target data points among the data points, and the area formed by the target data points in the fragment point cloud represents the fracture surface.

[0012] According to an embodiment of the present application, taking the centroid of the initial fragment point cloud as the origin, recalculating the coordinate information of each data point in the initial fragment point cloud to obtain the fragment point cloud includes:

[0013] If the centroid of the fragment point cloud does not belong to the data point of the fragment point cloud, the centroid of the fragment point cloud is used as the data point of the fragment point cloud;

[0014] Subtract the coordinate information of each data point from the coordinate information of the centroid to obtain new coordinate information of each data point;

[0015] The coordinate information of each of the data points is changed to new coordinate information to obtain a fragment point cloud with the centroid as the origin.

[0016] According to an embodiment of the present application, after subtracting the coordinate information of each data point from the coordinate information of the centroid to obtain new coordinate information of each data point, the method further includes:

[0017] The data point farthest from the origin in the fragment point cloud is used as a reference data point, and the distance between the reference data point and the origin is a reference distance;

[0018] The ratio of the reference distance to the preset target distance is used as a scaling factor, and the fragment point cloud is scaled according to the scaling factor.

[0019] According to an embodiment of the present application, scaling the fragment point cloud according to the scaling factor includes:

[0020] Multiplying the coordinate values ​​of the coordinate information of the data point on each axis by the scaling factor to obtain the target coordinate information of the data point;

[0021] Move the data point to the location of the target coordinate information.

[0022] According to an embodiment of the present application, determining the spatial distribution feature of the data point according to the relative position relationship between the data point and the neighboring points of the data point includes:

[0023] Randomly select a data point as a query point, and use data points whose distance to the query point is less than a set distance as neighboring points of the query point;

[0024] arbitrarily selecting a neighboring point of the query point as a feature reference point of the query point, and calculating the local spatial distribution feature of the query point according to the relative position relationship between the query point and the feature reference point;

[0025] When obtaining neighboring points as feature reference points of the query point respectively, each local spatial distribution feature of the query point;

[0026] All local spatial distribution features of the query point are integrated to obtain the spatial distribution features of the query point.

[0027] According to an embodiment of the present application, calculating the local spatial distribution feature of the query point according to the relative position relationship between the query point and the feature reference point includes:

[0028] Using the vector formed between the query point and the feature reference point as the first vector, using the normal vector of the query point as the second vector, and using the normal vector of the feature reference point as the third vector;

[0029] A vector which is perpendicular to the plane where the first vector and the second vector are located and passes through the query point is taken as a fourth vector;

[0030] A vector perpendicular to the plane where the second vector and the fourth vector are located and passing through the query point is used as a fifth vector;

[0031] The second vector, the fourth vector and the fifth vector are translated to obtain a second translation vector, a fourth translation vector and a fifth translation vector, wherein the intersection of the second translation vector, the third translation vector and the fourth translation vector is the feature reference point;

[0032] The angle between the first vector and the second vector is taken as the first angle, the angle between the fourth translation vector and the third vector is taken as the second angle, the angle formed by the third vector and the projection line is taken as the third angle, and the projection of the third vector on the fifth translation vector and the second translation vector is taken as the projection line;

[0033] A local feature histogram between the query point and the feature reference point, that is, the local spatial distribution feature, is calculated according to the first angle, the second angle, and the third angle.

[0034] According to an embodiment of the present application, all local spatial distribution features of the query point are integrated to obtain the spatial distribution features of the query point, including:

[0035] The inverse of the distance between each of the feature reference points and the query point is used as the feature weight of each of the feature reference points;

[0036] The feature weight of each feature reference point is multiplied by the local spatial distribution feature corresponding to each feature reference point, and the obtained products are integrated to obtain the spatial distribution feature of the query point.

[0037] According to an embodiment of the present application, all local spatial distribution features of the query point are integrated to obtain the spatial distribution features of the query point, further comprising:

[0038] Integrating all local spatial distribution features of the query point to obtain the initial spatial distribution features of the query point;

[0039] Calculate the initial spatial distribution characteristics of each neighboring point as the neighboring initial spatial distribution characteristics;

[0040] Calculate the feature weight corresponding to each neighboring point according to the distance between the neighboring point and the query point;

[0041] The neighboring initial spatial distribution feature of each of the neighboring points is multiplied by the feature weight corresponding to the neighboring point, and the sum of the products is added to the initial spatial distribution feature to obtain the spatial distribution feature.

[0042] According to an embodiment of the present application, determining a data point belonging to a fracture surface among the data points as a target data point according to the spatial distribution characteristics of the data points includes:

[0043] The spatial distribution characteristics of the data point and the coordinate information of the data point are input into a preset judgment neural network to determine whether the data point is a target data point.

[0044] According to one embodiment of the present application, the present application provides a fracture surface identification device for object fragments, the device comprising:

[0045] The acquisition module is used to collect point clouds of object fragments to obtain initial fragment point clouds;

[0046] An updating module, configured to recalculate the coordinate information of each data point in the initial fragment point cloud with the centroid of the initial fragment point cloud as the origin, so as to obtain the fragment point cloud;

[0047] A neighboring module, for determining neighboring points of the data points with each data point in the fragment point cloud as the center and a set distance as the radius;

[0048] A feature module, used to determine the spatial distribution feature of the data point according to the relative position relationship between the data point and the neighboring points of the data point;

[0049] The screening module is used to determine the data points belonging to the fracture surface among the data points as target data points according to the spatial distribution characteristics of the data points, and the area formed by the target data points in the fragment point cloud represents the fracture surface.

[0050] In the present application, the initial fragment point cloud is obtained by collecting point clouds of object fragments; the coordinate information of each data point in the initial fragment point cloud is recalculated with the centroid of the initial fragment point cloud as the origin to obtain the fragment point cloud; the neighboring points of the data point are determined with each data point in the fragment point cloud as the center and the set distance as the radius; the spatial distribution characteristics of the data point are determined according to the relative position relationship between the data point and the neighboring points of the data point; according to the spatial distribution characteristics of each data point, the data point belonging to the fracture surface is determined as the target data point in each data point, and the area formed by the target data point in the fragment point cloud represents the fracture surface. The present application does not rely on manual marking by the user, and realizes automatic recognition of the area representing the fracture surface in the fragment point cloud.

[0051] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by the practice of the present application.

[0052] It should be understood that the foregoing general description and the following detailed description are exemplary only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present application, and together with the specification are used to explain the principles of the present application. Obviously, the drawings described below are only some embodiments of the present application, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0054] Figure 1 A flow chart of a method for identifying fracture surfaces of object fragments according to an embodiment of the present application is shown.

[0055] Figure 2 A flowchart of converting an initial fragment point cloud into a fragment point cloud according to an embodiment of the present application is shown.

[0056] Figure 3 A flowchart of scaling a fragment point cloud according to an embodiment of the present application is shown.

[0057] Figure 4 A flow chart is shown of using a ratio of a reference distance to a preset target distance as a zoom factor and scaling a fragmented point cloud according to the zoom factor according to an embodiment of the present application.

[0058] Figure 5 A flow chart is shown for determining spatial distribution characteristics of data points based on the relative position relationship between the data points and their neighboring points according to an embodiment of the present application.

[0059] Figure 6A flow chart of calculating the local spatial distribution features of a query point according to the relative position relationship between the query point and a feature reference point according to an embodiment of the present application is shown.

[0060] Figure 7 A schematic diagram of the distribution of various vectors according to an embodiment of the present application is shown.

[0061] Figure 8 A flow chart of integrating all local spatial distribution features of a query point to obtain the spatial distribution features of the query point according to an embodiment of the present application is shown.

[0062] Fig. 9 A flow chart of integrating all local spatial distribution features of a query point to obtain the spatial distribution features of the query point according to another embodiment of the present application is shown.

[0063] Fig.10 A schematic diagram of a fracture surface identification device for object fragments according to an embodiment of the present application is shown.

[0064] Fig.11 The figure shows a hardware structure diagram of a method for identifying fracture surfaces of object fragments according to an embodiment of the present application. DETAILED DESCRIPTION

[0065] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this application will be more comprehensive and complete and fully convey the concept of the example embodiments to those skilled in the art.

[0066] In addition, described feature, structure or characteristic can be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present application. However, those skilled in the art will appreciate that the technical scheme of the present application can be put into practice without one or more of the specific details, or other methods, components, devices, steps, etc. can be adopted. In other cases, known methods, devices, realizations or operations are not shown or described in detail to avoid blurring the various aspects of the application.

[0067] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0068] The flowcharts shown in the accompanying drawings are only exemplary and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps can be decomposed, and some operations / steps can be combined or partially combined, so the actual execution order may change according to actual conditions.

[0069] In the embodiments of the present application, the term "module" or "unit" refers to a computer program or a part of a computer program with a predetermined function, and works together with other related parts to achieve a predetermined goal, and can be implemented in whole or in part by using software, hardware (such as processing circuits or memories) or a combination thereof. Similarly, a processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be part of an overall module or unit that includes the function of the module or unit.

[0070] See also Figure 1 , Figure 1 A flowchart of a method for identifying a fracture surface of an object fragment according to an embodiment of the present application is shown. The present application embodiment provides an execution step of a method for identifying a fracture surface of an object fragment, including:

[0071] Step S110, collecting point clouds of object fragments to obtain initial fragment point clouds;

[0072] Step S120, taking the centroid of the initial fragment point cloud as the origin, recalculating the coordinate information of each data point in the initial fragment point cloud to obtain the fragment point cloud;

[0073] Step S130, taking each data point in the fragment point cloud as the center and the set distance as the radius, determining the neighboring points of the data point;

[0074] Step S140, determining the spatial distribution characteristics of the data point according to the relative position relationship between the data point and the neighboring points of the data point;

[0075] Step S150, according to the spatial distribution characteristics of each data point, the data points belonging to the fracture surface are determined as target data points among the data points, and the area formed by the target data points in the fragment point cloud represents the fracture surface.

[0076] The above five steps are described in detail below.

[0077] In step S110, point cloud collection is performed on the fragments of the object to obtain an initial point cloud of the fragments. The point cloud collection of the fragments of the object refers to determining multiple sampling points on the fragments of the object, and then using a measuring device to obtain the position information of these sampling points in a preset coordinate system. It should be clear that the preset coordinate system can be constructed with any point as the origin. Then, based on the position information of each sampling point, the corresponding data point is found in the virtual space, and the initial point cloud of the fragment can be obtained by materialization.

[0078] The data points corresponding to the sampling points in the virtual space can be found in the following ways.

[0079] In some embodiments, a data point having the same coordinate value as the sampling point may be searched in the virtual space as the data point corresponding to the sampling point in the virtual space.

[0080] In some embodiments, a sampling point can be arbitrarily selected as a selected point, and a point can be arbitrarily selected in the virtual space as a data point corresponding to the selected point. Based on the relative position relationship between the selected point and other sampling points, the corresponding data point is determined in the virtual space, and the initial fragment point cloud is obtained by materialization. The relative position relationship between each data point of the initial fragment point cloud is the same as the relative position relationship between the sampling points corresponding to the data point.

[0081] In step S120, the centroid of the initial fragment point cloud is calculated. Specifically, the average value of the coordinate information of all data points is obtained according to the coordinate values ​​of each data point in the initial fragment point cloud, which is the centroid of the initial fragment point cloud. Taking it as the origin and the relative position relationship between each data point, the coordinate information of each data point is recalculated, and the original coordinate information of the data point is replaced with the recalculated coordinate information to obtain the fragment point cloud.

[0082] In step S130, taking a data point as an example, the neighboring points of the data point are calculated for each data point in the following manner. A spherical space is formed with the data point as the center and the set distance as the radius. The data point itself as the center is removed from the spherical space, and the other data points are the neighboring points of the data point as the center.

[0083] In step S140, the spatial distribution characteristics of the data points are determined based on the relative positional relationship between the data points and their neighboring points. The reason for calculating the spatial distribution characteristics of the data points is that the data points belonging to the fracture surface and the data points not belonging to the fracture surface have different spatial distribution characteristics.

[0084] For example, for data points that do not belong to the fracture surface, since the area that does not belong to the fracture surface is relatively smooth, the data point and its neighboring points are almost distributed in the same plane, which will make the curvature of the area formed by the data point and its neighboring points change less. Or the coordinate value of at least one dimension between the data point and its neighboring points is relatively close. It should be noted that the coordinate information includes multiple coordinate values, each coordinate value refers to the value corresponding to the data point on a data axis, and each data axis corresponds to a dimension of the virtual space.

[0085] The spatial distribution characteristics of the data points belonging to the fracture surface are opposite to those of the data points not belonging to the fracture surface. Since there are many irregular areas in the fracture surface, it is difficult for the data point cloud belonging to the fracture surface and its neighboring points to be distributed in the same plane, which will cause the curvature of the area formed by the data point and its neighboring points to change greatly. Or there is no dimension with close coordinate values ​​between the data point and its neighboring points.

[0086] It should be further explained that the curvature change described above is only a description of the spatial distribution characteristics. It does not mean that the spatial distribution characteristics only include the curvature change of the area formed by the data point and its neighboring points. It is only for the purpose of illustrating that the spatial distribution characteristics can distinguish the data points belonging to the fracture surface from the data points not belonging to the fracture surface. The spatial distribution characteristics can be expressed in the form of various numerical values ​​or tables, which are obtained by calculating the characteristic values ​​based on the relative position relationship between the data point and its neighboring points.

[0087] In step S150, according to the spatial distribution characteristics of each data point, the target data point belonging to the fracture surface is determined in each data point, and the area formed by the target data point is based on the area representing the fracture surface in the fragment point cloud. Therefore, the user does not need to manually mark, but can directly perform point cloud registration on multiple fragment point clouds based on the area representing the fracture surface in the fragment point cloud, thereby providing an accurate reference for the restoration of cultural relics in the real world.

[0088] See also Figure 2 , Figure 2 A flowchart of converting an initial fragment point cloud into a fragment point cloud according to an embodiment of the present application is shown. The present application embodiment provides a step S120 of converting an initial fragment point cloud into a fragment point cloud, including:

[0089] Step S121, if the centroid of the fragment point cloud does not belong to the data point of the fragment point cloud, the centroid of the fragment point cloud is used as the data point of the fragment point cloud;

[0090] Step S122, subtracting the coordinate information of each data point from the coordinate information of the centroid to obtain new coordinate information of each data point;

[0091] Step S123, changing the coordinate information of each data point to new coordinate information, and obtaining a fragment point cloud with the centroid as the origin.

[0092] The above three steps are described in detail below.

[0093] In step S121, if the centroid of the fragment point cloud does not belong to the data point of the fragment point cloud, the centroid of the fragment point cloud is used as the data point of the fragment point cloud. This is to ensure that the coordinate information of the centroid also changes to the origin when the unified coordinate information of the data points is calculated in subsequent steps.

[0094] In step S122, the coordinate information of the centroid is subtracted from the coordinate information of each data point, and the coordinate information of the centroid as the origin is also subtracted from the coordinate information of the centroid, so that the coordinate information of the centroid as the origin becomes (0,0,0), which truly becomes the origin. At the same time, the coordinate information of the centroid is also subtracted from the coordinate information of other data points.

[0095] In step S123, the coordinate information of the data points is replaced with the new coordinate information of each data point obtained by calculation, so as to obtain a fragment point cloud.

[0096] See also Figure 3 , Figure 3 A flowchart of scaling a fragment point cloud according to an embodiment of the present application is shown. The present application embodiment provides steps for scaling a fragment point cloud, including:

[0097] Step S201, taking the data point farthest from the origin in the fragment point cloud as the reference data point, and the distance between the reference data point and the origin as the reference distance;

[0098] Step S202: taking the ratio of the reference distance to the preset target distance as a scaling factor, scaling the fragmented point cloud according to the scaling factor.

[0099] The above steps are described in detail below.

[0100] In step S201, the data point farthest from the origin in the fragment point cloud is first used as a reference data point, and the distance between the reference data point and the origin is the reference distance. The reference distance measures the size of the fragment point cloud. If the size of the fragment point cloud is too large, it means that the fracture surface may also be relatively large. Therefore, when calculating the local spatial distribution characteristics based on the neighboring points of the data point, it is necessary to determine the range of the neighboring points to be reset to avoid setting the distance too small, which will result in the inability to better express the local spatial distribution characteristics of the data point.

[0101] Therefore, in an embodiment of the present application, the origin of the fragment point cloud is firstly re-determined, the coordinate information of each data point is updated, and the fragment point cloud is scaled so that the fragment point cloud always maintains a set size.

[0102] In step S202, the ratio of the reference distance to the preset target distance is used as a scaling factor, and the fragment point cloud is scaled according to the scaling factor. If the fragment point cloud is too small, the fragment point cloud is enlarged, and if the fragment point cloud is too large, the fragment point cloud is reduced.

[0103] See also Figure 4 , Figure 4 A flowchart of scaling the fragment point cloud according to an embodiment of the present application using the ratio of the reference distance to the preset target distance as a scaling factor is shown. The present application embodiment provides a step S202 of scaling the fragment point cloud according to the scaling factor using the ratio of the reference distance to the preset target distance as a scaling factor, including:

[0104] Step S2021, multiplying the coordinate value of the coordinate information of the data point on each axis by the zoom factor to obtain the target coordinate information of the data point;

[0105] Step S2022, moving the data point to the location of the target coordinate information.

[0106] The above two steps are described in detail below.

[0107] In step S2021, because the fragment point cloud takes the centroid as the origin, the coordinate value of the coordinate information of the data point on each axis can be directly multiplied by the scaling factor to obtain the target coordinate information of the data point.

[0108] In step S2022, the data point is moved to the location of the target coordinate information.

[0109] In some embodiments, the straight line between the data point and the origin is used as the scaling trajectory, the distance between the data point and the origin is obtained, and the product of the distance between the data point and the origin and the scaling factor is calculated to obtain the target distance. The origin is used as the starting point, and the position on the scaling trajectory that is the target distance from the origin is used as the target position, and the data point is moved to the target position. By moving each data point to the target position, the scaling of the fragmented point cloud can be achieved.

[0110] In some embodiments, the initial fragment point cloud is transformed into a fragment point cloud and scaled using the following formula.

[0111] First, find the centroid of the initial fragment point cloud:

[0112]

[0113] Among them, c i refers to the coordinate information of the data point in the initial fragment point cloud, i refers to the number of data points, N refers to the number of data points in the initial fragment point cloud, pc centroidIt refers to the centroid of the initial fragment point cloud.

[0114] Secondly, the initial fragment point cloud is transformed into a fragment point cloud with the centroid as the origin.

[0115] PC normalized =C-pc centroid

[0116] Among them, PC normalized refers to the fragment point cloud, and C refers to the initial fragment point cloud.

[0117] Next, calculate the zoom factor:

[0118]

[0119] Among them, Norm max refers to the zoom factor, L t is the target distance, It refers to the distance between the data point farthest from the origin in the fragment point cloud and the origin. x, y, z refer to the coordinate values ​​of the data point farthest from the origin in the fragment point cloud on each coordinate axis.

[0120] Finally, scale the fragmented point cloud:

[0121]

[0122] See also Figure 5 , Figure 5 A flow chart of determining the spatial distribution characteristics of a data point according to the relative position relationship between the data point and its neighboring points according to an embodiment of the present application is shown. The present application embodiment provides a step S140 of determining the spatial distribution characteristics of a data point according to the relative position relationship between the data point and its neighboring points, including:

[0123] Step S141, arbitrarily selecting a data point as a query point, and taking data points whose distance to the query point is less than a set distance as neighboring points of the query point;

[0124] Step S142, arbitrarily selecting a neighboring point of the query point as a feature reference point of the query point, and calculating the local spatial distribution feature of the query point according to the relative position relationship between the query point and the feature reference point;

[0125] Step S143, obtaining local spatial distribution features of the query point when the neighboring points are used as feature reference points of the query point respectively;

[0126] Step S144: Integrate all local spatial distribution features of the query point to obtain the spatial distribution features of the query point.

[0127] The above four steps are described in detail below.

[0128] In step S141, a data point is randomly selected as a query point, and data points whose distance to the query point is less than a set distance are regarded as neighboring points of the query point. That is, data points other than the query point in a spherical space formed with the query point as the center and the set distance as the radius are the neighboring points of the query point.

[0129] It should be noted that the spatial distribution characteristics of each data point in the fragment point cloud are determined using the method described in steps S141 to S144.

[0130] In step S142, a neighboring point of the query point is randomly selected as a feature reference point of the query point, and the local spatial distribution feature of the query point is calculated according to the relative position relationship between the query point and the feature reference point, that is, there are as many local spatial distribution features as there are neighboring points of the query point.

[0131] In step S143, each local spatial distribution feature of the query point is obtained when the neighboring points are respectively used as feature reference points of the query point.

[0132] In step S144, all local spatial distribution features of the query point are integrated to obtain the spatial distribution features of the query point. The integration method can be simple addition, or it can be multiplied with each local spatial distribution feature according to a preset weight coefficient, and finally the obtained products are added.

[0133] See also Figure 6 , Figure 6 A flowchart of calculating the local spatial distribution features of a query point according to the relative position relationship between the query point and the feature reference point according to an embodiment of the present application is shown. The embodiment of the present application provides a step S142 of calculating the local spatial distribution features of the query point according to the relative position relationship between the query point and the feature reference point, including:

[0134] Step S1421, taking the vector formed between the query point and the feature reference point as the first vector, taking the normal vector of the query point as the second vector, and taking the normal vector of the feature reference point as the third vector;

[0135] Step S1422, taking a vector perpendicular to the plane where the first vector and the second vector are located and passing through the query point as a fourth vector;

[0136] Step S1423, taking a vector perpendicular to the plane where the second vector and the fourth vector are located and passing through the query point as the fifth vector;

[0137] Step S1424, translating the second vector, the fourth vector, and the fifth vector to obtain a second translation vector, a fourth translation vector, and a fifth translation vector, and the intersection of the second translation vector, the third translation vector, and the fourth translation vector is a feature reference point;

[0138] Step S1425, taking the angle between the first vector and the second vector as the first angle, taking the angle between the fourth translation vector and the third vector as the second angle, taking the angle formed by the third vector and the projection line as the third angle, and the projection of the third vector on the fifth translation vector and the second translation vector as the projection line;

[0139] Step S1426, calculating a local feature histogram between the query point and the feature reference point according to the first angle, the second angle, and the third angle, that is, a local spatial distribution feature.

[0140] The above six steps are described in detail below.

[0141] See also Figure 7 , Figure 7 A schematic diagram of the distribution of various vectors according to an embodiment of the present application is shown.

[0142] In step S1421, a vector formed between the query point and the feature reference point is used as a first vector, a normal vector of the query point is used as a second vector, and a normal vector of the feature reference point is used as a third vector;

[0143] The normal vector of the query point is calculated by performing plane fitting on the query point and its neighboring points to obtain a fitting plane, and taking the vector perpendicular to the fitting plane and passing through the query point as the normal vector, i.e., the second vector.

[0144] For the calculation method of the normal vector of the feature reference point, first, with the feature reference point as the center and the set distance as the radius, the neighboring points of the feature reference point are determined, and a fitting plane is obtained by plane fitting based on the feature reference point and the neighboring points of the feature reference point. The vector perpendicular to the fitting plane and passing through the feature reference point is used as the normal vector of the feature reference point, that is, the third vector.

[0145] In step S1422, a vector perpendicular to the plane where the first vector and the second vector are located and passing through the query point is taken as a fourth vector;

[0146] In step S1423, a vector perpendicular to the plane where the second vector and the fourth vector are located and passing through the query point is taken as the fifth vector;

[0147] In step S1424, the second vector, the fourth vector and the fifth vector are translated to obtain a second translation vector, a fourth translation vector and a fifth translation vector, and the intersection of the second translation vector, the third translation vector and the fourth translation vector is a feature reference point;

[0148] In step S1425, the angle between the first vector and the second vector is taken as the first angle, the angle between the fourth translation vector and the third vector is taken as the second angle, the angle formed by the third vector and the projection line is taken as the third angle, and the projection of the third vector on the fifth translation vector and the second translation vector is taken as the projection line;

[0149] In step S1426, a local feature histogram, that is, a local spatial distribution feature, between the query point and the feature reference point is calculated according to the first angle, the second angle, and the third angle.

[0150] For the construction of local feature histogram, it is exemplary:

[0151] 1. Assign each angle value to the bins of the histogram:

[0152] -Map α, φ, and θ to their respective 11 bins.

[0153] -For example, assuming that the range of α is [0,π], we divide it into 11 intervals (bins), each interval corresponds to a bin.

[0154] 2. Statistical frequency:

[0155] -For the query point and feature reference point, count the intervals in which the above three angles fall, and the bins in which the α, φ, and θ values ​​fall, increase the counts of the corresponding bins, and obtain the local feature histogram, which is the local spatial distribution feature, used to indicate the relative position of the feature reference point to the query point.

[0156] See also Figure 8 , Figure 8 The flowchart of integrating all local spatial distribution features of the query point to obtain the spatial distribution features of the query point according to one embodiment of the present application is shown. The present application embodiment provides a step S1426 of integrating all local spatial distribution features of the query point to obtain the spatial distribution features of the query point according to one embodiment of the present application, including:

[0157] Step S301, taking the inverse of the distance between each feature reference point and the query point as the feature weight of each feature reference point;

[0158] Step S302: multiply the feature weight of each feature reference point by the local spatial distribution feature corresponding to each feature reference point, integrate the obtained products, and obtain the spatial distribution feature of the query point.

[0159] The above two steps are described in detail below.

[0160] In step S301, the inverse of the distance between each feature reference point and the query point is used as the feature weight of each feature reference point. This is because the local spatial distribution features corresponding to the feature reference points that are closer to the query point have greater reference significance to the query point. Therefore, the feature weight corresponding to the feature reference points that are closer to the query point is greater.

[0161] In step S302, the feature weight of each feature reference point is multiplied by the local spatial distribution feature corresponding to each feature reference point, and the obtained products are integrated to obtain the spatial distribution feature of the query point.

[0162] See also Fig. 9 , Fig. 9 A flowchart of integrating all local spatial distribution features of a query point to obtain the spatial distribution features of the query point according to another embodiment of the present application is shown. The present application embodiment provides a step S1426 of integrating all local spatial distribution features of the query point to obtain the spatial distribution features of the query point, including:

[0163] Step S401, integrating all local spatial distribution features of the query point to obtain the initial spatial distribution features of the query point;

[0164] Step S402, calculating the initial spatial distribution characteristics of each neighboring point as the neighboring initial spatial distribution characteristics;

[0165] Step S403, calculating the feature weight corresponding to each neighboring point according to the distance between the neighboring point and the query point;

[0166] Step S404, multiplying the neighboring initial spatial distribution feature of each neighboring point by the feature weight corresponding to the neighboring point, and adding the sum of the products to the initial spatial distribution feature to obtain the spatial distribution feature.

[0167] The above four steps are described in detail below.

[0168] In step S401, all local spatial distribution features of the query point are integrated to obtain the initial spatial distribution features of the query point. Alternatively, all local spatial features may be transformed into feature values ​​and then integrated to obtain the initial spatial distribution features.

[0169] In some embodiments, the local spatial distribution characteristics are obtained in the following manner.

[0170] 1. Assign each angle value to the bins of the histogram:

[0171] -Map α, φ, and θ to their respective 11 bins.

[0172] -For example, assuming that the range of α is [0,π], we divide it into 11 intervals (bins), each interval corresponds to a bin.

[0173] 2. Statistical frequency:

[0174] -For the query point and feature reference point, count the intervals in which the above three angles fall, and the bin in which the α, φ, and θ values ​​fall, increase the count of the corresponding bin, and obtain the local feature histogram, which is also the local spatial distribution feature.

[0175] The frequencies of the same bins in each local feature histogram are superimposed to obtain the initial feature histogram and the initial spatial distribution features.

[0176] In step S402, the initial spatial distribution features of each neighboring point are calculated in the same manner as the query, and are the neighboring initial spatial distribution features.

[0177] Among them, the characteristics of the neighboring initial spatial distribution are also represented by feature histograms.

[0178] In step S403, the feature weight corresponding to each neighboring point is calculated according to the distance between the neighboring point and the query point. The closer the neighboring point is to the query point, the greater the feature weight corresponding to the neighboring point is, because it has a greater impact on the spatial distribution characteristics of the query point.

[0179] In step S404, the neighboring initial spatial distribution feature of each neighboring point is multiplied by the feature weight corresponding to the neighboring point, and the sum of the products is added to the initial spatial distribution feature to obtain the spatial distribution feature.

[0180] In some embodiments, the frequency in each bin in the feature histogram corresponding to the adjacent initial spatial feature is multiplied by the feature weight, and then the frequency in each bin multiplied by the feature weight is superimposed with the frequency in the bin in the initial feature histogram, and finally the feature histogram of the query point, i.e., the spatial distribution feature, is obtained. It is used to represent the distribution of the neighboring points of the query point around the query point, and then determine whether the query point belongs to the fracture surface based on the distribution characteristics of the neighboring points of the data points located on the fracture surface and the data points not located on the fracture surface.

[0181] For example, for data points on non-fracture surfaces, if the neighboring points are only distributed on the same plane, the corresponding feature histogram may only indicate distribution up, down, left, and right, while for data points on fracture surfaces, the corresponding feature histogram may correspond to high, low, up, down, left, and right. Therefore, it is possible to determine whether the data is on a fracture surface based on the feature histogram.

[0182] In some embodiments, the spatial feature distribution of the query point may be calculated using the following formula.

[0183] First calculate the second vector, the fourth vector, and the fifth vector:

[0184]

[0185] Among them, because the query point can be any data point, the c of the i-th data point is represented by i represents the query point, c ij Indicates that the jth neighboring point corresponding to the query point is used as the feature reference point of the query point. i is the normal vector of the query point, that is, the second vector u ij , v ij is the fourth vector, w ij is the fifth vector. ij is the normal vector of the feature reference point, that is, the third vector, and × represents the outer product.

[0186] Calculate the local spatial distribution characteristics of the query point based on the current feature reference point according to the first angle, the second angle, and the third angle:

[0187]

[0188] Among them, α ij is the second angle, φ ij is the first angle, θ ij is the third angle, where d ij =||c i -c ij ||2 indicates the query point c i and feature reference point c ij The Euclidean distance between them, "." represents the inner product. α and Φ are both dot products between vectors, that is, angles. Their actual physical meaning is the cosine value of the angle between 3D vectors, so the value range is ±1. θ ij Yes i and n ij The angle formed by projecting the third vector onto the plane formed by the fifth translation vector and the second translation vector has a value range of ±π. The three parameters α, Φ, and θ solved by the formula constitute the local spatial distribution characteristics.

[0189] Finally, the local spatial distribution characteristics of each neighboring point when used as a feature reference point are calculated and integrated through the following formula to obtain the spatial distribution characteristics of the query point.

[0190] First, calculate the feature weights of each local spatial distribution feature:

[0191]

[0192] Among them, T ij Indicates c ij The feature weight of the local spatial distribution feature obtained when it is used as the feature reference point of the query point. K represents the number of feature reference points.

[0193] Secondly, all query points c i All local spatial distribution features are integrated to obtain the spatial distribution features of the query point.

[0194]

[0195] Among them, FPFH(c i ) refers to the spatial distribution characteristics of the query points, SPFH(c i ) refers to the initial local spatial distribution characteristics of the query point. SPFH(c ij ) refers to the neighboring initial local spatial features of the neighboring points of the query point.

[0196] In some embodiments, the spatial distribution characteristics of the data point and the coordinate information of the data point are input into a preset judgment neural network to determine whether the data point is a target data point. Compared with directly inputting the coordinate information of the data point into the judgment neural network, the spatial distribution characteristics of the data point and the coordinate information of the data point are input at the same time with higher accuracy.

[0197] The judgment neural network outputs the probability that the data point belongs to the fracture surface and the probability that it does not belong to the fracture surface according to the spatial distribution characteristics of the data input and the coordinate information of the data point, and takes the result with a larger probability as the judgment result of the data point.

[0198] In some embodiments, a neural network may be trained in the following manner.

[0199] First calculate the loss function:

[0200]

[0201] Among them, p i It refers to the accuracy of the neural network when testing the i-th sample point. The output of the neural network is the probability that the sample point belongs to the fracture surface category or the non-fracture surface category. (s i，y ) represents the probability that the sample point belongs to the category indicated by y. For example, the result of the neural network output is: the probability of belonging to the fracture surface is 0.8, and the probability of belonging to the non-fracture surface is 0.4. Then the correct result of the data point is that it belongs to the fracture surface, so (s i，y ) means 0.8, which is the probability of the correct output result. It represents the sum of the probabilities that the sample point belongs to the fracture surface category or the non-fracture surface category. Referring to the example above, it is 1.2.

[0202] Loss i =-logp i

[0203] According to p i Calculate sample points s i Damage function.

[0204]

[0205] Where B is the number of all sample points.

[0206] According to the loss functions corresponding to all sample points, the total damage function is calculated.

[0207]

[0208] According to the final loss function, the final loss function helps the judgment neural network to adjust the weights through the back propagation algorithm to increase the accuracy of the judgment neural network.

[0209] See also Fig.10 , Fig.10 A device for identifying fracture surfaces of object fragments according to an embodiment of the present application is shown, comprising:

[0210] The acquisition module is used to collect point clouds of object fragments to obtain initial fragment point clouds;

[0211] An updating module is used to recalculate the coordinate information of each data point in the initial fragment point cloud with the centroid of the initial fragment point cloud as the origin to obtain the fragment point cloud;

[0212] The neighboring module is used to determine the neighboring points of each data point in the fragment point cloud with each data point as the center and a set distance as the radius;

[0213] A feature module is used to determine the spatial distribution characteristics of the data point based on the relative position relationship between the data point and the neighboring points of the data point;

[0214] The screening module is used to determine the data points belonging to the fracture surface as target data points among the data points according to the spatial distribution characteristics of the data points, and the area formed by the target data points in the fragment point cloud represents the fracture surface.

[0215] The fracture surface identification method of the object fragments in the embodiment of the present application can be performed by Fig.11 The hardware device shown is implemented. Fig.11 To describe the hardware device according to the embodiment of the present application. Fig.11 The hardware device shown is only an example and should not bring any limitation to the functions and scope of application of the embodiments of the present application.

[0216] like Fig.11 As shown, the hardware device may be in the form of a general computing device. The components of the hardware device may include, but are not limited to: at least one processing unit 810, at least one storage unit 820, and a bus 830 connecting different system components (including the storage unit 820 and the processing unit 810).

[0217] The storage unit stores program codes, which can be executed by the processing unit 810, so that the processing unit 810 performs the steps according to various exemplary embodiments of the present invention described in the description of the exemplary method described above in this specification. For example, the processing unit 810 can perform the following steps: Figure 1 The steps shown in .

[0218] The storage unit 820 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 8201 and / or a cache memory unit 8202 , and may further include a read-only memory unit (ROM) 8203 .

[0219] The storage unit 820 may also include a program / utility 8204 having a set (at least one) of program modules 8205, such program modules 8205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0220] Bus 830 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0221] The hardware device may also communicate with one or more external devices 700 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may communicate with one or more devices that enable a user to interact with the hardware device, and / or may communicate with any device that enables the hardware device to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 850. Furthermore, the hardware device may also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 860. As shown, the network adapter 860 communicates with other modules via a bus 830. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the hardware device, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0222] Through the description of the above implementation methods, it is easy for those skilled in the art to understand that the example implementation methods described here can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the implementation methods of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the implementation methods of the present application.

[0223] In one implementation of the present application, a storage medium for carrying the methods described in the various embodiments described above is also provided. The storage medium shown in the embodiment of the present application may be a readable signal medium or a readable storage medium or any combination of the above two. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a readable storage medium may be any tangible medium containing or storing a program, which may be used by an instruction execution system, device or device or used in combination with it. In the present application, a readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, wherein a readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The readable signal medium may also be any storage medium other than a readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the storage medium may be transmitted using any suitable medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0224] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the above-mentioned module, program segment or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0225] It should be noted that, although several modules or units of the equipment for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into being embodied by multiple modules or units.

[0226] Through the description of the above implementation methods, it is easy for those skilled in the art to understand that the example implementation methods described here can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the implementation methods of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the implementation methods of the present application.

[0227] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary technical means in the art that are not disclosed in the present application.

[0228] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A method for identifying fracture surfaces of object fragments, characterized in that: The method comprises: Collect point clouds of object fragments to obtain initial fragment point clouds; Taking the centroid of the initial fragment point cloud as the origin, recalculating the coordinate information of each data point in the initial fragment point cloud to obtain the fragment point cloud; Taking each data point in the fragment point cloud as the center and a set distance as the radius, determine the neighboring points of the data point; Determining the spatial distribution characteristics of the data point according to the relative position relationship between the data point and the neighboring points of the data point; According to the spatial distribution characteristics of each of the data points, the data points belonging to the fracture surface are determined as target data points among the data points, and the area formed by the target data points in the fragment point cloud represents the fracture surface.

2. The method according to claim 1, characterized in that Taking the centroid of the initial fragment point cloud as the origin, recalculating the coordinate information of each data point in the initial fragment point cloud to obtain the fragment point cloud, including: If the centroid of the fragment point cloud does not belong to the data point of the fragment point cloud, the centroid of the fragment point cloud is used as the data point of the fragment point cloud; Subtract the coordinate information of each data point from the coordinate information of the centroid to obtain new coordinate information of each data point; The coordinate information of each of the data points is changed to new coordinate information to obtain a fragment point cloud with the centroid as the origin.

3. The method according to claim 2, characterized in that: After subtracting the coordinate information of each data point from the coordinate information of the centroid to obtain new coordinate information of each data point, the method further includes: The data point farthest from the origin in the fragment point cloud is used as a reference data point, and the distance between the reference data point and the origin is a reference distance; The ratio of the reference distance to the preset target distance is used as a scaling factor, and the fragment point cloud is scaled according to the scaling factor.

4. The method according to claim 3, characterized in that Scaling the fragmented point cloud according to the scaling factor includes: Multiplying the coordinate values ​​of the coordinate information of the data point on each axis by the scaling factor to obtain the target coordinate information of the data point; Move the data point to the location of the target coordinate information.

5. The method according to claim 1, characterized in that Determining the spatial distribution characteristics of the data point according to the relative position relationship between the data point and the neighboring points of the data point includes: Randomly select a data point as a query point, and use data points whose distance to the query point is less than a set distance as neighboring points of the query point; arbitrarily selecting a neighboring point of the query point as a feature reference point of the query point, and calculating the local spatial distribution feature of the query point according to the relative position relationship between the query point and the feature reference point; When obtaining neighboring points as feature reference points of the query point respectively, each local spatial distribution feature of the query point; All local spatial distribution features of the query point are integrated to obtain the spatial distribution features of the query point.

6. The method according to claim 5, characterized in that Calculating the local spatial distribution feature of the query point according to the relative position relationship between the query point and the feature reference point includes: Using the vector formed between the query point and the feature reference point as the first vector, using the normal vector of the query point as the second vector, and using the normal vector of the feature reference point as the third vector; A vector which is perpendicular to the plane where the first vector and the second vector are located and passes through the query point is taken as a fourth vector; A vector perpendicular to the plane where the second vector and the fourth vector are located and passing through the query point is used as a fifth vector; The second vector, the fourth vector and the fifth vector are translated to obtain a second translation vector, a fourth translation vector and a fifth translation vector, wherein the intersection of the second translation vector, the third translation vector and the fourth translation vector is the feature reference point; The angle between the first vector and the second vector is taken as the first angle, the angle between the fourth translation vector and the third vector is taken as the second angle, the angle formed by the third vector and the projection line is taken as the third angle, and the projection of the third vector on the fifth translation vector and the second translation vector is taken as the projection line; A local feature histogram, that is, a local spatial distribution feature, between the query point and the feature reference point is calculated according to the first angle, the second angle, and the third angle.

7. The method according to claim 5, characterized in that All local spatial distribution features of the query point are integrated to obtain the spatial distribution features of the query point, including: The inverse of the distance between each of the feature reference points and the query point is used as the feature weight of each of the feature reference points; The feature weight of each feature reference point is multiplied by the local spatial distribution feature corresponding to each feature reference point, and the obtained products are integrated to obtain the spatial distribution feature of the query point.

8. The method according to claim 5, characterized in that Integrating all local spatial distribution features of the query point to obtain the spatial distribution features of the query point, further comprising: Integrating all local spatial distribution features of the query point to obtain the initial spatial distribution features of the query point; Calculate the initial spatial distribution characteristics of each neighboring point as the neighboring initial spatial distribution characteristics; Calculate the feature weight corresponding to each neighboring point according to the distance between the neighboring point and the query point; The neighboring initial spatial distribution feature of each of the neighboring points is multiplied by the feature weight corresponding to the neighboring point, and the sum of the products is added to the initial spatial distribution feature to obtain the spatial distribution feature.

9. The method according to claim 1, characterized in that: According to the spatial distribution characteristics of each of the data points, determining the data points belonging to the fracture surface among the data points as target data points includes: The spatial distribution characteristics of the data point and the coordinate information of the data point are input into a preset judgment neural network to determine whether the data point is a target data point.

10. A device for identifying fracture surfaces of object fragments, characterized in that: The device comprises: The acquisition module is used to collect point clouds of object fragments to obtain initial fragment point clouds; An updating module, configured to recalculate the coordinate information of each data point in the initial fragment point cloud with the centroid of the initial fragment point cloud as the origin, so as to obtain the fragment point cloud; A neighboring module, for determining neighboring points of the data points with each data point in the fragment point cloud as the center and a set distance as the radius; A feature module, used to determine the spatial distribution feature of the data point according to the relative position relationship between the data point and the neighboring points of the data point; The screening module is used to determine the data points belonging to the fracture surface among the data points as target data points according to the spatial distribution characteristics of the data points, and the area formed by the target data points in the fragment point cloud represents the fracture surface.

Images