An interference filtering method, device and electronic equipment

Abstract

Embodiments of the present application provide an interference filtering method and device and electronic equipment. In the embodiments, the interference in the target stockpile point cloud of the specified area obtained by scanning is filtered by using the geometric filter model matched in advance for the interference of the specified area, the interference is filtered, and the statistical accuracy of the object in the specified area is improved. Further, by matching the combination of different geometric filter models in advance for the interference of the specified area, different areas can be adapted, the statistical requirements of different areas can be met, and the interference filtering of complex areas can be realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of radar detection, and in particular to an interference filtering method and device and electronic equipment. BACKGROUND

[0002] In intelligent manufacturing applications, it is often necessary to count objects in a specified area. For example, in a production plant, it is often necessary to count the volume, weight, and other parameters of material piles (used for stacking materials), which can also be referred to simply as material counting.

[0003] Currently, when counting objects in a specified area, the objects are first scanned to obtain a scanning point cloud, and then the counting parameters such as volume, weight, and the like are determined based on the scanning point cloud. Taking the above production plant as an example, each material pile can be scanned to obtain a scanning point cloud, and the volume, weight, and other parameters of the material piles can be determined based on the scanning point cloud of each material pile.

[0004] However, when counting objects in a specified area based on a scanning point cloud, there are often many interferences, such as the interferences of steel frames, columns, bucket wheel machines, and stacker-reclaimers in the above production plant, which can affect the accuracy of the counting. SUMMARY

[0005] Therefore, embodiments of the present application provide an interference filtering method, device, and electronic equipment to improve the accuracy of counting objects in a specified area by filtering interference.

[0006] According to a first aspect of embodiments of the present application, an interference filtering method is provided, which is applied to an electronic device, and the method comprises:

[0007] Obtaining target material pile point clouds scanned by each scanning device deployed in a specified area; the target material pile point cloud corresponds to a current material pile in the specified area; at least one target material pile point cloud contains the positions of materials in the current material pile corresponding to the target material pile point cloud and the positions of points on at least one interference object in the specified area;

[0008] Mapping each target material pile point cloud to the same target coordinate system to obtain a first mapping point cloud; the first mapping point cloud includes the positions of materials in each current material pile and the positions of points on all interference objects in the specified area;

[0009] obtain the filtering parameters of at least one calibrated geometric filter model for filtering the interference objects in the specified area; the geometric filter model is determined based on a second mapped point cloud, the second mapped point cloud is obtained by mapping reference stockpile point clouds scanned by the scanning devices to the target coordinate system; the reference stockpile point clouds correspond to historical stockpiles in the specified area, at least one reference stockpile point cloud contains the positions of materials in the historical stockpile corresponding to the reference stockpile point cloud and the positions of points on at least one interference object in the specified area, and the second mapped point cloud includes the positions of materials in each historical stockpile and the positions of points on each interference object; the filtering parameters of any geometric filter model are used to adjust the pose of the geometric filter model to filter at least one interference object in the specified area;

[0010] adjust the poses of the geometric filter models in the first mapped point cloud according to the filtering parameters of the geometric filter models, to filter the interference object point cloud in the first mapped point cloud, the interference object point cloud is composed of the positions of points on all interference objects in the specified area.

[0011] According to a second aspect of the embodiments of the present application, an interference filtering device is provided, which comprises:

[0012] a target stockpile point cloud obtaining module, configured to obtain target stockpile point clouds scanned by scanning devices deployed in a specified area; the target stockpile point clouds correspond to current stockpiles in the specified area; at least one target stockpile point cloud contains the positions of materials in the current stockpile corresponding to the target stockpile point cloud and the positions of points on at least one interference object in the specified area;

[0013] a mapping module, configured to map each target stockpile point cloud to the same target coordinate system to obtain the first mapped point cloud; the first mapped point cloud includes the positions of materials in each current stockpile and the positions of points on all interference objects in the specified area;

[0014] a filtering parameter obtaining module, configured to obtain the filtering parameters of at least one calibrated geometric filter model for filtering the interference objects in the specified area; the geometric filter model is determined based on a second mapped point cloud, the second mapped point cloud is obtained by mapping reference stockpile point clouds scanned by the scanning devices to the target coordinate system; the reference stockpile point clouds correspond to historical stockpiles in the specified area, at least one reference stockpile point cloud contains the positions of materials in the historical stockpile corresponding to the reference stockpile point cloud and the positions of points on at least one interference object in the specified area, and the second mapped point cloud includes the positions of materials in each historical stockpile and the positions of points on each interference object; the filtering parameters of any geometric filter model are used to adjust the pose of the geometric filter model to filter at least one interference object in the specified area;

[0015] a filtering module, configured to adjust poses of the geometric filter models in the first mapping point cloud according to filtering parameters of the geometric filter models, so as to filter an interference point cloud in the first mapping point cloud, the interference point cloud being composed of positions of points on all interferences in the specified area.

[0016] According to a third aspect of the embodiments of the present application, an electronic device is provided, which includes a processor and a memory.

[0017] The memory is configured to store machine executable instructions.

[0018] The processor is configured to read and execute the machine executable instructions stored in the memory, so as to implement the method according to the first aspect.

[0019] The technical solutions provided by the embodiments of the present application can include the following beneficial effects:

[0020] In the embodiments of the present application, the interference in the target stockpile point cloud of the specified area obtained by scanning is filtered by the geometric filter model matched in advance for the interference of the specified area, so that the interference is filtered, and the statistical accuracy of the objects in the specified area is improved.

[0021] Further, by matching combinations of different geometric filter models in advance for the interference of the specified area, different areas can be adapted, the statistical needs of different areas can be met, and the interference filtering of complex areas can be implemented. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 is a method flowchart provided by the embodiments of the present application.

[0023] Figure 2 is an example diagram of the geometric filter model provided by the embodiments of the present application.

[0024] Figure 3 is an example diagram of the pose adjustment provided by the embodiments of the present application.

[0025] Figure 4 is a block diagram of at least one geometric filter model provided by the embodiments of the present application.

[0026] Figure 5 is an example diagram of the geometric filter model selection provided by the embodiments of the present application.

[0027] Figure 6 is another example diagram of the geometric filter model selection provided by the embodiments of the present application.

[0028] Figure 7 is a filtering result determination flowchart provided by the embodiments of the present application.

[0029] Figure 8 is an example diagram of projection transformation provided by an embodiment of the present application.

[0030] Figure 9 is an example diagram of circumscribed rectangle filtering corresponding to model shape provided by an embodiment of the present application.

[0031] Figure 10 is an example diagram of model shape filtering provided by an embodiment of the present application.

[0032] Figure 11 is an example diagram of model internal point cloud determination provided by an embodiment of the present application.

[0033] Figure 12 is an apparatus diagram provided by an embodiment of the present application.

[0034] Figure 13 is an apparatus hardware structure schematic diagram provided by an embodiment of the present application. DETAILED DESCRIPTION

[0035] The exemplary embodiments will be described in detail herein with reference to the attached drawings. The description of the exemplary embodiments is intended to apply to various alternative embodiments as well. The following description is not limited to the exemplary embodiments, but rather, is applicable to any apparatus and method consistent with the present application as described in the claims.

[0036] The terminology used in the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the present application. As used in the description of the embodiments and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It also will be understood that the term “and / or” as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0037] It should be understood that although the terms first, second, third, etc. can be used herein to describe various information, these terms are not intended to denote a particular order or hierarchy. For example, a first information can be termed a second information, and similarly, a second information can be termed a first information, without departing from the scope of the present application. As used herein, the term “if’ can be construed to mean “when” or “in response to determining” or “in response to a determination” depending on the context.

[0038] In order to better understand the technical solutions provided by the embodiments of the present application by the skilled in the art, and to make the above-mentioned purposes, features and advantages of the embodiments of the present application more obvious and easy to understand, the technical solutions in the embodiments of the present application will be further described in detail below with reference to the drawings.

[0039] The method provided by the embodiments of the present application will be described below:

[0040] Referring to Figure 1 , Figure 1 The method flowchart provided by the embodiments of the present application. The method is applied to an electronic device, for example, a terminal, a server, etc., and the embodiments of the present application are not specifically limited.

[0041] As Figure 1 shown, the flowchart can include the following steps:

[0042] S110: Obtain each target stockpile point cloud scanned by each scanning device deployed in the specified area.

[0043] Exemplarily, in the embodiments of the present application, the specified area can be a production plant, and at least one stockpile is stacked in the production plant. Here, the stockpile can be a coal stockpile, a cement stockpile, a sand and stone stockpile, etc., and the embodiments of the present application are not specifically limited.

[0044] In the embodiments of the present application, the scanning device can be deployed on the top of the production plant. Specifically, a scanning device can be deployed at a specified distance (for example, 5m), or a scanning device can be deployed above each stockpile. Here, the scanning device can be a radar imaging device, and the embodiments of the present application are not specifically limited.

[0045] In the embodiments of the present application, the target stockpile point cloud corresponds to the current stockpile in the specified area. The target stockpile point cloud can correspond to one current stockpile, or can correspond to two current stockpiles, and the embodiments of the present application are not specifically limited. The number of current stockpiles in the field of view of the scanning device scanning the target point cloud can be determined.

[0046] At least one target stockpile point cloud in the target stockpile point cloud includes the position of the material in the current stockpile corresponding to the target stockpile point cloud and the position of each point on at least one interference object in the specified area. Here, the position of the current material refers to the coordinates of the material in the scanning device coordinate system. Similarly, the position of each point on the interference object refers to the coordinates of each point on the interference object in the scanning device coordinate system.

[0047] Exemplarily, in this step S110, each scanning device simultaneously scans the stockpile in the specified area to obtain each target stockpile point cloud.

[0048] S120: Map each target stockpile point cloud to the same target coordinate system to obtain a first mapped point cloud.

[0049] Exemplarily, in the embodiment, taking the scanning device as a radar imaging device as an example, the coordinate system in which the target stockpile point cloud is located is a radar coordinate system, in this case, the target coordinate system can be another coordinate system different from the radar coordinate system, such as a world coordinate system, and the like, which is not limited in the embodiment of the application.

[0050] In the embodiment, the mapping of each target stockpile point cloud to the same target coordinate system to obtain the first mapping point cloud can adopt a conventional radar calibration algorithm, which will not be described herein.

[0051] Exemplarily, in the embodiment, the first mapping point cloud includes the positions of the materials in each current stockpile and the positions of the points on all the interference objects in the specified area, where the position of the material in the stockpile is the coordinate of the material in the target coordinate system, and similarly, the position of the point on the interference object is also the coordinate of the point on the interference object in the target coordinate system.

[0052] In the embodiment, in step S120, the mapping of each target stockpile point cloud to the same target coordinate system to obtain the first mapping point cloud realizes the splicing processing of each target stockpile point cloud.

[0053] S130: Obtain the filtering parameter corresponding to the at least one geometric filter model for filtering the interference objects in the specified area.

[0054] Exemplarily, in the embodiment, the geometric filter model is calibrated based on the second mapping point cloud. Here, the second mapping point cloud is obtained by mapping the reference stockpile point clouds scanned by each scanning device to the target coordinate system; the reference stockpile point cloud corresponds to a historical stockpile in the specified area, and at least one reference stockpile point cloud includes the positions of the materials in the historical stockpile corresponding to the reference stockpile point cloud and the positions of the points on at least one interference object in the specified area, where the position of the material refers to the coordinate of the material in the scanning device coordinate system, and similarly, the position of the point on the interference object refers to the coordinate of the point on the interference object in the scanning device coordinate system.

[0055] Exemplarily, in the embodiment, the historical stockpile is the stockpile corresponding to the stockpile point cloud scanned by each scanning device deployed in the specified area for the first time. It should be noted that when the layout or the number of the interference objects in the specified area changes, the stockpile point cloud scanned by each scanning device deployed in the specified area will be obtained again, at this time, the stockpile corresponding to the stockpile point cloud obtained for the first time after the change is taken as the historical stockpile. Wherein, the change in the layout of the interference object can be a change in the position of the existing interference object; the change in the number of the interference object can be the addition or reduction of at least one interference object.

[0056] In this embodiment, the second mapping point cloud includes: the position of the material in each historical material pile and the position of each point on each interference object. Here, the position of the material in the historical material pile is the coordinate of the material in the target coordinate system. Similarly, the position of each point on the interference object is also the coordinate of each point on the interference object in the target coordinate system.

[0057] For details on how to calibrate the geometric filter model, please refer to the examples described in the following embodiments, which will not be repeated here.

[0058] For example, in this embodiment, the geometric filter model described above can be of many types, such as Figure 2 The cone model, cylinder model, sphere model, cuboid model, etc. shown in the embodiments of this application are not specifically limited.

[0059] In this embodiment, each geometric filter model is configured with corresponding model parameters, such as Figure 2 As shown, for example, for a cone model, the corresponding model parameters may include the coordinates of the cone's vertices, the coordinates of the center of its base, and the cone's angle; for a cylinder model, the corresponding model parameters may include the coordinates of the center of the top surface, the coordinates of the center of the base, and the cylinder's radius; for a sphere model, the corresponding model parameters may include the coordinates of the sphere's center and its radius; for a cuboid model, the corresponding parameters may include the coordinates of one of its vertices, the coordinates of the opposite vertices, and the length, width, and height of the cuboid.

[0060] For example, in this embodiment, the filtering parameters of the above-mentioned geometric filter model may include at least: a rotation calibration matrix and / or an offset calibration matrix. Of course, the above-mentioned rotation calibration matrix and offset calibration matrix are only examples, and the embodiments of this application are not specifically limited.

[0061] It should be noted that the rotation calibration matrix and / or offset calibration matrix here can be determined through the calibration process of the geometric filter model.

[0062] In this embodiment, the filtering parameters of any geometric filter model are used to adjust the pose of the geometric filter model to filter at least one interference object in a specified area. As for how to use the filtering parameters to adjust the pose of the geometric filter model to filter at least one interference object in a specified area, please refer to the description of step S140, which will not be repeated here.

[0063] S140: Based on the filtering parameters of each geometric filter model, adjust the pose of each geometric filter model in the first mapped point cloud to filter out the interference point cloud in the first mapped point cloud.

[0064] For example, in this embodiment, the above-mentioned interference point cloud consists of the positions of all points on all interference objects within a specified area.

[0065] In the embodiment, for the same specified area, the interference contained therein is fixed, and therefore in the embodiment, the interference point cloud in the first mapped point cloud can be filtered based on the filtering parameters corresponding to the at least one geometric filter model calibrated for filtering the interference in the specified area.

[0066] As to how to adjust the pose of each geometric filter model in the first mapped point cloud according to the filtering parameters of each geometric filter model, the following Figure 3 is exemplarily described, which is not described herein.

[0067] At this point, the following Figure 1 is completed.

[0068] As can be seen from the following Figure 1 , in the embodiment, the interference in the target stockpile point cloud of the specified area scanned is filtered by the geometric filter model matched in advance for the interference of the specified area, so that the interference is filtered, and the statistical accuracy of the object in the specified area is improved; further, by matching the combination of different geometric filter models for the interference of the specified area in advance, different regions can be adapted, the statistical needs of different regions can be met, and the interference filtering of complex regions can be implemented.

[0069] The calibration process of the geometric filter is described in detail as follows.

[0070] Step a: obtain the reference stockpile point cloud scanned by each scanning device deployed in the specified area; the target stockpile point cloud corresponds to the historical stockpile in the specified area; at least one target stockpile point cloud contains the positions of the materials in the historical stockpile corresponding to the target stockpile point cloud and the positions of the points on at least one interference in the specified area. For specific embodiments, refer to the description of step S110 above, which is not described herein.

[0071] Step b: map each target stockpile point cloud to the same target coordinate system to obtain a second mapped point cloud; the second mapped point cloud includes the positions of the materials in each historical stockpile and the positions of the points on all interferences in the specified area; for specific embodiments, refer to the description of step S120 above, which is not described herein.

[0072] Step c: for the second mapped point cloud, in response to a geometric filter model selection operation, obtain at least one geometric filter model for filtering the interference point cloud in the second mapped point cloud, wherein the geometric filter model is determined according to the shape of the interference and / or the shape of the stockpile in the specified area.

[0073] Exemplarily, in the present embodiment, in the present step c, in response to the geometric filter model selected by the operator on the operation interface of the electronic device according to the shape of the interference object and / or the shape of the material pile in the specified area, at least one geometric filter model filtering the interference object point cloud in the second mapping point cloud is obtained.

[0074] For example, as shown in the second mapping point cloud, Figure 5 determined according to the shape of the interference object point cloud. Specifically, as shown in the second mapping point cloud, Figure 5 the interference objects existing in the second mapping point cloud include the control room, the rotating large arm and the conveyor belt, and the geometric filter model configured for the control room and the rotating large arm is a conical model. In the present embodiment, by continuously adjusting the parameters such as the position, pose, radius and height of the geometric filter model, the conical model is made to include as many control room point clouds and rotating large arm point clouds as possible, but to wrap as few other point clouds as possible, and the filtering direction of the geometric filter model is set to filter the point clouds inside the cone, so that the filtering of the control room and the rotating large arm is realized. And for the conveyor belt, the geometric filter model configured therefor is a cuboid model. In the present embodiment, by continuously adjusting the parameters such as the position, pose, length, width and height of the cuboid model, the cuboid model is made to include as many conveyor belt point clouds as possible, but to wrap as few other point clouds as possible, and the filtering direction of the geometric filter is set to filter the point clouds inside the cuboid, so that the filtering of the conveyor belt is realized.

[0075] For example, as shown in the second mapping point cloud, Figure 6 determined according to the shape of the material pile. Specifically, as shown in the second mapping point cloud, Figure 6 for the material pile on the left, Figure 6 the geometric filter model configured therefor is a cuboid model. In the present embodiment, by adjusting the cuboid parameters, the cuboid is made to include as many target material pile point clouds as possible, and to wrap as few non-material pile point clouds as possible, and the filtering direction of the geometric filter is set to filter the external point clouds, so that the point clouds outside the material pile are filtered.

[0076] Step d: in response to the operation of moving the above-mentioned geometric filter model to the second mapping point cloud, the positions of the moved geometric filter models are obtained, and for each geometric filter model, the filtering parameters are determined according to the initial position of the geometric filter model and the position of the moved geometric filter.

[0077] Exemplarily, in the present embodiment, when the geometric filter model is selected, the geometric filter model has an initial position, and in order to filter the interference point cloud, the geometric filter model needs to be moved to the second candidate point cloud, and in response to the moving operation of the operator on the operation interface of the electronic device, the positions of the moved geometric filter models are obtained.

[0078] And using the initial position of the geometric filter model and the position of the geometric filter after moving to determine the filter parameter is a conventional technique, which will not be described here.

[0079] Referring to Figure 3 , Figure 3 The pose adjustment flowchart provided in the embodiment of the application. As shown in Figure 3 , the flowchart can include the following steps:

[0080] S310: traverse the geometric filter model according to the use order of the calibrated geometric filter model, and take the traversed geometric filter model as the current filter model; call the current geometric filter model and add it to the current filtering area.

[0081] Exemplarily, in the embodiment, the use order of the geometric filter model is determined in advance after the geometric filter model is matched, and the determination method of the use order of the geometric filter model can be various, for example, the use order of the geometric filter model can be determined randomly.

[0082] For another example, the use order of the geometric filter model can be determined according to a preset order, for example, according to the calibrated filtering direction of each geometric filter model, that is, the use order of the geometric filter model with the first filtering direction is arranged in front of the geometric filter model with the second filtering direction, and the use order of the geometric filter model with the first filtering direction or the second filtering direction can be determined randomly. For the first filtering direction and the second filtering direction, the following embodiments are exemplarily described, which will not be described here.

[0083] Exemplarily, in the embodiment, as shown in Figure 4 , the geometric filter models are M1-Mn respectively, and the use order of the calibrated geometric filter model is M1-Mn. When filtering, the geometric filter model is traversed according to the use order of the calibrated geometric filter model, and the traversed geometric filter model is taken as the current filter model. For example, the first mapping point cloud is input into the first geometric filter model in the use order, and the first geometric filter model in the use order is taken as the current geometric filter model.

[0084] After the current geometric filter model is determined, the current geometric filter model is directly called from the local electronic device or the database and added to the current filtering area.

[0085] Here, the current filtering region is a region in the first mapped point cloud. Specifically, if the current geometric filter model is the first geometric filter model in the use sequence, the current filtering region is the first mapped point cloud; if the current geometric filter model is a geometric filter model other than the first geometric filter model in the use sequence, the current filtering region is a region in the first mapped point cloud other than the filtering result corresponding to the last geometric filter model that has been used.

[0086] As to how to determine the filtering result of the geometric filter, refer to the following example description, which is not described herein. Figure 7

[0087] S320: Adjust the pose of the current geometric filter model in the current filtering region according to the rotation calibration matrix and / or the offset calibration matrix of the current geometric filter model, so that the adjusted geometric filter model wraps the point cloud to be processed.

[0088] Exemplarily, in the embodiment, the position where the current geometric filter model called in the step S310 is added to the current filtering region is taken as the initial position of the current geometric filter model, and in the step S320, the pose of the current geometric filter model at the initial position is transformed, i.e., the pose of the current geometric filter model in the current filtering region is adjusted according to the rotation calibration matrix and / or the offset calibration matrix of the current geometric filter model, so that the adjusted geometric filter model wraps the point cloud to be processed. How to adjust the pose of the current geometric filter model in the current filtering region according to the rotation calibration matrix and / or the offset calibration matrix of the current geometric filter model is a conventional technique, which is not described herein.

[0089] Here, the point cloud to be processed contains the positions of points on at least one interference object in the specified region. For example, as shown in Figure 5 , if the current geometric filter model is filter 1 (cone model), the adjusted geometric filter model includes the positions of points on the control chamber and the positions of points on the rotating arm; if the current geometric filter model is filter 2 (cylinder model), the adjusted geometric filter model includes the positions of points on the conveyor belt.

[0090] Alternatively, the point cloud to be processed contains the positions of materials on at least one current material pile in the specified region. For example, as shown in Figure 6 , if the current geometric filter model is filter 3 (cuboid model), the adjusted geometric filter model includes the positions of materials on the material piles.

[0091] At this point, the description of the flow shown in Figure 3 is completed.

[0092] By Figure 3 ​The shown flow realizes the pose adjustment of the current geometric filter model, so that the adjusted geometric filter model wraps the point cloud to be processed.

[0093] As an optional implementation of the embodiment of the present application, after adjusting the pose of each geometric filter model in the first mapping point cloud, the interference filtering method further comprises: determining the filtering result corresponding to the current geometric filter model, see the example description of the specific embodiment. Figure 7

[0094] See Figure 7 , Figure 7 The filtering result determination flowchart provided by the embodiment of the present application is shown in FIG. 7. As shown in the figure, the flow can comprise the following steps: Figure 7

[0095] S710: mapping the current geometric filter model from the current non-XY plane coordinate system to the XY plane coordinate system to obtain the first region corresponding to the current geometric filter model in the XY plane coordinate system; and mapping the current filtering region from the current non-XY plane coordinate system to the XY plane coordinate system to obtain the second region corresponding to the current filtering region in the XY plane coordinate system.

[0096] Exemplarily, in the embodiment, the current geometric filter model and the current filtering region are both in the non-XY plane target coordinate system, in order to facilitate subsequent filtering processing, the current geometric filter model and the current filtering region are both mapped to the XY plane.

[0097] Specifically, taking the current non-XY plane coordinate system as the world coordinate system (XYZ coordinate system) as an example, mapping the current geometric filter model from the current non-XY plane coordinate system to the XY plane coordinate system can be simply setting the Z coordinate value of each point on the current geometric filter model to 0. The processing process of the current filtering region is the same, which will not be described here.

[0098] Exemplarily, the first region corresponding to the current geometric filter model in the XY plane coordinate system can be only the model shape formed by mapping the current geometric filter model to the XY plane coordinate system, or can comprise the model shape formed by mapping the current geometric filter model to the XY plane coordinate system and the circumscribed rectangle corresponding to the model shape, and the embodiment of the present application is not specifically limited. The following embodiment exemplarily describes how to filter the current filtering region based on the first region, which will not be described here.

[0099] Here, the circumscribed rectangle corresponding to the model shape can be the minimum circumscribed rectangle of the model shape, or can be other rectangle containing the model shape, and the embodiment of the present application is not specifically limited.

[0100] The following takes Figure 8 ​​To illustrate the first region mentioned above, for example... Figure 8 As shown, taking the geometric filter model as a cone model as an example, the cone model in the XYZ target coordinate system ( Figure 8 The leftmost image is projected onto the XY plane to obtain the first region. When the first region is simply the shape of the model formed by mapping the current geometric filter model to the XY plane coordinate system, the first region can be seen in [reference needed]. Figure 8 The convex polygon diagram in the middle. When the first region includes the model shape formed by mapping the current geometric filter model to the XY plane coordinate system and the corresponding bounding rectangle of the model shape, the first region can be seen as... Figure 8 The convex polygon in the middle and Figure 8 The rightmost rectangle.

[0101] S720: Select points within the first region within the second region to obtain a candidate point cloud; for each candidate point in the candidate point cloud, map the candidate point from the XY plane coordinate system to a non-XY plane coordinate system to obtain the corresponding mapped point.

[0102] For example, in this embodiment, the candidate point cloud refers to points within a first region. There are many methods for selecting points within the first region within a second region to obtain the candidate point cloud. For instance, when the first region is the model shape formed by mapping the current geometric filter model to the XY plane coordinate system, step S720, selecting points within the first region within the second region to obtain the candidate point cloud, includes: selecting points within the model shape within the second region to obtain the candidate point cloud.

[0103] For example, when the first region includes at least: the model shape formed by mapping the current geometric filter model to the XY plane coordinate system, and the bounding rectangle corresponding to the model shape, in step S720, points within the first region are selected in the second region to obtain a candidate point cloud, including:

[0104] First, select points within the bounding rectangle corresponding to the model shape in the second region to obtain the point cloud within the bounding rectangle corresponding to the model. Second, select points within the model shape within the bounding rectangle corresponding to the model to obtain the candidate point cloud.

[0105] For example, Figure 9 The points within the rectangular area shown are the point cloud within the bounding rectangle corresponding to the model selected from the second area; Figure 10 The points (C, D, E, F) within the triangular region shown are the points within the model shape selected from the bounding rectangle corresponding to the model, i.e., the aforementioned candidate point cloud.

[0106] For example, in this embodiment, the candidate point is mapped from the XY plane coordinate system to a non-XY plane coordinate system in order to restore the Z coordinate of the candidate point back, that is, to obtain the mapped point corresponding to the candidate point.

[0107] S730: Obtain the mapped point cloud, and determine the filtering result corresponding to the current geometric filter model based on the mapped point cloud and the calibrated filtering direction of the current geometric filter model.

[0108] For example, the aforementioned mapping point cloud is composed of mapping points within the current geometric filter model. These mapping points can be determined by the mapping points corresponding to each candidate point within the candidate region and the model parameters of the current geometric filter model. For instance, when the current geometric filter model is a cone model, it can be determined by the angle formed between the mapping point corresponding to the candidate point and the cone axis AB, and the cone angle α.

[0109] Taking the current geometric filter model as a cone model as an example, such as Figure 11 As shown, based on the mapping points corresponding to each candidate point in the candidate point cloud, the angle formed by each mapping point and the cone axis AB is compared with the cone angle α. When the angle formed by the mapping point and the cone axis AB is less than the cone angle α, the mapping point is considered to be a mapping point inside the current geometric filter model (cone model).

[0110] For example, for Figure 11 Since ∠CAB < α, mapping point C is determined to be an internal mapping point of the current geometric filter model (conical model). Similarly, mapping points E and F are both external to the current geometric filter model, while mapping point D is internal to the current geometric filter model.

[0111] As for how to determine the mapping point within the current geometric filter model, please refer to the example description in the following embodiment, which will not be repeated here.

[0112] For example, in this embodiment, the filtering direction of the geometric filter model can be either a first filtering direction or a second filtering direction, and this embodiment of the application does not specifically limit it.

[0113] The first filtering direction is used to indicate the filtering out of the mapped point cloud, and the second filtering direction is used to indicate the retention of the mapped point cloud. In this embodiment, when calibrating the filtering direction of the geometric filter model, if the point cloud enclosed by the geometric filter model is an interfering point cloud, the filtering direction of the geometric filter model is calibrated as the first filtering direction; if the point cloud enclosed by the geometric filter model is a material pile point cloud, the filtering direction of the geometric filter model is calibrated as the second filtering direction.

[0114] In the embodiment, the filtering result corresponding to the current geometric filter model is determined according to the mapping point cloud and the filtering direction of the calibrated current geometric filter model in step S730, which can be specifically:

[0115] If the filtering direction is the first filtering direction, the mapping point cloud is filtered out from the current filtering region to obtain the filtering result; if the filtering direction is the second filtering direction, the point cloud in the candidate point cloud except the mapping point cloud is filtered out from the current filtering region to obtain the filtering result.

[0116] The implementation can be specifically achieved by the following formula:

[0117]

[0118]

[0119]

[0120] Wherein, C out is the output point cloud after filtering; C src is the first mapping point cloud; i represents the i th geometric filter model; n represents the number of geometric filter models; M (p, d) represents the point cloud set to be filtered out by the filtering model; C i represents the output point cloud of the i th geometric filter model; C i-1 represents the output point cloud of the i-1 th geometric filter model; M prj represents the candidate point cloud; M c represents the point cloud in the current geometric filter model; p represents the model parameter set of the geometric filter model; d represents the filtering direction of the geometric filter model, d < 0 represents that the filtering direction of the geometric filter is the second filtering direction, otherwise represents that the filtering direction of the geometric filter is the first filtering direction.

[0121] Thus, the description of the flowchart shown in Figure 7 is completed.

[0122] The filtering result of the current geometric filter on the current filtering region is determined through the flowchart shown in Figure 7 .

[0123] As an optional implementation manner of the embodiment, after filtering the interference point cloud in the first mapping point cloud, the interference filtering method further includes:

[0124] First, the segmented each segmented pile in the first mapping point cloud after filtering out the interference point cloud is obtained.

[0125] Exemplarily, in this embodiment, after filtering out the interference point cloud in the first mapping point cloud, a filtered point cloud corresponding to the first mapping point cloud is obtained. After the electronic device detects that the user performs segmentation on the filtered point cloud on the operation interface, each segmented pile is obtained.

[0126] Secondly, the statistical parameters corresponding to each segmented pile are calculated according to each segmented pile.

[0127] Exemplarily, in this embodiment, the statistical parameters can be volume, weight, etc., and the embodiments of the present application are not specifically limited.

[0128] And calculating the statistical parameters corresponding to each segmented pile according to each segmented pile is a conventional technology, which will not be described here.

[0129] As an optional implementation of the embodiments of the present application, after each segmented pile is obtained, the blind area of each segmented pile is compensated by point cloud, and the statistical parameters of the compensated segmented pile are calculated. Here, the point cloud compensation method can be a conventional linear interpolation compensation, which will not be described here.

[0130] In this embodiment, the method of point cloud compensation can also be to approximate each pile to a cone, and compensate the point cloud of the blind area by using the point cloud of the non-blind area according to the symmetry of the cone.

[0131] The above describes the method provided by the embodiments of the present application, and the device provided by the embodiments of the present application is described below:

[0132] Reference is made to Figure 12 , Figure 12 The device structure diagram provided by the embodiments of the present application. The device can include:

[0133] The target pile point cloud acquisition module is configured to acquire target pile point clouds scanned by each scanning device deployed in a specified area; the target pile point cloud corresponds to a current pile in the specified area; and at least one target pile point cloud includes positions of materials in the current pile corresponding to the target pile point cloud and positions of points on at least one interference object in the specified area;

[0134] The mapping module is configured to map each target pile point cloud to the same target coordinate system to obtain a first mapping point cloud; the first mapping point cloud includes positions of materials in each current pile and positions of points on all interference objects in the specified area;

[0135] The filter parameter obtaining module is configured to obtain a filter parameter corresponding to at least one calibrated geometric filter model for filtering interference objects in the specified area, wherein the geometric filter model is determined based on a second mapped point cloud, and the second mapped point cloud is obtained by mapping reference stockpile point clouds scanned by each scanning device to a target coordinate system, wherein the reference stockpile point cloud corresponds to a historical stockpile in the specified area, and at least one reference stockpile point cloud includes positions of materials in the historical stockpile corresponding to the reference stockpile point cloud and positions of points on at least one interference object in the specified area, and the second mapped point cloud includes the positions of the materials in each historical stockpile and the positions of the points on each interference object; and the filter parameter of any geometric filter model is used to adjust a pose of the geometric filter model to filter at least one interference object in the specified area.

[0136] The filter module is configured to adjust the pose of each geometric filter model in the first mapped point cloud according to the filter parameter of each geometric filter model, so as to filter an interference object point cloud in the first mapped point cloud, wherein the interference object point cloud is composed of the positions of points on all interference objects in the specified area.

[0137] As an optional implementation manner of the embodiment of the present application, the filter parameter of any geometric filter model at least includes a rotation calibration matrix and / or an offset calibration matrix, and the filter module is specifically configured to:

[0138] The geometric filter models are traversed according to the use order of the calibrated geometric filter models, and a traversed geometric filter model is taken as a current filter model; the current geometric filter model is called and added to a current filtering area; if the current geometric filter model is the first geometric filter model in the use order, the current filtering area is the first mapped point cloud; if the current geometric filter model is a geometric filter model other than the first one in the use order, the current filtering area is an area in the first mapped point cloud except for a filtering result corresponding to a previous geometric filter model that has been used.

[0139] The pose of the current geometric filter model in the current filtering area is adjusted according to the rotation calibration matrix and / or the offset calibration matrix of the current geometric filter model, so that the adjusted geometric filter model wraps a to-be-processed point cloud; the to-be-processed point cloud includes the positions of points on at least one interference object in the specified area, or the to-be-processed point cloud includes positions of materials on at least one current stockpile in the specified area.

[0140] As an optional implementation manner of the embodiment of the present application, the interference filtering device further includes a filter result determining module, which is specifically configured to:

[0141] mapping the current geometric filter model from the current non-XY plane coordinate system to the XY plane coordinate system to obtain a first region corresponding to the current geometric filter model in the XY plane coordinate system; and mapping the current filter region from the current non-XY plane coordinate system to the XY plane coordinate system to obtain a second region corresponding to the current filter region in the XY plane coordinate system;

[0142] selecting, in the second region, points within the first region to obtain a candidate point cloud; and mapping each candidate point in the candidate point cloud from the XY plane coordinate system to the non-XY plane coordinate system to obtain a mapping point corresponding to the candidate point;

[0143] obtaining a mapping point cloud composed of the mapping points within the current geometric filter model; and determining a filter result corresponding to the current geometric filter model according to the mapping point cloud and a filter direction calibrated for the current geometric filter model.

[0144] As an optional implementation of the embodiment of the present application, the first region at least includes a model shape formed by mapping the current geometric filter model to the XY plane coordinate system and a circumscribed rectangle corresponding to the model shape.

[0145] The above selecting, in the second region, points within the first region to obtain a candidate point cloud includes:

[0146] selecting, in the second region, points within the circumscribed rectangle corresponding to the model shape to obtain a point cloud within the circumscribed rectangle corresponding to the model shape.

[0147] selecting, in the circumscribed rectangle corresponding to the model shape, points within the model shape to obtain a candidate point cloud.

[0148] As an optional implementation of the embodiment of the present application, the above determining a filter result corresponding to the current geometric filter model according to the mapping point cloud and a filter direction calibrated for the current geometric filter model includes:

[0149] if the filter direction is a first filter direction, the first filter direction is used to indicate that the mapping point cloud is filtered out, and the mapping point cloud is filtered out from the current filter region to obtain the filter result.

[0150] if the filter direction is a second filter direction, the second filter direction is used to indicate that the mapping point cloud is retained, and points in the candidate point cloud other than the mapping point cloud are filtered out from the current filter region to obtain the filter result.

[0151] As an optional implementation of the embodiment of the present application, the interference filtering device further includes:

[0152] a statistical parameter determination module configured to obtain each segmented stockpile segmented from a first mapping point cloud after the interference point cloud is filtered out.

[0153] According to each segmented pile, a statistical parameter corresponding to each segmented pile is calculated.

[0154] The implementation process of the functions and roles of each unit in the above device is specifically described in the implementation process of the corresponding steps in the above method, which will not be repeated here.

[0155] At this point, the description of the structure of the device shown in Figure 12 is completed.

[0156] Correspondingly, the embodiments of the present application also provide Figure 12 the hardware structure diagram of the device, as shown in Figure 13 , the electronic device can be the device of the above-mentioned implementation method. As shown in Figure 13 , the hardware structure includes a processor and a memory.

[0157] The memory is configured to store machine executable instructions.

[0158] The processor is configured to read and execute the machine executable instructions stored in the memory, so as to realize the corresponding interference filtering method embodiment as shown above.

[0159] As an embodiment, the memory can be any electronic, magnetic, optical or other physical storage device, which can contain or store information such as executable instructions, data, etc. For example, the memory can be a volatile memory, a non-volatile memory or a similar storage medium. Specifically, the memory can be a RAM (Random Access Memory), a flash memory, a storage drive (such as a hard disk drive), a solid state disk, any type of storage disk (such as an optical disk, a DVD, etc.), or a similar storage medium, or a combination thereof.

[0160] At this point, the description of the electronic device shown in Figure 13 is completed.

[0161] The above is only the preferred embodiment of the present application, and does not limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. An interference filtering method, characterized by, The method is applied to an electronic device, and the method comprises: Obtaining target stockpile point clouds scanned by each scanning device deployed in a specified area; the target stockpile point clouds correspond to current stockpiles in the specified area; at least one target stockpile point cloud contains the positions of materials in the current stockpile corresponding to the target stockpile point cloud and the positions of points on at least one interference object in the specified area; Mapping each target stockpile point cloud to the same target coordinate system to obtain a first mapped point cloud; the first mapped point cloud comprises the positions of materials in each current stockpile and the positions of points on all interference objects in the specified area; Obtaining filtering parameters corresponding to at least one calibrated geometric filter model for filtering interference objects in the specified area; the geometric filter model is determined based on a second mapped point cloud, the second mapped point cloud is obtained by mapping reference stockpile point clouds scanned by each scanning device to the target coordinate system; the reference stockpile point clouds correspond to historical stockpiles in the specified area; at least one reference stockpile point cloud contains the positions of materials in the historical stockpile corresponding to the reference stockpile point cloud and the positions of points on at least one interference object in the specified area; the second mapped point cloud comprises the positions of materials in each historical stockpile and the positions of points on each interference object; the filtering parameters of any geometric filter model are used to adjust the pose of the geometric filter model to filter at least one interference object in the specified area; According to the filtering parameters of each geometric filter model, the pose of each geometric filter model in the first mapped point cloud is adjusted to filter interference object point clouds in the first mapped point cloud, the interference object point clouds are composed of the positions of points on all interference objects in the specified area.

2. The method of claim 1, wherein, The filtering parameters of any geometric filter model at least comprise a rotation calibration matrix and / or an offset calibration matrix; The adjustment of the pose of each geometric filter model in the first mapped point cloud according to the filtering parameters of each geometric filter model comprises: According to the use order of the calibrated geometric filter models, the geometric filter models are traversed, and the traversed geometric filter model is taken as a current filter model; the current geometric filter model is called and added to a current filtering area; if the current geometric filter model is the first geometric filter model in the use order, the current filtering area is the first mapped point cloud; if the current geometric filter model is not the first geometric filter model in the use order, the current filtering area is an area in the first mapped point cloud except the filtering result corresponding to the last used geometric filter model; According to the rotation calibration matrix and / or the offset calibration matrix of the current geometric filter model, the pose of the current geometric filter model in the current filtering area is adjusted, so that the adjusted geometric filter model wraps a to-be-processed point cloud; the to-be-processed point cloud contains the positions of points on at least one interference object in the specified area, or the to-be-processed point cloud contains the positions of materials on at least one current stockpile in the specified area.

3. The method of claim 2, wherein, After adjusting the poses of the geometric filter models in the first mapping point cloud, the method further comprises determining the filtering result corresponding to the current geometric filter model according to the following steps: mapping the current geometric filter model from the current non-XY plane coordinate system to an XY plane coordinate system to obtain a first region corresponding to the current geometric filter model in the XY plane coordinate system; and mapping the current filtering region from the current non-XY plane coordinate system to the XY plane coordinate system to obtain a second region corresponding to the current filtering region in the XY plane coordinate system; selecting points in the first region in the second region to obtain a candidate point cloud; for each candidate point in the candidate point cloud, mapping the candidate point from the XY plane coordinate system to the non-XY plane coordinate system to obtain a mapping point corresponding to the candidate point; obtaining a mapping point cloud composed of the mapping points in the current geometric filter model; and determining the filtering result corresponding to the current geometric filter model according to the mapping point cloud and a filtering direction calibrated for the current geometric filter model.

4. The method of claim 3, wherein, The first region at least includes a model shape formed by mapping the current geometric filter model to the XY plane coordinate system and a circumscribed rectangle corresponding to the model shape; The step of selecting points in the first region in the second region to obtain a candidate point cloud comprises: selecting points in the circumscribed rectangle corresponding to the model shape in the second region to obtain a point cloud in the circumscribed rectangle corresponding to the model shape; selecting points in the model shape in the circumscribed rectangle corresponding to the model shape to obtain the candidate point cloud.

5. The method of claim 3, wherein, The step of determining the filtering result corresponding to the current geometric filter model according to the mapping point cloud and a filtering direction calibrated for the current geometric filter model comprises: if the filtering direction is a first filtering direction, the first filtering direction is used to indicate filtering out the mapping point cloud, and the mapping point cloud is filtered out from the current filtering region to obtain the filtering result; if the filtering direction is a second filtering direction, the second filtering direction is used to indicate retaining the mapping point cloud, and points in the candidate point cloud except the mapping point cloud are filtered out from the current filtering region to obtain the filtering result.

6. The method of claim 1, wherein, After filtering the interference point cloud in the first mapping point cloud, the method further comprises: obtaining each segmented stockpile segmented from the first mapping point cloud after the interference point cloud is filtered out; calculating statistical parameters corresponding to each segmented stockpile according to the segmented stockpiles.

7. An interference filter device, characterized by The device comprises: a target stockpile point cloud acquisition module configured to acquire target stockpile point clouds scanned by each scanning device deployed in a specified region; the target stockpile point clouds correspond to current stockpiles in the specified region; at least one target stockpile point cloud includes positions of materials in a current stockpile corresponding to the target stockpile point cloud and positions of points on at least one interference object in the specified region. The mapping module is configured to map the target stockpile point clouds to the same target coordinate system to obtain first mapping point clouds, and the first mapping point clouds include positions of materials in the current stockpiles and positions of points on all interference objects in the specified area; The filtering parameter obtaining module is configured to obtain filtering parameters corresponding to at least one geometric filter model used for filtering interference objects in the specified area, and the geometric filter model is determined based on second mapping point clouds obtained by mapping reference stockpile point clouds scanned by the scanning devices to the target coordinate system, the reference stockpile point clouds correspond to historical stockpiles in the specified area, at least one reference stockpile point cloud includes positions of materials in the historical stockpile corresponding to the reference stockpile point cloud and positions of points on at least one interference object in the specified area, the second mapping point clouds include positions of materials in the historical stockpiles and positions of points on the interference objects, and the filtering parameters of any geometric filter model are used to adjust a pose of the geometric filter model to filter at least one interference object in the specified area. The filtering module is configured to adjust poses of the geometric filter models in the first mapping point clouds according to the filtering parameters of the geometric filter models to filter interference object point clouds in the first mapping point clouds, and the interference object point clouds are composed of positions of points on all interference objects in the specified area.

8. The apparatus of claim 7, wherein, The filtering parameters of any geometric filter model at least include a rotation calibration matrix and / or an offset calibration matrix, and the filtering module is specifically configured to: traverse the geometric filter models according to a use order of the calibrated geometric filter models, take a geometric filter model traversed as a current filter model, call the current geometric filter model and add it to a current filtering area, if the current geometric filter model is the first geometric filter model in the use order, the current filtering area is the first mapping point clouds, if the current geometric filter model is a geometric filter model other than the first in the use order, the current filtering area is an area in the first mapping point clouds except a filtering result corresponding to a previous geometric filter model used; adjust a pose of the current geometric filter model in the current filtering area according to the rotation calibration matrix and / or the offset calibration matrix of the current geometric filter model, so that the adjusted geometric filter model wraps a to-be-processed point cloud, and the to-be-processed point cloud includes positions of points on at least one interference object in the specified area or the to-be-processed point cloud includes positions of materials on at least one current stockpile in the specified area.

9. The apparatus of claim 8, wherein, The device further includes a filtering result determining module configured to: map the current geometric filter model from a current non-XY plane coordinate system to an XY plane coordinate system to obtain a first area corresponding to the current geometric filter model in the XY plane coordinate system, and map the current filtering area from the current non-XY plane coordinate system to the XY plane coordinate system to obtain a second area corresponding to the current filtering area in the XY plane coordinate system. selecting a point in the first region in the second region to obtain a candidate point cloud; for each candidate point in the candidate point cloud, mapping the candidate point from the XY plane coordinate system to the non-XY plane coordinate system to obtain a mapping point corresponding to the candidate point; obtaining a mapping point cloud composed of mapping points in the current geometric filter model; determining a filtering result corresponding to the current geometric filter model according to the mapping point cloud and a filtering direction calibrated for the current geometric filter model.

10. An electronic device, comprising: An electronic device includes a processor and a memory; The memory is configured to store machine executable instructions. The processor is configured to read and execute the machine executable instructions stored in the memory to implement any one of the methods in claims 1 to 6.

Images