Radiation source positioning method and device based on virtual fingerprint, equipment and storage medium

Abstract

The invention discloses a radiation source positioning method, device and equipment based on a virtual fingerprint and a storage medium, belongs to the technical field of data processing, is used for radiation source positioning in an outdoor scene, and comprises the following steps: establishing a three-dimensional model of a target area based on three-dimensional point cloud data of the target area; dividing the three-dimensional model into a plurality of grid points by taking N meters as a step; based on the multiple grid points, a first point location set is determined, the first point location set comprises multiple point locations, all the point locations are located on different grid points, and the multiple point locations are located outside the building; the radiation sources are sequentially placed in multiple point locations of the first point location set to construct a virtual fingerprint database, and the virtual fingerprint database comprises virtual fingerprint data of M receiving points under the condition that the radiation sources are sequentially placed in the point locations of the first point location set; and positioning the radiation source based on the virtual fingerprint database. According to the method, the virtual fingerprint positioning technology can be applied to the outdoor scene.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing technology, and in particular to a radiation source positioning method, device, equipment and storage medium based on virtual fingerprint. Background Art

[0002] Virtual fingerprint positioning technology includes two stages: offline fingerprint collection and online positioning. Since outdoor scenes are often large in area, the cost of fingerprint collection in the offline stage is too high, which makes it difficult to apply virtual fingerprint positioning technology to outdoor scenes. Summary of the invention

[0003] The present disclosure provides a radiation source positioning method, device, equipment and storage medium based on virtual fingerprint, which can apply virtual fingerprint positioning technology to outdoor scenes. The technical solution includes at least the following solutions: In a first aspect, a radiation source positioning method based on virtual fingerprint is provided, which is used to position the radiation source in an outdoor scene, and the method includes: establishing a three-dimensional model of the target area based on three-dimensional point cloud data of the target area; dividing the three-dimensional model into multiple grid points with N meters as a step; determining a first point set based on the multiple grid points, the first point set including multiple points, each point located on a different grid point, and the multiple points are all located outside the building; placing the radiation source in multiple points of the first point set in sequence to construct a virtual fingerprint database, the virtual fingerprint database including virtual fingerprint data of M receiving points when the radiation source is placed in the points of the first point set in sequence, the virtual fingerprint data including the total receiving power, average delay and average angle of each receiving point; positioning the radiation source based on the virtual fingerprint database; wherein M is a positive integer, M is greater than or equal to 3, and N is a positive number.

[0004] Optionally, the three-dimensional model of the target area is established based on the three-dimensional point cloud data of the target area, including: performing semantic segmentation on the three-dimensional point cloud data to obtain multiple types of point cloud data, the types including ground, buildings, vegetation and small structures, and the small structures including street lights and fences; for different types of point cloud data, using different algorithms to perform surface reconstruction to obtain the three-dimensional model.

[0005] Optionally, different algorithms are used to perform surface reconstruction on different types of point cloud data to obtain the three-dimensional model, including: for ground type point cloud data, a bilinear interpolation algorithm is used to perform surface reconstruction; for building type point cloud data, a polyfit algorithm is used to perform surface reconstruction; for vegetation type point cloud data and small structure type point cloud data, a BallPivoting algorithm is used to perform surface reconstruction.

[0006] Optionally, placing the radiation sources in the points of the first point set in sequence to construct a virtual fingerprint database includes: based on a ray tracing algorithm, obtaining the power, delay, arrival angle and phase of each ray from the first radiation source to the first receiving point, the first radiation source is the case where the radiation source is placed in the first point in the first point set, the first point is any point in the first point set, and the first receiving point is any one of the M receiving points; based on the power, delay, arrival angle and phase of each ray of the first receiving point, determining the total receiving power, average delay and average angle of the first receiving point, thereby obtaining the virtual fingerprint data of the first radiation source for the first receiving point.

[0007] Optionally, determining the total received power, average delay, and average angle of the first receiving point based on the power, delay, arrival angle, and phase of each ray of the first receiving point includes: The total received power of the first receiving point is determined using the following formula:

[0008] in, is the total received power of the first receiving point, For the The power of the ray, For the The phase of the ray, there are Rays, and are all positive integers; The average delay of the first receiving point is determined by the following formula:

[0009] in, is the average delay of the first receiving point, For the The time delay of the ray; The average angle of the first receiving point is determined using the following formula:

[0010] in, is the average angle of the first receiving point; It is calculated using the following formula:

[0011] in, For the The angle of arrival of the ray; It is calculated using the following formula: .

[0012] Optionally, locating the radiation source based on the virtual fingerprint database includes: inputting the virtual fingerprint data in the virtual fingerprint database into a radiation source positioning model to locate the radiation source; wherein the radiation source positioning model includes an embedding layer connected in sequence, a Transformer encoder with a multi-head attention mechanism, a one-dimensional convolutional neural network, a fully connected layer and an output layer.

[0013] In the second aspect, a radiation source positioning device based on virtual fingerprint is also provided, including: used for positioning the radiation source in an outdoor scene, the device including: a modeling module, used to establish a three-dimensional model of the target area based on the three-dimensional point cloud data of the target area; a division module, used to divide the three-dimensional model into multiple grid points with N meters as a step; a point determination module, used to determine a first point set based on the multiple grid points, the first point set including multiple points, each point located on a different grid point, and the multiple points are all located outside the building; a virtual fingerprint data modeling module, used to place the radiation source in multiple points of the first point set in sequence to construct a virtual fingerprint database, the virtual fingerprint database includes the virtual fingerprint data of M receiving points when the radiation source is placed in the points of the first point set in sequence, and the virtual fingerprint data includes the total receiving power, average delay and average angle of each receiving point; a radiation source positioning module, used to locate the radiation source based on the virtual fingerprint database; wherein M is a positive integer, M is greater than or equal to 3, and N is a positive number.

[0014] Optionally, the modeling module is also used to perform semantic segmentation on the three-dimensional point cloud data to obtain multiple types of point cloud data, including ground, buildings, vegetation and small structures, and the small structures include street lights and fences; for different types of point cloud data, different algorithms are used to reconstruct the surface to obtain the three-dimensional model.

[0015] Optionally, the modeling module is also used to reconstruct the surface of ground type point cloud data using a bilinear interpolation algorithm; to reconstruct the surface of building type point cloud data using a polyfit algorithm; and to reconstruct the surface of vegetation type point cloud data and small structure type point cloud data using a BallPivoting algorithm.

[0016] Optionally, the virtual fingerprint data modeling module is also used to obtain the power, delay, arrival angle and phase of each ray from the first radiation source to the first receiving point based on a ray tracing algorithm, where the first radiation source is the case where the radiation source is placed in the first point position in the first point position set, the first point position is any point position in the first point position set, and the first receiving point is any one of the M receiving points; based on the power, delay, arrival angle and phase of each ray of the first receiving point, the total receiving power, average delay and average angle of the first receiving point are determined, thereby obtaining the virtual fingerprint data of the first radiation source for the first receiving point.

[0017] Optionally, the virtual fingerprint data modeling module is further used to determine the total received power of the first receiving point by using the following formula:

[0018] in, is the total received power of the first receiving point, For the The power of the ray, For the The phase of the ray, there are Rays, and are all positive integers; The average delay of the first receiving point is determined by the following formula:

[0019] in, is the average delay of the first receiving point, For the The time delay of the ray; The average angle of the first receiving point is determined using the following formula:

[0020] in, is the average angle of the first receiving point; It is calculated using the following formula:

[0021] in, For the The angle of arrival of the ray; It is calculated using the following formula: .

[0022] Optionally, the radiation source positioning module is also used to input the virtual fingerprint data in the virtual fingerprint database into the radiation source positioning model to locate the radiation source; wherein the radiation source positioning model includes an embedding layer connected in sequence, a Transformer encoder with a multi-head attention mechanism, a one-dimensional convolutional neural network, a fully connected layer and an output layer.

[0023] In a third aspect, a computer device is also provided, including: a memory and a processor, wherein at least one computer program is stored in the memory, and the at least one computer program is loaded and executed by the processor, so as to execute the radiation source positioning method based on virtual fingerprint described in the above embodiment.

[0024] In a fourth aspect, a computer-readable storage medium is also provided, wherein at least one computer program is stored in the computer-readable storage medium, and the at least one computer program is loaded and executed by a processor to execute the radiation source positioning method based on virtual fingerprint described in the above embodiment.

[0025] In a fifth aspect, a computer program product is provided, comprising a computer program / instruction, wherein when the computer program / instruction is executed by a processor, the method described in the first aspect is implemented.

[0026] The beneficial effects brought by the technical solution provided by the embodiments of the present disclosure include at least: In the disclosed embodiment, a three-dimensional model of the target area is established, the three-dimensional model is divided into multiple grid points, a first point set is determined based on the multiple grid points, and then a virtual fingerprint database of the outdoor scene is constructed by sequentially placing the radiation source in the multiple points of the first point set. The radiation source can then be accurately located based on the virtual fingerprint database of the outdoor scene. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0028] Figure 1 A flowchart of a radiation source positioning method based on virtual fingerprint provided by an exemplary embodiment of the present disclosure is shown; Figure 2 A flowchart of a radiation source positioning method based on virtual fingerprint provided by another exemplary embodiment of the present disclosure is shown; Figure 3 It is a schematic diagram of surface reconstruction of ground type point cloud data; Figure 4 It is a schematic diagram of surface reconstruction of point cloud data of building types; Figure 5 It is a schematic diagram of surface reconstruction of point cloud data of vegetation type and point cloud data of small structure type; Figure 6 is a schematic diagram of the ray tracing algorithm; Figure 7 It is a structural diagram of the radiation source positioning model; Figure 8 A schematic structural diagram of a radiation source positioning device based on virtual fingerprint provided by an exemplary embodiment of the present disclosure is shown; Fig. 9 It is a schematic diagram of the structure of a computer device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0029] Unless otherwise defined, the technical or scientific terms used herein shall have the usual meanings understood by persons of ordinary skill in the field to which the present disclosure belongs. The words "first", "second", "third" and similar words used in the patent application specification and claims of the present disclosure do not indicate any order, quantity or importance, but are only used to distinguish different components. Similarly, words such as "one" or "one" do not indicate a quantitative limitation, but rather indicate the presence of at least one. Words such as "include" or "comprise" and similar words mean that the elements or objects appearing before "include" or "comprise" cover the elements or objects listed after "include" or "comprise" and their equivalents, and do not exclude other elements or objects. Words such as "connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect.

[0030] In order to make the objectives, technical solutions and advantages of the present disclosure more clear, the embodiments of the present disclosure will be further described in detail below with reference to the accompanying drawings.

[0031] Figure 1 A flowchart of a method for locating a radiation source based on virtual fingerprints provided by an exemplary embodiment of the present disclosure is shown. The method is used to locate a radiation source in an outdoor scene. The method can be executed by a computer device, see Figure 1 , the method comprising: In step 101, a three-dimensional model of the target area is established based on the three-dimensional point cloud data of the target area.

[0032] Here, the target area is an outdoor scene where radiation source positioning is required, and the target area includes at least one radiation source.

[0033] In step 102, the three-dimensional model is divided into a plurality of grid points at a step of N meters.

[0034] Wherein, N is a positive number. Optionally, the value range of N is 3 to 7, for example, 3, 5, or 7.

[0035] Divide the 3D model into multiple grid points with N meters as the step. In essence, divide the 3D model into multiple cubes with a side length of N meters connected in sequence. Each vertex of the cube is a grid point. At the same time, the 3D coordinates of each grid point can also be obtained. .

[0036] In step 103, a first point set is determined based on a plurality of grid points.

[0037] The first point set includes multiple points, each point is located at a different grid point, and the multiple points are all located outside the building.

[0038] After multiple grid points are determined in step 102, it can be determined whether each grid point is located inside the building. The grid points located inside the building are discarded, and the grid points located outside the building are retained in the first point set.

[0039] Optionally, the following two steps are used to determine whether the first grid point is located inside the building, wherein the first grid point is any one of the multiple grid points: The first step is to determine the first grid point Whether the coordinates are within the projection range of any building; If the first grid point If the coordinates of the first grid point are not within the projection range of any building, then the first grid point is not inside the building. If the coordinates are within the projection range of a certain building, execute the second step.

[0040] The projection range of the first building can be expressed as , if the coordinates of the first grid point exist The coordinates of the first grid point within the range exist range, it means that the first grid point The coordinates are within the projection range of the first building. If the coordinates of the first grid point exist The coordinates of the first grid point within the range exist Out of range; or the coordinates of the first grid point exist Coordinates of the first grid point outside the range exist range; or if the coordinates of the first grid point are exist Coordinates of the first grid point outside the range exist range, it means that the first grid point The coordinates are not within the projection range of the first building.

[0041] The first building is any building in the target area. For other buildings in the target area except the first building, the above method can also be used to determine whether the first grid point is within the projection range of these buildings, thereby achieving the above first step.

[0042] The second step is to determine the first grid point Whether the coordinate is greater than the height of the building.

[0043] If the first grid point The coordinates are greater than the height of the building, indicating that the first grid point is located outside the building; if the first grid point If the coordinates are less than or equal to the height of the building, it means that the first grid point is located inside the building.

[0044] In step 104, the radiation sources are sequentially placed at a plurality of points in the first point set to construct a virtual fingerprint database.

[0045] The virtual fingerprint database includes virtual fingerprint data of M receiving points when the radiation sources are sequentially placed in the points of the first point set, and the virtual fingerprint data includes the total receiving power, average delay and average angle of each receiving point.

[0046] In the case where there are M receiving points, M virtual fingerprint data can be determined for any point in the first point set. Then, assuming that there are k points in the first point set, there are k times M virtual fingerprint data in the virtual fingerprint database, where k is a positive integer.

[0047] In step 105, the radiation source is located based on the virtual fingerprint database.

[0048] In the disclosed embodiment, a three-dimensional model of the target area is established, the three-dimensional model is divided into multiple grid points, a first point set is determined based on the multiple grid points, and then a virtual fingerprint database of the outdoor scene is constructed by sequentially placing the radiation source in the multiple points of the first point set. The radiation source can then be accurately located based on the virtual fingerprint database of the outdoor scene.

[0049] Figure 2A flowchart of a method for locating a radiation source based on virtual fingerprints provided by another exemplary embodiment of the present disclosure is shown. The method is used to locate a radiation source in an outdoor scene. The method can be executed by a computer device, see Figure 2 , the method comprising: In step 201, three-dimensional point cloud data of a target area is obtained.

[0050] Optionally, step 201 includes: acquiring three-dimensional point cloud data of the target area through unmanned aerial vehicle oblique photography technology.

[0051] Drone oblique photography technology breaks through the limitation of traditional orthophotos that only have vertical angles. By carrying multiple sensors on drones, images can be collected from five different angles at the same time, including one vertical angle, four oblique angles, etc. Drone oblique photography technology can obtain detailed terrain data of the target area, and the three-dimensional point cloud data of the target area can be obtained by processing the terrain data. When implementing it, Smart3D software can be used to process terrain data based on the principle of aerial triangulation to obtain the three-dimensional point cloud data of the target area.

[0052] There are many related technologies regarding the implementation method of obtaining three-dimensional point cloud data of the target area based on drone oblique photography technology, which will be omitted here.

[0053] In some embodiments, laser radar scanning technology can be used to obtain point cloud data of the target area, which is not limited in the embodiments of the present disclosure.

[0054] In step 202, a three-dimensional model of the target area is established based on the three-dimensional point cloud data of the target area.

[0055] Optionally, step 202 includes the following steps ab.

[0056] Step a: semantically segment the three-dimensional point cloud data to obtain multiple types of point cloud data.

[0057] Types include ground, buildings, vegetation, and small structures, such as street lights and fences.

[0058] Among them, semantic segmentation of 3D point cloud data includes the following two steps: In the first step, the SensatUrban dataset is used as a pre-training dataset to train the RandLA-Net network to obtain a point cloud semantic segmentation network model.

[0059] The SensatUrban dataset is a city-scale photometric point cloud dataset that includes approximately three billion richly annotated points, each of which is labeled as one of 13 semantic categories. Using this dataset as a pre-training dataset can improve the accuracy of the trained point cloud semantic segmentation model.

[0060] The RandLA-Net network is suitable for efficient semantic segmentation of large-scale point clouds, such as semantic segmentation of a three-dimensional point cloud of a target area in the embodiment of the present disclosure.

[0061] Therefore, in the first step, the SensatUrban dataset is used as a pre-training dataset to train the RandLA-Net network. The obtained point cloud semantic segmentation network model has the advantage of high robustness.

[0062] The second step is to perform semantic segmentation on the three-dimensional point cloud data based on the point cloud semantic segmentation network model.

[0063] The second step is to input the three-dimensional point cloud data of the target area into the point cloud semantic segmentation network model to obtain the type of each point in the point cloud data.

[0064] Step b: for different types of point cloud data, different algorithms are used to reconstruct the surface to obtain a three-dimensional model.

[0065] Optionally, step b includes: for ground type point cloud data, using a bilinear interpolation algorithm to perform surface reconstruction; for building type point cloud data, using a polyfit algorithm to perform surface reconstruction; for vegetation type point cloud data and small structure type point cloud data, using a BallPivoting algorithm to perform surface reconstruction.

[0066] There are many related technologies for implementing surface reconstruction of point cloud data using bilinear interpolation algorithm, polyfit algorithm and BallPivoting algorithm, which will be omitted here.

[0067] Since different types of point cloud data have different importance and complexity, different algorithms are used for surface reconstruction for different types of point cloud data, so as to reduce the complexity of the 3D model while ensuring the geometric accuracy of the reconstructed 3D model.

[0068] Figure 3 This is a schematic diagram of surface reconstruction of ground point cloud data. Figure 3 As shown in FIG. 1 , after the surface reconstruction is completed, the obtained surface model has a certain simplification effect compared with the original point cloud data.

[0069] Figure 4This is a schematic diagram of surface reconstruction of point cloud data of building types. Figure 4 It can be seen that Figure 3 Compared with the surface model built by ground-type point cloud data, the surface model obtained by surface reconstruction of building-type point cloud data is simpler and has fewer details.

[0070] Figure 5 It is a schematic diagram of surface reconstruction of vegetation type point cloud data and small structure type point cloud data. Figure 5 Part (a) and part (b) are schematic diagrams showing reconstruction of point cloud data of vegetation types; Figure 5 Part (c) and part (d) are surface reconstruction of point cloud data of fence type; Figure 5 Part (f) and part (f) are the surface reconstruction of the point cloud data of the street lamp type. It can be seen that after the surface reconstruction of the point cloud data of the vegetation type and the point cloud data of the small structure type, the obtained three-dimensional model is closer to the real form of the vegetation and the small structure, that is, the modeling accuracy is higher.

[0071] In step 203, the three-dimensional model is divided into a plurality of grid points at a step of N meters.

[0072] In step 204, a first point set is determined based on a plurality of grid points.

[0073] The first point set includes multiple points, each point is located at a different grid point, and the multiple points are all located outside the building.

[0074] The relevant contents of steps 203 to 204 refer to the aforementioned steps 102 to 103, and the detailed description is omitted here.

[0075] In step 205, the radiation sources are sequentially placed at a plurality of points in the first point set to construct a virtual fingerprint database.

[0076] The virtual fingerprint database includes virtual fingerprint data of M receiving points when the radiation sources are sequentially placed in the points of the first point set, and the virtual fingerprint data includes the total receiving power, average delay and average angle of each receiving point. Wherein, M is a positive integer, and M is greater than or equal to 3.

[0077] In the case where there are M receiving points, M virtual fingerprint data can be determined for any point in the first point set. Then, assuming that there are k points in the first point set, there are k times M virtual fingerprint data in the virtual fingerprint database, where k is a positive integer.

[0078] Optionally, step 205 includes the following steps cd.

[0079] Step c: based on a ray tracing algorithm, obtaining the power, delay, arrival angle and phase of each ray from the first radiation source to the first receiving point.

[0080] The first radiation source is a case where the radiation source is placed in a first point in the first point set, the first point is any point in the first point set, and the first receiving point is any one of the M receiving points.

[0081] The basic principle of the ray tracing algorithm is to track a single ray emitted from a radiation source. The direct path and reflected path of the ray are tracked by applying Snell's law and geometric optics; the diffraction path of the ray is tracked by the Uniform Theory of Diffraction (UTD); and the reception of the ray is determined by whether the receiving point is within the ray tube formed by the ray.

[0082] Figure 6 is a schematic diagram of the ray tracing algorithm. Figure 6 As shown, the red sphere represents the first radiation source, the yellow sphere represents the first receiving point, the black rays represent the paths of direct, reflected, and diffracted rays emitted from the first radiation source and propagated to the first receiving point; the red rays represent the paths of scattered rays emitted from the first radiation source and propagated to the first receiving point.

[0083] Optionally, when using the ray tracing algorithm, a GPU (Graphics Processing Unit, image processor) parallel acceleration method can be used to improve the operating efficiency of the ray tracing algorithm.

[0084] There are many methods for implementing the ray tracing algorithm in the related art, so the detailed description is omitted here.

[0085] Step d, based on the power, delay, arrival angle and phase of each ray of the first receiving point, determine the total receiving power, average delay and average angle of the first receiving point, so as to obtain the virtual fingerprint data of the first radiation source for the first receiving point.

[0086] Optionally, the total receiving power of the first receiving point is determined using formula (1).

[0087] (1) In formula (1), is the total received power at the first receiving point, For the The power of the ray, For the The phase of the ray, there are Rays, and are all positive integers; Optionally, the average delay of the first receiving point is determined using formula (2).

[0088] (2) In formula (2), is the average delay of the first receiving point, For the The meanings of other parameters in formula (2) are the same as those in formula (1), and their detailed description is omitted here.

[0089] Optionally, the average angle of the first receiving point is determined using formula (3).

[0090] (3) In formula (3), is the average angle of the first receiving point, It is calculated using formula (4). The meanings of other parameters in formula (3) are the same as those in formula (1), and detailed description is omitted here.

[0091] (4) In formula (4), For the The angle of arrival of the ray, It is calculated using formula (5); the meanings of other parameters in formula (4) are the same as those in formula (3), and detailed description is omitted here.

[0092] (5) The meanings of the parameters in formula (5) are the same as those in formula (1) and formula (4), and their detailed description is omitted here.

[0093] The virtual fingerprint data of other receiving points except the first receiving point corresponding to the first radiation source (a total of M-1 virtual fingerprint data) can also be determined using the above steps cd to obtain M virtual fingerprint data corresponding to the first radiation source.

[0094] Finally, the M virtual fingerprint data corresponding to the first radiation source can be represented as fingerprint 1 in sequence: ; Fingerprint 2: ; ...Fingerprint M: .

[0095] The M virtual fingerprint data corresponding to the radiation sources at other points except the first radiation source can also be determined in the above manner, so that the M virtual fingerprint data corresponding to each point can be determined when the radiation sources are sequentially placed in the k points of the first point set. These virtual fingerprint data constitute a virtual fingerprint database.

[0096] In step 206, the radiation source is located based on the virtual fingerprint database.

[0097] Optionally, step 206 includes: inputting the virtual fingerprint data in the virtual fingerprint database into the radiation source positioning model to locate the radiation source.

[0098] Figure 7 It is a schematic diagram of the structure of the radiation source positioning model. Figure 7 As shown in the figure, the radiation source localization model includes an embedding layer connected sequentially, a Transformer encoder with a multi-head attention mechanism, a one-dimensional convolutional neural network, a fully connected layer, and an output layer.

[0099] The embedding layer includes sub-embedding and position encoding for the input virtual fingerprint data; the Transformer encoder with a multi-head attention mechanism is used to extract features from the output of the embedding layer, and the one-dimensional convolutional neural network and the fully connected layer are used to perform position matching on the extracted data.

[0100] Suppose the M virtual fingerprint data corresponding to a point in the first point set of the radiation source is a group of virtual fingerprint data. There are k points in the first point set, so there are k groups of virtual fingerprint data. The radiation source positioning model is essentially a classification model. The result of the classification is whether the coordinates of the point corresponding to a certain group of virtual fingerprint data are the real position of the radiation source. Through the radiation source positioning model, the k groups of virtual fingerprint data can be classified in sequence, so as to determine whether the coordinates of the points corresponding to the first group, the second group...the k-1 group, and the kth group of virtual fingerprint data are the real positions of the radiation source.

[0101] If the coordinates of the points corresponding to a set of virtual fingerprint data are the real positions of the radiation source, it means that the coordinates are one of the positions of the radiation source finally located.

[0102] The following are device embodiments of the present application. For details not described in detail in the device embodiments, reference may be made to the above method embodiments.

[0103] Figure 8 The schematic diagram of the structure of a radiation source positioning device based on virtual fingerprint provided by an exemplary embodiment of the present disclosure is shown, and the device is used to position the radiation source in an outdoor scene. Figure 8 The radiation source positioning device 800 based on virtual fingerprint includes: a modeling module 801, a division module 802, a point determination module 803, a virtual fingerprint data modeling module 804, and a radiation source positioning module 805.

[0104] The modeling module 801 is used to establish a three-dimensional model of the target area based on the three-dimensional point cloud data of the target area.

[0105] The division module 802 is used to divide the three-dimensional model into a plurality of grid points at a step of N meters.

[0106] The point determination module 803 is used to determine a first point set based on multiple grid points. The first point set includes multiple points, each point is located at a different grid point, and the multiple points are all located outside the building.

[0107] The virtual fingerprint data modeling module 804 is used to place the radiation source in sequence at multiple points in the first point set to construct a virtual fingerprint database, which includes the virtual fingerprint data of M receiving points when the radiation source is placed in sequence at the points in the first point set, and the virtual fingerprint data includes the total receiving power, average delay and average angle of each receiving point.

[0108] The radiation source positioning module 805 is used to locate the radiation source based on the virtual fingerprint database; wherein M is a positive integer, M is greater than or equal to 3, and N is a positive number.

[0109] Optionally, the modeling module 801 is also used to perform semantic segmentation on the three-dimensional point cloud data to obtain multiple types of point cloud data, including ground, buildings, vegetation and small structures, small structures including street lights and fences; for different types of point cloud data, different algorithms are used to reconstruct the surface to obtain a three-dimensional model.

[0110] Optionally, the modeling module 801 is also used to perform surface reconstruction using a bilinear interpolation algorithm for ground type point cloud data; to perform surface reconstruction using a polyfit algorithm for building type point cloud data; and to perform surface reconstruction using a BallPivoting algorithm for vegetation type point cloud data and small structure type point cloud data.

[0111] Optionally, the virtual fingerprint data modeling module 804 is also used to obtain the power, delay, arrival angle and phase of each ray from the first radiation source to the first receiving point based on a ray tracing algorithm, where the first radiation source is the case where the radiation source is placed at the first point position in the first point position set, the first point position is any point position in the first point position set, and the first receiving point is any one of the M receiving points; based on the power, delay, arrival angle and phase of each ray from the first receiving point, the total receiving power, average delay and average angle of the first receiving point are determined, thereby obtaining the virtual fingerprint data of the first radiation source for the first receiving point.

[0112] Optionally, the virtual fingerprint data modeling module 804 is further configured to determine the total received power of the first receiving point using the following formula:

[0113] in, is the total received power at the first receiving point, For the The power of the ray, For the The phase of the ray, there are Rays, and All are positive integers; The average delay of the first receiving point is determined by the following formula:

[0114] in, is the average delay of the first receiving point, For the The delay of the ray; The average angle of the first receiving point is determined using the following formula:

[0115] in, is the average angle of the first receiving point; It is calculated using the following formula:

[0116] in, For the The angle of arrival of the ray; It is calculated using the following formula: .

[0117] Optionally, the radiation source positioning module 805 is also used to input the virtual fingerprint data in the virtual fingerprint database into the radiation source positioning model to locate the radiation source; wherein the radiation source positioning model includes an embedding layer connected in sequence, a Transformer encoder with a multi-head attention mechanism, a one-dimensional convolutional neural network, a fully connected layer and an output layer.

[0118] It should be noted that: when the radiation source positioning device based on virtual fingerprint provided in the above embodiment performs radiation source positioning, only the division of the above functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the radiation source positioning device based on virtual fingerprint provided in the above embodiment and the radiation source positioning method embodiment based on virtual fingerprint belong to the same concept. The specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0119] The division of modules in the embodiments of the present disclosure is schematic and is only a logical function division. There may be other division methods in actual implementation. In addition, each functional module in each embodiment of the present disclosure may be integrated into a processor, or may exist physically separately, or two or more modules may be integrated into one module. The above-mentioned integrated modules may be implemented in the form of hardware or in the form of software functional modules.

[0120] If the integrated module is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present disclosure is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a terminal device (which can be a personal computer, mobile phone, or communication device, etc.) or a processor (processor) to execute all or part of the steps of the method of each embodiment of the present disclosure. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.

[0121] Fig. 9 Schematic diagram of the structure of the computer device provided by the embodiment of the present disclosure. Fig. 9 As shown, the computer device 900 includes: a processor 901 and a memory 902 .

[0122] The processor 901 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 901 may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 901 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 901 may be integrated with a GPU, which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 901 may also include an AI (Artificial Intelligence) processor, which is used to process computing operations related to machine learning.

[0123] The memory 902 may include one or more computer-readable storage media, which may be non-transitory. The memory 902 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 902 is used to store at least one instruction, which is used to be executed by the processor 901 to implement the radiation source positioning method based on virtual fingerprints provided in the embodiments of the present disclosure.

[0124] Those skilled in the art will understand that Fig. 9 The structure shown in the figure does not constitute a limitation on the computer device 900, and the computer device 900 may include more or less components than those shown in the figure, or combine some components, or adopt a different arrangement of components.

[0125] The embodiment of the present disclosure also provides a non-temporary computer-readable storage medium. When the instructions in the storage medium are executed by a processor of a computer device, the computer device can execute the radiation source positioning method based on virtual fingerprint provided in the embodiment of the present disclosure.

[0126] The embodiments of the present disclosure also provide a computer program product, including a computer program / instruction, which, when executed by a processor, implements the radiation source positioning method based on virtual fingerprint provided in the embodiments of the present disclosure.

[0127] The above description is only an optional embodiment of the present disclosure and is not intended to limit the present disclosure. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A radiation source positioning method based on virtual fingerprint, It is characterized in that For locating a radiation source in an outdoor scene, the method comprises: Building a three-dimensional model of the target area based on the three-dimensional point cloud data of the target area; Dividing the three-dimensional model into a plurality of grid points with N meters as a step; Based on the multiple grid points, determine a first point set, the first point set including multiple points, each point is located at a different grid point, and the multiple points are all located outside the building; Placing the radiation source in sequence at multiple points of the first point set to construct a virtual fingerprint database, wherein the virtual fingerprint database includes virtual fingerprint data of M receiving points when the radiation source is placed in sequence at the points of the first point set, and the virtual fingerprint data includes total receiving power, average delay, and average angle of each receiving point; Based on the virtual fingerprint database, locating the radiation source; Wherein, M is a positive integer, M is greater than or equal to 3, and N is a positive number.

2. The method according to claim 1, It is characterized in that The step of establishing a three-dimensional model of the target area based on the three-dimensional point cloud data of the target area comprises: Performing semantic segmentation on the three-dimensional point cloud data to obtain multiple types of point cloud data, the types including ground, buildings, vegetation and small structures, the small structures including street lights and fences; For different types of point cloud data, different algorithms are used to perform surface reconstruction to obtain the three-dimensional model.

3. The method according to claim 1, It is characterized in that The method of using different algorithms to perform surface reconstruction on different types of point cloud data to obtain the three-dimensional model includes: For ground-type point cloud data, bilinear interpolation algorithm is used for surface reconstruction; For point cloud data of building types, the polyfit algorithm is used for surface reconstruction; For vegetation type point cloud data and small structure type point cloud data, BallPivoting algorithm is used for surface reconstruction.

4. The method according to any one of claims 1 to 3, It is characterized in that The placing of radiation sources in the points of the first point set in sequence to construct a virtual fingerprint database includes: Based on a ray tracing algorithm, the power, delay, arrival angle and phase of each ray from a first radiation source to a first receiving point are obtained, where the first radiation source is a case where the radiation source is placed in a first point position in the first point position set, the first point position is any point position in the first point position set, and the first receiving point is any one of the M receiving points; Based on the power, delay, arrival angle and phase of each ray of the first receiving point, the total receiving power, average delay and average angle of the first receiving point are determined, thereby obtaining virtual fingerprint data of the first radiation source for the first receiving point.

5. The method according to claim 4, It is characterized in that Determining a total received power, an average delay, and an average angle of the first receiving point based on the power, delay, arrival angle, and phase of each ray of the first receiving point includes: The total received power of the first receiving point is determined using the following formula: in, is the total received power of the first receiving point, For the The power of the ray, For the The phase of the ray, there are Rays, and are all positive integers; The average delay of the first receiving point is determined by the following formula: in, is the average delay of the first receiving point, For the The delay of the ray; The average angle of the first receiving point is determined using the following formula: in, is the average angle of the first receiving point; It is calculated using the following formula: in, For the The angle of arrival of the ray; It is calculated using the following formula: 。 6. The method according to any one of claims 1 to 3, It is characterized in that The locating the radiation source based on the virtual fingerprint database includes: Inputting the virtual fingerprint data in the virtual fingerprint database into the radiation source positioning model to locate the radiation source; The radiation source positioning model includes an embedding layer connected in sequence, a Transformer encoder with a multi-head attention mechanism, a one-dimensional convolutional neural network, a fully connected layer and an output layer.

7. A radiation source positioning device based on virtual fingerprint, It is characterized in that Used for locating radiation sources in outdoor scenes, the device comprises: A modeling module, used for establishing a three-dimensional model of the target area based on the three-dimensional point cloud data of the target area; A division module, used for dividing the three-dimensional model into a plurality of grid points with N meters as a step; A point determination module, configured to determine a first point set based on the plurality of grid points, wherein the first point set includes a plurality of points, each point being located at a different grid point, and the plurality of points are all located outside the building; a virtual fingerprint data modeling module, used to sequentially place the radiation source at a plurality of points in the first point set to construct a virtual fingerprint database, wherein the virtual fingerprint database includes virtual fingerprint data of M receiving points when the radiation source is sequentially placed at the points in the first point set, and the virtual fingerprint data includes a total receiving power, an average delay, and an average angle of each receiving point; A radiation source positioning module, used for positioning the radiation source based on the virtual fingerprint database; Wherein, M is a positive integer, M is greater than or equal to 3, and N is a positive number.

8. A computer device, It is characterized in that The computer device comprises: a memory and a processor, wherein at least one computer program is stored in the memory, and the at least one computer program is loaded and executed by the processor to implement the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores at least one computer program, and the at least one computer program is loaded and executed by a processor to implement the method according to any one of claims 1 to 6.

10. A computer program product comprising a computer program / instructions, It is characterized in that When the computer program / instructions are executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images