An underwater weak magnetic moving target detection method and system based on unmanned dual machines

Abstract

The present invention belongs to the technical field of underwater weak magnetic target detection, and proposes an underwater weak magnetic moving target detection method and system based on unmanned dual aircraft. First, two unmanned aircraft equipped with geomagnetic gradient tensor instruments are used to cooperatively locate underwater magnetic targets, initially eliminating the interference of the geomagnetic field on measurement; then, the rough positioning coordinates are projected onto three planes, and binary Hough detection including point accumulation and power accumulation is performed on the tracks of the three planes respectively, so as to maximize the extraction of the true track of the weak magnetic target from the positioning data; finally, when the Hough line detection is unsuccessful, the amount of data to be detected is reduced, so that the possible tracks can be approximately linearized in a short time. The detection of underwater weak magnetic moving targets is ultimately realized through the above method.

Description

Technical Field

[0001] The present invention relates to the technical field of magnetic target detection, and particularly to a method and system for detecting underwater weak magnetic maneuvering targets based on an unmanned aerial vehicle geomagnetic gradient tensor meter. Background Art

[0002] Unmanned aerial vehicles have advantages such as low cost, long range, and long hovering time compared to manned aircraft when searching for underwater submarines. They can closely search for "escaped fish" missed by manned aircraft and can also block various maritime choke points all-weather, which are incomparable to manned aircraft. However, unmanned aerial vehicles are not without disadvantages. Currently, the payload carried by unmanned aerial vehicles is relatively small, and they cannot deploy sonar buoys like manned aircraft.

[0003] Aeronautical magnetic exploration equipment is less affected by media such as air, seawater, sediment, and soil, and has advantages such as continuous searchability, simple and reliable use, small weight and volume, high positioning accuracy, rapid response, high search efficiency, and good concealment. It has become an important means for underwater target detection and is particularly suitable as a payload for unmanned aerial vehicles to search for underwater targets, effectively making up for the disadvantages of low accuracy, large weight, and significant influence by sea conditions of acoustic sensors.

[0004] However, in actual use, due to factors such as magnetic background anomalies in the detection sea area, non-classical magnetic interference of the carrier platform, magnetic anomaly differences in the target's heading, detection threshold settings of the magnetic exploration instrument, and target maneuvering, there are serious false alarm and missed detection phenomena in the magnetic exploration instrument, and there is still much room for improvement in actual use efficiency. When detecting using the target's magnetic signal, it is very important to enhance the signal-to-noise ratio and improve the performance of the detection algorithm. In particular, the magnetic field intensity of underwater targets is several orders of magnitude lower than the geomagnetic background field intensity and is often submerged by the geomagnetic background field, making feature extraction very difficult. The detection accuracy is greatly affected by the measurement error of the geomagnetic field, and the false alarm rate is high. At the same time, most current detection and tracking methods are based on linear motion, and how to achieve low false alarm rate detection of underwater weak magnetic maneuvering targets is a very challenging problem. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for detecting underwater weak magnetic maneuvering targets based on an unmanned double aircraft, so as to solve the problem that it is relatively difficult for traditional magnetic sensors to detect underwater weak magnetic maneuvering targets, and thus achieve low false alarm rate detection of underwater weak magnetic maneuvering targets based on an unmanned double aircraft.

[0006] To achieve the above purpose, the present invention provides the following solutions:

[0007] A method for detecting underwater weak magnetic maneuvering targets based on an unmanned double aircraft, comprising:

[0008] Step 1: Utilize the geomagnetic gradient measurement data of the first unmanned aerial vehicle (UAV) and the second UAV to collaboratively locate the underwater weak magnetic target in three-dimensional space, and obtain the rough positioning position of the underwater weak magnetic target; the first UAV and the second UAV are UAVs equipped with geomagnetic gradient tensor meters.

[0009] Step 2: Project the above-mentioned rough positioning positions at l moments onto the xoz plane, xoy plane, and zoy plane respectively, and form position data vector groups in the three planes respectively. Then, simultaneously perform binary hough track detection using the three position data vector groups. If the target motion track can be detected, restore the corresponding track points to three-dimensional space to obtain the target three-dimensional motion track.

[0010] Step 3: If the binary hough track detection fails to detect the target, adopt the curve hough track detection method based on subdivided time slices for track detection.

[0011] Optionally, Step 1 specifically includes:

[0012] Set an arbitrary point in space as the origin to establish a coordinate system. Then, at the kth moment, the rough positioning coordinates r d of the underwater weak magnetic target in the coordinate system are

[0013] r d = S1 - r1

[0014] where r1 is the relative position vector between the first UAV and the underwater weak magnetic target, and S1 is the position vector of the first UAV in the coordinate system.

[0015] Optionally, the calculation formula for r1 is:

[0016] r1 = -(G2 - G1) -1 (3G1 + G2)dr

[0017] where G1 and G2 are the magnetic field gradient tensors measured by the first UAV and the second UAV at the same moment respectively, and dr is the distance vector between the first UAV and the second UAV at the current moment.

[0018] Optionally, the binary hough track detection in Step 2 specifically includes:

[0019] Construct a binary data vector using the collaborative positioning information and magnetic field intensity information of the first UAV and the second UAV; the binary data vector includes position information and power information. The specific method is as follows:

[0020] In the xoy plane, at the kth moment, the rough positioning position information r d,xoy,k of the underwater weak magnetic target obtained by the collaborative positioning of the two UAVs is

[0021] r d,xoy,k = [x k y k

[0022] The power E at time k k :

[0023]

[0024] where δ k (B k ) is Gaussian white noise and is independent at each time, satisfying B k is the contribution value of the underwater weak magnetic target to the magnetic field intensity at the k-th time;

[0025] The standard Hough transform method is used to discretize the parameter space, and a binary accumulation matrix of the parameter space is established. The binary accumulation matrix of the parameter space includes a point number accumulation matrix D(·) and a power storage matrix E(·):

[0026] D(θ t , ρ m ) = D(θ t , ρ m ) + 1

[0027] E(θ t , ρ m ) += E(θ t , ρ m )

[0028] where θ t , ρ m represent the central point coordinates of the parameter unit, D(θ t , ρ m ) represents the point number accumulation matrix of the parameter unit, E(θ t , ρ m ) represents the power storage matrix of the parameter unit, and "+=" is the accumulation operation;

[0029] The cooperative positioning position information is mapped to the parameter space through the Hough transform, the position data vector group is traversed, and the accumulation operation is ended to obtain the point number accumulation histogram and the power accumulation histogram of the reference space;

[0030] Set the point number accumulation threshold and the power accumulation threshold, and based on the point number accumulation threshold and the power accumulation threshold, respectively perform threshold detection on the accumulation results of the parameter space parameter units, and perform inverse Hough mapping on the parameter units that pass the point number accumulation threshold detection and the power accumulation threshold detection to obtain the initial detection track;

[0031] ​Merge the initial detection tracks with the same point traces, and on the basis of the merge, delete the false initial tracks by means of speed and angle constraint conditions to obtain valid detection tracks corresponding one by one to the initial detection tracks.

[0032] Optionally, the curve hough detection method based on sub-divided time slices in step three specifically includes:

[0033] When the binary hough detection in step three is unsuccessful, reduce the data volume by a multiple of l / n, re-divide the data length, continue to project onto three planes, and perform track detection within each time slice. Connect the straight line detection results of multiple time slices to finally achieve the detection of the curve maneuvering track. Where l represents the number of rough positioning coordinates in the position data vector group, and n represents the multiple by which the length of the position data vector group is reduced.

[0034] The present invention also provides an underwater weak magnetic maneuvering target detection system based on two unmanned aircraft, and the system includes:

[0035] A rough positioning position acquisition module, configured to use the geomagnetic gradient measurement data of the first unmanned aircraft and the second unmanned aircraft to cooperatively locate an underwater weak magnetic target in a three-dimensional space to obtain the rough positioning position of the underwater weak magnetic target; the first unmanned aircraft and the second unmanned aircraft are unmanned aircraft carrying geomagnetic gradient tensor meters;

[0036] A target three-dimensional motion track determination module, configured to project the above-mentioned rough positioning positions at l moments onto the xoz plane, the xoy plane, and the zoy plane respectively, and form position data vector groups in the three planes respectively. Use the three position data vector groups to simultaneously perform binary hough track detection. If a target motion track can be detected, restore the corresponding track points to the three-dimensional space to obtain the target three-dimensional motion track;

[0037] A sub-divided time slice module, configured to, if the binary hough track detection fails to detect the target, use a curve hough track detection method based on sub-divided time slices to perform track detection.

[0038] Optionally, the rough positioning position acquisition module specifically includes:

[0039] Set an arbitrary point in space as the origin and establish a coordinate system. Then, at the kth moment, the rough positioning coordinate r of the underwater weak magnetic target in the coordinate system d is

[0040] r d = S1 - r1

[0041] where r1 is the relative position vector between the first unmanned aircraft and the underwater weak magnetic target, and S1 is the position vector of the first unmanned aircraft in the coordinate system.

[0042] Optionally, the calculation formula of r1 is:

[0043] r1 = -(G2 - G1) -1 (3G1 + G2)dr

[0044] where G1 and G2 are the magnetic field gradient tensors measured by the first unmanned aerial vehicle and the second unmanned aerial vehicle at the same moment, respectively, and dr is the distance vector between the first unmanned aerial vehicle and the second unmanned aerial vehicle at the current moment.

[0045] Optionally, the underwater weak magnetic target three-dimensional motion track determination module specifically includes:

[0046] A rough positioning position information acquisition unit, configured to construct a binary data vector by using the collaborative positioning information and the magnetic field intensity information of the first unmanned aerial vehicle and the second unmanned aerial vehicle; the binary data vector includes position information and power information;

[0047] In the xoy plane, at the k-th moment, the rough positioning position information r of the underwater weak magnetic target obtained by the cooperative positioning of the two unmanned aerial vehicles is d,xoy,k as

[0048] r d,xoy,k = [x k y k

[0049] A power acquisition unit, configured to acquire the power E at the k-th moment k :

[0050]

[0051] where δ k (B k ) is Gaussian white noise and is independent at each moment, satisfying B k is the contribution value of the underwater weak magnetic target to the magnetic field intensity at the k-th moment;

[0052] A binary accumulation matrix acquisition unit, configured to discretize the parameter space by using the standard hough transform method, and simultaneously establish a binary accumulation matrix in the parameter space, where the binary accumulation matrix in the parameter space includes a point number accumulation matrix D(·) and a power storage matrix E(·):

[0053] D(θ t , ρ m ) = D(θ t , ρ m ) + 1

[0054] E(θ t , ρ m ) += E(θ t , ρ m )

[0055] Among them, θ t , ρ m represent the central point coordinates of the parameter unit, D(θ t , ρ m ) represents the point accumulation matrix of the parameter unit, E(θ t , ρ m ) represents the power storage matrix of the parameter unit, and "+=" is the accumulation operation;

[0056] The histogram acquisition unit is used to cooperatively locate the position information mapped to the parameter space through the hough transform, traverse the position data vector group, end the accumulation operation, and obtain the point accumulation histogram and power accumulation histogram of the reference space;

[0057] The initial detection track acquisition unit is used to set the point accumulation threshold and the power accumulation threshold, and respectively perform threshold detection on the accumulation results of the parameter units in the parameter space based on the point accumulation threshold and the power accumulation threshold, and perform inverse hough mapping on the parameter units that pass the point accumulation threshold detection and the power accumulation threshold detection to obtain the initial detection track;

[0058] The effective detection track acquisition unit is used to merge the initial detection tracks with the same point traces, and delete the false initial tracks by means of speed and angle constraint conditions on the basis of the merger, so as to obtain the effective detection tracks corresponding one by one to the initial detection tracks.

[0059] Optionally, the sub - divided time - slice module specifically includes:

[0060] When the binary hough detection in step three is unsuccessful, reduce the data volume by a multiple of l / n, re - divide the data length, continue to project onto three planes, and perform track detection respectively within each time - slice. Connect the straight - line detection results of multiple time - slices to finally achieve the detection of the curved maneuvering track. Among them, l represents the number of rough - positioning coordinates in the position data vector group, and n represents the multiple by which the length of the position data vector group is reduced.

[0061] According to the specific embodiments provided by the present invention, the following technical effects are disclosed: The underwater weak magnetic moving target detection method and system based on unmanned dual drones provided by the present invention include: using the geomagnetic gradient measurement data of the first drone and the second drone to cooperatively locate the underwater weak magnetic target in three-dimensional space to obtain the rough positioning position of the underwater weak magnetic target; the first drone and the second drone are drones equipped with geomagnetic gradient tensor meters; projecting the rough positioning position onto the xoz plane, the xoy plane, and the zoy plane respectively, and l form a position data vector group in each of the three planes. Use the three position data vector groups to simultaneously perform binary hough track detection. If the target motion track can be detected, the corresponding track points are restored to three-dimensional space to obtain the accurate three-dimensional motion track of the target; if the binary hough track detection fails to detect the target, a curve hough track detection method based on subdivided time slices is used for track detection. The present invention cooperatively locates underwater magnetic targets through two drones equipped with geomagnetic gradient tensor meters, initially eliminating the interference of the geomagnetic field on the measurement; projects the rough positioning coordinates onto three planes, and performs binary hough detection on the tracks of the three planes respectively, including point number accumulation and power accumulation, so as to maximize the extraction of the true track of the weak magnetic target from the positioning data, and extract the linear motion component from the three-dimensional maneuvering track; when the hough line detection is unsuccessful, reduce the amount of data to be detected, so that the possible tracks can be approximately linearized in a short time, realizing the detection of underwater weak magnetic moving targets. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0063] Figure 1 Schematic flow chart of the underwater weak magnetic moving target detection method based on unmanned dual drones provided in Embodiment 1 of the present invention;

[0064] Figure 2 Implementation flowchart of the underwater weak magnetic moving target detection method based on unmanned dual drones provided in Embodiment 1 of the present invention;

[0065] Figure 3 Principle diagram of cooperative positioning of underwater weak magnetic targets by dual drones provided in Embodiment 1 of the present invention

[0066] Figure 4 Principle diagram of hough transform provided in Embodiment 1 of the present invention;

[0067] Figure 5Schematic diagram of the parameter space threshold-crossing parameter unit provided in Embodiment 1 of the present invention;

[0068] Figure 6 Method diagram of curve hough detection based on subdivided time slices provided in Embodiment 1 of the present invention;

[0069] Figure 7 Block diagram of an underwater weak magnetic mobile target detection system based on unmanned dual aircraft provided in Embodiment 2 of the present invention. Detailed implementation manners

[0070] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0071] The disadvantages of traditional magnetic sensors detecting underwater weak magnetic mobile targets in the prior art are as follows:

[0072] (1) When a manned aircraft detects an underwater weak magnetic target, the flight range is short, the hovering time is short, and the cost-effectiveness is low.

[0073] (2) Due to the influence of the magnetic background field in the detection sea area, the non-typical magnetic interference of the aircraft platform, the difference in the magnetic anomaly of the target heading, and the setting of the detection threshold of the magnetic detector, the false alarm and missed detection phenomena of the magnetic detector are relatively serious.

[0074] (3) The intensity of the geomagnetic background field is much greater than the target magnetic field intensity, often submerging the target magnetic field, and the measurement error of the geomagnetic field has a greater impact on the final target detection accuracy.

[0075] (4) The classical hough detection method can only detect the linear motion of the target, and the detection performance for mobile targets decreases greatly and is easy to be lost.

[0076] (5) The classical detection and tracking method can only detect the track in a two-dimensional plane.

[0077] Aiming at the above disadvantages of the prior art, the purpose of the present invention is to provide an underwater weak magnetic mobile target detection method and system based on unmanned dual aircraft to eliminate the interference of the geomagnetic field on the positioning of underwater weak magnetic mobile targets and improve the authenticity of the detected track of mobile targets.

[0078] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0079] Embodiment 1

[0080] This embodiment provides an underwater weak magnetic moving target detection method based on unmanned dual aircraft. Refer to Figure 1 and Figure 2 , the method includes:

[0081] Step 1: Utilize the geomagnetic gradient measurement data of the first unmanned aircraft and the second unmanned aircraft to collaboratively locate an underwater weak magnetic target in three-dimensional space, and obtain the rough positioning position of the underwater weak magnetic target; the first unmanned aircraft and the second unmanned aircraft are unmanned aircraft equipped with geomagnetic gradient tensor meters;

[0082] Step 2: Project the above rough positioning positions at l moments onto the xoz plane, xoy plane, and zoy plane respectively, and form position data vector groups in the three planes respectively. Simultaneously perform binary hough track detection using the three position data vector groups. If the target motion track can be detected, restore the corresponding track points to three-dimensional space to obtain the target three-dimensional motion track;

[0083] Step 3: If the binary hough track detection fails to detect the target, adopt a curve hough track detection method based on subdivided time slices for track detection.

[0084] Specifically, Steps 1 to 3 can be as follows:

[0085] Obtain the first magnetic gradient tensor at multiple moments; the first magnetic gradient tensor is obtained by measuring the underwater weak magnetic target with a geomagnetic gradient tensor instrument located on a first unmanned aerial vehicle (UAV); obtain the second magnetic gradient tensor at multiple moments; the second magnetic gradient tensor is obtained by measuring the underwater weak magnetic target with a geomagnetic gradient tensor instrument located on a second UAV; according to the first magnetic gradient tensor, the second magnetic gradient tensor and the cooperative positioning formula at the same moment, obtain the rough positioning coordinates, and the rough positioning coordinates at multiple moments form a position data vector group; the cooperative positioning UAV is the first UAV or the second UAV; project each of the rough positioning coordinates in the position data vector group onto the three planes of xoz, xoy, and zoy respectively to obtain the projection coordinates of the target on the three planes at each moment; perform Hough transformation on the projection coordinates of the target on the three planes respectively to obtain a number of Hough curves in the parameter space; discretize the parameter space into a number of parameter units; calculate the number of Hough curves passing through each parameter unit to obtain a point accumulation matrix for each parameter unit; obtain the magnetic field intensity of the target at each moment; the magnetic field intensity is obtained by measuring the target with a geomagnetic gradient tensor instrument; calculate the power value of each Hough curve at each moment according to the magnetic field intensity; the power value of the Hough curve is the power value of the target at the corresponding moment; perform an accumulation operation on the power values of the Hough curves passing through each parameter unit to obtain a power storage matrix for each parameter unit; determine whether the point accumulation matrix and the power storage matrix of each parameter unit meet the preset threshold detection decision condition. If they meet, perform inverse Hough mapping on the Hough curve corresponding to the parameter unit to obtain an initial detection track; the preset threshold detection decision condition includes a preset point accumulation matrix and a preset power storage matrix. When the binary Hough detection is unsuccessful, group the position data vector group by time to obtain a number of position data vector subgroups, replace the above position data vector group with the position data vector subgroups, and return to the step "project each of the rough positioning coordinates in the position data vector group onto the three planes of xoz, xoy, and zoy respectively to obtain the projection coordinates of the target on the three planes at each moment;". If after replacing the position data vector group with the position data vector subgroups, the point accumulation matrix and the power accumulation result of the preset number of parameter units still do not reach the preset threshold detection decision condition, discard the position data vector group, return to step one, re-perform the cooperative positioning of the two UAVs to obtain new rough positioning coordinates, and then use the new rough positioning coordinates to continue the next round of binary Hough detection.

[0086] The process of obtaining the rough positioning coordinates of the target in this embodiment is introduced below:

[0087] As Figure 3As shown in the figure, let an arbitrary point in space be set as the origin to establish a coordinate system. At time k, the rough positioning coordinates r of the underwater weak magnetic target in the coordinate system are d is

[0088] r d = S1 - r1 (1)

[0089] where r1 is the relative position vector between the first unmanned aerial vehicle (UAV) and the underwater weak magnetic target, and S1 is the position vector of the first UAV in the coordinate system. Among them, the calculation formula of r1 is:

[0090] r1 = -(G2 - G1) -1 (3G1 + G2)dr (2)

[0091] where G1 and G2 are the magnetic field gradient tensors measured by the first UAV and the second UAV at the same moment respectively, and dr is the distance vector between the first UAV and the second UAV at the current moment.

[0092] In this embodiment, the process of obtaining the collaborative positioning formula (1) can be as follows:

[0093] For the magnetic field vector B = (B x , B y , B z ) at any point (x, y, z), the rate of change of the three components B x , B y , B z in the space direction is the magnetic gradient tensor, which can be directly measured by a magnetic gradient tensor instrument and can be expressed as G, which contains nine-dimensional components:

[0094]

[0095] Among them,

[0096] When the distance between the magnetic target and the observation point is greater than 2.5 times the target length, the magnetic target can be regarded as a magnetic dipole. Therefore, the magnetic field vector at any point can be expressed as

[0097]

[0098] where μ0 is the magnetic permeability of vacuum, and in air μ0 ≈ 4π × 10 -7 H / m; r is the modulus of the distance between the magnetic target and the observation point; m is the magnetic moment of the magnetic target; r0 = r / r is the unit vector of r.

[0099] The magnetic field at the position r' = r + dr·r0 is B', where B' = (B' x , B' y , B' z), Equation (4) can be used to obtain

[0100]

[0101] According to Formulas (2) and (5)

[0102]

[0103] In Equation (6),

[0104]

[0105] Substituting Formula (7) into (6) gives:

[0106]

[0107] In Equation (8), B'-B = (B' x -B x , B' y -B y , B' z -B z ) can be obtained from the differential equation:

[0108]

[0109] Therefore, (B'-B) can be expressed as:

[0110]

[0111] Therefore, it can be obtained

[0112] r = -3G -1 B (11)

[0113] The above Formula (11) is the single-point magnetic gradient tensor positioning method. When a single unmanned aerial vehicle carries a geomagnetic gradient tensor instrument and flies over the sea surface, the magnetic gradient tensor G and the magnetic field vector B at a certain underwater point can be measured. In theory, the position of the underwater magnetic target can be determined through Formula (11). However, in actual measurements, the magnetic field vector is the sum of the geomagnetic field and the target magnetic field. The target magnetic field is only 0.1 - 1 μT, while the intensity of the total geomagnetic field is usually between 30 - 70 μT. Therefore, a small measurement error of the geomagnetic field can cause a large target positioning error. Moreover, when the observation point is far from the target point, the target magnetic field vector is very likely to be submerged by the geomagnetic field, resulting in unreliable measurement data information and positioning results. The measurement error of the geomagnetic field seriously affects the single-point magnetic tensor positioning accuracy, thus greatly limiting the application of this positioning method.

[0114] To eliminate the influence of the magnetic field vector on positioning and further expand the application range of the positioning method, this embodiment proposes a two-point magnetic gradient tensor positioning method, that is, using two unmanned aerial vehicles (UAVs) equipped with geomagnetic gradient tensor meters to cooperatively locate underwater weak magnetic targets. Without the geomagnetic field vector, only the full tensor information of two measurement points is required to obtain the position information of the target point. The specific process is as follows:

[0115] The position of the underwater weak magnetic target is r d , there are two UAVs both carrying geomagnetic gradient tensor meters, numbered 1 and 2 respectively. The relative position vector between the first UAV and the underwater weak magnetic target is r1; the relative position between the second UAV and the underwater weak magnetic target is r2. The position vector relationship among the two UAVs and the target point is

[0116] r2 = r1 + dr (12)

[0117] where dr is the position vector between the first UAV and the second UAV. Substituting the magnetic gradient tensors G1, G2 and the magnetic field vector data B1, B2 measured by the two UAVs into Equation (11) respectively, we can get

[0118] G1r1 = -3B1 (13)

[0119] G2r2 = -3B2 (14)

[0120] where G1 and G2 are the magnetic gradient tensors measured by UAV No. 1 and UAV No. 2 at the same moment respectively; B1 and B2 are the magnetic field vector data measured by UAV No. 1 and UAV No. 2 at the same moment respectively.

[0121] Substituting the magnetic tensor G1 measured by the first UAV, the magnetic field vector B1 and the magnetic field vector B2 measured by the second UAV into Equation (10), we can get

[0122] B2 - B1 = G1dr (15)

[0123] Through Formulas (12)-(15), Formula (2) can be obtained. Combining Formulas (2), (10) and Figure 3 , Formula (1) can be obtained, that is, the rough positioning coordinates r of the target d .

[0124] When there are rough positioning positions of the target at multiple moments, a position data vector group r d1 , r d2 , …, r dl, a rough target track is obtained. In this method, there is no need to measure the geomagnetic field value, which greatly reduces the positioning error caused by the geomagnetic field measurement error. At the same time, this method uses a linear model to complete the solution, and the process is simple, and an analytical solution can be directly obtained. Through the above method, the position of the underwater weak magnetic target can be obtained. When the unmanned aerial vehicle is flying, multiple measurements can be used to obtain a batch of underwater weak magnetic target position point data.

[0125] Although the above method can eliminate most of the errors of the geomagnetic background field, due to the influence of the platform, airborne equipment, and crustal anomaly field, a large amount of false information will still be included in the obtained batch of data, and the phenomena of false alarms and missed alarms are serious, making it difficult to accurately estimate the true position and track of the underwater weak magnetic target. In order to eliminate the influence of electromagnetic signals such as the platform, airborne equipment, and crustal anomaly field on magnetic field measurement and obtain true and accurate target track information, the data will be further processed using the Hough transform below.

[0126] The Hough transform is a method for detecting straight lines. When the underwater target is moving, it is arbitrary and does not necessarily move in a straight line. However, during the maneuvering process, it is possible that the navigation depth remains unchanged and only the heading changes; another possible situation is that the navigation depth changes but the heading remains unchanged. Considering these common maneuvering methods, the rough positioning coordinate data of the target are projected in the xoz, xoy, and zoy planes respectively to obtain the projection data of the underwater weak magnetic target in the three planes:

[0127]

[0128] where k is the current moment, k = 1,..., l. represents the projection coordinates in the xoz plane, represents the projection coordinates in the xoy plane, represents the projection coordinates in the zoy plane.

[0129] Therefore, in order to further extract the true track from the batch of target measurements in this embodiment, the binary Hough transform method is used to perform a secondary detection on the above-obtained rough coordinates, that is, three-channel binary Hough detection. The binary Hough transform method is introduced below:

[0130] Through Equation (16), the observation point (x k , y k ) in the data space can be transformed into a curve in the parameter space, as shown in the appendix Figure 4As shown in the figure. Among them, ρ is the normal distance from the origin to the straight line, θ is the included angle between the normal and the positive direction of the horizontal axis, and θ ∈ [0, π]. For the points (x, y) on the same straight line, all the curves after transformation to the parameter space will intersect at one point, and this point is the eigenvalue of the straight line in the parameter space. This shows that the Hough transform has a good accumulation effect on the data on the spatial straight line. Therefore, the Hough transform can be used to detect the target moving in a straight line. The Hough transform is a one-to-many mapping that maps each point in the data space to a curve in the parameter space, and each curve corresponds to multiple parameter units.

[0131] For the points (x k , y k ) on a straight line in the plane, there must be a unique (ρ, θ) corresponding to it:

[0132] ρ = x k cosθ + y k sinθ (16)

[0133] In order to further extract the true track from the positioning data in this embodiment, a binary hough track detection method including point accumulation and power accumulation is designed. The following takes the xoy plane as an example to specifically introduce the binary Hough track transformation:

[0134] Step 1: Construct a data matrix. At the k-th moment, the rough positioning coordinate data of the target obtained by the cooperative positioning of the UAV is

[0135] r d,xoy,k = [x k y k (17)

[0136] Among them, r d,xoy,k represents the rough positioning position of the underwater weak magnetic target in the xoy plane at the k-th moment, and k = 1,..., l.

[0137] Obtain the power value E k at the k-th moment as

[0138]

[0139] Among them, δ k (B k ) is Gaussian white noise and is independent at each moment, satisfying B k is the contribution value of the underwater weak magnetic target to the magnetic field intensity at the k-th moment, which is obtained by any magnetometer in the geomagnetic gradient tensor instrument.

[0140] Step 2: Discretize the parameter space. Considering the influence of the angle measurement error, in order to effectively detect the underwater weak target information, the hough parameter space is discretely divided into Nθ ×N p parameter units, and the center point (θ t , ρ m ) of each parameter unit is

[0141]

[0142] wherein, is the parameter unit size.

[0143] Step 3: Establish a parameter space accumulation matrix and initialize the parameter space accumulation matrix. The parameter space point number accumulation matrix is used to store the point number accumulation data of each parameter unit, and the parameter space power storage matrix is used to store the power accumulation data. Before storage, each accumulation matrix should be initialized. The matrix initialization is specifically to set each type of accumulation matrix as a zero matrix.

[0144] Step 4: Map the rough positioning coordinate data of the underwater weak magnetic target obtained in Step 1 to the curve in the parameter space through the Hough transform. Specifically, according to the Hough mapping formula, map the rough positioning position of the underwater weak magnetic target. Then, perform point number accumulation and power accumulation on the parameter units covered by the curve:

[0145] wherein, the formulas for point number accumulation and power accumulation are (20) and (21) respectively, and the formulas are as follows:

[0146] D(θ t , ρ m ) = D(θ t , ρ m ) + 1 (20)

[0147] E(θ t , ρ m ) += E(θ t , ρ m ) (21)

[0148] wherein, θ t , ρ m represent the center point coordinates of the parameter unit, D(θ t , ρ m ) represents the point number accumulation matrix of the parameter unit, E(θ t , ρ m ) represents the power storage matrix of the parameter unit, and "+=" is the cumulative operation.

[0149] Step 5: Repeat Step 4. When k > l, end the accumulation operation to obtain the reference space point number accumulation histogram and the power accumulation histogram.

[0150] Step 6: Set the point number accumulation threshold δ D and the power accumulation threshold δE Threshold detection is respectively performed on the accumulation results of parameter units in the parameter space. In this embodiment, the decision condition for threshold detection is as follows:

[0151] D(θ t ,ρ m )>δ D (22)

[0152] E(θ t ,ρ m )>δ E (23)

[0153] Wherein, D(θ t ,ρ m ) represents the point accumulation matrix with the center point coordinates being the parameter unit, and E(θ t ,ρ m ) represents the power storage matrix of the parameter unit, δ D is the preset point accumulation threshold, and δ E is the preset power accumulation threshold.

[0154] When the accumulation value of the parameter unit simultaneously satisfies formula (22) and formula (23), the corresponding parameter unit is obtained On the contrary, when any one of the conditions in formula (22) and formula (23) is not satisfied, the detection fails, that is, no target track is found. In this embodiment, when the number of parameter units that meet the threshold detection decision condition is less than the preset number w, it is determined that the detection fails.

[0155] Step 7: When the detection is successful, perform an inverse Hough mapping on the parameter units that pass the point accumulation threshold detection and the power accumulation threshold detection to obtain the initial detected track.

[0156] For example, when there are w parameter units (parameter units) that meet the preset threshold detection decision condition, as shown in the shaded part of Figure 5 , where each parameter unit corresponds to a set of measurement point traces in the data space i, j represents the position of the parameter unit in the parameter space, and w represents the serial number of the parameter unit corresponding to all the parameter units that pass the threshold detection decision condition. The corresponding measurement can be deduced backward according to the serial number, and multiple measurements form a point trace set.

[0157] Step 8: Perform reliability processing on the initial detected track. Merge the initial detected tracks with the same point traces, and on the basis of the merge, delete the false initial tracks by means of speed and angle constraint conditions to obtain high-confidence detected tracks corresponding one-to-one to the initial detected tracks, that is, it is considered that the target is effectively detected, thereby realizing the detection of underwater weak magnetic moving targets. The speed constraint condition and the heading constraint condition are specifically introduced as follows:

[0158] 1) Speed constraint

[0159] For an underwater weakly magnetic maneuvering target, the flight speed is between the minimum speed v min and the maximum speed v max . Therefore, using the positions of the target at two moments and v min , v max to constrain the target detection track:

[0160]

[0161] where t k+1 -t k is the time interval between two measurements, (x k , y k ) is the rough positioning coordinates of the target at the k-th moment, and (x k+1 , y k+1 ) are the rough positioning coordinates of the target at the (k + 1)-th moment.

[0162] 2) Heading constraint

[0163] For an underwater weakly magnetic maneuvering target, it generally does not suddenly stop and reverse strongly. That is, the change in the heading between frames of the target is relatively smooth. The track of the real target within three consecutive scanning periods can be approximated as a straight line. Therefore, the false track can be eliminated by the change in the target maneuvering angle to perform heading constraint. The specific process for determining the heading constraint conditions is as follows:

[0164] The rough positioning coordinates of the target at the k-th moment are (x k , y k ). The rough positioning coordinates of the target at the subsequent (k + 1)-th and (k + 2)-th moments are (x k+1 , y k+1 ) and (x k+2 , y k+2 ) respectively. Define the vector R k+1,k = [x k+1 - x k , y k+1 - y k T , R k+1,k+2 = [x k+1 - x k+2 , y k+1 - y k+2 T , then the target heading angle is defined as

[0165]

[0166] In this embodiment, the maximum maneuvering angle of the underwater weakly magnetic maneuvering target is defined as β0, then the heading constraint is​​

[0167] |β k,k+1,k+2 |≤β0 (26)

[0168] In this embodiment, the curve Hough detection method based on sub-divided time slices in step three specifically includes:

[0169] The prerequisite for successful Hough detection and the ability to extract the true target track is the existence of a straight track in three planes. However, during the actual movement of the target, it may maneuver in all three projection planes. Therefore, in this embodiment, when the Hough detection is unsuccessful, the curve Hough detection method based on sub-divided time slices shown in Figure 6 is adopted. The data volume is reduced by a multiple of l / n (where l represents the number of target rough positioning coordinates in the position data vector group, and n represents the multiple by which the length l of the position data vector group is reduced), the data length is re-divided, and the data is projected onto three planes again. In each time slice, binary Hough track detection is performed separately.

[0170] For example, the l rough positioning coordinates obtained in step one are grouped according to time to obtain n groups of rough positioning coordinates for time slices. Each group of rough positioning coordinates for time slices includes g (g = l / n) rough positioning coordinates. The l rough positioning coordinates are replaced with the groups of rough positioning coordinates for time slices, and binary Hough track detection is performed on each group of rough positioning coordinates for time slices. When the detection is still unsuccessful, these l rough positioning data are discarded, and step one is returned to re-perform the cooperative positioning of the two UAVs to obtain new rough positioning coordinates, and then the new rough positioning coordinates are used to continue the next round of binary Hough detection.

[0171] Compared with the prior art, the underwater maneuvering weak magnetic target detection method based on two UAVs of the present invention has the following beneficial effects:

[0172] (1) When the present invention uses a UAV carrying a magnetic load to detect underwater targets, the platform has a long range and long hovering time, and the cost-effectiveness is relatively high.

[0173] (2) The present invention directly uses the geomagnetic gradient tensor data of two UAVs to cooperatively locate underwater weak magnetic targets without introducing the geomagnetic field, avoiding the influence of geomagnetic field measurement errors on the final positioning result.

[0174] (3) The present invention uses the binary Hough transform for pre-detection tracking, reducing the influence of factors such as magnetic background anomalies in the detection sea area, non-typical magnetic interference of the carrier platform, magnetic anomaly differences in the target heading, and the setting of the detection threshold of the magnetic detector, and can effectively reduce the false alarm and missed detection probabilities of magnetic sensors.

[0175] (4) The present invention still has good detection accuracy for curved maneuvering targets.

[0176] (5) The present invention can achieve the trajectory detection of maneuvering targets in the underwater three-dimensional space.

[0177] Embodiment 2

[0178] This embodiment provides an underwater weak magnetic maneuvering target detection system based on two unmanned aircrafts. Referring to Figure 7 , the system includes:

[0179] A rough positioning position acquisition module T1, which is used to utilize the geomagnetic gradient measurement data of the first unmanned aircraft and the second unmanned aircraft to cooperatively position the underwater weak magnetic target in the three-dimensional space and obtain the rough positioning position of the underwater weak magnetic target; the first unmanned aircraft and the second unmanned aircraft are unmanned aircrafts equipped with geomagnetic gradient tensor meters.

[0180] A target three-dimensional motion trajectory determination module T2, which is used to project the above-mentioned rough positioning positions at l moments onto the xoz plane, the xoy plane, and the zoy plane respectively, and form position data vector groups in the three planes respectively. The binary hough trajectory detection is simultaneously performed using the three position data vector groups. If the target motion trajectory can be detected, the corresponding trajectory points are restored to the three-dimensional space to obtain the target three-dimensional motion trajectory.

[0181] A sub-divided time slice module T3, which is used to perform trajectory detection using the curve hough trajectory detection method based on sub-divided time slices if the binary hough trajectory detection fails to detect the target.

[0182] In this embodiment, the rough positioning position acquisition module T1 specifically includes:

[0183] Let any point in space be set as the origin to establish a coordinate system. Then, at the kth moment, the rough positioning coordinate r d of the underwater weak magnetic target in the coordinate system is

[0184] r d = S1 - r1

[0185] where r1 is the relative position vector between the first unmanned aircraft and the underwater weak magnetic target, and S1 is the position vector of the first unmanned aircraft in the coordinate system.

[0186] Among them, the calculation formula of r1 is:

[0187] r1 = -(G2 - G1) -1 (3G1 + G2)dr

[0188] where G1 and G2 are the magnetic field gradient tensors measured by the first unmanned aircraft and the second unmanned aircraft at the same moment respectively, and dr is the distance vector between the first unmanned aircraft and the second unmanned aircraft at the current moment.

[0189] In this embodiment, the underwater weak magnetic target three-dimensional motion trajectory determination module T2 specifically includes:

[0190] A rough positioning position information acquisition unit, configured to construct a binary data vector by using the cooperative positioning information of the first unmanned aerial vehicle and the second unmanned aerial vehicle and the magnetic field intensity information; the binary data vector includes position information and power information;

[0191] In the xoy plane, at the k-th moment, the rough positioning position information r of the underwater weak magnetic target obtained by the cooperative positioning of the two unmanned aerial vehicles d,xoy,k is

[0192] r d,xoy,k =[x k y k

[0193] A power acquisition unit, configured to acquire the power E at the k-th moment k :

[0194]

[0195] wherein, δ k (B k ) is Gaussian white noise and is independent at each moment, satisfying B k is the contribution value of the underwater weak magnetic target to the magnetic field intensity at the k-th moment;

[0196] A binary accumulation matrix acquisition unit, configured to discretize the parameter space by using the standard hough transform method, and simultaneously establish a binary accumulation matrix of the parameter space, the binary accumulation matrix of the parameter space including a point number accumulation matrix D(·) and a power storage matrix E(·):

[0197] D(θ t ,ρ m ) = D(θ t ,ρ m ) + 1

[0198] E(θ t ,ρ m ) += E(θ t ,ρ m )

[0199] wherein, θ t ,ρ m represent the central point coordinates of the parameter unit, D(θ t ,ρ m ) represents the point number accumulation matrix of the parameter unit, E(θ t ,ρ m ) represents the power storage matrix of the parameter unit, and "+=" is an accumulation operation;

[0200] ​A histogram acquisition unit, configured to cooperatively locate the position information, map it to the parameter space through Hough transform, traverse the position data vector group, end the accumulation operation, and obtain the reference space point number accumulation histogram and the power accumulation histogram;

[0201] An initial detection track acquisition unit, configured to set a point number accumulation threshold and a power accumulation threshold, and respectively perform threshold detection on the accumulation results of the parameter space parameter units based on the point number accumulation threshold and the power accumulation threshold, and perform inverse Hough mapping on the parameter units that pass the point number accumulation threshold detection and the power accumulation threshold detection to obtain an initial detection track;

[0202] An effective detection track acquisition unit, configured to merge the initial detection tracks with the same spot, and on the basis of the merge, delete the false initial tracks by means of speed and angle constraint conditions to obtain effective detection tracks corresponding one by one to the initial detection tracks.

[0203] In this embodiment, the sub - divided time - slice module T3 specifically includes:

[0204] When the binary Hough detection in step three is unsuccessful, reduce the data volume by a multiple of l / n, re - divide the data length, continue to project onto three planes, and respectively perform track detection within each time - slice. By connecting the straight - line detection results of multiple time - slices, the detection of the curved maneuvering track is finally realized. Here, l represents the number of rough - positioning coordinates in the position data vector group, and n represents the multiple by which the length of the position data vector group is reduced.

[0205] For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple. For related parts, refer to the description in the method part.

[0206] In this article, specific examples are used to elaborate on the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. An underwater weak magnetic moving target detection method based on unmanned dual-machines, characterized in that, Including: Step 1: Using the geomagnetic gradient measurement data of the first unmanned aerial vehicle (UAV) and the second UAV, cooperatively locate the underwater weak magnetic target in three-dimensional space to obtain the rough positioning position of the underwater weak magnetic target; the first UAV and the second UAV are UAVs equipped with geomagnetic gradient tensor meters. Step 2: Project the rough positioning positions at l moments onto the xoz plane, xoy plane, and zoy plane respectively, and form position data vector groups in the three planes respectively. Simultaneously perform binary Hough track detection using the three position data vector groups. If the target motion track can be detected, restore the corresponding track points to three-dimensional space to obtain the target three-dimensional motion track. Step 3: If the binary Hough track detection fails to detect the target, adopt a curve Hough track detection method based on subdivided time slices for track detection.

2. The underwater weak magnetic moving target detection method based on unmanned dual-machine according to claim 1, characterized in that, Step 1 specifically includes: Let an arbitrary point in space be set as the origin to establish a coordinate system. Then, at time k, the rough positioning coordinates r of the underwater weak magnetic target in the coordinate system d be r d ＝S1-r1 Among them, r1 is the relative position vector between the first UAV and the underwater weak magnetic target, and S1 is the position vector of the first UAV in the coordinate system.

3. The underwater weak magnetic moving target detection method based on unmanned dual machines according to claim 2, wherein The calculation formula of r1 is: r1 = -(G2 - G1) -1 (3G1 + G2)dr Among them, G1 and G2 are the magnetic field gradient tensors measured by the first UAV and the second UAV at the same moment respectively, and dr is the distance vector between the first UAV and the second UAV at the current moment.

4. The underwater weak magnetic moving target detection method based on unmanned dual-machine according to claim 1, characterized in that The binary Hough track detection in Step 2 specifically includes: Construct a binary data vector using the cooperative positioning information and magnetic field intensity information of the first UAV and the second UAV; the binary data vector includes position information and power information. The specific method is as follows: In the xoy plane, at the k-th moment, the rough positioning position information r of the underwater weak magnetic target obtained by the cooperative positioning of two UAVs d,xoy,k is r d,xoy,k = [x k y k ​ Power E at time k k : where, δ k (B k ) is Gaussian white noise and is independent at each moment, satisfying B k is the contribution value of the underwater weak magnetic target to the magnetic field intensity at the k-th moment; Adopt the standard Hough transform method to discretize the parameter space, and simultaneously establish a binary accumulation matrix in the parameter space. The binary accumulation matrix in the parameter space includes a point accumulation matrix D(·) and a power storage matrix E(·): D(θ t ,ρ m ) = D(θ t ,ρ m ) + 1 E(θ t ,ρ m )+ = E(θ t ,ρ m ) Among them, θ t , ρ m represent the center point coordinates of the parameter unit, D(θ t , ρ m ) represents the point accumulation matrix of the parameter unit, E(θ t , ρ m ) represents the power storage matrix of the parameter unit, and "+=” is the accumulation operation; The cooperative positioning position information is mapped to the parameter space through the Hough transform, traverse the position data vector group, end the accumulation operation, and obtain the reference space point accumulation histogram and power accumulation histogram. Set a point accumulation threshold and a power accumulation threshold, and respectively perform threshold detection on the accumulation results of the parameter space parameter units based on the point accumulation threshold and the power accumulation threshold. Perform inverse Hough mapping on the parameter units that pass the point accumulation threshold detection and the power accumulation threshold detection to obtain the initial detection track. Merge the initial detection tracks with the same point traces, and delete the false initial tracks on the basis of the merge by means of speed and angle constraint conditions to obtain valid detection tracks corresponding one by one to the initial detection tracks.

5. The underwater weak magnetic moving target detection method based on unmanned dual-machine according to claim 1, characterized in that, The curve Hough detection method based on subdivided time slices in Step 3 specifically includes: When the binary Hough detection in Step 3 is unsuccessful, reduce the data volume by a multiple of l / n, re-divide the data length, continue to project onto the three planes, and perform track detection in each time slice. Connect the straight line detection results of multiple time slices to finally achieve curve maneuvering track detection, where l represents the number of rough positioning coordinates in the position data vector group, and n represents the multiple by which the length of the position data vector group is reduced.

6. An underwater weak magnetic moving target detection system based on unmanned dual-vehicles, characterized in that, Including: The rough positioning position acquisition module is used to cooperatively locate underwater weak magnetic targets in three-dimensional space by using the geomagnetic gradient measurement data of the first unmanned aerial vehicle (UAV) and the second UAV, and obtain the rough positioning position of the underwater weak magnetic target; the first UAV and the second UAV are UAVs equipped with geomagnetic gradient tensor meters. The target three-dimensional motion trajectory determination module is used to project the above-mentioned rough positioning positions at l moments onto the xoz plane, xoy plane, and zoy plane respectively, and form position data vector groups in the three planes respectively. The binary Hough trajectory detection is performed simultaneously using the three position data vector groups. If the target motion trajectory can be detected, the corresponding trajectory points are restored to the three-dimensional space to obtain the target three-dimensional motion trajectory. The sub-divided time slice module is used to, if the binary Hough trajectory detection fails to detect the target, perform trajectory detection using the curve Hough trajectory detection method based on sub-divided time slices.

7. The underwater weak magnetic moving target detection method based on unmanned dual-robots according to claim 6, characterized in that, The rough positioning position acquisition module specifically includes: Let an arbitrary point in space be set as the origin to establish a coordinate system. Then, at time k, the rough positioning coordinates r of the underwater weak magnetic target in the coordinate system are d be r d ＝S1-r1 Among them, r1 is the relative position vector between the first UAV and the underwater weak magnetic target, and S1 is the position vector of the first UAV in the coordinate system.

8. The underwater weak magnetic moving target detection method based on unmanned dual machines according to claim 7, characterized in that, The calculation formula of r1 is: r1 = -(G2 - G1) -1 (3G1 + G2)dr Among them, G1 and G2 are the magnetic field gradient tensors measured by the first UAV and the second UAV at the same moment respectively, and dr is the distance vector between the first UAV and the second UAV at the current moment.

9. The underwater weak magnetic moving target detection method based on unmanned dual-robots according to claim 6, wherein The underwater weak magnetic target three-dimensional motion trajectory determination module specifically includes: Coarse positioning position information acquisition unit, taking the xoy plane as an example, at the k-th moment, the coarse positioning position information r of the underwater weak magnetic target obtained by the cooperative positioning of the two UAVs d,xoy,k be r d,xoy,k = [x k y k ​ Power acquisition unit, used to acquire the power E at time k k ; where, δ k (B k ) is Gaussian white noise and is independent at each moment, satisfying B k is the contribution value of the underwater weak magnetic target to the magnetic field intensity at the k-th moment; The binary accumulation matrix acquisition unit is used to discretize the parameter space by using the standard Hough transform method, and establish a binary accumulation matrix in the parameter space. The binary accumulation matrix in the parameter space includes a point number accumulation matrix D(·) and a power storage matrix E(·): D(θ t ,ρ m ) = D(θ t ,ρ m ) + 1 E(θ t ,ρ m )+=E(θ t ,ρ m ) Among them, θ t , ρ m represent the center point coordinates of the parameter unit, D(θ t , ρ m ) represents the point accumulation matrix of the parameter unit, E(θ t , ρ m ) represents the power storage matrix of the parameter unit, and "+=" is the accumulation operation; The histogram acquisition unit is used to map the cooperative positioning position information to the parameter space through the Hough transform, traverse the position data vector group, end the accumulation operation, and obtain the reference space point number accumulation histogram and the power accumulation histogram. The initial detection trajectory acquisition unit is used to set a point number accumulation threshold and a power accumulation threshold to perform threshold detection on the accumulation results of the parameter unit in the parameter space, and perform inverse Hough mapping on the parameter units passing through the binary threshold to obtain the initial detection trajectory. The high-confidence detection trajectory acquisition unit is used to merge the initial detection trajectories with the same point traces, and on the basis of the merger, delete the false initial trajectories by means of speed and angle constraint conditions to obtain the corresponding high-confidence detection trajectories, that is, it is considered that the underwater weak magnetic target is effectively detected.

10. The underwater weak magnetic moving target detection method based on unmanned dual-machine according to claim 6, characterized in that The sub-divided time slice module specifically includes: When the binary Hough detection in step three is unsuccessful, the data volume is reduced by a multiple of l / n, the data length is re-divided, and the data is projected onto the three planes again. In each time slice, trajectory detection is performed respectively, and the straight line detection results of multiple time slices are connected to finally realize the curve maneuver trajectory detection, where l represents the number of rough positioning coordinates in the position data vector group, and n represents the multiple by which the length of the position data vector group is reduced.

Images