Acoustic-magnetic composite intersection feature fusion-based region identification method

Abstract

The invention discloses a region identification method based on acoustic-magnetic composite intersection feature fusion, and the method comprises the following steps: 1, collecting n-dimensional electromagnetic and acoustic composite intersection feature quantities through an electromagnetic and underwater acoustic sensor, and forming n-dimensional intersection feature data S = {S1, S2,... Sn}; 2, performing consistency preprocessing on the n-dimensional intersection feature data S, and converting the n-dimensional intersection feature data S into a target area identification support degree X = {X1, X2,... Xn}; 3, calculating a feature combination reliability coefficient lambda according to the given optimal confidence Cbest = {C1, C2,... Cn} of each feature quantity; 4, solving the reliability M ({Ai}) of all possible feature combinations {Ai}; 5, calculating a composite value Fj (Ai) of all feature combinations {Ai} under the composite feature data according to the jth (j = 1, 2,... m) target area identification support degree Xi, j (i = 1, 2,... n); 6, calculating the credibility Rj (Ai) of all feature combinations {Ai} under the composite feature data according to the jth (j = 1, 2,... m) target area identification support degree Xi, j (i = 1, 2,... n); step 7, solving a fusion result Wj (S) of the jth (j = 1, 2,... m) n-dimensional intersection feature data of electromagnetic and acoustic composite detection; and 8, skipping to the step 5 until j is equal to m, and obtaining a final fusion result W (S).

Description

Technical Field

[0001] The present invention relates to the field of autonomous target area recognition algorithms, and particularly to a method for area recognition based on the fusion of acoustic and magnetic composite intersection features. Background Art

[0002] Underwater target area recognition technology uses physical field information such as electromagnetic and acoustic fields to achieve the detection and area recognition functions of objects within a certain distance underwater. This technology can be used in engineering fields such as sunken ship salvage or the detection and area positioning of specific underwater targets. Affected by the conductivity and permeability characteristics of seawater, current target area recognition mainly uses physical field detection technologies such as electromagnetic or underwater acoustic. The ability and fault tolerance of traditional single physical field detection means to obtain target information are relatively limited, and often the target detection or area recognition fails due to the influence of complex underwater environments or the failure of some sensors, making it difficult to meet the growing application requirements. Summary of the Invention

[0003] To solve the problems raised in the above background art, the technical solution adopted by the present invention is as follows:

[0004] A method for area recognition based on the fusion of acoustic and magnetic composite intersection features, comprising the following steps:

[0005] Step 1: Use electromagnetic and underwater acoustic sensors to collect n-dimensional electromagnetic and acoustic composite intersection feature quantities, forming n-dimensional intersection feature data S = {S1, S2, LS n}, and the data length of each feature quantity is m;

[0006] Step 2: According to the local decision principle of feature quantities, perform consistency preprocessing on the n-dimensional intersection feature data S and convert it into the target area recognition support degree X = {X1, X2, LX n};

[0007] Step 3: Calculate the reliability coefficient λ of the feature combination according to the optimal confidence degrees C best = {C1, C2, LC n} of the given feature quantities;

[0008] Step 4: Solve the reliability M({A i}) of all possible feature combinations {A i};

[0009] Step 5: Calculate the composite value F i,j (A i ) of all feature combinations {A j (A i ) under the composite feature data of the j (j = 1, 2,..., m) - th bit target area recognition support degree X

[0010] Step 6: From the support degree X of the j-th (j = 1, 2, … m) target area recognition i,j (i = 1, 2, Ln), calculate the credibility R of all feature combinations {A i} under the composite feature data of this position j (A i );

[0011] Step 7: According to the credibility R of the feature combination A i and the reliability M(A j ), solve the fusion result W i of the n-dimensional intersection feature data of the j-th (j = 1, 2, … m) position of electromagnetic and acoustic composite detection i (S); j ;

[0012] Step 8: Jump to Step 5 until j = m, that is, complete the fusion of all intersection feature data to obtain the final fusion result W(S).

[0013] Furthermore, in Step 1, the n-dimensional electromagnetic and acoustic composite intersection feature quantities include electromagnetic field intersection feature quantities: the electromagnetic field intensity S1 and phase information S2 of the z-axis received by the electromagnetic sensor, and the electromagnetic field intensity S3 of the x-axis;

[0014] It also includes acoustic field intersection feature data: the sound pressure amplitude S4 of the target noise and the target azimuth interval S5.

[0015] Furthermore, in Step 1, the electromagnetic field intersection feature quantities are obtained in the following way: a rod-shaped spiral coil type electromagnetic transmitting antenna is adopted, its axis is perpendicular to the longitudinal axis of the detection system, and x, y, z three-axis electromagnetic receiving coils are used, where the x-axis direction is the same as the longitudinal axis direction of the detection system, the z-axis is the vertical direction, and the electromagnetic field intensity S1 and phase information S2 of the z-axis, and the electromagnetic field intensity S3 of the x-axis are extracted respectively;

[0016] The acoustic field intersection feature quantities are obtained in the following way: a low-frequency underwater acoustic receiving transducer is adopted to collect the target radiation noise during the intersection process, the sound pressure amplitude S4 is extracted, and a medium-frequency active sonar is used to collect the target echo signal and calculate and generate the target azimuth interval S5.

[0017] Furthermore, in Step 2, when performing the consistency preprocessing operation on the n-dimensional intersection feature data S, the formula is as follows:

[0018]

[0019] where, X i,j is the consistency processing result of the j-th data point of the i-th dimension feature, S i,j represents the j-th data point of the i-th dimension feature, lj Denotes the j-th intersection point.

[0020] Furthermore, in step three, when calculating the reliability coefficient λ of the feature combination, the formula is as follows:

[0021]

[0022] Furthermore, in step four, when solving the reliability M({A i}), the formula is as follows:

[0023]

[0024] Wherein, assuming k is the number of types of feature quantities of the feature combination A i}.

[0025] Furthermore, in step five, when calculating the composite value of all feature combinations {A i}, the formula is as follows:

[0026]

[0027] Furthermore, in step six, when calculating the credibility of all feature combinations {A i}, the formula is as follows:

[0028]

[0029] Wherein, X i,j and X l,j are the target region recognition supports of any two feature quantities of the feature combination A i}, and a1, a2 are preset parameters.

[0030] Furthermore, in step seven, when solving the fusion result W j (S) of the n-dimensional intersection feature data, the formula is as follows:

[0031] G j ({A i ) = min(R j ({A i}), M({A i}))

[0032] W j (S) = F j (max(G j ({A i )})) index ) )

[0033] Wherein, G j ({A i ) is the feature combination {A i} of the fusion acceptance value, max(G j ({A i})) index is the feature combination A corresponding to the maximum fusion acceptance value i sequence number.

[0034] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0035] The regional recognition method based on acoustic-magnetic composite intersection feature fusion provided by the present invention, based on the composite detection of electromagnetic and underwater acoustic, collects multi-source information, can effectively fuse the multi-physical field features of the target, improve the detectability and recognition credibility of the target, increase the fault tolerance and self-adaptability of the system, and thus obtain a more reliable underwater target regional recognition effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1a and Figure 1b are the setting diagrams when collecting intersection feature quantities in the specific embodiment of the regional recognition method based on acoustic-magnetic composite intersection feature fusion provided by the present invention;

[0037] Figure 2a and Figure 2b are the schematic diagrams of the intersection process in the specific embodiment;

[0038] Figure 3 are the schematic diagrams of the change of the target area recognition support degree of each dimension feature with the intersection distance in the specific embodiment;

[0039] Figure 4 are the schematic diagrams of the change of the final fusion result of the visible intersection feature data with the intersection distance in the specific embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] In order to make the technical means, creative features, achieved purposes and effects of the present invention easy to understand, the present invention will be further described below in conjunction with the drawings and specific embodiments.

[0041] In a specific embodiment, the present invention provides a regional recognition method based on acoustic-magnetic composite intersection feature fusion, including the following steps:

[0042] Step 1: Use electromagnetic and underwater acoustic sensors to collect n-dimensional electromagnetic and acoustic composite intersection feature quantities, forming n-dimensional intersection feature data S = {S1, S2, LS n}, and the data length of each feature quantity is m.

[0043] In this embodiment, the n-dimensional electromagnetic and acoustic composite intersection feature quantities include electromagnetic field intersection feature quantities: the electromagnetic field intensity S1 and phase information S2 of the z-axis received by the electromagnetic sensor, and the electromagnetic field intensity S3 of the x-axis; and also include acoustic field intersection feature data: the sound pressure amplitude S4 of the target noise and the target azimuth interval S5.

[0044] Refer to Figure 1a as shown, Figure 1a is a side view. The detection system can be installed on an underwater vehicle. The electromagnetic field intersection feature quantities are obtained in the following way: a rod-shaped spiral coil electromagnetic transmitting antenna is used, whose axis is perpendicular to the longitudinal axis of the detection system. X, Y, and Z three-axis electromagnetic receiving coils are used, where the x-axis direction is the same as the longitudinal axis direction of the detection system, and the z-axis is the vertical direction. The electromagnetic field intensity S1 and phase information S2 of the z-axis, and the electromagnetic field intensity S3 of the x-axis are extracted respectively.

[0045] Refer to Figure 1b as shown, Figure 1b is a top view. The acoustic field intersection feature quantities are obtained in the following way: a low-frequency underwater acoustic receiving transducer is used to collect the target radiation noise during the intersection process, and the sound pressure amplitude S4 is extracted. A medium-frequency active sonar is used to collect the target echo signal and calculate and generate the target azimuth interval S5.

[0046] Refer to Figure 2a and Figure 2b as shown. In this embodiment, the intersection distance h between the detection system and the typical target is taken as 5 m. The intersection distance refers to the shortest distance between the detection system and the typical target during the intersection process. The intersection path considers a straight line interval range of -8 m to +8 m with the vertical projection point of the typical target on the intersection path as the origin, the approach process being negative and the departure process being positive, and the intersection angle is taken as 60°.

[0047] Step 2: According to the local decision principle of feature quantities, perform consistency preprocessing on the n-dimensional intersection feature data S and convert it into the target area recognition support degree X = {X1, X2, LX n}.

[0048] Specifically, when performing the consistency preprocessing operation on the n-dimensional intersection feature data S, the formula is as follows:

[0049]

[0050] where X i,j is the consistency processing result of the j-th data point of the i-th dimension feature, S i,j represents the j-th data point of the i-th dimension feature, and l j represents the j-th intersection point. Refer to Figure 3 as shown, which is a schematic diagram of the change of the target area recognition support degree of each dimension feature with the intersection distance in this embodiment.

[0051] Step 3: According to the optimal confidence levels C of the given characteristic quantities best ={C1, C2, LC n}, calculate the reliability coefficient λ of the characteristic combination. In this embodiment, C1 = 0.87, C2 = 0.84, C3 = 0.75, C4 = 0.82, and C5 = 0.92 are taken respectively.

[0052] Specifically, when calculating the reliability coefficient λ of the characteristic combination, the formula is as follows:

[0053]

[0054] Step 4: Solve the reliability M({A i}) of all possible characteristic combinations {A i}.

[0055] Specifically, when solving the reliability M({A i}), the formula is as follows:

[0056]

[0057] Among them, assume that k is the number of types of characteristic quantities of the characteristic combination A i .

[0058] Step 5: Calculate the composite value F i,j (A i ) of all characteristic combinations {A j (A i ) under the composite characteristic data of the j-th (j = 1, 2,..., m) target area recognition support X

[0059] Specifically, when calculating the composite value of all characteristic combinations {A i}, the formula is as follows:

[0060]

[0061] Step 6: Calculate the credibility R i,j (A i ) of all characteristic combinations {A j (A i ) under the composite characteristic data of the j-th (j = 1, 2,..., m) target area recognition support X

[0062] Specifically, when calculating the credibility of all characteristic combinations {A i}, the formula is as follows:

[0063]

[0064] Among them, X i,j and X l,j are the target area recognition supports of any two characteristic quantities of the feature combination A i . a1 and a2 are preset parameters. In this embodiment, a1 can take 0.4 and a2 can take 0.7.

[0065] Step Seven: According to the credibility R i (A j ) and the reliability M(A i ), solve the fusion result W i (S) of the j-th (j = 1, 2,... m) n-dimensional intersection characteristic data of electromagnetic and acoustic composite detection. j (S).

[0066] Specifically, when solving the fusion result W j (S) of the n-dimensional intersection characteristic data, the formula is as follows:

[0067] G j ({A i}) = min(R j ({A i}), M({A i}))

[0068] W j (S) = F j (max(G j ({A i})) index )

[0069] Among them, G j ({A i}) is the fusion acceptance value of the feature combination {A i}, and max(G j ({A i})) index [[ID="]] is the serial number of the feature combination A i corresponding to the maximum fusion acceptance value.

[0070] Step Eight: Jump to Step Five until j = m, that is, complete the fusion of all intersection characteristic data to obtain the final fusion result W(S). Figure 4 This is the schematic diagram of the change of the final fusion result of the visible intersection characteristic data with the intersection distance in this embodiment.

[0071] In summary, the regional recognition method based on the fusion of acoustic-magnetic composite intersection features provided by the present invention collects multi-source information based on the combined detection of electromagnetic and underwater acoustic signals, can effectively fuse the multi-physical field features of the target, improve the detectability and recognition credibility of the target, increase the fault tolerance and self-adaptability of the system, and thus obtain a more reliable underwater target regional recognition effect.

[0072] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A regional recognition method based on the fusion of acoustic and magnetic composite intersection features, characterized in that It includes the following steps: Step 1: Use electromagnetic and underwater acoustic sensors to collect n-dimensional electromagnetic and acoustic composite intersection feature quantities, forming n-dimensional intersection feature data S = {S1, S2, ...} n , where the data length of each feature quantity is m; Step 2: According to the local decision principle of feature quantity, perform consistency preprocessing on the n-dimensional intersection feature data S and convert it into the target area recognition support degree X = {X1, X2, LX n}; Step 3: Calculate the reliability coefficient λ of the feature combination according to the optimal confidence levels C of the given feature quantities best ={C1, C2, LC n}, Step 4: Solve for the reliability M({A i}) of all possible feature combinations {A i}); Step Five: Calculate the composite value F i,j (i = 1, 2, …, n) of all feature combinations {A i} under the composite feature data of the j-th (j = 1, 2, …, m) target region recognition support degree X j (A i ); Step 6: Calculate the credibility R i,j (i = 1, 2, …, n) of all feature combinations {A i} under the composite feature data of the j-th (j = 1, 2, …, m) target region recognition support degree X j (A i ); Step Seven: According to the confidence level R i of the feature combination A j (A i ) and the reliability M(A i ), solve the fusion result W j (S) of the j-th (j = 1, 2, … m) n-dimensional intersection feature data of electromagnetic and acoustic composite detection; Step Eight: Jump to Step Five until j = m, that is, complete the fusion of all intersection feature data to obtain the final fusion result W(S).

2. The regional recognition method based on acoustic-magnetic composite intersection feature fusion according to claim 1, characterized in that In Step One, the n-dimensional electromagnetic and acoustic composite intersection feature quantity includes an electromagnetic field intersection feature quantity: the electromagnetic field intensity S1 and phase information S2 of the z-axis received by an electromagnetic sensor, and the electromagnetic field intensity S3 of the x-axis; It also includes acoustic field intersection feature data: the sound pressure amplitude S4 of the target noise and the target azimuth interval S5.

3. The regional recognition method based on acoustic-magnetic composite intersection feature fusion according to claim 2, characterized in that In Step One, the electromagnetic field intersection feature quantity is obtained in the following way: a rod-shaped spiral coil type electromagnetic transmitting antenna is adopted, whose axis is perpendicular to the longitudinal axis of the detection system, and x, y, z three-axis electromagnetic receiving coils are used, where the x-axis direction is the same as the longitudinal axis direction of the detection system, the z-axis is the vertical direction, and the electromagnetic field intensity S1 and phase information S2 of the z-axis, and the electromagnetic field intensity S3 of the x-axis are extracted respectively; The acoustic field intersection feature quantity is obtained in the following way: a low-frequency underwater acoustic receiving transducer is adopted to collect the target radiation noise during the intersection process, the sound pressure amplitude S4 is extracted, and an intermediate-frequency active sonar is used to collect the target echo signal and calculate and generate the target azimuth interval S5.

4. The regional recognition method based on acoustic-magnetic composite intersection feature fusion according to claim 2, wherein In Step Two, when performing the consistency preprocessing operation on the n-dimensional intersection feature data S, the formula is as follows: Among them, X i,j is the consistency processing result of the j-th data point of the i-th dimensional feature, S i,j represents the j-th data point of the i-th dimensional feature, l j represents the j-th intersection point.

5. The regional recognition method based on acoustic-magnetic composite intersection feature fusion according to claim 4, characterized in that In Step Three, when calculating the feature combination reliability coefficient λ, the formula is as follows:

6. The regional recognition method based on acoustic-magnetic composite intersection feature fusion according to claim 5, wherein In Step 4, when solving the reliability M({A i}), the formula is as follows: Among them, assume that k is the number of types of feature quantities of the feature combination A i of.

7. The regional recognition method based on acoustic-magnetic composite intersection feature fusion according to claim 6, characterized in that In step five, when calculating the composite values of all feature combinations {A i}, the formula is as follows:

8. The regional recognition method based on acoustic-magnetic composite intersection feature fusion according to claim 7, wherein In Step 6, when calculating the credibility of all feature combinations {A i}, the formula is as follows: Among them, X i,j and X l,j are the target region recognition supports of any two feature quantities of the feature combination A i , and a1 and a2 are preset parameters.

9. The regional recognition method based on acoustic-magnetic composite intersection feature fusion according to claim 8, wherein In step seven, solve the fusion result W of the n-dimensional intersection feature data j (S), the formula is as follows: G j ({A i}) = min(R j ({A i}), M({A i})) W j (S) = F j (max(G j ({A i})) index ) Among them, G j ({A i}) is the fusion acceptance value of the feature combination {A i}, and max(G j ({A i})) index is the serial number of the feature combination A corresponding to the maximum fusion acceptance value i .

Images