Angle measurement data association method under position unknown condition based on minimum angle distance and

Abstract

The application discloses a position-unknown condition-based angle measurement data association method based on minimum angle distance sum, comprising the following steps: inputting angle measurements of all optical sensors; calculating the angle distance between the angle measurements of any two optical sensors based on the angle measurements of the optical sensors; taking the minimum angle distance sum corresponding to the association relationship as the target, and constructing an association relationship target function; and solving the target function to obtain the best association result. The calculation process of the application does not involve the position information of the optical sensors, thus breaking the dependence of the classical method on the position information of the optical sensors, and being applicable to multi-to-multi combat scenes based on low-cost mobile platforms, such as unmanned aerial vehicle swarm confrontation, and avoiding the problem of performance decline of the data association algorithm caused by the failure or insufficient precision of the platform navigation positioning function in the electromagnetic interference environment. The result of the data association can provide information support for target allocation and cooperative attack.

Description

Technical Field

[0001] This invention relates to the field of sensor information processing, and in particular to a method for associating angle measurement data under unknown position conditions based on minimum angular distance and position. Background Technology

[0002] Utilizing multi-platform collaborative defense against target swarms such as drone swarms and warhead swarms can effectively improve interception efficiency. Data association technology can determine the correspondence between target data from different platforms, laying the foundation for target allocation and coordinated strikes. Considering factors such as accuracy, size, weight, cost, and power consumption, many platforms choose optical sensors (infrared or visible light) for target detection. Due to the small size and long distance of targets, they often only represent a few pixels (i.e., point targets) on the imaging plane of optical sensors. Furthermore, swarms or multiple warhead targets often have similar appearances, making it difficult to associate them using features such as color, shape, and texture. Angle measurement becomes a crucial method for target data association. Researchers typically determine the correlation between angle measurements based on the distance between lines of sight. A line of sight refers to a ray originating from the optical sensor position and determined by the target angle measurement direction. Under ideal conditions with no measurement errors, lines of sight corresponding to the same target intersect at the target's location, with zero distance between them. Therefore, the closer the distance between lines of sight, the greater the likelihood that they correspond to the same target. Calculating target lines of sight requires precise knowledge of the sensor's position coordinates; the higher the spatial density of targets, the higher the requirement for the sensor's own positional accuracy. For mobile platforms, satellite navigation or inertial navigation systems are typically used to determine their position coordinates. Satellite signals can be interfered with or blocked, and inertial navigation positioning errors accumulate over time, often failing to meet the accuracy requirements for line-of-sight intersection. If the correlation between target data can be determined even when the sensor positions are unknown, it not only avoids problems such as interference with the navigation positioning system or the introduction of measurement errors, but also saves the high cost of high-precision positioning systems.

[0003] For multi-sensor, multi-target data association problems under conditions where sensor locations are unknown, researchers typically rely on the topological relationships between target data for association, proposing data association algorithms based on reference patterns (REP). A reference pattern refers to the topological structure between a given target location and its neighboring target locations. The problem with these methods is that they all assume the sensors have complete target location information (e.g., xyz coordinates in Cartesian coordinates, or distance and angle information in polar coordinates), essentially solving a correlation problem between point sets. However, this invention aims to solve the correlation problem between angles (lines of sight), inferring the association between three-dimensional targets from two-dimensional measurements—inferring a high-dimensional state from low-dimensional information—which is significantly more challenging. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method for associating angle measurement data under conditions where the minimum angular distance and position are unknown.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] This invention provides a method for associating angle measurement data under unknown position conditions based on minimum angular distance, including:

[0007] Input the angle measurements from all optical sensors;

[0008] Based on the angle measurements of each optical sensor, calculate the angular distance between any two angle measurements of the optical sensors.

[0009] To find the minimum angular distance sum corresponding to the association relationship, an objective function for the association relationship is constructed.

[0010] Solve the objective function to obtain the optimal correlation result.

[0011] Furthermore, for the j-th and k-th optical sensors, j = 1,...,N S k = 1, ..., N S And j≠k, the angular distance between any two angle measurements of the j-th optical sensor and the k-th optical sensor is calculated through the following steps:

[0012] Obtain the p-th angle measurement of the j-th optical sensor and the q-th angle measurement of the k-th optical sensor Where p = 1,...,M j , q=1,...,M k M j M represents the number of angle measurements in the j-th optical sensor angle measurement set. k This represents the number of angle measurements in the k-th optical sensor angle measurement set;

[0013] Get Corresponding angle measurement direction vector Their corresponding equivalent target lines of sight are respectively

[0014] based on calculate Angular distance between;

[0015]

[0016] Where dot represents the dot product of vectors, and |·| represents the magnitude of the vector; Represents the equivalent target line of sight The angle formed.

[0017] Furthermore, suppose there exist N in the current scenario. S There are N optical sensors, of which N S =2, with the goal of finding the minimum angular distance sum corresponding to the association relationship, the association relationship objective function is constructed as follows:

[0018]

[0019]

[0020] Where a* represents the best association result; It is a binary indicator function, which is defined when (i1, i2) form a tuple. otherwise

[0021] Furthermore, the methods for solving the objective function of the association relationship include the Hungarian algorithm and the auction algorithm.

[0022] Furthermore, suppose there exist N in the current scenario. S There are N optical sensors, of which N S =3, with the goal of finding the minimum angular distance sum corresponding to the association relationship, construct the association relationship objective function, including:

[0023] Based on the angular distance between any two optical sensor angle measurements, construct the correlation cost matrix between any two optical sensor measurements.

[0024] Calculate the m-best correlation results of the correlation cost matrix between any two optical sensor measurements;

[0025] Based on the m-best association results, a set of tuples containing the most likely associations between angle measurements from any two optical sensors is obtained.

[0026] For each tuple in the tuple set, calculate the association cost corresponding to each tuple;

[0027] A correlation cost matrix between multiple optical sensor measurements is constructed based on the correlation cost corresponding to each tuple.

[0028] Furthermore, the construction of the correlation cost matrix between any two optical sensor measurements based on the angular distance between the angle measurements of any two optical sensors includes:

[0029] Construct the correlation cost matrix CM between any two sensor measurements. (1,2) CM (1,3) and CM (2,3) ,as follows:

[0030]

[0031] Further, the calculation of the m-best correlation results between any two optical sensor measurements includes:

[0032] Calculate CM using the Murty method (1,2) CM (1,3) and CM (2,3) m-best association results (j,k)∈{(1,2),(1,3),(2,3)}; This represents the i-th correlation result between the measurements of sensors j and k.

[0033] Furthermore, the set of tuples obtained based on the m-best correlation results, which contains elements with a high probability of correlation between angle measurements from any two optical sensors, includes:

[0034] For each Perform the following operations

[0035]

[0036] Obtain the union ε of the m-best correlation results measured in optical sensors j and k. (j,k) .

[0037] Furthermore, for each tuple in the tuple set, calculating the association cost corresponding to each tuple includes:

[0038] Iterate through all (i1, i2, i3) and calculate their corresponding association costs using the following formula.

[0039]

[0040] in, The cost represents the tuple (i1, i2, i3).

[0041] Furthermore, the correlation cost matrix between multi-optical sensor measurements, constructed based on the correlation cost corresponding to each tuple, is as follows:

[0042]

[0043] in, It is an indicator function that indicates when (i1, i2, i3) form a tuple. otherwise

[0044] Compared with the prior art, the beneficial technical effects of the present invention are as follows:

[0045] The method for associating angle measurement data from unknown sensors based on minimum angular distance provided by this invention does not involve the position information of the optical sensor itself in its calculation process, thus eliminating the dependence of classical methods on the optical sensor's own position information. The method described in this invention is applicable to many-to-many combat scenarios based on low-cost mobile platforms, such as UAV swarm warfare. It avoids the performance degradation problem of data association algorithms caused by the failure or insufficient accuracy of platform navigation and positioning functions under electromagnetic interference environments. The data association results provided by this invention can provide information support for target allocation and coordinated attacks. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0047] Figure 1 A schematic flowchart of a method for associating angle measurement data of an unknown position sensor based on minimum angular distance and one embodiment is provided.

[0048] Figure 2 Figure (a) shows the effect of the difference in the observation angle of the optical sensor on the data association algorithm based on the reference mode. Figure (b) shows the effect when the observation angle of the optical sensor is almost the same, and Figure (a) shows the effect when the observation angle of the optical sensor is significantly different.

[0049] Figure 3 This is a schematic diagram of an equivalent data association scenario that does not rely on optical sensor position information, provided in one embodiment. Figure 3 (a) is a classic data association scenario. Figure 3 (b) is a scenario for equivalent data association;

[0050] Figure 4 A schematic diagram illustrating the definition of angular distance in one embodiment;

[0051] Figure 5 This is a schematic diagram of target angle measurement and calculation provided in one embodiment;

[0052] Figure 6 This is a design diagram of a two-optical-sensor scenario experiment, provided as an example. Figure 6 (a) is a side view of pose combination 1. Figure 6 (b) is a front view of pose combination 1. Figure 6 (c) is a front view of pose combination 2. Figure 6 (d) is a front view of pose combination 3. Figure 6 (e) is a front view of pose combination 4;

[0053] Figure 7 This is an experimental scenario diagram of two optical sensors provided in one embodiment, wherein... Figure 7 (a) is a schematic diagram of sensor pose combination 1. Figure 7 (b) is a schematic diagram of sensor pose combination 2. Figure 7 (c) is a schematic diagram of sensor pose combination 3. Figure 7 (d) is a schematic diagram of sensor pose combination 4;

[0054] Figure 8 for Figure 7 The attached diagram shows the actual hanging rack (labeled 1).

[0055] Figure 9 for Figure 7 The attached image shows a picture of the luminous beads labeled 2.

[0056] Figure 10 for Figure 7 The attached figure, labeled 3, shows a physical image of an industrial camera equipped with an attitude sensor. Figure 10 (a) is a schematic diagram of an industrial camera equipped with an attitude sensor. Figure 10 (b) is a picture of an industrial camera. Figure 10 (c) is a picture of the actual attitude sensor;

[0057] Figure 11 A target photograph taken by a camera as provided in one embodiment;

[0058] Figure 12 This is an experimental result diagram of two optical sensors provided in one embodiment, wherein... Figure 12 (a) is an example diagram of the association result of pose combination 1. Figure 12 (b) is an example diagram of the pose combination 2 association result. Figure 12 (c) is an example diagram of the pose combination 3 association result. Figure 12 (d) is an example diagram of the pose combination 4 association results;

[0059] Figure 13 This is a schematic diagram of an experimental design for a three-optical-sensor scenario, provided as an example.

[0060] Figure 14 An experimental scenario diagram of three optical sensors provided in one embodiment;

[0061] Figure 15 An experimental result diagram of a three-optical-sensor system is provided for one embodiment;

[0062] Attached image captions:

[0063] 1. Hanging rack; 2. Luminous beads; 3. Industrial camera equipped with attitude sensor. Detailed Implementation

[0064] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0065] Existing reference-mode methods use the relative positions of targets on the imaging plane of an optical sensor as a reference mode to calculate the correlation between targets. However, this method is only applicable to scenes where the optical sensors are oriented almost identically. As the sensor orientations differ, the relative positions of the targets on the imaging plane undergo perspective transformation, altering the original positional relationships and causing the method to fail. Figure 2 ,exist Figure 2 In (a), the observation angles of the two optical sensors are almost the same, and the reference patterns of target A on the imaging planes of the two sensors are also almost the same, making the reference pattern method effective. However, in Figure 2 In (b), the observation angles of the two sensors differ significantly, and the reference patterns of target A on the imaging planes of the two sensors differ significantly, thus the reference pattern method fails.

[0066] Reference Figure 1 In one embodiment, a method for associating position-unknown sensor angle measurement data based on minimum angular distance sum is provided, including:

[0067] Input the angle measurements from all optical sensors;

[0068] Based on the angle measurements of each optical sensor, calculate the angular distance between any two angle measurements of the optical sensors.

[0069] To find the minimum angular distance sum corresponding to the association relationship, an objective function for the association relationship is constructed.

[0070] Solve the objective function to obtain the optimal correlation result.

[0071] Reference Figure 5 Obtaining an angle measurement involves the following steps:

[0072] Assuming the target is in the optical sensor S s The coordinates on the imaging plane are First, calculate the deflection angle of the target in the optical sensor coordinate system. Pitch angle

[0073]

[0074] Where cu and cv represent the camera principal point coordinates; fu and fv represent the units of converting the focal length into pixel coordinates in the u and v directions.

[0075] The orientation vector of the optical sensor's optical axis in the world coordinate system is calculated using the attitude sensor; assuming the attitude sensor reading is φ. x φ y φ z Rotate the y-axis direction vector in the world coordinate system around the z-axis, x-axis, and y-axis by φ respectively. x φ y φ z This yields the direction vector of the optical sensor's optical axis in the world coordinate system; the direction vector of the optical sensor's principal axis is then rotated around the z′ axis. Then rotate around the x′ axis This allows us to obtain the direction vector of the optical sensor's optical axis in the world coordinate system.

[0076] The angle measurement of the target, that is, the deflection angle of the target in the world coordinate system. Pitch angle Calculated using the following formula

[0077]

[0078] As can be seen from the above, calculating the target angle measurement through an optical sensor requires knowing the target's coordinates on the imaging plane, the sensor's internal parameters, and the sensor's attitude, but does not require knowing the sensor's own position information.

[0079] The set of angle measurements from all optical sensors is denoted as The set of angle measurements from the s-th optical sensor is denoted as .

[0080]

[0081] Where, N S Represents the number of optical sensors; Representing set Z (s) The i-th s Measurement at each angle; M s Representing Z (s) Number of mid-angle measurements;

[0082]

[0083] in, These represent the target's yaw and pitch angles in the world coordinate system, respectively; w A This represents the error in angle measurement.

[0084] The correlation result 'a' divides the set of angle measurements from all optical sensors into n T A subset, using tuples Record the i-th T Number of individual sub-concentrated angle measurements

[0085]

[0086] Where ξ records i T and its corresponding tuple The mapping relationship between them.

[0087] The association result is denoted as a

[0088]

[0089] Where, n T This represents the number of target assumptions.

[0090] Since each target can only generate one angle measurement on a single optical sensor, and each tuple can only correspond to one target, the indices of the angle measurements must satisfy a one-to-one mapping constraint. If an optical sensor misses a target corresponding to a tuple, a dummy measurement number 0 is added to the corresponding index position of that tuple.

[0091] If a certain association result 'a' and its contained tuples satisfy the above constraints, then 'a' is called a feasible solution, and the set of all feasible solutions is denoted as . The correctness of 'a' is usually evaluated using the cost function Cost(a) of the association result. Generally speaking, the higher the probability of the association result being correct, the lower its corresponding cost. Based on this, solving the data association problem is equivalent to finding a feasible solution 'a' that minimizes the cost of the association result. The solution process can be expressed as follows:

[0092]

[0093] Among them, a * This indicates the best correlation result.

[0094] For the j-th and k-th optical sensors, j = 1,...,N S k = 1, ..., N S And j≠k, the angular distance between any two angle measurements of the j-th optical sensor and the k-th optical sensor is calculated through the following steps:

[0095] Obtain the p-th angle measurement of the j-th optical sensor and the q-th angle measurement of the k-th optical sensor Where p = 1,...,Mj , q=1,...,M k M j M represents the number of angle measurements in the j-th optical sensor angle measurement set. k This represents the number of angle measurements in the k-th optical sensor angle measurement set;

[0096] Get Corresponding angle measurement direction vector Their corresponding equivalent target lines of sight are respectively

[0097] The so-called equivalent target line of sight originates from an equivalent data association scenario provided by this invention that does not rely on optical sensor position information. Classic data association scenarios include... Figure 3 As shown in (a), the equivalent scenario is as follows: Figure 3 As shown in (b), the classic scenario can be understood as follows: the target remains stationary, and the sensor generates a measurement set Z at position S1. (1) Then, maintaining the same orientation, it moves along the vector. The direction is shifted to position S2 to generate the measurement set Z. (2) The equivalent scenario can be understood as follows: Assume the sensor generates a measurement set Z at position S1. (1) Then the sensor remains stationary, and the target moves along the vector. directional movement distance At the new location, generate the measurement set Z. (2) More generally, in equivalent scenarios, the sensor position is denoted as S; the target's initial position is denoted as T1. (1) , The set of targets located at these positions is called "equivalent target set 1", corresponding to the angle measurement generated by sensor S1; the targets are arranged according to vectors. The positions after the movement are denoted as T1. (2) , The set of targets located at these positions is called "equivalent target set 2", corresponding to the angle measurements generated by sensor S2. Clearly, solving the data association problem in scenario 1 is equivalent to solving the data association problem in scenario 2.

[0098] based on calculate Angular distance between;

[0099]

[0100] Where dot represents the dot product of vectors, and |·| represents the magnitude of the vector; Represents the equivalent target line of sight The angle formed.

[0101] In one embodiment, reference is made to Figure 4 Assuming The corresponding angle measurement direction vectors are respectively Their corresponding equivalent target lines of sight are ST1 (1) ST1 (2) ,So and The angular distance between them is:

[0102]

[0103] The angular distance and SumAD(a) corresponding to the associated result 'a' are defined as follows:

[0104]

[0105] The angle distance SumAD(a) corresponding to the association result 'a' refers to the sum of the angle distances corresponding to all associations in the association result.

[0106] In one embodiment, assume that there are N in the current scenario. S There are N optical sensors, of which N S =2, with the goal of finding the minimum angular distance sum corresponding to the association relationship, the association relationship objective function is constructed as follows:

[0107]

[0108] Among them, a * This indicates the best correlation result; It is a binary indicator function, which is defined when (i1, i2) form a tuple. otherwise

[0109] Equations (12) and (13) are standard two-dimensional assignment problems, and the cost between any two data points can be represented by angular distance. The methods for solving the objective function of the association relationship include the Hungarian algorithm and the auction algorithm.

[0110] In one embodiment, assume that there are N in the current scenario. S There are N optical sensors, of which N S =3, with the goal of finding the minimum angular distance sum corresponding to the association relationship, construct the association relationship objective function, including:

[0111] Based on the angular distance between any two optical sensor angle measurements, construct the correlation cost matrix between any two optical sensor measurements.

[0112] Calculate the m-best correlation results of the correlation cost matrix between any two optical sensor measurements;

[0113] Based on the m-best association results, a set of tuples containing the most likely associations between angle measurements from any two optical sensors is obtained.

[0114] For each tuple in the tuple set, calculate the association cost corresponding to each tuple;

[0115] A correlation cost matrix between multiple optical sensor measurements is constructed based on the correlation cost corresponding to each tuple.

[0116] The correlation cost matrix between multiple optical sensor measurements, constructed based on the correlation cost matrix corresponding to each tuple, is as follows:

[0117]

[0118] in, It is an indicator function that indicates when (i1, i2, i3) form a tuple. otherwise Let represent the cost of the tuple (i1, i2, i3).

[0119]

[0120] Equation (14) is the standard generalized 3D assignment problem, which can be solved using the Lagrange relaxation technique. The cost function expressed by Equation (16) can effectively resolve the association ambiguity problem.

[0121] The so-called cost matrix contains three dimensions, with each element recording the cost corresponding to a certain association (i.e., a tuple). Obviously, by traversing all (i1, i2, i3) and using equations (10) and (16), the association cost matrix CM can be directly constructed. However, the association cost matrix constructed in this way is dense. Here, "dense" means that all elements in the matrix have values. Although theoretically, this "denseness" does not affect the attribution of the correct association result, for the Lagrange relaxation technique, this "denseness" seriously affects its quality search. This is mainly because the "dense" matrix makes the search space of feasible solutions huge, resulting in greater uncertainty in the optimization result.

[0122] The method for constructing a correlation cost matrix between any two optical sensor measurements based on the angular distance between the angle measurements of any two optical sensors includes:

[0123] Construct the correlation cost matrix CM between any two sensor measurements. (1,2) CM (1,3) and CM (2,3) ,as follows:

[0124]

[0125] The calculation of the m-best correlation results between any two optical sensor measurements, respectively, includes:

[0126] Calculate CM using the Murty method (1,2) CM (1,3) and CM (2,3) m-best association results (j,k)∈{(1,2),(1,3),(2,3)}; This represents the i-th correlation result between the measurements of sensors j and k.

[0127] The set of tuples obtained based on m-best correlation results, containing those with a high probability of correlation between angle measurements from any two optical sensors, includes:

[0128] For each Perform the following operations

[0129]

[0130] Obtain the union ε of the m-best correlation results measured in optical sensors j and k. (j,k) ; the obtained ε (j,k) It contains a set of elements that are likely to be correlated with the measurements of sensors j and k, and this likelihood is set by the value of m-best.

[0131] For each tuple in the set of tuples, the association cost corresponding to each tuple is calculated, including:

[0132] Iterate through all (i1, i2, i3) and calculate their corresponding association costs using the following formula.

[0133]

[0134] in, The cost represents the tuple (i1, i2, i3).

[0135] After the above processing, the multi-optical sensor association cost matrix CM becomes sparse, with a large number of elements equal to Infinity. This significantly reduces the number of feasible solutions, lowering the optimization difficulty and effectively improving the association accuracy of the algorithm. Generally, the smaller the value of m-best, the easier the optimization algorithm becomes. When m-best equals 1, the number of feasible solutions in the solution space is 1, and the association result can be obtained without optimization. However, this may miss some possible association results (e.g., in scenarios where there is ambiguity in the association between two sensors), leading to a decrease in association accuracy. Therefore, the value of m-best is usually set to be greater than 1, and gradually increases as the number of targets increases.

[0136] In one embodiment, reference is made to Figure 6 , Figure 6 This is a design diagram for a two-sensor scenario experiment, which includes two industrial cameras, denoted as S1 and S2. (Refer to...) Figure 10 Each camera is equipped with one attitude sensor to measure camera attitude. (See reference...) Figure 9 Use glow-in-the-dark beads to simulate the target, for reference. Figure 8 It is attached to the hanging rack with a thin, transparent thread. The hanging rack is 1.8m away from the camera. See the side view of the scene. Figure 6 (a). To verify the effectiveness of the algorithm under different observation angle conditions, four different sensor pose combinations were designed in the experiment, as shown below. Figure 6 (b)-(e). In pose combination 1, the two cameras face the target group, with a distance of approximately 0.1m between them and a height of approximately 0.3m above the ground. Compared to pose combination 1, in pose combination 2, sensor S1 rolls approximately 30 degrees around the optical axis. Pose combinations 3 and 4 progressively increase the angle between the optical axes of the sensors.

[0137] Actual experimental scenarios such as Figure 7 As shown. The image resolution of both cameras is 1600 pixels * 1200 pixels. The internal parameters were obtained using the Zhang Zhengyou calibration method, and the specific values ​​are shown in Table 1. Here, cu and cv represent the principal point coordinates of the camera, and fu and fv represent the units of the focal length converted into pixel coordinates in the u and v directions. The attitude sensor used is the WT9011DCL-BT50 attitude sensor manufactured by Shenzhen Weite Intelligent Technology Co., Ltd. The readings of the S1 and S2 attitude sensors in pose combinations 1-4 are shown in Table 2. The diameter of the target luminous beads is 0.08 μm, and they are distinguished by circular stickers of different colors.

[0138] Table 1 Camera Internal Parameters

[0139]

[0140] Table 2 Attitude sensor readings

[0141]

[0142] The experimental process includes three steps.

[0143] Step 1: Turn off all the lights in the room, and take pictures of the target with both cameras (example photos shown). Figure 11 As shown in the figure, the target on the image is then detected by threshold segmentation and connected component analysis, and the target angle measurement is calculated using the method described in this invention.

[0144] Step 2: Use the algorithm proposed in this invention to solve the correlation between the target angle measurements obtained by the two sensors, and number the targets in the scene according to the correlation results.

[0145] Step 3: Turn on the room lights and compare the data association results with the real scene. If the stickers corresponding to the same number are the same color, it means that the association results are correct.

[0146] Figure 11 The experimental process and results are shown for four pose combination scenarios. The following example uses sensor pose combination 4 (i.e., Figure 12 (d) will be described in detail.

[0147] Step 1 Experimental Results: In a dark environment, S1 and S2 respectively detected targets in their fields of view and obtained a set of target angle measurements.

[0148] Step 2 Experimental Results: After calculation by the algorithm proposed in this patent, the association result a = {(1,3),(2,1),(3,4),(4,2),(5,5)} is obtained. Based on this, the targets in the scene are numbered, and the correspondence between the angle measurements of each sensor and the targets is shown in Table 3. The targets are numbered according to the sorting of the binary pairs in the association result.

[0149] Table 3. Correspondence between Angle Measurements and Targets

[0150]

[0151] Step 3 Experimental Results: Check whether the sticker colors corresponding to the same numbered targets in the two sensors are the same. The results are shown in Table 4.

[0152] Table 4. Checklist for Correctness of Association Results

[0153]

[0154] The inspection results show that stickers corresponding to the same numbered targets have the same color, indicating that the association result is correct. The experimental results demonstrate that the data association algorithm assigns the same number to the same target detected by different sensors, enabling different sensors to form a unified understanding of targets in the scene. Using the same method, the association results for the other three pose combinations are also correct.

[0155] In one embodiment, reference is made to Figure 13 , Figure 13 This diagram illustrates the experimental design for a 3-sensor scenario, which includes three industrial cameras, denoted as S1, S2, and S3. The experimental method is similar to that of the 2-sensor scenario and will not be repeated here. The actual experimental scenario is as follows. Figure 14 As shown in Table 5, the internal parameters of the three cameras are shown in Table 6, and their corresponding attitude sensor readings are shown in Table 6.

[0156] The same steps are used as in the two-sensor scenario.

[0157] Step 1 Experimental Results: In a dark environment, three cameras detected targets in their respective fields of view, obtaining a set of target angle measurements.

[0158] Step 2 Experimental Results: After calculation by the algorithm proposed in this patent, the association result a = {(1,2,2),(2,4,4),(3,1,1),(4,3,3),(5,5,5)} is obtained. Based on this, the targets in the scene are numbered, and the correspondence between the angle measurements of each sensor and the targets is shown in Table 5. The targets are numbered according to the order of the tuples in the association result.

[0159] Step 3 Experimental Results: Check whether the sticker colors corresponding to the same numbered targets in the three sensors are the same. The results are shown in Table 6.

[0160] Table 5. Correspondence between Angle Measurements and Targets

[0161]

[0162] Table 6. Checklist for Correctness of Association Results

[0163]

[0164] Reference Figure 15 The inspection results showed that the stickers corresponding to the same number were the same color, and the association result was correct.

[0165] Matters not covered in this invention are common knowledge.

[0166] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0167] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

[0168] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for data association based on angle measurement under position unknown condition with minimum angular distance and, characterized in that, The method comprises the following steps: inputting angle measurements of all optical sensors; calculating an angular distance between angle measurements of any two optical sensors based on the angle measurements of the optical sensors; constructing a correlation relationship objective function aiming at finding a minimum angular distance corresponding to a correlation relationship; solving the objective function to obtain an optimal correlation result; For the first j and the second k optical sensor, , , and , the angular distance between any two angular measurements of the first j optical sensor and the second k optical sensor is calculated by the steps of: Get the j The first optical sensor p Measurement of each angle and the k The first optical sensor q Measurement of each angle ,in, , ; Representing the j The number of angle measurements in a set of optical sensor angle measurements Representing the k The number of angle measurements in a set of optical sensor angle measurements; acquisition , corresponding angular measurement direction vectors , ; their corresponding equivalent target line of sight are , ; based on , calculating , the angular distance between where dot represents the dot product of vectors, represents the magnitude of a vector; represents the equivalent target line of sight , the angle formed. Suppose that there are optical sensors in the current scene, wherein To find the minimum angular distance corresponding to the association relationship as the target, an association relationship target function is constructed as follows: wherein, represents the best association result; is a binary indicator function, when composes a binary tuple, , otherwise ; It is found that in the current scene There are In order to find the minimum angular distance corresponding to the correlation relationship, the correlation relationship objective function is constructed, including: constructing a correlation cost matrix between measurements of any two optical sensors based on the angular distance between the angle measurements of any two optical sensors; calculating the correlation cost matrix between measurements of any two optical sensors m -best correlation results based on m a set of multi-tuples containing the angle measurements between any two optical sensors with higher correlation likelihood is obtained from the best correlation results; calculating a correlation cost corresponding to each multi-element group for each multi-element group in a multi-element group set; constructing a correlation cost matrix between measurements of multiple optical sensors based on the correlation cost corresponding to each multi-element group.

2. The minimum angular distance based angle measurement data association method under position unknown condition according to claim 1, wherein, The solving method of the correlation relationship objective function comprises a Hungarian algorithm and an auction algorithm.

3. The minimum angular distance based angle measurement data association method under position unknown condition according to claim 1, wherein, The construction of the correlation cost matrix between measurements of any two optical sensors based on the angular distance between the angle measurements of any two optical sensors comprises the following steps: The correlation cost matrix between any two sensor measurements is constructed as follows: as follows:​​ 。 4. The minimum angular distance based angle measurement data association method under position unknown condition according to claim 1, wherein, said calculating a correlation cost matrix between any two optical sensor measurements m the -best correlation results, comprising: Calculate using Murty method respectively , and of m -best related results ; ; For sensors j , k The first measurement between i One related result.

5. The minimum angular distance based angle measurement data association method under position unknown condition according to claim 1, wherein, The method comprises the following steps: m The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the for each the following operations are performed Obtaining optical sensor j , k measured m union of tuples in the best k correlation results .

6. The minimum angular distance based angle measurement data association method under position unknown condition according to claim 1, wherein, The calculation of the correlation cost corresponding to each multi-element group for each multi-element group in a multi-element group set comprises the following steps: traversing all of and their respective associated costs are calculated by the following equations wherein representing a tuple of costs.

7. The minimum angular distance based angle measurement data association method under position unknown condition according to claim 1, wherein, The correlation cost matrix between measurements of multiple optical sensors constructed based on the correlation cost corresponding to each multi-element group is as follows: wherein is an indicator function, when when the components form a multimer, otherwise .

Images