Binocular photoelectric searching and tracking method and system

Abstract

The invention provides a binocular photoelectric searching and tracking method and system, and the method comprises the steps: 1, carrying out the true north calibration of each observation station of a binocular photoelectric searching and tracking system, enabling the calibrated stations to select target tracking detection according to preset target features, and obtaining the multi-cycle target observation data; 2, performing data fusion processing and direction-finding cross positioning calculation to obtain cross point information; and step 3, unreasonable cross points are eliminated through first-round screening of spatial logic, false points are eliminated by combining target prior information comparison, then multi-dimensional information is integrated to obtain an association combination, false points are preliminarily screened out, statistics are constructed according to a geometric law to further eliminate false points, and a final correct association combination is screened out in a layered manner to obtain a target tracking trajectory. The method has the advantages of high positioning precision, strong false point discrimination capability and the like.

Description

Technical Field

[0001] The present invention relates to the technical fields of optoelectronic detection and laser ranging, and particularly relates to a binocular optoelectronic search and tracking method and system. Background Art

[0002] In the field of modern optoelectronic countermeasures, a binocular optoelectronic search and tracking system with the function of multi-station target positioning occupies a crucial position. Its core function is to accurately provide key information such as the target azimuth and distance for optoelectronic tracking and optoelectronic strikes. Traditional target positioning means mainly rely on laser rangefinders or radars. However, such methods have significant drawbacks. When the system is operating, they need to radiate energy pulses outward, and this characteristic makes it extremely easy to be exposed to the enemy's view, and thus face the severe threats of optoelectronic anti-reconnaissance and electromagnetic anti-radiation attacks, greatly limiting their effective application in complex combat environments. In contrast, a binocular optoelectronic search and tracking system with the function of multi-station target positioning adopts a passive reconnaissance and positioning method. By relying on the active radiation, self-radiation or reflection characteristics of the target itself, it uses single-station or multi-station optoelectronic detectors to achieve passive ranging and positioning. It has outstanding advantages such as strong concealment and excellent anti-electromagnetic interference ability. It can not only independently undertake the detection and capture of medium and short-range targets, but also cooperate with radar detection systems in complex electromagnetic environments to significantly improve the confrontation efficiency. Therefore, it has become a key research direction at present.

[0003] Although the multi-station target positioning system has many potentials, there are still a series of problems to be solved in this field currently: 1. When calibrating the true north of the observation site in the prior art, complex factors such as geomagnetic anomalies and surrounding environmental interferences are often not fully considered, resulting in a large calibration error, which makes the starting direction reference of the subsequent entire target tracking process deviate. This deviation will accumulate continuously with the tracking process, ultimately seriously affecting the accuracy of target positioning, making it difficult to accurately lock the target position. Especially in scenarios with long-distance tracking or high-precision positioning requirements, it cannot meet the actual application requirements.

[0004] 2. When eliminating false points in the prior art, the first-round screening usually only relies on basic judgment conditions such as simple distance and angle thresholds, lacking comprehensive consideration of various factors such as the target kinematic model, geographical environment characteristics, and dynamic changes in the system observation field of view. It is difficult to accurately identify the intersection points that are generated due to complex interferences and deviate from the reasonable observation range. The efficiency of false point elimination is low. The retention of a large number of false points will increase the subsequent calculation amount and is also prone to false alarms, interfering with the tracking of real targets. Summary of the Invention

[0005] Technical problems to be solved by the present invention: Aiming at the above problems of the prior art, a binocular optoelectronic search and tracking method and system with high precision and high efficiency are provided, which can achieve more accurate positioning and tracking of the target.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is as follows: A binocular optoelectronic search and tracking method, comprising the following steps: Step S1: Calibrate each observation station of the binocular optoelectronic search and tracking system to due north to obtain each calibrated observation station; each calibrated observation station selects a target for tracking detection according to preset target feature information to obtain target observation data in different periods; Step S2: Perform data fusion processing on the multi-period target observation data collected by each observation station, and perform direction-finding cross-positioning calculation on the data after fusion processing to obtain multiple cross-point information of the target; Step S3: Perform a first-round screening on the multiple cross-point information of the target based on spatial logic to eliminate cross-points outside the reasonable observation range; By comparing the preset target prior information with the cross-point information, false points that do not meet the preset conditions are eliminated; Combining the multi-dimensional information of the cross-points to obtain the associated combination of the target, and initially screening out false points; According to the geometric law of the azimuth data of the real target at different observation stations, construct a statistic for describing the change characteristics of the azimuth measurement data, and compare it with the preset prior information to exclude some false points; stratify the remaining cross-point combinations according to the specified rules, so that each data combination presenting the target characteristics can only be uniquely assigned to one radiation target until the final correct associated combination is screened out to obtain the target tracking trajectory.

[0007] As a further improvement of the method of the present invention: In the step S1, the method for calibrating each observation station to due north includes: Step S101: Use the infrared optical axis to adjust the central axis feature points of each observation station to the center of the optoelectronic field of view to achieve an accurate corresponding relationship between the observation station and optoelectronic observation; Step S102: Adjust the azimuth angle and elevation angle of each observation station at the same time to make the optoelectronic equipment of each observation station aim at the central axis feature points of other observation stations, and record and store the measurement data of each aiming of each observation station during the aiming process; Step S103: After completing the aiming operations between all stations, calculate the directional deviation value of each observation station relative to the due north direction according to the stored measurement data to achieve the due north calibration of each observation station.

[0008] As a further improvement of the method of the present invention: in the step S1, when the target is a small target, the detection method for the small target is as follows: Step S111: Set the circular neighborhood radius and the number of sampling points according to the target image scale information, compare the gray values of the sampling pixels in the circular neighborhood with the gray value of the central pixel. If the gray value of the sampling pixel is greater than the gray value of the central pixel, mark the position of the gray value of the sampling pixel as 1; otherwise, mark the position of the gray value of the sampling pixel as 0, so as to obtain a binary coding value based on the circular neighborhood, which is used to describe the texture features of the target image; Step S112: In the circular neighborhood corresponding to each pixel, calculate the local signal-to-noise ratio according to the pixel gray value, and convert the calculated local signal-to-noise ratio into a local signal-to-noise ratio criterion coding value according to a preset coding rule, which is used to describe the signal intensity contrast features of the target image; Step S113: Compare the binary coding value and the local signal-to-noise ratio criterion coding value with the preset target feature information. When the binary coding value and the local signal-to-noise ratio criterion coding value meet the preset target feature information, it is determined that there is a small target in the central pixel; otherwise, there is no small target in the central pixel.

[0009] As a further improvement of the method of the present invention: in the step S1, while tracking and detecting the target at each observation site, multi-station data synchronization operation is performed through the satellite timing information received by the Beidou positioning module of each observation site.

[0010] As a further improvement of the method of the present invention: in the step S2, the method for solving the direction-finding cross-positioning includes: Step 201: Determine the geographical coordinates of each observation site based on the spherical model of the space geodetic coordinate system, and construct an equation set according to the direction-finding cross-positioning principle in combination with the obtained target observation data after fusion processing to obtain a preliminary position candidate point of the target; Step 202: Correct the preliminary target position candidate point according to the actual curvature parameter of the earth to obtain a corrected target position candidate point.

[0011] As a further improvement of the method of the present invention: in the step S3, the operation of eliminating false targets in direction-finding cross-positioning further includes: Step S301: Select a subjective observation site, select one of the direction-finding lines of the observation target from the subjective observation site as a reference line, and calculate the intersection points of the direction-finding lines of all targets of the auxiliary observation sites and the reference line to form an intersection point position subset; Step S302: Arbitrarily select a reference point from the intersection position subset. With the reference point as the center, select the nearest intersection from the intersection position subset of the auxiliary observation stations according to the minimum distance principle to form a true positioning point subset with the reference point; Step S303: Evaluate the clustering degree of each true positioning point subset, select the subset with the highest clustering degree as the target positioning subset, and the intersection points corresponding to the remaining subsets are determined as false target points; Repeat steps S221 to S223 until the false targets are eliminated.

[0012] As a further improvement of the method of the present invention: Step S3 further includes, after the system automatically tracks the target, correcting the target trajectory. The method for correcting the target trajectory includes: Step S313: Obtain the trajectory point set corresponding to the target motion trajectory and the parameter information of each observation station related to the trajectory. When a new trajectory point is added to the trajectory point set, calculate the ground distance interval value between the last added point and the newly added point in the trajectory point set, continuously record and count all the distance interval values in the trajectory point set, and calculate the average value of the ground distance interval values; Step S312: Calculate the position interval value between the point to be detected and the last added point of the smoothed and corrected historical detection point set, and compare the position interval value with the average value of the ground distance interval values multiplied by a specified multiple. When the position interval value is less than the average value of the ground distance interval values multiplied by the specified multiple, it is determined that the newly added point is not a drift point, and the moving smoothing operation is performed; When the position interval value is greater than the average value of the ground distance interval values multiplied by the specified multiple, it is determined that the newly added point is a drift point and the position of the drift point is corrected.

[0013] As a further improvement of the method of the present invention: The target trajectory correction further includes selecting several key points on the target trajectory according to a preset rule and extracting them in a specific order, and judging whether the key points are the feature points of the target trajectory by calculation. When the key points are the feature points of the target trajectory, the key points are retained; otherwise, the key points are deleted. The feature points include inflection points and polyline points.

[0014] The present invention also provides a binocular optoelectronic search and tracking system, including: Two or more observation stations, each of the observation stations includes: A true north calibration module for calibrating the true north of the observation station so that each observation station obtains a unified direction reference; A target detection module for receiving a search instruction and selecting a target for tracking and detection according to preset target feature information, and simultaneously collecting multi-period target observation data; A data transmission module, configured to transmit the acquired target observation data to a data fusion processing device; A data fusion processing device, configured to receive the data transmitted by each observation station and perform fusion processing on the multi-period target observation data; the data fusion processing device includes a false target elimination module for eliminating false targets by direction-finding cross-positioning of multiple geographic coordinates of the target, and a target trajectory deviation correction module connected to the false target elimination module for correcting the deviation of the target tracking trajectory.

[0015] As a further improvement of the method of the present invention: the observation station further includes an antenna module for receiving signals in a specific working frequency band, the antenna module is a multi-section segmented structure, and a rabbet is provided between each section and fixed and connected by screws.

[0016] Compared with the prior art, the advantages of the present invention are as follows: 1. The present invention ensures that each station can accurately point to the true north by accurately calibrating the true north of each observation station, and uses the calibrated observation stations to screen and track the target to obtain the observation data of the target in different periods and perform fusion processing on it to calculate multiple intersection point information of the target. After performing multi-stage false point elimination operations on the multiple intersection point information and combining the multi-dimensional information of the intersection points, a statistic is constructed and compared with the limit angle, and the remaining intersection point combinations are processed layer by layer to ensure that each data combination is uniquely assigned to a radiation target, and finally the correct association combination is selected to form a target tracking trajectory. The present invention effectively improves the false point elimination efficiency, reduces unnecessary calculation burdens, improves the data processing efficiency, ensures the real-time performance and accuracy of target tracking, and improves the anti-interference ability and stability of the system.

[0017] 2. The present invention further solves the problem of orientation accuracy caused by insufficient baseline length in the traditional dual-antenna Beidou orientation system through the infrared optical axis absolute mutual aiming calibration technology and multi-station passive measurement, and improves the anti-interference ability and stability of the system. At the same time, the direction-finding cross-positioning solution method based on the space geodetic coordinate system takes into account the influence of the earth's curvature, making the positioning result more accurate and reliable.

[0018] 3. The present invention further realizes the deviation correction processing of the target trajectory by the methods of moving smoothing and thinning, which not only retains important contour feature points but also reduces the position drift phenomenon, making the trajectory display effect fast and smooth, and enhancing the readability and interpretability of the data. Description of the Drawings

[0019] Figure 1 It is a block diagram of the working principle of dual-station communication in an embodiment of the present invention.

[0020] Figure 2Schematic diagram of the direction finding and intersection positioning principle in the embodiment of the present invention.

[0021] Figure 3 Flowchart of the binocular optoelectronic search and tracking method in the embodiment of the present invention.

[0022] Figure 4 Schematic diagram of target tracking by two observation stations in the specific application embodiment of the present invention.

[0023] Figure 5 Schematic diagram of the dual-station Beidou time service synchronization information interaction in the embodiment of the present invention.

[0024] Figure 6 Flowchart of target tracking in the embodiment of the present invention.

[0025] Figure 7 Schematic diagram of the direction finding and intersection positioning based on the spherical model of the space geodetic coordinate system in the embodiment of the present invention.

[0026] Figure 8 Flowchart of the operation of the dual-station system in the embodiment of the present invention.

[0027] Figure 9 Composition diagram of the binocular measurement system in the embodiment of the present invention.

[0028] Figure 10 Schematic diagram of the turntable structure in the embodiment of the present invention.

[0029] Legend Explanation 1. Infrared thermal imager; 2. Turntable; 3. Visible light camera; 4. Laser rangefinder; 5. Dual-antenna Beidou module; 6. Antenna mount; 7. Tripod; 8. Pitch frame; 9. Azimuth frame. Detailed Implementation Manner

[0030] 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 of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0031] In this embodiment, the binocular optoelectronic search and tracking method can achieve the functions of single-station measurement target positioning and multi-station passive measurement target positioning. When performing single-station measurement target positioning, the laser ranging of a single observation platform is used to continuously measure the target to calculate the position of the target, and the optical payload is coordinated to complete the functions of target detection, tracking, and positioning. Multi-station passive measurement target positioning is to measure the direction angle of the target by using more than two observation stations, and perform intersection positioning calculation based on the measured target azimuth information and the distance between stations to obtain the target positioning information. The equipment technical indicators of the two stations are basically the same. The passive positioning accuracy of multi-station measurement depends on the size of the station spacing. When the spacing exceeds a certain value, the maximum positioning width and the maximum positioning distance will rapidly decrease. Therefore, the observation station spacing directly affects the target positioning accuracy, positioning error, and the degree of dispersion of the error value. Among them, the error first decreases and then increases as the spacing gradually increases. If you want to better exert the effectiveness of multi-station measurement and positioning, you need to combine the climate conditions and observation perspectives, and comprehensively consider the requirements of target position, positioning range, and positioning accuracy according to the characteristic parameters of each observation station, and optimize the configuration of the observation station positions.

[0032] Among them, the working principle of the single system is as follows: The optoelectronic payload is installed on the pitch frame 8 (inner frame) of the turret, and the axis of the optical imaging system is changed by the rotation of the azimuth and pitch frames 8 of the turret. When the system works, the control software sends working instructions to the turret servo controller through the Ethernet to drive the axis of the optoelectronic payload to perform panoramic search or point to the specified observation area. The ground and aerial scenes within the axis area are imaged on the image sensor through the optical system of the optoelectronic payload, and the generated digital image video signal is connected to the image processing module for processing. The image processing module extracts the threat targets in the image, selects the targets for tracking and monitoring, and calculates the geographical location of the target according to the positioning and orientation information of the turntable 2 and the laser ranging information.

[0033] The single-station positioning steps include: Step 1: Use Beidou positioning to obtain the positioning latitude , longitude , target azimuth , and use laser ranging to obtain the straight-line distance D of the target point; Step 2: Calculate the latitude of the target point: (1) Calculate the longitude of the target point. If the target is on the east side of the station, the function expression of the target point longitude is: (2) If the target point is on the west side, the function expression of the target point longitude is: (3) If the target is in the southern latitude, take a negative value. If the longitude exceeds 180°,

[0034] as Figure 1 shown, the bistatic working principle is as follows: The binocular optoelectronic search and tracking system measures the direction of the target through the optoelectronic turrets 2 of two observation stations. As Figure 2 shown, A and B represent the observation stations, and C represents the target point. According to the data measured by the turrets 2 of each station and the distance between AB, the target position is obtained through geometric calculation. The system performs data fusion on the two direction-finding stations and calculates the intersection point of the directions of the same signal to obtain the target position information.

[0035] Based on the above working principle, the detailed scheme of the binocular optoelectronic search and tracking method of the present invention is as follows: as Figure 3 shown, the binocular optoelectronic search and tracking method of this embodiment includes the following steps: Step S1: Calibrate each observation station of the binocular optoelectronic search and tracking system to due north to obtain each observation station after calibration; each observation station after calibration selects a target for tracking and detection according to the preset target feature information to obtain target observation data in different periods.

[0036] In this embodiment, the method for calibrating each observation station to due north includes: Step S101: Use the infrared optical axis to adjust the central axis feature point of each observation station to the center of the optoelectronic field of view to achieve an accurate correspondence between the observation station and the optoelectronic observation; Step S102: Adjust the azimuth angle and elevation angle of each observation station simultaneously to enable the optoelectronic equipment of each observation station to aim at the central axis feature point of other observation stations, and record and store the measurement data of each aiming of each observation station during the aiming process; Step S103: After completing the aiming operations between all stations, calculate the directional deviation value of each observation station relative to the due north direction according to the stored measurement data to achieve the due north calibration of each observation station.

[0037] In a specific application embodiment, according to the absoluteness of the infrared optical axis, the central axis characteristic points of each observation station are adjusted to the center of the optoelectronic field of view, and each observation station is mutually aimed and calibrated, and the calibration results are stored in real time. After the calibration is completed, the least squares statistical processing is performed on the stored measurement results to obtain the dynamic random error. In this embodiment, the minimum variance method is used as the basis for multi-measurement parameter data processing, that is, for each spatial point, each combination of original measurement data is substituted into the orientation algorithm for error calculation, and it is judged which orientation algorithm and which original measurement data have the smallest error at the same spatial point. At this point, the orientation algorithm and measurement parameters with the smallest error are used to calculate the north orientation deviation, solving the orientation deviation caused by the insufficient length of the Beidou dual-antenna orientation baseline during multi-station measurement target positioning.

[0038] Specifically, as Figure 4 shown, points A and B are any two points on the earth, representing the main observation station and the auxiliary observation station respectively, point C is the North Pole, the earth's center is O, the distance between points A and B is the arc length of Lab, and the azimuth between A and B is the angle between the tangent of arc Lab and the earth's latitude. The position of the horizontal cross-section of the sphere where point A is located is the latitude of point A, and the position of the longitudinal cross-section is the longitude of point A.

[0039] The longitude and latitude of the main station (Aw, Aj) and the longitude and latitude of the auxiliary station (Bw, Bj) are known; a = ∠BOC; b = ∠AOC; c = ∠AOB; A = ∠CAB; B = ∠CBA; C = ∠ABC.

[0040] The main steps for north direction calibration include: Step 1: After expanding the main observation station and the auxiliary observation station, when the distance AB > 500m, the two stations are mutually aimed using the absoluteness of the infrared optical axis, and the central axis characteristic points of each station are adjusted to the center of the infrared image field of view, and the servo azimuths Az and Bz of the turntable 2 after aiming are obtained respectively; According to the cosine formula of a trihedral angle: (4) AOCB is the dihedral angle between plane AOC and plane BOC, which is the longitude difference between the two points; Substitute the known data to get: (5) Step 2: According to the cosine value of angle C, find its sine value (6) The third step: Use the spherical sine formula (7) Substitute the known data and slightly transform it to get: (8) Use the arcsine function to find the angle. Among them, A is the true north azimuth of the line connecting points A and B, and the function expression of A is: (9) Step 4: Site 0 position correction. Az - A is the true north deviation angle of the master station A. Substitute the deviation value into the system calculation to correct the master station site 0 position. The slave station can be obtained in the same way. According to different site environments, the third point calibration can also be used, and the calibration method is the same as the above calculation method.

[0041] In step S1 of this embodiment, when the target is a small target, the detection method for the small target is: Step S111: Set the circular neighborhood radius and the number of sampling points according to the target image scale information. Compare the gray values of the sampling pixels in the circular neighborhood with the gray value of the central pixel. If the gray value of the sampling pixel is greater than the gray value of the central pixel, mark the position of the gray value of the sampling pixel as 1; otherwise, mark the position of the gray value of the sampling pixel as 0, so as to obtain the binary coding value based on the circular neighborhood, which is used to describe the texture characteristics of the target image; Step S112: In the circular neighborhood corresponding to each pixel, calculate the local signal-to-noise ratio according to the pixel gray value, and convert the calculated local signal-to-noise ratio into a local signal-to-noise ratio criterion coding value according to the preset coding rule, which is used to describe the signal intensity contrast characteristics of the target image; Step S113: Compare the binary coding value and the local signal-to-noise ratio criterion coding value with the preset target feature information. When the binary coding value and the local signal-to-noise ratio criterion coding value meet the preset target feature information, it is determined that there is a small target in the central pixel; otherwise, there is no small target in the central pixel.

[0042] In this embodiment, a small target refers to a non-uniform light spot with pixels ≤ 3x3 matrix pixels in infrared imaging. From an optical perspective, the gray distribution of the target can be regarded as a discrete cosine pulse in two-dimensional space: (10) In the above formula, T is the target intensity, R is the target size, and i, j are the coordinates of the target in the two-dimensional image. represents the intensity and size of the target image.

[0043] The method for detecting small and weak targets using the LBP operator is to compare the 8 pixels in the neighborhood of the window center with the window center pixel respectively. The positions where the neighborhood pixel values are greater than the center pixel value are marked as 1, otherwise marked as 0, thus obtaining an 8-bit binary value, and taking this value as the LBP value of the window center pixel. One defect of the LBP operator in the prior art is that it only covers a small area within a fixed radius range, which obviously cannot meet the needs of different sizes and frequencies of textures. In order to adapt to texture features of different scales and meet the requirements of gray-scale and rotation invariance, the LBP operator is improved as follows in this embodiment: the 3×3 neighborhood is extended to an arbitrary neighborhood, and the square neighborhood is replaced by a circular neighborhood. This method eliminates the problem of illumination change to a certain extent, has rotation invariance, low-dimensional texture features, accurate suppression of target maxima, fast calculation speed. This method encodes the gray-scale texture of the image using the local signal-to-noise ratio criterion of the image, and the obtained encoded value is used as the feature of the infrared small and weak target to complete target detection. This method can quickly and effectively detect infrared small and weak targets without suppressing the background.

[0044] In step S1 of this embodiment, while tracking and detecting the target at each observation site, multi-station data synchronization operation is performed through the satellite time synchronization information received by the Beidou positioning module at each observation site.

[0045] Specifically, the Beidou positioning module is an important device for system time synchronization. Its main function is to receive satellite signals and Beidou time synchronization information, and use the high-precision Beidou time synchronization information to achieve data synchronization of multiple sites. In the actual use of the system, the synchronization requirement is high. The system uses bridge microwave wireless communication. Microwave communication uses electromagnetic waves between 1mm and 1m as the communication medium, with stable data transmission and long transmission distance. The Beidou positioning module includes components such as a GPS antenna, a Beidou board, and a Beidou antenna rack. The Beidou receives satellite data, outputs time information and the second pulse signal PPS. The turntable 2FPGA receives the second pulse and then divides it into synchronous signals with the same frame rate as the camera and sends them to the infrared detector, the image processing board, and the servo control board to achieve the system time synchronization function. The time synchronization accuracy of this system can reach 20ns (rms), meeting the synchronization requirements of engineering applications.

[0046] Such as Figure 5As shown, the master station information includes a synchronization pulse based on the time synchronization information and a 100 Hz trigger synchronization signal. The 1 Hz time synchronization signal and the synchronization pulse based on the time synchronization information act on four information sources: Beidou positioning information (longitude and latitude), the attitude of the inclinometer and the Beidou direction information (attitude angle), the servo pointing information (encoder value, angle), and the image recognition information (target position, pixel). These information sources respectively point to and track the target in the geographical coordinate system, the servo platform coordinate system, and the detector coordinate system. After that, the information of these four information sources is aggregated to the link of calculating the position result. After calculating the position result, the Beidou positioning information and the inclinometer attitude information are connected through the synchronization pulse based on the time synchronization information and the 1 Hz time synchronization signal, while the servo pointing information and the image recognition information are connected to the slave station information through the 100 Hz trigger synchronization signal, thus completing the entire information processing and target positioning.

[0047] As Figure 6 shown, in this embodiment, the accuracy control method for tracking the target includes: the turntable 2 tracking closed-loop control method and the image processing target tracking method.

[0048] Among them, for the optoelectronic search and tracking system, the commonly used control method in engineering for the turntable 2 tracking closed-loop control method is generally lead-lag correction. The open-loop transfer function of the corrected system is usually a typical type I or type II system. Here, taking a typical type II system with a miss distance time delay as an example, the influence of the miss distance time delay on the system performance is analyzed. The transfer function of a typical type II system with a miss distance time delay can be expressed as: , (11) where K is the open-loop gain of the system, , are the system time constants, is the miss distance time delay, s is the standard deviation of the miss distance time delay. Then the amplitude-frequency characteristic of the above formula is: , (12) The phase-frequency characteristic is: (13) It can be seen from the amplitude-frequency characteristic and the phase-frequency characteristic that the miss distance time delay has no influence on the amplitude-frequency characteristic and has an influence on the phase-frequency characteristic. This is because the amplitude of the time delay link is always 1, and the phase angle is proportional to , and the proportionality coefficient is . Let the phase margin loss caused by the miss distance time delay be , then , (14) where is the cut-off frequency, is the sampling frequency, Let \(T\) be the sampling period and \(N\) be the filtering order.

[0049] Generally speaking, the larger \(T\) is, that is, the time delay of the miss distance relative to the sampling period of the system the larger it is, the larger the phase loss margin caused by the time delay of the miss distance is, and the greater the impact on the stability of the system. The smaller \(N\) is, that is, the sampling frequency relative to the open-loop cut-off frequency of the system the larger it is, the smaller the phase loss margin caused by the time delay of the miss distance is, and the smaller the impact on the stability of the system. Therefore, the sampling frequency of the system should be selected by making a compromise according to the time delay of the miss distance of the system, the requirements of the system phase margin, and the system technical requirements.

[0050] The compensation of the tracking loop of turntable 2 is mainly based on the predictive tracking control technology to overcome the time delay of the miss distance and ensure the real-time and rapid tracking of the system. In the image processing board, there may occasionally be a large jump in the miss distance or no miss distance when detecting weak targets. In this case, the servo system must process the received miss distance. When the target is temporarily blocked by a building or cloud layer, the optoelectronic search and tracking system should be able to continue to track by memorizing the current trajectory of the target.

[0051] The image processing target tracking method uses the KCF tracking algorithm improved based on the dynamic learning rate. KCF is a tracking algorithm based on online learning, and the size of the learning rate will affect the performance of the algorithm. An overly low learning rate will prevent the classifier from quickly integrating the correspondence between new samples and features, resulting in a decreased resistance to minor deformations, scale changes, etc., thus causing poor tracking effects. While an overly high learning rate will cause the target to learn too much noise information and be more vulnerable to interference from similar objects, resulting in incorrect positioning when passing by similar obstacles. Therefore, it is unreasonable to update the model with a fixed learning rate under different response intensities.

[0052] By setting the method of the dynamic learning rate, while ensuring that the classifier can effectively integrate new feature information, it is possible to avoid the model from integrating too much noise information. When designing this dynamic learning rate, the reference criterion is: when the response intensity is relatively large, the response intensity - learning rate curve is relatively smooth; when the response intensity is less than a certain threshold, the learning rate drops rapidly; when it is less than the lowest threshold, the learning rate is 0. Based on this idea, the hyperbolic tangent function formula is used to set the learning rate.

[0053] Step S2: Perform data fusion processing on the multi-period target observation data collected by each observation station, and perform direction-finding cross-positioning calculation on the data after the fusion processing to obtain multiple cross-point information of the target.

[0054] In this embodiment, the method for direction-finding cross-positioning calculation includes: Step 201: Determine the geographical coordinates of each observation station based on the spherical model of the space geodetic coordinate system, and construct an equation set according to the direction-finding cross-positioning principle in combination with the obtained target observation data after fusion processing to obtain the preliminary position candidate points of the target; Step 202: Correct the preliminary target position candidate points according to the actual curvature parameters of the earth to obtain the corrected target position candidate points.

[0055] In a specific application embodiment, as Figure 7 shown, is the target radiation source, latitude, longitude, is the direction-finding station coordinate, is the direction-finding angle of P. The coordinate of is , and the two points and the North Pole form a spherical triangle , represents the great circular arc of the spherical triangle, and the three angles in the spherical triangle are represented by the spherical angle (15) It can be known from the definition of longitude and latitude that: , (16) It can be known from the definition of the direction-finding angle that: Substitute and equation (16) into equation (15) to get , (17) That is: , (18) Sum-to-product formula: , (19) Let the known number , (20) Let the unknown number , (21) Substitute (20) and (21) into (18) to get , (22) Let the known number , (23) Substitute (23) into (22) and simplify to get , construct an equation set with the coordinates of two direction-finding stations and basic trigonometric formulas: , (24) Let the known number , if , the target longitude is 0 or 180° or the three points are in the same meridian circle. If it is not 0, two sets of longitude solutions are obtained: , (25) Let the known number , we can know from (25) that B has two values, and we can also use the coordinates of observation point 1 or 2 to calculate, to avoid ,if It means P is at the poles. After elimination, we get two sets of latitude solutions: , (26) There are 2 solutions for each A, a total of 4 solutions, respectively use The function calculates the longitude and latitude to obtain 4 sets of longitude and latitude coordinates. Two sets of latitude coordinates are not within the range of (-π / 2,π / 2). After removing them, we get two sets of coordinates, which are symmetrical points on the sphere passing through the center of the earth. We calculate the distances from the two lateral stations to the two sets of coordinates and get 4 values.

[0056] In the process of positioning calculation, when measuring multiple targets, many direction finding lines will be formed and a large number of false points will be generated. As the number of observed targets and observation stations increases, the number of false points will also increase. This will inevitably make the data transmission and time synchronization between stations very complicated. Therefore, how to quickly eliminate the large number of false points generated in the direction finding cross positioning has always been a difficult problem in multi-station measurement passive positioning.

[0057] Multi-station passive positioning performs data association for the received azimuth information. How to obtain correct data association is a critical issue. If there is only one radiation source target in the detection area, this problem is relatively simple. It only requires simple mathematical operations on the measured azimuth and the distance between each direction-finding station to obtain the position of the radiation source. If there are multiple targets in the detection area, it is first necessary to distinguish whether the measured azimuth information comes from the same target. After distinguishing, the azimuth data from the same target are associated and combined for positioning calculation. Since some large measurement errors will occur during engineering use, the target position information must be fused after the positioning calculation is completed so that the random errors caused by the measurement results can be eliminated, thereby obtaining more accurate position information.

[0058] Step S3: Perform a first round of screening on multiple intersection information of the target based on spatial logic to eliminate intersections that are not within a reasonable observation range; By comparing the preset target prior information with the intersection information, false points that do not meet the preset conditions are eliminated; Combine the multi-dimensional information of the intersection points to obtain the associated combination of the target and realize the preliminary screening of false points; According to the geometric law of the azimuth data of real targets at different observation stations, a statistic for describing the variation characteristics of azimuth measurement data is constructed, and the statistic is compared with the preset prior information to exclude some false points; the remaining intersection point combinations are stratified according to the specified azimuth and elevation thresholds, so that each data combination presenting target characteristics can only be uniquely assigned to one radiation target until the final correct association combination is selected to obtain the target tracking trajectory.

[0059] In a specific application embodiment, false points are removed by using the multi-period data correlation. First, the intersection points in the cross-positioning are preprocessed to remove some intersection points that do not conform to the spatial logic. Secondly, an angle limit and prior information about the approximate distance of the target are set for rough association, which can eliminate a large number of false points and reduce the calculation amount. Then, the correct association combination is selected by using the idea of multi-dimensional assignment, so as to achieve the purpose of removing false points.

[0060] Secondly, through the geometric law of the azimuth measurement of the same target, a statistic is constructed and compared with the limit angle for association to exclude some false points. Then, the remaining combinations are stratified and combined. According to the fact that the target can only belong to one radiation target during the association process, the correct association combination is obtained, thus realizing the elimination of false targets.

[0061] In this embodiment, the operation of removing false targets in direction-finding cross-positioning further includes: Step S301: Select a subjective observation station, select one of the direction-finding lines of the observation target from the subjective observation station as the reference line, and calculate the intersection points of the direction-finding lines of all targets by the auxiliary observation stations and the reference line to form a subset of intersection positions; Step S302: Arbitrarily select a reference point in the subset of intersection positions. With the reference point as the center, according to the principle of the minimum distance, select the nearest intersection point from the subset of intersection positions of the auxiliary observation stations to form a subset of true positioning points with the reference point; Step S303: Evaluate the clustering degree of each subset of true positioning points, select the subset with the highest clustering degree as the positioning subset of the target, and the intersection points corresponding to the remaining subsets are determined as false target points; repeat steps S221 to S223 until the elimination of false targets.

[0062] Specifically, there are N observation targets. The direction-finding line of one of the targets by the main station is used as the reference line. Then, the direction-finding lines of the N targets by the auxiliary stations and the reference line form a subset of 1 intersection position. Then, any intersection point is selected as the reference point to form a reference subset. With each reference point of the main station as the center, using the minimum distance, select the nearest intersection point of each intersection point subset of the auxiliary station and form a subset of true positioning points with the reference point. Use the subset with the highest clustering degree as the positioning subset of the target, and then by analogy, repeat the above operations to achieve the elimination of false targets.

[0063] After the elimination of false targets, this embodiment further includes more accurately determining the true position of the target according to the error ellipse principle to improve the positioning accuracy of the optoelectronic search and tracking system. Among them, the error ellipse is a geometric figure reflecting the point position accuracy distribution obtained based on the measurement adjustment theory. In the optoelectronic search and tracking system, after observing the target through multiple observation stations, there are certain errors in the estimated position of each target point, and these errors can be represented by the error ellipse. The specific method includes: Step 1: According to the data fusion result obtained by observing the target through multiple observation stations, calculate the covariance matrix of the target position estimate, solve for the eigenvalues and corresponding eigenvectors of the covariance matrix, and calculate the major semi-axis, minor semi-axis and major axis direction angle of the error ellipse according to the eigenvalues and eigenvectors; Step 2: For each error ellipse, calculate the distance from the target point to each observation station and find the minimum value among all the distances. The observation station corresponding to this minimum value is the observation station closest to the target point. At this time, the estimated coordinate of the target point corresponding to the minimum value is the finally determined target point coordinate.

[0064] After data association, target position calculation is performed. For moving targets, target movement trajectories will be formed. Due to the influence of Beidou positioning accuracy of the station, horizontal accuracy of the station, and angular resolution of the infrared detector, a northward calibration deviation of the bistatic station is caused. When the observation angle of the target approaches 0° or 180°, the calculation error increases, directly affecting the target positioning accuracy. The target trajectory will have position drift, and the drift phenomenon will cause a significant increase in track data and large statistical errors; at the same time, when the trajectory position is displayed on the front-end UI map, there will be disordered connections, affecting the UI aesthetic effect. Therefore, it is necessary to correct the drift points.

[0065] Generally, the frame rate of the infrared detector is 50 or 100 Hz, and a large amount of data will be generated during cross-positioning calculation. For low-speed targets, the display of a large number of points on the UI interface will affect the software aesthetics. It is necessary to screen out key valuable points and retain them, and eliminate points that do not affect the overall trajectory to improve the trajectory display effect. Through research on position clustering and geometric features, the system forms a method that can correct and thin the trajectory points, making the trajectory display effect fast, smooth and retaining important contour feature points, making the data more readable and interpretable.

[0066] Therefore, step S3 of this embodiment further includes correcting the target trajectory after the system automatically tracks the target. The method for correcting the target trajectory includes: Step S313: Obtain the set of trajectory points corresponding to the target motion trajectory and the parameter information of each observation station related to the trajectory. When a new trajectory point is added to the set of trajectory points, calculate the ground distance interval value between the last added point and the newly added point in the set of trajectory points, continuously record and count all the distance interval values in the set of trajectory points, and calculate the average value of the ground distance interval values. Step S312: Calculate the position interval value between the point to be detected and the last added point in the set of historical detection points after smoothing and correction, and compare the position interval value with the average value of the ground distance interval values multiplied by a specified multiple. When the position interval value is less than the average value of the ground distance interval values multiplied by the specified multiple, it is determined that the newly added point is not a drift point, and the moving smoothing operation is performed; when the position interval value is greater than the average value of the ground distance interval values multiplied by the specified multiple, it is determined that the newly added point is a drift point, and the position correction of the drift point is performed.

[0067] In this embodiment, the target trajectory correction further includes selecting several key points on the target trajectory according to a preset rule and extracting them in a specific order, and determining whether the key points are the feature points of the target trajectory by calculation. When the key points are the feature points of the target trajectory, the key points are retained; otherwise, the key points are deleted. The feature points include inflection points and polyline points.

[0068] In a specific application embodiment, the trajectory correction is performed by calculating the ground distance interval between the last added point and the newly added point in the trajectory point set W. The average value of the distance intervals in the trajectory set is denoted as A, and the position interval P between the point to be detected and the last smoothed point is calculated; if P is less than a set multiple of A, it is considered not a drift point, and the method of moving both the smoothing and the detection point forward by one position is continued. If P is greater than the set multiple of A, it is considered a scattered point, and the position correction is performed by taking the difference. The correction method is to correct the last two points in the trajectory point set W. Correction formula: (27) (28) In the above formula, x and y are the longitude and latitude respectively.

[0069] During the trajectory correction, a thinning operation is performed simultaneously. Thinning mainly retains the feature points of the trajectory trend and contour, usually inflection points or polyline points. An inflection point, also known as a point of inflection, refers to a point that changes the upward or downward direction of a curve in mathematics. Intuitively, an inflection point is a point where the tangent line crosses the curve, that is, the demarcation point between the concave arc and the convex arc of a continuous curve. If the function of the curve graph has a second derivative at the inflection point, the second derivative has different signs or does not exist at the inflection point. Detection is performed through four adjacent points (A, B, C, D). For the line segment set (AB, BC, CD), it is judged whether the cross product results of (AB X BC) and (BC X CD) are different.

[0070] (29) For the polyline points, three points are used to detect the included angle ∠ABC for judgment. ∠ABC is calculated by computing the lengths of line segments |AB|, |BC|, and |AC| and using the arccosine function. The system is given two angle threshold hyperparameters. One is the larger angle threshold, and the other is the smaller angle threshold. One of them is adaptively selected as the current threshold for angle detection. When ∠ABC is less than the current angle threshold, point B is considered a polyline point and should be retained during the thinning process. The strategy for adaptively selecting the current angle threshold is as follows: When the current point B is thinned out, it indicates that the next point is relatively important for maintaining the contour. Therefore, the current angle takes the smaller angle threshold to increase the possibility of the next point being retained; when the current point B is retained as an important point, the next point is not important for maintaining the contour, and the current angle takes the larger angle threshold to reduce the possibility of the next point being retained.

[0071] The following further illustrates the present invention by taking the above method for optoelectronic search and tracking in a specific application embodiment as an example, as Figure 8 shown. Its detailed steps are as follows: Step 1: The entire system starts the power-on operation to provide power support for subsequent operations, starts the display control terminal, and runs the integrated display control software. This software will be used to monitor and control the operation of the entire system, including the operations and data display of the main observation station and the auxiliary observation station.

[0072] Step 2: The main observation station performs self-check and initialization operations. This includes checking the status of the internal hardware devices (such as sensors, processors, etc.) of the main observation station to ensure that each device is working properly and initializing the system parameters. Similar to the main observation station, the auxiliary observation station also performs self-check and initialization operations to ensure that the hardware and software of the auxiliary observation station are in a normal working state.

[0073] Step 3: Perform true north calibration on the main observation station and the auxiliary observation station, and check whether the cooling of the thermal imaging device is completed. Thermal imaging cooling is to ensure that the thermal imager can work properly and obtain clear thermal imaging images. If the thermal imaging cooling is completed, the main observation station and the auxiliary observation station perform calibration and orientation operations. Calibration and orientation may involve calibrating parameters such as the angle and direction of the thermal imager to ensure the accuracy of its observation data. If the thermal imaging cooling is not completed, continue to wait until the cooling is completed before performing calibration and orientation.

[0074] Step 4: The main observation station and the auxiliary observation station observe the specified area according to the task requirements. During the observation process, the main observation station uses various sensors (such as optical sensors, thermal imaging sensors, etc.) equipped with it to collect data within the area.

[0075] Step 5: During the observation process, determine whether a target is detected. If a target is detected, the subjective observation station transmits the detected data to the subsequent data association and synchronization processing module. If no target is detected, continue with the observation of the specified area. When both the subjective and auxiliary observation stations detect a target, the data of the main and auxiliary observation stations are associated and synchronized. This includes integrating and calibrating data such as the position, speed, and direction of the target to ensure that the obtained target information is accurate and consistent. After data association and synchronization processing, the target data is screened to determine whether it is a single target or multiple targets.

[0076] Step 6: If it is a single target, the system enters the automatic tracking mode. During the automatic tracking process, the status of the target is monitored in real time to ensure that the target is always within the tracking range. If the target is lost, the system performs trajectory correction and track prediction and tracking operations. Trajectory correction is to analyze and correct the trajectory before the target is lost to determine the possible movement direction of the target. Track prediction and tracking is to predict the possible position of the target based on the previous movement trajectory and related parameters of the target and perform tracking operations. If the target is recaptured during the track prediction and tracking process, continue with the automatic tracking. If the target is not recaptured, determine whether to re-detect. If re-detection is selected, return to the specified area observation step. If re-detection is not selected, close the display and control software and perform the power-down and withdrawal operation.

[0077] If it is determined that there are multiple targets after data association and synchronization processing, the system performs multi-target screening to eliminate false alarms. Multi-target screening is to classify and identify multiple targets and eliminate the false alarm targets (i.e., non-target objects misjudged as targets) among them. After screening, the real targets are processed according to the single-target processing flow, that is, enter the related operations such as automatic tracking and target loss processing.

[0078] Step 7: During the target tracking process, whether it is single-target or multi-target tracking, the system stores the relevant data of the target (such as trajectory, position, speed, etc.). When the system completes the task, or when the target is not re-detected and there is no need to continue working, close the display and control software and perform the power-down and withdrawal operation to ensure the safe shutdown of the system.

[0079] This embodiment also provides a binocular optoelectronic search and tracking system, including: More than two observation stations, each observation station includes: A true north calibration module for calibrating the true north of the observation station so that each observation station obtains a unified direction reference; A target detection module for receiving a search instruction and selecting a target for tracking and detection according to the preset target feature information, and simultaneously collecting multi-period target observation data; A data transmission module for transmitting the collected target observation data to the data fusion processing device; A data fusion processing device is used to receive data transmitted by each observation station and perform fusion processing on multi-period target observation data. The data fusion processing device includes a false target elimination module for eliminating false targets by direction-finding cross-positioning of multiple geographical coordinates of the target, and a target trajectory correction module connected to the false target elimination module for correcting the target tracking trajectory.

[0080] As Figure 9 shown, in a specific application embodiment, the binocular optoelectronic search and tracking system uses 2 observation stations. The structures of each observation station are the same and include a turret, a display and control terminal, and a Beidou positioning module. Among them, as Figure 10 shown, the turret adopts an azimuth frame 8 and a pitch frame 9 structure, and internally loads an infrared thermal imager 1, a visible light camera 3, a laser rangefinder 4, and an image processing module. It mainly consists of a servo controller, a driver, a motor, an encoder, and a conductive slip ring. The azimuth and pitch transmission mechanisms select permanent magnet servo motors, use high-precision encoders as the speed and position detection devices of the system, and use slip rings for internal electrical signal transmission to achieve 360-degree rotation of the azimuth. The display and control terminal is a ruggedized computer, which is the display and control center of the system. It exchanges image data and control commands with the turret through Ethernet, and performs multi-station networking and interconnection with wireless image transmission devices through Ethernet during binocular measurement and positioning. The Beidou positioning module is used to receive satellite signals and Beidou timing information, including a dual-antenna Beidou module 5, a Beidou board card, and an antenna mount 6. Its output time information and second pulse signal PPS are sent to the FPGA of the turntable 2. The FPGA of the turntable 2 divides the second pulse into a synchronous signal with the same frame rate as the camera and sends it to the infrared detector, the image processing board, and the servo control board to achieve the system time synchronization function, and the timing accuracy reaches 20 ns (rms).

[0081] In this embodiment, the GPS antenna is used to receive signals in a specific working frequency band. The antenna module is a multi-section segmented structure, and there are stop ports between each section, and they are fixedly connected by screws.

[0082] The above are only the preferred embodiments of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A binocular electrogram search and tracking method, characterized in that: The following steps are involved: Step S1: performing true north calibration operation on each observation station equipped with the binocular electroscopic search and tracking system to obtain each calibrated observation station; Each calibrated observation station selects a target for tracking and detection according to preset target feature information to obtain target observation data of different periods; Step S2: performing data fusion processing on the multi-period target observation data collected by each observation station, and performing direction finding cross positioning solution on the fused data to obtain multiple intersection point information of the target; Step S3: performing a first round of screening on the multiple intersection information of the target based on spatial logic to eliminate the intersections that are not within a reasonable observation range; By comparing the preset target prior information with the intersection information, false points that do not meet the preset conditions are eliminated; Combining the multi-dimensional information of the intersection points to obtain the associated combination of the targets, and realizing the preliminary screening of false points; According to the geometric laws of the azimuth data of the real target at different observation sites, a statistic is constructed to describe the changing characteristics of the azimuth measurement data, and the statistic is compared with the preset prior information to exclude some false points; the remaining intersection point combinations are layered according to specified rules, so that each data combination showing the target characteristics can only be uniquely assigned to one radiation target, until the final correct associated combination is screened out to obtain the target tracking trajectory.

2. The binocular electrogram search and tracking method according to claim 1, characterized in that: In step S1, the method for calibrating each observation station to the true north includes: Step S101: using the infrared optical axis to adjust the central axis feature point of each observation site to the center of the optical television field, so as to achieve an accurate correspondence between the observation site and the photoelectric observation; Step S102: adjusting the azimuth and elevation of each observation site simultaneously, so that the optoelectronic equipment of each observation site can aim at the central axis feature point of other observation sites, and recording and storing the measurement data of each aiming of each observation site during the aiming process; Step S103: After completing the aiming operation between all the sites, the orientation deviation value of each observation site relative to the true north direction is calculated based on the stored measurement data to achieve true north calibration of each observation site.

3. The binocular electrographic search and tracking method according to claim 1, characterized in that: In step S1, when the target is a small target, the detection method for the small target is: Step S111: the circular neighborhood radius and the number of sampling points are set according to the scale information of the target image, and the grayscale value of the sampled pixel in the circular neighborhood is compared with the grayscale value of the central pixel. If the grayscale value of the sampled pixel is greater than the grayscale value of the central pixel, the position of the grayscale value of the sampled pixel is marked as 1; otherwise, the position of the grayscale value of the sampled pixel is marked as 0, thereby obtaining a binary code value based on the circular neighborhood to describe the texture characteristics of the target image; Step S112: in the circular neighborhood corresponding to each pixel, the local signal-to-noise ratio is calculated according to the pixel gray value, and the calculated local signal-to-noise ratio is converted into a local signal-to-noise ratio criterion coding value according to a preset coding rule to describe the signal intensity contrast feature of the target image; Step S113: Compare the binary code value and the local signal-to-noise ratio criterion code value with the preset target feature information. When the binary code value and the local signal-to-noise ratio criterion code value meet the preset target feature information, it is determined that a small target exists in the central pixel; otherwise, no small target exists in the central pixel.

4. The binocular electrogram search and tracking method according to claim 1, characterized in that: In the step S1, while each observation site is tracking and detecting the target, a multi-site data synchronization operation is performed through the satellite timing information received by the Beidou positioning module of each observation site.

5. The binocular electrocoagulation search and tracking method according to claim 1, characterized in that: In step S2, the method for direction finding cross positioning solution includes: Step 201: determining the geographic coordinates of each observation site based on a spherical model of a spatial geodetic coordinate system, and constructing a set of equations based on a direction finding cross positioning principle in combination with the acquired fused target observation data to obtain a preliminary candidate position point of the target; Step 202: Correct the preliminary target position candidate point according to the actual curvature parameter of the earth to obtain a corrected target position candidate point.

6. The binocular electrocoagulation search and tracking method according to claim 1, characterized in that: In step S3, the direction finding cross positioning false target elimination operation also includes: Step S301: Select a main observation site, select one of the direction finding lines of the main observation site to the observed target as a baseline, and calculate the intersection points of the direction finding lines of the auxiliary observation site to all targets and the baseline to form an intersection position subset; Step S302: randomly selecting a reference point from the intersection position subset, taking the reference point as the center, selecting the nearest intersection point from the intersection position subset of the auxiliary observation site according to the minimum distance principle and the reference point to form a real positioning point subset; Step S303: Evaluate the degree of clustering of each of the real positioning point subsets, select the subset with the highest degree of clustering as the positioning subset of the target, and the intersection points corresponding to the remaining subsets are determined to be false target points; repeat steps S221 to S223 until the false targets are eliminated.

7. The binocular electrogram search and tracking method according to claim 1, characterized in that: The step S3 also includes correcting the target trajectory after the system automatically tracks the target, and the method for correcting the target trajectory includes: Step S313: obtaining a trajectory point set corresponding to the target motion trajectory and parameter information of each observation site related to the trajectory, and when a new trajectory point is added to the trajectory point set, calculating the ground distance interval value between the last point added to the trajectory point set and the newly added point, continuously recording and counting all distance interval values ​​in the trajectory point set, and calculating the average value of the ground distance interval value; Step S312: Calculate the position interval value of the last point added to the point to be detected and the historical detection point set after smoothing correction, compare the position interval value with the average value of the ground distance interval value of the specified multiple, when the position interval value is less than the average value of the ground distance interval value of the specified multiple, determine that the newly added point is not a drift point, and perform a moving smoothing operation; when the position interval value is greater than the average value of the ground distance interval value of the specified multiple, determine that the newly added point is a drift point and perform drift point position correction.

8. The binocular electrogram search and tracking method according to claim 7, characterized in that: The target trajectory correction also includes selecting a number of key points on the target trajectory according to preset rules, extracting them in a specific order, and determining by calculation whether the key points are feature points of the target trajectory. When the key points are feature points of the target trajectory, the key points are retained; otherwise, the key points are deleted. The feature points include inflection points and broken line points.

9. A binocular electroscopic search and tracking system, characterized in that: include: Two or more observation sites, each of which comprises: The true north calibration module is used to calibrate the true north of the observation station so that each observation station can obtain a unified direction reference; The target detection module is used to receive the search instruction and select the target for tracking and detection according to the preset target feature information, and collect multi-cycle target observation data at the same time; A laser ranging module, used to measure the distance between the target and the observation site; A data transmission module, used to transmit the collected target observation data to a data fusion processing device; A data fusion processing device is used to receive the data transmitted by each observation site and perform fusion processing on the multi-period target observation data; the data fusion processing device includes a false target elimination module for performing direction finding and cross-positioning on multiple geographic coordinates of the target to eliminate false targets, and a target trajectory correction module connected to the false target elimination module for correcting the target tracking trajectory.

10. The binocular electroscopic search and tracking system according to claim 9, characterized in that: The observation site also includes an antenna module for receiving signals in a specific working frequency band. The antenna module is a multi-section segmented structure, each section is provided with a stopper, and is fixedly connected by screws.

Images