A method for identifying false targets with multiple reflections in navigation radar based on arithmetic progression extraction

Abstract

The present invention discloses a navigation radar multiple reflection false target identification technology based on arithmetic sequence extraction. The technology utilizes the rule that false echoes formed by multiple reflections appear at equal intervals along the same azimuth and their energy decays successively. First, a group of equally spaced extreme points is extracted for each pulse echo. Extreme points distributed according to the arithmetic sequence rule are extracted using a distance sampling rate δr as a stepping interval. Then, "distance-azimuth-equally spaced value" clustering and "group number-class number" clustering are performed in sequence. Finally, multiple reflection false echoes are identified and marked, thereby eliminating or reducing their impact on the navigation radar's detection of real surface targets. The present invention can effectively identify and mark false targets formed by multiple reflection echoes in navigation radar echoes, avoiding false alarms and missed detections caused by false echoes on surface target observations. Furthermore, ship scale information is extracted from the echoes, facilitating subsequent surface target identification and tracking.

Description

Technical Field

[0001] The invention belongs to the technical field of navigation radar signal processing, and in particular relates to a navigation radar multiple reflection false target recognition technology. Background Art

[0002] Navigation radars are primarily used to detect and track surface targets, assisting ships in navigation and collision avoidance, and ensuring safe navigation. When a cargo ship, especially an unladen one, sails near the line connecting the ship's navigation radar, multiple reflections of electromagnetic waves off the front and rear walls of the ship's hold can cause multiple echoes of the target to appear on the radar screen. These echoes are typically located in the same direction, and the echo energy gradually decreases with distance. Echoes can also sometimes appear overlapping, easily leading to misjudgment.

[0003] When ships navigate narrow waterways, the conditions for multiple reflections of false targets are very likely to be met due to the curvature of the waterway and the dense presence of ships. In severe cases, this can result in a long tail behind the true echo of the ship, affecting the observation and tracking of real targets on the water surface. This effect becomes more significant when the range resolution of navigation radar is improved (such as with solid-state pulse compression and continuous wave operating modes).

[0004] While navigation radar operators can identify these false targets based on experience, this inevitably reduces attention to other targets, increasing the risk to ships in increasingly complex navigation environments. Furthermore, since multiple reflection echoes are only generated by specific vessels, such as unladen cargo ships, extracting the corresponding echo characteristics can also be used for target identification and other purposes. Therefore, it is extremely necessary to distinguish and extract navigation radar multiple reflection false targets from the complex received signals. Summary of the Invention

[0005] The purpose of the present invention is to overcome the defect of the existing navigation radar in identifying surface moving targets, especially multiple reflection false targets formed by empty cargo ships at sea, and propose a navigation radar multiple reflection false target identification technology based on arithmetic progression extraction to eliminate or reduce the impact of the multiple reflection false targets on the navigation radar's detection of real targets on the water surface, thereby improving the navigation radar's target recognition capability.

[0006] The purpose of the present invention is achieved through the following technical solutions:

[0007] A navigation radar multiple reflection false target identification technology based on arithmetic sequence extraction is characterized by utilizing the rule that false echoes formed by multiple reflections appear at equal intervals along the same azimuth and distance and their energy decays successively. First, a group of equally spaced extreme value points of each pulse echo is extracted. Extreme value points distributed according to the arithmetic sequence rule are extracted with a distance sampling rate δr as a step interval. Then, "range-azimuth-equal interval value" clustering and "group number-class number" clustering are performed in sequence. Finally, multiple reflection false echoes are identified and marked, thereby eliminating or reducing their influence on the navigation radar detection of real surface targets.

[0008] In a preferred embodiment, the clustering is performed when the radar RF echo signal enters the receiver after being received by the antenna, is amplified, down-converted and filtered by the receiver to become an intermediate frequency signal, is then converted into a digital signal by AD, and is stored in a data buffer after distance dimension processing including matched filtering; after smoothing and filtering the data on each azimuth line, a group of equally spaced extreme points is extracted, and the distance coordinates, azimuth coordinates, spacing value and extreme point group number of each extreme point are recorded; the neighborhood radius is calculated by the distance, azimuth and spacing value, and the extracted extreme points are clustered once using a density-based clustering algorithm such as DBSCAN, and the class number to which each extreme point belongs is recorded; the neighborhood radius is calculated by the group number and class number, and the extreme points are clustered twice again using a density-based clustering algorithm; and then multiple reflection false targets are extracted from the clustering results and integrated with the radar video on the integrated display module.

[0009] The present invention can effectively identify and mark false targets formed by multiple reflection echoes in navigation radar echoes, avoiding false alarms and missed judgments caused by false echoes in surface target observations; and extract hull scale information from them, facilitating subsequent surface target identification and tracking.

[0010] The preferred solution is to extract the extreme point by eliminating the noise or interference in the radar echo, performing mean filtering on the original echo data, and then extracting the maximum point that exceeds the threshold. The threshold is controlled by the navigation radar gain knob.

[0011] The preferred solution is to extract arithmetic progressions with the distance sampling rate δr as the step interval, in [L min , L max ] interval traverses the equally spaced values, extracts the extreme point group with similar azimuth coordinates and distance coordinates distributed according to the law of arithmetic progression, adds two dimensions of equally spaced value and group number to each point in the extreme point group, and forms a five-dimensional point cloud of "amplitude-distance-azimuth-equally spaced value-group number", where L min , L max are the minimum and maximum possible ship lengths in the scenario, respectively.

[0012] The preferred solution is to calculate the neighborhood radius through distance, azimuth and spacing values, use the density-based clustering algorithm to cluster the extracted extreme points and eliminate noise points, obtain the echo nodes of multiple reflection false targets, and record the class number of each extreme point.

[0013] The preferred solution is that the secondary clustering calculates the neighborhood radius by the group number and the class number, and uses the density-based clustering algorithm to associate the echo nodes extracted by the primary clustering into a chain. Except for the starting node of the chain, which is the real target echo, the remaining nodes of the chain are all false echoes formed by multiple reflections.

[0014] The preferred solution is to display the extracted multiple reflection echoes in the form of a chain composed of points and lines superimposed on the navigation radar video image to assist the radar operator in identifying multiple reflection false targets. The multiple reflection false echo information can also be used to eliminate false targets in radar videos and dot traces, and add tank length information to the dot traces formed by the ship's real echoes to improve tracking performance.

[0015] The beneficial effects of the present invention are:

[0016] 1. The present invention extracts the original echo extreme points based on the law of arithmetic progression with the distance sampling rate δr as the step interval, which can effectively identify and mark the false targets formed by multiple reflection echoes in the navigation radar echo, avoiding false alarms and missed judgments caused by false echoes in the observation of surface targets;

[0017] 2. Obtain the minimum and maximum possible ship length data in the scene from the five-dimensional point cloud of "amplitude-range-azimuth-equidistant value-group number" to facilitate subsequent surface target identification and tracking;

[0018] 3. Calculate the neighborhood radius using distance, azimuth, and the optimal spacing value. Use a density-based clustering algorithm to cluster the extracted extreme points and remove noise points to obtain the echo nodes of multiple-reflection false targets. Record the class number of each extreme point.

[0019] 4. Calculate the neighborhood radius using the group number and class number, and use a density-based clustering algorithm to link the echo nodes extracted by the first clustering into chains to obtain a second clustering to distinguish between true target echoes and false echoes formed by multiple reflections;

[0020] 5. The extracted multiple reflection echoes are superimposed on the navigation radar video image in the form of a chain composed of points and lines to achieve comprehensive display, which is used to identify and eliminate multiple reflection false targets for easy observation. At the same time, the tank length information is added to the dot traces formed by the real echo of the ship to improve tracking performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a schematic diagram of the principle of forming multiple reflection false echoes of the navigation radar of the present invention.

[0022] Figure 2 This is a block diagram of the navigation radar multiple reflection false target recognition algorithm based on arithmetic progression extraction.

[0023] Figure 3 Schematic diagram of the one-dimensional range image of multiple reflection echoes.

[0024] Figure 4 Extract a flowchart for an arithmetic progression.

[0025] Figure 5 This is a schematic diagram of "distance-azimuth-equal interval value-group number".

[0026] Figure 6 This is a schematic diagram of the results of a single clustering, namely the "distance-azimuth-equal interval value" clustering.

[0027] Figure 7 This is a schematic diagram of the results of secondary clustering, namely "group number-class number" clustering.

[0028] Figure 8 It is a schematic diagram of the comprehensive display effect. DETAILED DESCRIPTION

[0029] The following is a detailed description of an embodiment of the present invention in conjunction with the accompanying drawings: This embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation method and specific operation process are given, but the protection scope of the present invention is not limited to the following embodiment.

[0030] Example: A navigation radar multiple reflection false target identification technology based on arithmetic sequence extraction utilizes the rule that false echoes formed by multiple reflections appear at equal intervals along the same azimuth and their energy decays successively. First, a group 1 of equally spaced extreme points of each pulse echo is extracted. The extreme points distributed according to the arithmetic sequence rule are extracted using the distance sampling rate δr as the step interval. Then, a "distance-azimuth-equal interval value" primary clustering and a "group number-class number" secondary clustering are performed in sequence. Finally, the multiple reflection false echoes are identified and marked, thereby eliminating or reducing the impact of multiple reflection false echoes on the navigation radar's detection of real surface targets. The formation principle of multiple reflection false echoes of navigation radar can be found in [1]. Figure 1 .

[0031] The clustering is when the radar RF echo signal enters the receiver after being received by the antenna, and becomes an intermediate frequency signal after being amplified, down-converted and filtered by the receiver, and then converted into a digital signal by AD, and stored in the data buffer after being processed in the distance dimension including matched filtering; after smoothing and filtering the data on each azimuth line, the extreme point groups with equal spacing are extracted, and the distance coordinates, azimuth coordinates, spacing value and extreme point group number of each extreme point are recorded; the neighborhood radius is calculated by the distance, azimuth and spacing value, and the extracted extreme points are clustered once using a density-based clustering algorithm such as DBSCAN, and the class number to which each extreme point belongs is recorded; the neighborhood radius is calculated by the group number and class number, and the extreme points are clustered twice again using a density-based clustering algorithm; and then the multiple reflection false targets are extracted from the clustering results and displayed in fusion with the radar video on the integrated display module. The block diagram of the navigation radar multiple reflection false target recognition algorithm based on arithmetic progression extraction can be found in Figure 2 .

[0032] Extreme point extraction is to remove the burrs that may be generated by noise or interference in the radar echo by performing mean filtering on the original echo data, and then extract the maximum points that exceed the threshold. The threshold is controlled by the navigation radar gain knob.

[0033] The arithmetic sequence extraction is based on the distance sampling rate δr as the step interval, in [L min , L max ] interval traverses the equally spaced values, extracts the extreme point group with similar azimuth coordinates and distance coordinates distributed according to the law of arithmetic progression, adds two dimensions of equally spaced value and group number to each point in the extreme point group, and forms a five-dimensional point cloud of "amplitude-distance-azimuth-equally spaced value-group number", where L min , L max are the minimum and maximum possible ship lengths in the scenario, respectively.

[0034] Specifically, the first clustering is to calculate the neighborhood radius through the distance, azimuth and spacing values, and use the density-based clustering algorithm to cluster the extracted extreme points and eliminate noise points to obtain the echo nodes of multiple reflection false targets, and record the class number of each extreme point.

[0035] Secondary clustering calculates the neighborhood radius by group number and class number, and uses a density-based clustering algorithm to associate the echo nodes extracted by primary clustering into chains. Except for the starting node of the chain, which is the real target echo, the remaining nodes of the chain are all false echoes formed by multiple reflections.

[0036] The integrated display is to superimpose the extracted multiple reflection echoes in the form of a chain composed of points and lines on the navigation radar video image to assist the radar operator in identifying multiple reflection false targets. The multiple reflection false echo information can also be used to eliminate false targets in radar videos and dot traces, and add tank length information to the dot traces formed by the ship's real echoes to improve tracking performance.

[0037] After the echo data of each pulse of the marine radar is processed in the distance dimension including matched filtering, it is first smoothed to eliminate the burrs that may be caused by noise or interference. Then the data is detected to cross the threshold. The threshold can be preset according to the sensitivity of the navigation radar receiver and the constant false alarm index, or it can be adjusted in real time by reading the navigation radar gain control value. The maximum value point that crosses the threshold is extracted. The results of the extreme point extraction are as follows: Figure 3 shown.

[0038] If there is a time sensitivity control (STC) circuit in the navigation radar receiver, the energy of the extreme point needs to be compensated according to the STC attenuation value corresponding to the distance of the extreme point to preserve the energy distribution characteristics of the multiple reflection echoes.

[0039] For the set of over-threshold maximum points extracted from the single pulse echo data, the range sampling rate δr is used as the step interval, and [L min , L max ] Interval traversal of equally spaced values ​​(where L min , L max (where is the minimum and maximum possible ship length in the scene, respectively) and extracts the extreme point group whose distance coordinates are distributed according to the arithmetic progression. During the extraction process, the following characteristics of multiple reflection false echoes are used to eliminate related interference:

[0040] 1) The echo energy at the starting point of the multiple reflection false echo sequence should be higher than the threshold Th;

[0041] 2) The echo energy of point n+1 in a multiple reflection false echo sequence should not exceed 1 / 2 of that of point n (the energy decays gradually with the increase in the number of reflections);

[0042] 3) The length of the multiple reflection false echo sequence should be greater than P min .

[0043] Extraction flow chart as Figure 4 As shown, the array A[n] (n=1, 2, ..., N) is the amplitude corresponding to the echo data, the array R[n] (n=1, 2, ..., N) is the distance corresponding to the echo data, N is the echo data length, and θ is the antenna beam azimuth of the navigation radar when collecting echo data.

[0044] After extracting the extreme point group of an arithmetic progression that meets the requirements, all points in the extreme point group are added to the point cloud Q. Each point contains five-dimensional information: amplitude, distance, orientation, equal spacing value, and group number. When each point is added, the set Q is checked to see if there is already a point with the same distance and orientation as the newly added point, and an equal spacing value of 1 / M (M is an integer). If so, the newly added point is ignored to prevent duplicate extraction of the extracted arithmetic progression. Figure 5 The distribution of point cloud Q in the three-dimensional space of "distance-azimuth-equal spacing value" is given.

[0045] For point cloud Q, the neighborhood radius between two points is calculated using distance, azimuth coordinates, and equal interval values. When the difference between the equal interval values ​​of two points exceeds a threshold, it is considered that the two points cannot belong to the same group of multiple reflection false echoes, that is, the neighborhood radius between the two points is infinite; when the difference between the equal interval values ​​of two points is lower than the threshold, the Euclidean distance between the two points is calculated using the normalized distance and azimuth of the two points as the neighborhood radius between the two points. The specific calculation formula is:

[0046]

[0047] Where r is the distance, θ is the azimuth, l is the equal spacing value, Δr is the radar range resolution, and Δθ is the radar azimuth resolution.

[0048] After the neighborhood radius is calculated, the density-based clustering algorithm is used to cluster the point cloud Q and remove noise points. The “distance-azimuth-equal interval value” clustering result is as follows: Figure 6 As shown in Figure 1, different grayscales are used to distinguish class numbers. After removing noise, each point in the point cloud Q is given a "class number" dimension, meaning each point contains six dimensions of information: amplitude, distance, orientation, equal spacing value, group number, and class number.

[0049] For point cloud Q, the neighborhood radius between two points is calculated using the group number and class number. When the group number or class number of two points is the same, it is considered that the two points belong to the same group of multiple reflection false echoes, and the neighborhood radius is zero; when the group number and class number of two points are different, it is considered that the two points cannot belong to the same group of multiple reflection false echoes, and the neighborhood radius is infinite. The specific calculation formula is:

[0050]

[0051] Where C is the group number and D is the class number.

[0052] After the neighborhood radius is calculated, the density-based clustering algorithm is used to cluster the point cloud Q and remove noise points. The clustering results are as follows: Figure 7 As shown in the figure, each point in the dotted ellipse represents a group of multiple reflection false echoes formed by an empty tank ship, and then the tank length and echo nodes are extracted (the initial node corresponds to the real target echo, and the subsequent nodes correspond to the false target echo). The tank length is the average of the equally spaced values ​​of all points, and the calculation formula is as follows:

[0053]

[0054] Where U is a set of points contained in a group of multiple reflection false echoes, N is the total number of points contained in the point set U, and L i is the equally spaced value of point i in the point set U.

[0055] Points with the same class number in a group of multiple reflection false echoes belong to the same echo node. The distance and azimuth coordinates of the echo node are calculated as follows:

[0056]

[0057] Where V is the set of points contained in the echo node, r j ,θ j and A j are the distance, orientation and amplitude of point j in point set V respectively.

[0058] The above detection results are superimposed and drawn on the navigation radar video image, such as Figure 8 As shown in the figure, the automatic identification and marking of multiple reflection false echoes is completed, which serves the purpose of assisting observation. The initial echo node (real target echo) is marked with a solid dot, and the remaining echo nodes (false target echoes) are marked with hollow dots. The echo nodes are connected by lines, and a group of multiple reflection echoes are represented by the same point type.

[0059] The multi-reflection false echo information can be used to remove false targets in navigation radar videos and tracks, and to add tank length information to the identified ship tracks to improve tracking performance.

[0060] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A navigation radar multiple reflection false target identification method based on arithmetic progression extraction, characterized in that: Utilizing the rule that false echoes formed by multiple reflections appear at equal intervals along the same azimuth and their energy decays successively, the method first extracts a group of equally spaced extreme points from each pulse echo. Using the distance sampling rate δr as the stepping interval, the extreme points distributed according to the law of arithmetic progression are then extracted. Then, "range-azimuth-equally spaced value" clustering and "group number-class number" clustering are performed in sequence. Finally, the multiple reflection false echoes are identified and marked, thereby eliminating or reducing their impact on the navigation radar's detection of real surface targets. The clustering is that when the radar radio frequency echo signal enters the receiver after being received by the antenna, it becomes an intermediate frequency signal after being amplified, down-converted and filtered by the receiver, and then converted into a digital signal by AD, and stored in the data cache after distance dimension processing including matched filtering; after smoothing and filtering the data on each azimuth line, a group of equally spaced extreme points is extracted, and the distance coordinates, azimuth coordinates, spacing value and extreme point group number of each extreme point are recorded; the neighborhood radius is calculated based on the distance, azimuth and spacing values, and the extracted extreme points are clustered once using a density-based clustering algorithm, and the class number to which each extreme point belongs is recorded; the neighborhood radius is calculated based on the group number and class number, and the extreme points are clustered twice again using a density-based clustering algorithm; and then multiple reflection false targets are extracted from the clustering results and integrated with the radar video on the integrated display module.

2. the navigation radar multiple reflection false target identification method based on arithmetic progression extraction according to claim 1, is characterized in that, Extreme point extraction is to remove the burrs generated by noise or interference in the radar echo from the original echo data and perform mean filtering, and then extract the maximum points that exceed the threshold. The threshold is controlled by the navigation radar gain knob.

3. the navigation radar multiple reflection false target identification method based on arithmetic progression extraction according to claim 2, is characterized in that, The arithmetic sequence extraction is based on the distance sampling rate δr as the step interval, in [L min , L max ] interval traverses the equally spaced values, extracts the extreme point group with similar azimuth coordinates and distance coordinates distributed according to the law of arithmetic progression, adds two dimensions of equally spaced value and group number to each point in the extreme point group, and forms a five-dimensional point cloud of "amplitude-distance-azimuth-equally spaced value-group number", where L min , L max The minimum and maximum ship lengths in the scene respectively.

4. the navigation radar multiple reflection false target identification method based on arithmetic progression extraction according to claim 3, is characterized in that, The first clustering is to calculate the neighborhood radius through the distance, azimuth and spacing values, and use the density-based clustering algorithm to cluster the extracted extreme points and eliminate noise points to obtain the echo nodes of multiple reflection false targets, and record the class number of each extreme point.

5. the navigation radar multiple reflection false target identification method based on arithmetic progression extraction according to claim 4 is characterized in that, Secondary clustering calculates the neighborhood radius by group number and class number, and uses a density-based clustering algorithm to associate the echo nodes extracted by primary clustering into chains. Except for the starting node of the chain, which is the real target echo, the remaining nodes of the chain are all false echoes formed by multiple reflections.

6. The navigation radar multiple reflection false target identification method based on arithmetic progression extraction according to claim 5 is characterized in that, The integrated display is to superimpose the extracted multiple reflection echoes in the form of a chain composed of points and lines on the navigation radar video image to assist the radar operator in identifying multiple reflection false targets. The multiple reflection false echo information can also be used to eliminate false targets in radar videos and dot traces, and add tank length information to the dot traces formed by the ship's real echoes to improve tracking performance.

Images