CN113537347B - Unmanned aerial vehicle and flying bird target classification method based on track motion characteristics - Google Patents
Unmanned aerial vehicle and flying bird target classification method based on track motion characteristics Download PDFInfo
- Publication number
- CN113537347B CN113537347B CN202110802116.7A CN202110802116A CN113537347B CN 113537347 B CN113537347 B CN 113537347B CN 202110802116 A CN202110802116 A CN 202110802116A CN 113537347 B CN113537347 B CN 113537347B
- Authority
- CN
- China
- Prior art keywords
- track
- target
- class
- joint probability
- targets
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
- G06F18/24155—Bayesian classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Probability & Statistics with Applications (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses an unmanned aerial vehicle and a flying bird target classification method based on track motion characteristics, and belongs to the technical field of radar target tracking and recognition. The method comprises the following steps: extracting a motion characteristic vector from each target track, wherein the motion characteristic vector comprises an average speed, a speed standard deviation, a heading standard deviation, a maneuvering factor and a track oscillation factor; the motion feature vectors extracted from different target tracks form a training sample set, a hierarchical table is established for each feature element, and two feature elements are arbitrarily selected to construct a joint probability matrix of each type of target; and extracting motion characteristic vectors from unknown target tracks, and identifying the belonging target category according to the hierarchical table and the joint probability matrix. The method and the device realize effective identification of the unmanned aerial vehicle and the flying bird targets, avoid the problem of large identification difficulty of the two targets of the flying bird and the unmanned aerial vehicle caused by high overlapping degree of radar scattering cross section values, and have high identification efficiency and strong universality.
Description
Technical Field
The invention belongs to the technical field of radar target tracking and recognition, relates to radar target feature extraction and recognition classification, and in particular relates to an unmanned aerial vehicle based on track motion features and a flying bird target classification method.
Background
The flying bird and the unmanned aerial vehicle target belong to typical 'low-speed and small' targets, and the targets generally have the characteristics of low flying height, low flying speed, small radar scattering cross section, low detectability and the like. The bird target has a large safety threat to the civil airliner in the stage of entering and exiting ports, and the commercial small unmanned aerial vehicle is easier to be utilized by illegal molecules due to the rapid development of the commercial unmanned aerial vehicle, so that the low-altitude safety of key areas such as airports is seriously threatened. Therefore, accurate identification and tracking of bird and unmanned aerial vehicle targets has important application and research value.
Low-altitude surveillance radars are already applied to important areas such as airports, wind farms, borders and the like. The existing low-altitude monitoring radar generally adopts a Doppler signal processing technology based on a solid-state power amplifier, has small target detection capability in a complex low-altitude strong ground clutter environment, and realizes all-weather high-resolution continuous monitoring of a specific airspace. However, bird targets in low-altitude areas generally have high similarity to low-altitude unmanned aerial vehicles in terms of radar cross section, polarization characteristics, flight speed, altitude, and the like. The radar target micro-Doppler feature has the potential of accurately distinguishing the flying bird from the rotor unmanned aerial vehicle in theory, but the micro-Doppler feature quality is highly related to the monitoring distance in the actual application scene, the small target long-distance micro-Doppler feature extraction difficulty is high, and the application value in the actual radar monitoring needs to be further improved. Therefore, the existing low-altitude surveillance radar is difficult to realize effective discrimination for the bird and the unmanned aerial vehicle target.
Disclosure of Invention
The invention aims to solve the problems, and provides a rotor unmanned aerial vehicle and flying bird target classification and identification method based on radar track motion characteristics, which is suitable for effectively identifying flying birds and unmanned aerial vehicle targets in low-altitude complex environments.
A rotor unmanned aerial vehicle and flying bird target classification recognition method based on radar track motion characteristics utilizes radar target track information to extract target motion characteristic vectors, and adopts a naive Bayesian method to distinguish flying birds and unmanned aerial vehicle targets, comprising the following steps:
step one, extracting a motion characteristic vector from a target track acquired by a radar; the feature elements contained in the motion feature vector are: average speed, speed standard deviation, heading standard deviation, maneuver factor and track oscillation factor;
step two, extracting motion feature vectors from tracks of different targets by adopting the step 1 to form a training sample set, and then constructing a joint probability matrix of the motion features of the different targets;
and thirdly, judging the category of the unknown target track by using the obtained joint probability matrix.
Extracting motion characteristic vectors from a target track acquired by a radar, wherein the motion characteristic vectors comprise average speed, speed standard deviation, heading standard deviation, maneuvering factors and track oscillation factors; the maneuvering factor is calculated as the ratio of the average speed to the heading standard deviation; the track oscillation factor is calculated as follows: firstly, calculating a heading difference symbol vector O of a target track, wherein the kth heading difference symbol in O is calculatedWherein Δh (k) is the heading angle between the (k+1) th track sampling point and the (k) th track sampling pointDifference, delta e Is a measurement error threshold; then judging the course oscillation times in the target track according to the vector O, if one of the following two oscillation modes exists, judging that one course oscillation exists,
oscillation mode 1: o (i-1) +o (i) =0 and O (i-1) noteqo (i);
oscillation mode 2: o (i-1) +o (i+1) =0 and O (i-1) noteqo (i+1), O (i) =0;
finally, calculating the oscillation factorWherein, deltah (i) is the absolute value of course angle change under the ith oscillation, and w (i) is a weight factor.
The second step comprises the following steps: (1) first establishing a hierarchical table for each feature element; (2) Secondly, two characteristic elements in the motion characteristic vector are selected at will to construct a joint probability matrix of each type of target; when constructing a joint probability matrix of a t-th type target, initializing the joint probability matrix as a 0 matrix, traversing motion feature vectors of each track of the t-th type target in a training sample set, acquiring levels m and n corresponding to the values of the two selected feature elements according to a hierarchical table, increasing the value of an element (m, n) in the joint probability matrix by 1, and normalizing the elements in the joint probability matrix after traversing; the value of the element (m, n) in the obtained joint probability matrix represents the probability p that the first characteristic element belongs to the class m under the condition that the first characteristic element belongs to the class n for the class t target.
In the third step, motion feature vectors are extracted for unknown target tracks collected by a radar, the levels of the feature values are obtained according to a hierarchical table, then for each type of target, probability values of the unknown target tracks belonging to the type of target are searched from 20 joint probability matrixes, and finally the type with the highest probability is selected as the type of the unknown target tracks.
Compared with the prior art, the invention has the advantages and positive effects that:
(1) According to the method, the target track information is utilized to extract the target motion characteristics, the effective identification of the targets of the flying bird and the unmanned aerial vehicle is realized without depending on radar echo intensity information such as radar scattering cross sections, and the problem of high identification difficulty of the two targets caused by high overlapping degree of the values of the radar scattering cross sections is solved.
(2) The method has the advantages of high construction speed of the target feature vector, high recognition efficiency and high flexibility, and can be widely applied to recognition of other low-altitude interference targets such as precipitation clutter.
(3) According to the method, software and hardware equipment of the low-altitude airspace radar monitoring system are not required to be changed or upgraded, and effective identification of unmanned aerial vehicles and flying bird targets can be realized only by carrying out feature extraction and feature vector reconstruction on radar target tracks, so that universality is high.
Drawings
FIG. 1 is a schematic diagram of a radar target recognition process implemented by the method of the present invention;
fig. 2 is a diagram illustrating an example of the trajectory distribution of a lightweight unmanned aerial vehicle and a bird target according to an embodiment of the present invention.
Detailed Description
The invention will be described in further detail with reference to the drawings and examples.
The invention relates to an unmanned aerial vehicle and a flying bird target classification method based on track motion characteristics, which utilize the difference existing in the motion modes of flying birds and rotor unmanned aerial vehicle targets to realize the effective classification and identification of a light unmanned aerial vehicle and the flying bird targets in a complex low-altitude environment, and comprise the following steps:
step one, extracting a radar target track motion characteristic vector.
The radar acquires the track of a moving object, and defines the target track Z= [ Z ] 1 ,z 2 ,...,z N ]The extraction speed and heading information vectors V and H according to the space-time information of each sampling point in the flight path are as follows:
V＝[v 1 ,v 2 ,...,v N ]；H＝[h 1 ,h 2 ,...,h N ] (1)
where N represents the number of sampling points in the track, i.e. the track length. v i 、h i Indicating the speed and heading of the target at the ith sample point, i=1, 2, … N, respectively.The target track length is closely related to the target characteristics, the motion pattern and the radar tracking algorithm, and the track lengths are not generally consistent. The invention constructs the motion feature vector of the target track by extracting the statistical descriptors of the track motion features, realizes the consistency of the feature dimensions of the target track, and is favorable for realizing target identification by adopting a supervised machine learning algorithm. The invention extracts the following five types of track motion characteristics:
(1) Average velocity v mean The calculation is as follows:
(2) Standard deviation of velocity v std The calculation is as follows:
(3) Heading standard deviation h std The calculation is as follows:
parameter delta in equation (4) h (i) Is defined as:
wherein delta h For the threshold value of the heading angle difference, the invention sets the threshold value to be 50 degrees. h is a mean The average course value of the target track is obtained by averaging the course of N sampling points in the track.
(4) The maneuver factor σ, defined as shown in equation (6), describes the maneuver performance of the target. The larger the sigma value, the simpler the motion pattern representing the object in the track, and the lower the corresponding mobility.
(5) Trace oscillation factor ζ.
The heading angle difference between track sampling points is defined as Δh (k) =h k+1 -h k K=1, 2, …, N-1. Taking into account the actual radar measurement error, a measurement error threshold delta is defined e =0.5°, defining the kth heading difference sign J of the target track h (k) The method comprises the following steps:
J h (k) For the kth element O (k) in the heading difference sign vector O of the target track, the vector O may be described as o= [1, -1,0,1, …]In the form of (a). According to the heading difference symbol vector O, if one of the following two oscillation modes exists:
oscillation mode 1: o (k-1) +O (k) =0 and O (k-1) noteqO (k)
Oscillation mode 2: o (k-1) +o (k+1) =0 and O (k-1) noteqo (k+1), O (k) =0
It is determined that there is a course oscillation in the track. And judging the oscillation times of O by adopting the criteria. If there is one or more oscillations, defining an oscillation factor:
wherein, |Δh (i) | is the absolute value of the change of heading angle under the ith oscillation. The weight factor w (i) is defined as shown in table 1:
TABLE 1 weight factor definition
| Number of oscillations | 1 | 2 | 3 | 4 | Greater than or equal to 5 |
| Weight factor | 1 | 1.5 | 2 | 3 | 5 |
Constructing a track motion feature vector by adopting the five types of track motion feature descriptors:
S＝[v mean ,v std ,h std ,σ,ζ] (9)
as shown in fig. 2, examples of bird and drone tracking tracks based on the low altitude airspace observations about seattle tacoma international airport in the united states provided by the canadian actilter company bird detection CASCADE radar system. And constructing a target track motion characteristic vector according to the information such as the corresponding speed, direction and the like on each track point in the track.
And step two, constructing a target motion characteristic joint probability matrix.
The invention provides a rotor unmanned aerial vehicle and flying bird target classification recognition method based on radar track motion characteristics, which adopts a supervision type learning algorithm. And (3) assuming that the track motion feature vector elements are mutually independent, acquiring radar tracking tracks of different types of targets, and constructing the motion feature vector of the target track by adopting the method in the step one. A training database and a joint distribution probability matrix are then constructed based on the plurality of motion feature vector sample sets.
Establishing a hierarchical table for each motion feature descriptor in formula (9) by v mean The descriptor is an illustration of a descriptor hierarchy boundary value definition method. For v in a defined value range mean Is a value v of each class of target subset in the training sample set mean The corresponding probability value in the database isWherein->And (3) representing the probability that the target class is the T-th class under the condition that the average speed value is v, wherein T is the number of the classes of the targets in the training sample set. In the embodiment of the present invention, for example, the target categories are classified into two categories, that is, a bird and an unmanned aerial vehicle, i.e., t=2. Will v mean Dividing the value range into Q levels, defining the lower bound and the upper bound of the value range corresponding to each level as +.>i=1, 2,..q-1, t=1, 2,.. T, numerically satisfies the following conditions:
at the lower boundary of the determined value rangeAfter that, when the integral value satisfies the definition of the formula (10), the upper bound +.>The maximum value and the minimum value of the characteristic elements can be obtained by counting the values of the characteristic elements in the training data set, and the minimum value is used as the lower boundary of the level 1, namely +.>Then the upper bound ++is obtained according to equation (10)>I.e., the upper bound of level 1, and so on, and then sequentially acquiring the 3 rd to the Q th boundary values according to the formula (10).
For different classes of targets T (t=1, 2,., T), an array of boundary values consisting of Q levelsThere is a certain difference in values, and the final boundary value is obtained by adopting a method of carrying out statistical average on T-class targets, as shown in the following formula:
r v (i) Is the i-th boundary value of the average speed.
Other classes of feature descriptors can be calculated using similar methods as described in equations (10) and (11) above.
After the hierarchical table of all track motion feature descriptors is obtained, two feature elements { S (i), S (j) } in the motion feature vector S can be arbitrarily selected to construct a joint probability matrixThe dimension of the joint probability distribution matrix is 5 x 5, according to the definition of S. The two feature elements selected are different, i.e. i+.j, where i, j denote the i-th and j-th feature elements in the motion feature vector S, respectively. For training sample set with target class t, corresponding joint probability distribution matrix +.>The calculation can be performed by the following method:
will beArray initialization to a 5 x 5 dimension array with all cells of 0, traversing each leg in a training sample set of class tTrace and its corresponding feature vector. The value x corresponding to descriptors S (i) and S (j) 1 And x 2 The corresponding levels m and n in the hierarchical table are derived and the matrix is updated +.>Middle element->The following are provided:
after the traversal of all t-th class target training samples is completed, the matrix is obtainedThe matrix elements are normalized by the following method:
for normalized matrixIts matrix element value->Representing that for the t-th class target, the probability that the track characteristic element S (j) belongs to the class m is P under the condition that S (i) belongs to the class n, so that the conditional probability p=p can be further used ij (n|m) description. The order of the selected characteristic elements is different, and the obtained joint probability matrix is also different, namely the matrix +.>And->Representative ofTwo different matrices. Thus, for the feature space in which the feature vector is located, a +.>And a joint probability distribution matrix.
Tables 2 and 3 show five types of motion feature classification tables and joint probability distribution matrices of average speed and heading standard deviation features, respectively, based on experimental data of this example.
TABLE 2 track movement characteristics hierarchical form
TABLE 3 Joint probability matrix example of average speed-heading standard deviation
| Unmanned plane/bird | 1 | 2 | 3 | 4 | 5 |
| 1 | 0.25\0.18 | 0.27\0.35 | 0.43\0.23 | 0.02\0.14 | 0.03\0.1 |
| 2 | 0.19\0.11 | 0.24\0.31 | 0.41\0.26 | 0.12\0.2 | 0.02\0.12 |
| 3 | 0.21\0.18 | 0.33\0.41 | 0.32\0.28 | 0.07\0.08 | 0.07\0.05 |
| 4 | 0.15\0.21 | 0.37\0.35 | 0.38\0.31 | 0.06\0.08 | 0.04\0.05 |
| 5 | 0.12\0.23 | 0.35\0.42 | 0.29\0.27 | 0.15\0.06 | 0.09\0.02 |
As shown in table 2 and table 3, in the embodiment of the present invention, 5 levels are set for each feature element, table 2 gives a hierarchical table of 5 feature elements, and two feature elements { average speed, heading standard deviation } are selected in table 3 to establish a joint probability matrix, and values in the joint probability matrix are calculated for two types of targets, i.e., a bird and an unmanned plane, respectively.
And thirdly, judging the target track category by utilizing the joint probability matrix.
Unknown target track for radar acquisitionExtracting motion feature vectors according to the first method, obtaining the belonging level of each feature value according to the hierarchical table of the second method, searching corresponding probability values based on 20 joint probability matrixes constructed by various targets in the second method, and constructing probability vectors +.>Wherein P is l t The values are from the first joint probability distribution matrix, l=1, 2, …,20; t=1, 2,. -%, T.
According to the Bayesian theorem, and assuming characteristic elementsThe track is mutually independent, and the probability of belonging to the t-th class of targets is as follows:
assume thatIndependent of category t, it is determined that the main information of the category comes from +.>Based on the statistical independent assumption, the joint probability model satisfies:
wherein, the liquid crystal display device comprises a liquid crystal display device,representing unknown target track derived from the first joint probability matrix of the t-th class of targets>Probabilities belonging to class t objects, i.e. corresponding probabilities P l t . Assuming that the various objects appear with equal probability, i.e. +.>The category to which the track belongs is determined by the following method:
in the embodiment, the identification method is adopted for verification by using the flying bird observation track data and the experimental unmanned aerial vehicle track observation data of three months of 2016, 8, 9 and 10. The correct recognition rates for the bird and drone targets are shown in table 4.
Table 4 bird/unmanned aerial vehicle recognition accuracy
| Bird/unmanned aerial vehicle | Early morning 05:00-07:00 | Evening 17:00-19:00 |
| August of August | 82/87 | 86/88 |
| September (September) | 84/85 | 85/89 |
| October (October) | 80/90 | 88/92 |
According to the embodiment, the method can effectively identify the flying birds and the unmanned aerial vehicle, and solves the problem that the identification difficulty is high due to the fact that the number overlapping degree of radar scattering cross sections of the existing two targets is high.
Claims (7)
1. An unmanned aerial vehicle and a flying bird target classification method based on track motion characteristics are characterized by comprising the following steps:
step 1, extracting a motion characteristic vector from a target track acquired by a radar; the feature elements contained in the motion feature vector are: average speed, speed standard deviation, heading standard deviation, maneuver factor and track oscillation factor;
step 2, extracting motion feature vectors from tracks of different targets by adopting the step 1 to form a training sample set, and then constructing a joint probability matrix of motion features of different targets, wherein the method comprises the following steps: (1) first establishing a hierarchical table for each feature element; (2) Secondly, two characteristic elements in the motion characteristic vector are selected at will to construct a joint probability matrix of each type of target; when constructing a joint probability matrix of a t-th type target, initializing the joint probability matrix as a 0 matrix, traversing motion feature vectors of each track of the t-th type target in a training sample set, acquiring levels m and n corresponding to the values of the two selected feature elements according to a hierarchical table, increasing the value of an element (m, n) in the joint probability matrix by 1, and normalizing the elements in the joint probability matrix after traversing; the value of the element (m, n) in the obtained joint probability matrix represents the probability of the first characteristic element belonging to the class m under the condition that the first characteristic element belongs to the class n for the class t targetp; build get for each class of targetsA joint probability matrix;
and 3, judging the category of the unknown target track by using the obtained joint probability matrix, wherein the method comprises the following steps: extracting motion characteristic vectors from unknown target tracks acquired by a radar, acquiring the class to which each characteristic value belongs according to a hierarchical table, searching probability values of the unknown target tracks belonging to the class of targets from 20 joint probability matrixes according to each class of targets, and finally selecting the class with the highest probability as the class of the unknown target tracks.
2. The method of claim 1, wherein in step 1, the maneuver factor is denoted as σ and calculated as the average velocity v mean Standard deviation from heading h std Is a ratio of (2).
3. The method of claim 1, wherein in step 1, the track oscillation factor is denoted as ζ and is calculated as follows: firstly, calculating a heading difference symbol vector O of a target track, wherein a kth element O (k) in the O is a heading difference symbolWherein Δh (k) is the heading angle difference between the (k+1) th track sampling point and the (k) th track sampling point, and δ e Is a measurement error threshold; then judging the course oscillation times in the target track according to the vector O, and judging that one course oscillation exists if one of the following two oscillation modes exists;
oscillation mode 1: o (k-1) +o (k) =0 and O (k-1) noteqo (k);
oscillation mode 2: o (k-1) +o (k+1) =0 and O (k-1) noteqo (k+1), O (k) =0;
finally, calculating the oscillation factorWherein, deltah (i) is the absolute value of course angle change under the ith oscillation, and w (i) is a weight factor.
4. A method according to claim 3, wherein in step 1, the weight factor w (i) is set as:
when 1 course oscillation exists in the target track, w (i) =1;
w (i) =1.5 when there are 2 heading oscillations in the target track;
when 3 course oscillations exist in the target track, w (i) =2;
when 4 course oscillations exist in the target track, w (i) =3;
w (i) =5 when there are 5 or more heading oscillations in the target track.
5. The method according to claim 1, wherein in the step 2, a hierarchical table is built for the feature elements, and the method for determining the hierarchical boundary value is as follows:
at average velocity v mean For illustration, in the training sample set, setRepresenting the probability that the target class is the T-th class at an average speed value v, t=1, 2..t, T being the number of classes of targets in the training sample set; dividing the average speed into Q levels on the value range, and setting the lower bound and the upper bound of the value range corresponding to each level of the t-th class target as +.>Obtaining the upper and lower bound values by solving a definition satisfying the following formula;
and finally, determining the grading boundary value of the average speed for the statistical average value of the boundary values of each grade of the T-class targets, and obtaining a grading table of the average speed.
6. The method according to claim 1 or 5, wherein in the step 2, two feature elements { S (i), S (j) } are selected to construct a joint probability matrix of the t-th class objectAfter traversing the training sample of the t-th class object, the matrix obtained is +.>The normalization mode of each element is as follows:
wherein the matrixIs 5 x 5.
7. The method of claim 1, wherein in step 3, the unknown target track is set asSearching 20 joint probability matrixes of t-th class targets to obtain probability vectors +.>Let P be l t For probability values obtained from the first joint probability distribution matrix, which represent the track +.>Probability of belonging to class t object, l=1, 2, …,20;
assuming motion feature elementsMutually independent, then the trackProbability of belonging to class t object>
Wherein, the probability is setIndependent of category t, further yields:
let the various objects occur with equal probability,track->The category is determined by the following method:
t' represents the trackBelonging to the category.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110802116.7A CN113537347B (en) | 2021-07-15 | 2021-07-15 | Unmanned aerial vehicle and flying bird target classification method based on track motion characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110802116.7A CN113537347B (en) | 2021-07-15 | 2021-07-15 | Unmanned aerial vehicle and flying bird target classification method based on track motion characteristics |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113537347A CN113537347A (en) | 2021-10-22 |
| CN113537347B true CN113537347B (en) | 2023-07-18 |
Family
ID=78128220
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110802116.7A Active CN113537347B (en) | 2021-07-15 | 2021-07-15 | Unmanned aerial vehicle and flying bird target classification method based on track motion characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113537347B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2015121473A (en) * | 2013-12-24 | 2015-07-02 | 三菱電機株式会社 | Target tracking device and target tracking method |
| CN106546975A (en) * | 2016-10-14 | 2017-03-29 | 中国民航科学技术研究院 | A kind of small-sized unmanned plane based on radar data and flying bird classifying identification method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106970648B (en) * | 2017-04-19 | 2019-05-14 | 北京航空航天大学 | Unmanned plane multi-goal path plans combined method for searching under the environment of city low latitude |
-
2021
- 2021-07-15 CN CN202110802116.7A patent/CN113537347B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2015121473A (en) * | 2013-12-24 | 2015-07-02 | 三菱電機株式会社 | Target tracking device and target tracking method |
| CN106546975A (en) * | 2016-10-14 | 2017-03-29 | 中国民航科学技术研究院 | A kind of small-sized unmanned plane based on radar data and flying bird classifying identification method |
Non-Patent Citations (1)
| Title |
|---|
| 飞鸟与无人机目标雷达探测与识别技术进展与展望;陈小龙 等;《雷达学报》;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113537347A (en) | 2021-10-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112132042B (en) | SAR image target detection method based on contrast domain adaptation | |
| CN110018453B (en) | Intelligent model identification method based on aircraft track characteristics | |
| CN114111764B (en) | Navigation knowledge graph construction and reasoning application method | |
| CN109753874A (en) | A kind of low slow small classification of radar targets method based on machine learning | |
| CN111401168B (en) | Multilayer radar feature extraction and selection method for unmanned aerial vehicle | |
| CN111142085B (en) | External radiation source radar target classification and identification method based on track feature extraction | |
| Zhang et al. | SEFEPNet: Scale expansion and feature enhancement pyramid network for SAR aircraft detection with small sample dataset | |
| CN115943439A (en) | Multi-target vehicle detection and re-identification method based on radar vision fusion | |
| CN109117776A (en) | Aircraft and meteorological clutter classifying identification method based on track information | |
| CN113808174A (en) | Radar small target tracking method based on full convolution network and Kalman filtering | |
| CN113050797A (en) | Method for realizing gesture recognition through millimeter wave radar | |
| CN113191185A (en) | Method for classifying targets of unmanned aerial vehicle by radar detection through Dense2Net | |
| CN111736172A (en) | Air target detection method based on atmospheric disturbance coherent laser detection | |
| Cheng et al. | Moving Target Detection Technology Based on UAV Vision | |
| CN113537347B (en) | Unmanned aerial vehicle and flying bird target classification method based on track motion characteristics | |
| Vitiello et al. | Detection and tracking of non-cooperative flying obstacles using low SWaP radar and optical sensors: an experimental analysis | |
| Sévigny et al. | Unmanned aircraft (ua) telemetry data for track modelling and classification | |
| CN116304966A (en) | Track association method based on multi-source data fusion | |
| Yang et al. | Shape feature aided target detection method for micro-drone surveillance radar | |
| Han et al. | Aircraft target detection in polsar image based on region segmentation and multi-feature decision | |
| Liu et al. | Application of the fruit fly optimization algorithm to an optimized neural network model in radar target recognition | |
| Jordanov et al. | Intelligent radar signal recognition and classification | |
| Liang et al. | A Multi-Featured Detection Method for Small Target on Sea Surface Based on GSP | |
| Liu et al. | A YOLO-X Deep Learning Approach to Detect Traffic Targets from UAV Video with On-Board Vehicle Data Validation | |
| CN114359743B (en) | Low-slow small target identification method and system based on multiband |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |