CN108845302B - K-nearest neighbor transformation true and false target feature extraction method - Google Patents
K-nearest neighbor transformation true and false target feature extraction method Download PDFInfo
- Publication number
- CN108845302B CN108845302B CN201810964240.1A CN201810964240A CN108845302B CN 108845302 B CN108845302 B CN 108845302B CN 201810964240 A CN201810964240 A CN 201810964240A CN 108845302 B CN108845302 B CN 108845302B
- Authority
- CN
- China
- Prior art keywords
- neighbor
- matrix
- true
- equal
- nearest neighbor
- 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
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Radar Systems Or Details Thereof (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a K neighbor transformation true and false target feature extraction method, which belongs to the technical neighborhood of radar target identification.
Description
Technical Field
The invention belongs to the field of radar target identification technology, and particularly relates to a K nearest neighbor transformation true and false target feature extraction method.
Background
In radar target identification, the discrimination vector transformation method can increase the difference between different target features and reduce the difference between the same target features, so that the features with obvious difference are extracted, and therefore, the discrimination vector transformation method obtains good classification performance.
However, the discriminant vector transform method is only suitable for the case where the sample data is gaussian distributed, but the distribution of the sample data may be non-gaussian distributed in practice, and the recognition performance of the discriminant vector transform method is significantly degraded in the case of the non-gaussian distribution. There is room for further improvement in the recognition performance of the conventional discrimination vector conversion method.
Disclosure of Invention
The invention aims to: aiming at the existing problems, a K neighbor transformation feature extraction method is provided to overcome the defects of a conventional discrimination vector transformation method and effectively improve the classification performance of radar true and false targets.
The technical scheme of the K nearest neighbor transformation true and false target feature extraction method specifically comprises the following steps:
step 1: inputting a training sample set of one-dimensional range profile of radar target by xijRepresenting training samples, wherein subscript i is a class identifier, subscript j is a training sample identifier, i is greater than or equal to 1 and less than or equal to g, j is greater than or equal to 1 and less than or equal to NiG denotes the number of categories, NiA number of samples representing a corresponding category;
Constraint coefficient of similar K nearest neighbor ruleIs set as follows: if it isOrThenOtherwiseWhere the subscript k is a training sample discriminator of some kind, σ2The value of the gaussian parameter is represented,k representing a vector of the same kind1Set of neighbor vectors, k1Is a preset neighbor number;
Constraint coefficient of heterogeneous K neighbor ruleIs set as follows: if it isOr alternativelyThenOtherwiseWherein the subscript l is a category specifier,k representing a vector of a different class2Set of neighbor vectors, k2Is a preset neighbor number;
step (ii) of3: inputting radar true and false target one-dimensional range profile x of sub-image features to be extractedtAccording toObtaining a one-dimensional range profile xtCharacteristic vector y oft。
In summary, due to the adoption of the technical scheme, the invention has the beneficial effects that:
according to the method, the difference between the same-class sample characteristics is reduced based on the K neighbor constraint rule, the difference between different-class sample characteristics is increased for weighting, and the influence of other samples on the construction of the transformation matrix is reduced, so that the characteristics of non-Gaussian distribution sample data can be extracted, the defects of a conventional discrimination vector transformation method are overcome, and the classification performance of radar true and false targets is effectively improved.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments.
The K nearest neighbor transformation true and false target feature extraction method reduces the difference between the same type sample features based on the K nearest neighbor constraint rule, increases the weighting of the difference between the different type sample features, and reduces the influence of other samples on the construction of a transformation matrix, so that the features of non-Gaussian distribution sample data can be extracted, and the specific implementation process is as follows:
by xij(n-dimensional column vector) represents the iththJ-th of class true and false targetthI is more than or equal to 1 and less than or equal to g, j is more than or equal to 1 and less than or equal to Ni,Wherein N isiIs the iththThe number of training one-dimensional distance image samples of the true and false-like targets is N, and N is the total number of the training one-dimensional distance image samples. Will train a one-dimensional range profile xijThe following transformations are performed:
yij=ATxij (1)
where A is the transformation matrix, yijIs xijThe corresponding feature vector, T, represents the matrix transpose.
Calculating the sum of squares of differences between any two sample feature vectors of the same type in a feature spaceC:
wherein the Gaussian parameter σ2Is an empirical value, and is determined through experiments under the condition of meeting the requirement of processing precision,k representing a vector of the same kind1A set of neighbor vectors. Equation (3) shows that when two samples belonging to the same target class are k each other1When the samples are adjacent to each other, the constraint coefficient of the difference between the similar samples is not equal to zero, and the constraint coefficient of the difference between the other similar samples is zero.
Using the equation of the matrix trace, equation (2) can be converted into:
substitution of formula (1) for formula (4) gives:
by simplifying the formula (5), the following can be obtained:
SC=ATX(DC-WC)XTA (6)
wherein
Similarly, calculating the weighted distance square sum S between the heterogeneous target sample characteristics in the characteristic spaceB:
Wherein wij,lkConstraint coefficients based on heterogeneous K nearest neighbor rules:
Using the formula of the matrix trace, equation (10) can be converted into
Substituting formula (1) into formula (12)
The formula (13) is simplified to obtain
SB=ATX(DB-WB)XTA (14)
Wherein
The estimation value of the K nearest neighbor transformation matrix can be obtained by solving the following minimization problem
The right side of the formula (17) calculates the partial derivative of A and makes it equal to zero, and the estimated value of the K neighbor transformation matrix can be obtainedIs composed of a matrix (X (D)B-WB)XT)-1(X(DC-WC)XT) Is used for determining the characteristic vector corresponding to the M (n) maximum characteristic values.
Obtaining the estimated value of K adjacent transformation matrixThen, an arbitrary sample x can be obtained by using the formula (1)tCharacteristic vector y oftI.e. byAnd then radar true and false target identification processing is carried out based on the extracted feature vector, so that the classification performance of the radar true and false targets is effectively improved.
To verify the effectiveness of the proposed method, the following simulation experiments were performed.
Four point targets are set: true objects, debris, light baits, and heavy baits. The bandwidth of radar emission pulse is 1000MHZ (the range resolution is 0.15m, the radar radial sampling interval is 0.075m), the target is set as a uniform scattering point target, the scattering point of a true target is 7, and the number of the scattering points of the other three targets is 11. In the one-dimensional distance images of every 1 degree within the range of the target attitude angle of 0-80 degrees, the one-dimensional distance images of the target attitude angle of 0 degree, 2 degrees, 4 degrees, 6 degrees, and 90 degrees are taken for training, and the one-dimensional distance images of the rest attitude angles are taken as test data, so that each category of targets has 45 test samples.
For four targets (true target, fragment, light bait and heavy bait), in the range of 0-90 degrees of attitude angle, the K nearest neighbor transformation feature extraction method and the discrimination vector transformation-based feature extraction method are utilized to carry out recognition experiment, and the result is shown in the table I. Experimental nearest neighbor parameter k1=20,k210, gaussian parameter σ2=6.25。
From the results in table one, it can be seen that for the true target, the recognition rate of the discriminant vector transform feature extraction method is 83%, while the recognition rate of the K neighbor transform feature extraction method of the present invention is 90%; for the fragments, the recognition rate of the vector transformation feature extraction method is judged to be 78%, and the recognition rate of the K neighbor transformation feature extraction method is 85%; for light baits, the recognition rate of the discrimination vector transformation feature extraction method is 80 percent, while the recognition rate of the K neighbor transformation feature extraction method is 86 percent; the discrimination vector transform feature extraction method has a recognition rate of 82% for heavy baits, and the K-nearest neighbor transform feature extraction method of the present invention has a recognition rate of 83%. On average, for four types of targets, the correct recognition rate of the K neighbor transformation feature extraction method is higher than that of a discrimination vector transformation feature extraction method, and the K neighbor transformation feature extraction method provided by the invention is proved to improve the recognition performance of multiple types of targets.
Table one two methods of identification results
While the invention has been described with reference to specific embodiments, any feature disclosed in this specification may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise; all of the disclosed features, or all of the method or process steps, may be combined in any combination, except mutually exclusive features and/or steps.
Claims (4)
1. A K nearest neighbor transformation true and false target feature extraction method is characterized by comprising the following steps:
step 1: inputting a training sample set of one-dimensional range profile of radar target by xijRepresenting training samples, wherein subscript i is a class identifier, subscript j is a training sample identifier, i is more than or equal to 1 and less than or equal to g, j is more than or equal to 1 and less than or equal to NiG denotes the number of categories, NiA number of samples representing a corresponding category;
Constraint coefficient of similar K nearest neighbor ruleIs set as follows: if it isOrThenOtherwiseWhere the subscript k is the training sample specifier, σ2The value of the gaussian parameter is represented,k representing a vector of the same kind1Set of neighbor vectors, k1Is a preset neighbor number;
Constraint coefficient of heterogeneous K neighbor ruleIs set as follows: if it isOrl is not equal to i, thenOtherwiseWherein the subscript l is a category specifier,k representing a vector of a different class2Set of neighbor vectors, k2Is a preset neighbor number;
3. Method according to claim 1 or 2, characterized in that the number of neighbors k1、k2Is preferably taken to be k1=20,k2=10。
4. Method according to claim 1 or 2, characterized in that the preferred value of the gaussian parameter is σ2=6.25。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810964240.1A CN108845302B (en) | 2018-08-23 | 2018-08-23 | K-nearest neighbor transformation true and false target feature extraction method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810964240.1A CN108845302B (en) | 2018-08-23 | 2018-08-23 | K-nearest neighbor transformation true and false target feature extraction method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108845302A CN108845302A (en) | 2018-11-20 |
CN108845302B true CN108845302B (en) | 2022-06-03 |
Family
ID=64188337
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810964240.1A Active CN108845302B (en) | 2018-08-23 | 2018-08-23 | K-nearest neighbor transformation true and false target feature extraction method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108845302B (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110007286B (en) * | 2019-04-22 | 2022-05-24 | 电子科技大学 | Linear discriminant learning true and false target one-dimensional range profile feature extraction method |
CN110658507B (en) * | 2019-10-12 | 2022-07-29 | 电子科技大学 | Multi-class average maximization true and false target feature extraction method for radar target identification |
CN110687514B (en) * | 2019-10-16 | 2022-03-15 | 电子科技大学 | Nonlinear discrimination learning true and false target one-dimensional range profile feature extraction method |
CN112149061A (en) * | 2020-09-25 | 2020-12-29 | 电子科技大学 | Multi-class average maximization true and false target feature extraction method |
CN113191447B (en) * | 2021-05-17 | 2022-05-24 | 电子科技大学 | Method for extracting characteristics of sample distribution structure chart in unknown target discrimination |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2001014907A1 (en) * | 1999-08-26 | 2001-03-01 | Raytheon Company | Target acquisition system and radon transform based method for target azimuth aspect estimation |
US7218270B1 (en) * | 2003-02-10 | 2007-05-15 | The United States Of America As Represented By The Secretary Of The Air Force | ATR trajectory tracking system (A-Track) |
WO2008094172A2 (en) * | 2006-06-01 | 2008-08-07 | University Of Florida Research Foundation, Inc. | Radar microsensor for detection, tracking, and classification |
CN101561865A (en) * | 2009-05-22 | 2009-10-21 | 西安电子科技大学 | Synthetic aperture radar image target identification method based on multi-parameter spectrum feature |
CN103679162A (en) * | 2014-01-03 | 2014-03-26 | 苏州大学 | Human-face identifying method and system |
CN103679161A (en) * | 2014-01-03 | 2014-03-26 | 苏州大学 | Human-face identifying method and device |
CN105608471A (en) * | 2015-12-28 | 2016-05-25 | 苏州大学 | Robust transductive label estimation and data classification method and system |
CN106772307A (en) * | 2017-03-02 | 2017-05-31 | 电子科技大学 | A kind of true and false bullet target identification method based on many radar informations |
CN107238822A (en) * | 2017-06-13 | 2017-10-10 | 电子科技大学 | True and false target one-dimensional range profile Nonlinear Orthogonal subspace representation method |
CN107678007A (en) * | 2017-09-06 | 2018-02-09 | 电子科技大学 | A kind of radar true and false target one-dimensional range profile feature extracting method of the close subspace of pointer field |
CN107678006A (en) * | 2017-09-06 | 2018-02-09 | 电子科技大学 | A kind of true and false target one-dimensional range profile feature extracting method of the radar of largest interval subspace |
Family Cites Families (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
NL302562A (en) * | 1963-12-24 | |||
US6035057A (en) * | 1997-03-10 | 2000-03-07 | Hoffman; Efrem H. | Hierarchical data matrix pattern recognition and identification system |
WO2006087854A1 (en) * | 2004-11-25 | 2006-08-24 | Sharp Kabushiki Kaisha | Information classifying device, information classifying method, information classifying program, information classifying system |
CN101526995B (en) * | 2009-01-19 | 2011-06-29 | 西安电子科技大学 | Synthetic aperture radar target identification method based on diagonal subclass judgment analysis |
CN101807258B (en) * | 2010-01-08 | 2012-05-23 | 西安电子科技大学 | SAR (Synthetic Aperture Radar) image target recognizing method based on nuclear scale tangent dimensionality reduction |
CN102902979B (en) * | 2012-09-13 | 2015-08-19 | 电子科技大学 | A kind of method of synthetic-aperture radar automatic target detection |
CN103675787A (en) * | 2013-12-03 | 2014-03-26 | 电子科技大学 | One-dimension range profile optimal orthogonal nolinear subspace identification method for radar targets |
US10186123B2 (en) * | 2014-04-01 | 2019-01-22 | Avigilon Fortress Corporation | Complex event recognition in a sensor network |
CN103941244B (en) * | 2014-04-23 | 2016-12-07 | 电子科技大学 | A kind of radar target-range image local optimum subspace identification method |
CN103942572A (en) * | 2014-05-07 | 2014-07-23 | 中国标准化研究院 | Method and device for extracting facial expression features based on bidirectional compressed data space dimension reduction |
US10192117B2 (en) * | 2015-06-25 | 2019-01-29 | Kodak Alaris Inc. | Graph-based framework for video object segmentation and extraction in feature space |
CN105023006B (en) * | 2015-08-05 | 2018-05-04 | 西安电子科技大学 | Face identification method based on enhanced nonparametric maximal margin criterion |
CN106199544B (en) * | 2016-06-24 | 2018-07-17 | 电子科技大学 | Differentiate the Recognition of Radar Target Using Range Profiles method of local tangent space alignment based on core |
CN106257488B (en) * | 2016-07-07 | 2019-11-19 | 电子科技大学 | A kind of radar target identification method based on neighborhood characteristics space discriminatory analysis |
CN107037417B (en) * | 2017-06-13 | 2019-08-23 | 电子科技大学 | The true and false target of radar is one-dimensional as non-linear arest neighbors subspace representation method |
CN107271965B (en) * | 2017-06-13 | 2020-02-04 | 电子科技大学 | Method for extracting true and false target one-dimensional range profile features in cluster subspace |
CN107784293B (en) * | 2017-11-13 | 2018-08-28 | 中国矿业大学(北京) | A kind of Human bodys' response method classified based on global characteristics and rarefaction representation |
-
2018
- 2018-08-23 CN CN201810964240.1A patent/CN108845302B/en active Active
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2001014907A1 (en) * | 1999-08-26 | 2001-03-01 | Raytheon Company | Target acquisition system and radon transform based method for target azimuth aspect estimation |
US7218270B1 (en) * | 2003-02-10 | 2007-05-15 | The United States Of America As Represented By The Secretary Of The Air Force | ATR trajectory tracking system (A-Track) |
WO2008094172A2 (en) * | 2006-06-01 | 2008-08-07 | University Of Florida Research Foundation, Inc. | Radar microsensor for detection, tracking, and classification |
CN101561865A (en) * | 2009-05-22 | 2009-10-21 | 西安电子科技大学 | Synthetic aperture radar image target identification method based on multi-parameter spectrum feature |
CN103679162A (en) * | 2014-01-03 | 2014-03-26 | 苏州大学 | Human-face identifying method and system |
CN103679161A (en) * | 2014-01-03 | 2014-03-26 | 苏州大学 | Human-face identifying method and device |
CN105608471A (en) * | 2015-12-28 | 2016-05-25 | 苏州大学 | Robust transductive label estimation and data classification method and system |
CN106772307A (en) * | 2017-03-02 | 2017-05-31 | 电子科技大学 | A kind of true and false bullet target identification method based on many radar informations |
CN107238822A (en) * | 2017-06-13 | 2017-10-10 | 电子科技大学 | True and false target one-dimensional range profile Nonlinear Orthogonal subspace representation method |
CN107678007A (en) * | 2017-09-06 | 2018-02-09 | 电子科技大学 | A kind of radar true and false target one-dimensional range profile feature extracting method of the close subspace of pointer field |
CN107678006A (en) * | 2017-09-06 | 2018-02-09 | 电子科技大学 | A kind of true and false target one-dimensional range profile feature extracting method of the radar of largest interval subspace |
Non-Patent Citations (5)
Title |
---|
Automatic Modulation Recognition of Radar Signals Based on Manhattan Distance-Based Features;Y. Huang, W. Jin, B. Li, P. Ge and Y. Wu;《IEEE 》;20190325;第41193-41204页 * |
Du, C ; Zhou, S ; Zhao, J.Feature extraction for SAR target recognition based on supervised manifold learning.《35th International Symposium on Remote Sensing of Environment (ISRSE35)》.2014,第1-7页. * |
Meng, JC (Meng, JC) ; Yang, WL (Yang, WL).Nearest neighbor classifier based on Riemannian metric in radar target recognition.《2005 IEEE International Radar, Conference Record》.2005,第851-853页. * |
改进的ReliefF算法用于雷达距离像目标识别;廖阔,付建胜,杨万麟;《电子测量与仪器学报》;20100915;第24卷(第9期);第831-836页 * |
雷达目标距离像的识别方法研究;戴春杨;《中国优秀硕士学位论文全文数据库》;20081114;第1-80页 * |
Also Published As
Publication number | Publication date |
---|---|
CN108845302A (en) | 2018-11-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108845302B (en) | K-nearest neighbor transformation true and false target feature extraction method | |
CN108761411B (en) | True and false target one-dimensional range profile feature extraction method | |
CN110007286B (en) | Linear discriminant learning true and false target one-dimensional range profile feature extraction method | |
CN107238822B (en) | Method for extracting orthogonal nonlinear subspace characteristics of true and false target one-dimensional range profile | |
CN107271965B (en) | Method for extracting true and false target one-dimensional range profile features in cluster subspace | |
CN109064502B (en) | Multi-source image registration method based on combination of deep learning and artificial design features | |
CN108845303B (en) | Nonlinear robust subspace true and false target feature extraction method | |
CN110658507B (en) | Multi-class average maximization true and false target feature extraction method for radar target identification | |
CN108073895B (en) | Hyperspectral target detection method based on unmixing pretreatment | |
CN107037417B (en) | The true and false target of radar is one-dimensional as non-linear arest neighbors subspace representation method | |
CN110687514B (en) | Nonlinear discrimination learning true and false target one-dimensional range profile feature extraction method | |
CN110516525B (en) | SAR image target recognition method based on GAN and SVM | |
CN108594202B (en) | Neighborhood distribution structure nonlinear projection target feature extraction method | |
CN108828574B (en) | Inter-class separation enhancer space true and false target feature extraction method | |
CN114241233B (en) | Nonlinear class group sparse representation real and false target one-dimensional range profile identification method | |
CN114578293A (en) | Electric scanning radar signal identification method by using intercepted signal amplitude value | |
CN107678007B (en) | Method for extracting radar true and false target one-dimensional range profile features in exponential domain compact subspace | |
CN110068799B (en) | Sparse neighborhood center preserving RCS sequence feature extraction method | |
CN114037650A (en) | Ground target visible light damage image processing method for change detection and target detection | |
CN109948520B (en) | Crop classification method based on multi-temporal dual-polarization SAR characteristic curve | |
CN112418250A (en) | Optimized matching method for complex 3D point cloud | |
CN109344681B (en) | Geological target recognition method based on face recognition | |
CN108428226B (en) | Distortion image quality evaluation method based on ICA sparse representation and SOM | |
CN114280566B (en) | One-dimensional range profile identification method based on class label association | |
CN116309366A (en) | Early-stage determination method and device for nematode disease of pine tree |
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 |