[go: up one dir, main page]

CN108845302B - A feature extraction method of true and false target by K-nearest neighbor transform - Google Patents

A feature extraction method of true and false target by K-nearest neighbor transform Download PDF

Info

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
nearest neighbor
matrix
true
feature extraction
subscript
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.)
Expired - Fee Related
Application number
CN201810964240.1A
Other languages
Chinese (zh)
Other versions
CN108845302A (en
Inventor
周代英
沈晓峰
冯健
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
University of Electronic Science and Technology of China
Original Assignee
University of Electronic Science and Technology of China
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by University of Electronic Science and Technology of China filed Critical University of Electronic Science and Technology of China
Priority to CN201810964240.1A priority Critical patent/CN108845302B/en
Publication of CN108845302A publication Critical patent/CN108845302A/en
Application granted granted Critical
Publication of CN108845302B publication Critical patent/CN108845302B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

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)
  • Image Analysis (AREA)
  • Radar Systems Or Details Thereof (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

一种K近邻变换真假目标特征提取方法A feature extraction method of true and false target by K-nearest neighbor transform

技术领域technical field

本发明属于雷达目标识别技术邻域，具体涉及一种K近邻变换真假目标特征提取方法。The invention belongs to the technical neighborhood of radar target recognition, and in particular relates to a method for extracting true and false target features by K-nearest neighbor transformation.

背景技术Background technique

在雷达目标识别中，判别矢量变换法能够增大异类目标特征之间的差异，同时减小同类目标特征之间的差异，从而提取到差异明显的特征，因此，判别矢量变换法获得了良好的分类性能。In radar target recognition, the discriminant vector transformation method can increase the difference between the characteristics of heterogeneous targets and reduce the difference between the characteristics of similar targets, so as to extract features with obvious differences. Therefore, the discriminant vector transformation method has achieved good results. Classification performance.

但是，判别矢量变换法只适合于样本数据是高斯分布的情况，而实际中样本数据的分布可能是非高斯分布，针对非高斯分布情况，判别矢量变换法的识别性能显著降低。现有常规判别矢量变换法的识别性能有进一步改善的余地。However, the discriminant vector transformation method is only suitable for the case where the sample data is Gaussian distribution, and the distribution of the sample data may be non-Gaussian distribution in practice. For the non-Gaussian distribution, the recognition performance of the discriminant vector transformation method is significantly reduced. The recognition performance of the existing conventional discriminant vector transformation methods has room for further improvement.

发明内容SUMMARY OF THE INVENTION

本发明的发明目的在于：针对上述存在的问题，提出一种K近邻变换特征提取方法，以克服常规判别矢量变换法的缺陷，有效改善了对雷达真假目标的分类性能。The purpose of the present invention is to provide a feature extraction method of K-nearest neighbor transform to overcome the defects of the conventional discriminant vector transform method and effectively improve the classification performance of radar true and false targets.

本发明的K近邻变换真假目标特征提取方法的技术方案具体如下：The technical scheme of the K-nearest neighbor transformation true and false target feature extraction method of the present invention is specifically as follows:

步骤1：输入关于雷达目标一维距离像的训练样本集，用xij表示训练样本，其中下标i为类别区分符、下标j为训练样本区分符，且1≤i≤g，1≤j≤Ni，g表示类别数量，Ni表示对应类别的样本数；Step 1: Input the training sample set about the one-dimensional range image of the radar target, and use x ij to represent the training sample, where the subscript i is the category specifier, and the subscript j is the training sample specifier, and 1≤i≤g, 1≤ j≤N i , g represents the number of categories, and N i represents the number of samples of the corresponding category;

步骤2：计算K近邻变换矩阵A的估计值

Figure BDA0001774542280000011
Figure BDA0001774542280000012
Step 2: Calculate the estimated value of the K-nearest neighbor transformation matrix A
Figure BDA0001774542280000011
Figure BDA0001774542280000012

其中，样本矩阵

Figure BDA0001774542280000013
Among them, the sample matrix
Figure BDA0001774542280000013

同类K近邻规则的约束系数矩阵

Figure BDA0001774542280000014
Constraint coefficient matrix for homogeneous K-nearest neighbor rules
Figure BDA0001774542280000014

矩阵

Figure BDA0001774542280000015
matrix
Figure BDA0001774542280000015

其中同类K近邻规则的约束系数

Figure BDA0001774542280000021
的设置为：若
Figure BDA0001774542280000022
或者
Figure BDA0001774542280000023
则
Figure BDA0001774542280000024
否则
Figure BDA0001774542280000025
其中下标k为某类训练样本区分符，σ2表示高斯参数，
Figure BDA0001774542280000026
表示同类中某个矢量的k1个近邻矢量的集合，k1为预设近邻数；Among them, the constraint coefficient of the same K-nearest neighbor rule
Figure BDA0001774542280000021
is set to: if
Figure BDA0001774542280000022
or
Figure BDA0001774542280000023
but
Figure BDA0001774542280000024
otherwise
Figure BDA0001774542280000025
The subscript k is a certain type of training sample discriminator, σ 2 represents the Gaussian parameter,
Figure BDA0001774542280000026
Represents the set of k 1 neighbor vectors of a vector in the same class, where k 1 is the preset number of neighbors;

异类K近邻规则的约束系数矩阵

Figure BDA0001774542280000027
Constraint Coefficient Matrix of Heterogeneous K-Nearest Neighbor Rules
Figure BDA0001774542280000027

矩阵

Figure BDA0001774542280000028
matrix
Figure BDA0001774542280000028

其中异类K近邻规则的约束系数

Figure BDA0001774542280000029
的设置为：若
Figure BDA00017745422800000210
或者
Figure BDA00017745422800000211
则
Figure BDA00017745422800000212
否则
Figure BDA00017745422800000213
其中，其中下标l为类别区分符，
Figure BDA00017745422800000214
表示异类中某个矢量的k2个近邻矢量的集合，k2为预设近邻数；where the constraint coefficient of the heterogeneous K-nearest neighbor rule
Figure BDA0001774542280000029
is set to: if
Figure BDA00017745422800000210
or
Figure BDA00017745422800000211
but
Figure BDA00017745422800000212
otherwise
Figure BDA00017745422800000213
Among them, the subscript l is the category specifier,
Figure BDA00017745422800000214
Represents the set of k 2 nearest neighbor vectors of a vector in heterogeneous, k 2 is the preset number of neighbors;

步骤3：输入待提取子像特征的雷达真假目标一维距离像xt，根据

Figure BDA00017745422800000215
得到一维距离像xt的特征矢量yt。Step 3: Input the one-dimensional distance image x t of the radar true and false targets of the sub-image features to be extracted, according to
Figure BDA00017745422800000215
Get a eigenvector y t with a one-dimensional distance like x t .

综上所述，由于采用了上述技术方案，本发明的有益效果是：To sum up, due to the adoption of the above technical solutions, the beneficial effects of the present invention are:

本发明通过基于K近邻约束规则减小同类样本特征之间的差异，而加大异类样本特征之间的差异进行加权，降低其它样本对变换矩阵构建的影响，从而能够提取非高斯分布样本数据的特征，克服常规判别矢量变换法的缺陷，有效改善了对雷达真假目标的分类性能。The present invention reduces the difference between the features of the same samples based on the K-nearest neighbor constraint rule, and increases the difference between the features of the heterogeneous samples for weighting, so as to reduce the influence of other samples on the construction of the transformation matrix, so that the non-Gaussian distribution sample data can be extracted. It overcomes the shortcomings of the conventional discriminant vector transformation method and effectively improves the classification performance of radar true and false targets.

具体实施方式Detailed ways

为使本发明的目的、技术方案和优点更加清楚，下面结合实施方式，对本发明作进一步地详细描述。In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments.

本发明的K近邻变换真假目标特征提取方法，通过基于K近邻约束规则减小同类样本特征之间的差异，而加大异类样本特征之间的差异进行加权，降低其它样本对变换矩阵构建的影响，从而能够提取非高斯分布样本数据的特征，其具体实现过程如下：The method for extracting the true and false target features of the K-nearest neighbor transformation of the present invention reduces the difference between the features of the same samples based on the constraint rule of the K-nearest neighbor, and increases the difference between the features of the heterogeneous samples for weighting, thereby reducing the influence of other samples on the construction of the transformation matrix. Therefore, the characteristics of non-Gaussian distributed sample data can be extracted. The specific implementation process is as follows:

用xij(n维列矢量)表示第ith类真假目标的第jth个训练一维距离像，1≤i≤g，1≤j≤Ni，

Figure BDA0001774542280000031
其中Ni为第ith类真假目标的训练一维距离像样本数，N为训练一维距离像样本总数。将训练一维距离像xij进行如下变换：Use x ij (n-dimensional column vector) to represent the j th training one-dimensional distance image of the i th class of true and false targets, 1≤i≤g, 1≤j≤N i ,
Figure BDA0001774542280000031
Among them, N i is the number of training one-dimensional distance image samples of the ith class of true and false targets, and N is the total number of training one-dimensional distance image samples. Transform the training one-dimensional distance image x ij as follows:

yij＝ATxij (1)y ij =A T x ij (1)

其中A为变换矩阵，yij为xij对应的特征矢量，T表示矩阵转置。where A is the transformation matrix, y ij is the eigenvector corresponding to x ij , and T is the matrix transpose.

在特征空间计算同类任意两个样本特征矢量之间的差值平方和SC：Calculate the sum of squared differences S C between any two sample feature vectors of the same class in the feature space:

Figure BDA0001774542280000032
Figure BDA0001774542280000032

其中

Figure BDA0001774542280000033
为同类K近邻规则的约束系数：in
Figure BDA0001774542280000033
is the constraint coefficient of the same K-nearest neighbor rule:

Figure BDA0001774542280000034
Figure BDA0001774542280000034

其中高斯参数σ2为经验值，在满足处理精度需求的条件下，经过实验确定，

Figure BDA0001774542280000035
表示同类中某个矢量的k1个近邻矢量的集合。式(3)表明，当属于同一目标类的两个样本互为k1近邻子时，其同类样本间的差值约束系数不等于零，而其它同类样本间差值的约束系数为零。Among them, the Gaussian parameter σ 2 is an empirical value, which is determined through experiments under the condition that the processing accuracy requirements are met.
Figure BDA0001774542280000035
Represents the set of k 1 nearest neighbors of a vector in the same class. Equation (3) shows that when two samples belonging to the same target class are k 1 nearest neighbors to each other, the difference constraint coefficient between the samples of the same class is not equal to zero, and the constraint coefficient of the difference between other samples of the same class is zero.

利用矩阵迹的运算公式，式(2)可转换为：Using the operation formula of matrix trace, equation (2) can be converted into:

Figure BDA0001774542280000036
Figure BDA0001774542280000036

将式(1)代入式(4)可得：Substitute equation (1) into equation (4) to get:

Figure BDA0001774542280000037
Figure BDA0001774542280000037

对式(5)化简，可得：Simplifying equation (5), we can get:

SC＝ATX(DC-WC)XTA (6)S C =A T X(D C -W C )X T A (6)

其中in

Figure BDA0001774542280000041
Figure BDA0001774542280000041

Figure BDA0001774542280000042
Figure BDA0001774542280000042

Figure BDA0001774542280000043
Figure BDA0001774542280000043

同理，在特征空间计算异类目标样本特征间的加权距离平方和SB：In the same way, calculate the weighted sum of squared distances S B between the features of heterogeneous target samples in the feature space:

Figure BDA0001774542280000044
Figure BDA0001774542280000044

其中wij,lk为基于异类K近邻规则的约束系数：where w ij,lk are the constraint coefficients based on the heterogeneous K-nearest neighbor rule:

Figure BDA0001774542280000045
Figure BDA0001774542280000045

其中

Figure BDA0001774542280000046
表示异类中某个矢量的k2个近邻矢量的集合。in
Figure BDA0001774542280000046
Represents the set of k 2 nearest neighbors of a vector in the heterogeneous.

利用矩阵迹的运算公式，式(10)可转换为Using the formula of matrix trace, Equation (10) can be converted into

Figure BDA0001774542280000047
Figure BDA0001774542280000047

将式(1)代入式(12)Substitute equation (1) into equation (12)

Figure BDA0001774542280000048
Figure BDA0001774542280000048

对式(13)化简，可得Simplifying equation (13), we can get

SB＝ATX(DB-WB)XTA (14)S B =A T X(D B -W B )X T A (14)

其中in

Figure BDA0001774542280000051
Figure BDA0001774542280000051

Figure BDA0001774542280000052
Figure BDA0001774542280000052

通过求解以下极小化问题，即可获得K近邻变换矩阵的估计值

Figure BDA0001774542280000053
An estimate of the K-nearest neighbor transformation matrix can be obtained by solving the following minimization problem
Figure BDA0001774542280000053

Figure BDA0001774542280000054
Figure BDA0001774542280000054

式(17)右边对A求偏导并令其等于零，可得K近邻变换矩阵的估计值

Figure BDA0001774542280000055
是由矩阵(X(DB-WB)XT)-1(X(DC-WC)XT)的M(＜n)个最大特征值对应的特征向量组成的矩阵。Taking the partial derivative of A on the right side of equation (17) and making it equal to zero, the estimated value of the K-nearest neighbor transformation matrix can be obtained
Figure BDA0001774542280000055
is a matrix composed of eigenvectors corresponding to the M (<n) largest eigenvalues of the matrix (X(D B -W B )X T ) -1 (X(D C -W C )X T ).

在获得K近邻变换矩阵的估计值

Figure BDA0001774542280000056
后，利用式(1)即可得到任意样本xt的特征矢量yt，即
Figure BDA0001774542280000057
再基于所提取的特征矢量进行雷达真假目标识别处理，从而有效改善对雷达真假目标的分类性能。After obtaining the estimated value of the K-nearest neighbor transformation matrix
Figure BDA0001774542280000056
After that, the feature vector y t of any sample x t can be obtained by using formula (1), that is,
Figure BDA0001774542280000057
Then based on the extracted feature vector, the radar real and false target recognition processing is carried out, so as to effectively improve the classification performance of the radar real and false targets.

为了验证所提方法的有效性，进行如下仿真实验。In order to verify the effectiveness of the proposed method, the following simulation experiments are carried out.

设置四种点目标：真目标、碎片、轻诱饵和重诱饵。雷达发射脉冲的带宽为1000MHZ(距离分辨率为0.15m,雷达径向取样间隔为0.075m),目标设置为均匀散射点目标，真目标的散射点为7，其余三目标的散射点数均为11。在目标姿态角为0°～80°范围内每隔1°的一维距离像中，取目标姿态角为0°、2°、4°、6°、...、90°的一维距离像进行训练,其余姿态角的一维距离像作为测试数据,则每类目标有45个测试样本。Four types of point targets are set: True Target, Fragment, Light Decoy, and Heavy Decoy. The bandwidth of the radar transmit pulse is 1000MHZ (range resolution is 0.15m, and the radar radial sampling interval is 0.075m). The target is set as a uniform scattering point target. The scattering point of the real target is 7, and the scattering points of the other three targets are all 11. . In the one-dimensional range image every 1° in the range of the target attitude angle of 0° to 80°, take the one-dimensional distance of the target attitude angle of 0°, 2°, 4°, 6°, ..., 90° The one-dimensional distance images of the remaining attitude angles are used as test data, and there are 45 test samples for each type of target.

对四种目标(真目标、碎片、轻诱饵和重诱饵)，在姿态角0°～90°范围内，利用本发明的K近邻变换特征提取方法和基于判别矢量变换特征提取方法进行了识别实验，结果如表一所示。实验近邻参数k1＝20，k2＝10，高斯参数σ2＝6.25。For four kinds of targets (true target, debris, light decoy and heavy decoy), in the range of attitude angle 0°～90°, the recognition experiment was carried out using the K-nearest neighbor transformation feature extraction method and the discriminant vector transformation feature extraction method of the present invention. , the results are shown in Table 1. Experimental neighbor parameters k 1 =20, k 2 =10, Gaussian parameter σ 2 =6.25.

从表一的结果可以看到，对真目标，判别矢量变换特征提取法的识别率为83％，而本发明的K近邻变换识特征提取方法的识别率为90％；对碎片，判别矢量变换特征提取法的识别率为78％，而本发明的K近邻变换特征提取方法的识别率为85％；对轻诱饵，判别矢量变换特征提取法的识别率为80％，而本发明的K近邻变换特征提取方法的识别率为86％；对重诱饵，判别矢量变换特征提取法的识别率为82％，而本发明的K近邻变换特征提取方法的识别率为83％。平均而言，对四类目标，本发明的K近邻变换特征提取方法的正确识别率高于判别矢量变换特征提取法，表明本发明的K近邻变换特征提取方法确实改善了多类目标的识别性能。It can be seen from the results in Table 1 that for the real target, the recognition rate of the discriminant vector transformation feature extraction method is 83%, while the recognition rate of the K-nearest neighbor transformation recognition feature extraction method of the present invention is 90%; The recognition rate of the feature extraction method is 78%, while the recognition rate of the K-nearest neighbor transformation feature extraction method of the present invention is 85%. The recognition rate of the transformation feature extraction method is 86%; for heavy bait, the recognition rate of the discriminant vector transformation feature extraction method is 82%, and the recognition rate of the K-nearest neighbor transformation feature extraction method of the present invention is 83%. On average, for four types of targets, the correct recognition rate of the K-nearest neighbor transformation feature extraction method of the present invention is higher than that of the discriminant vector transformation feature extraction method, indicating that the K-nearest neighbor transformation feature extraction method of the present invention has indeed improved the recognition performance of multi-class targets. .

表一 两种方法的识别结果Table 1 Identification results of the two methods

Figure BDA0001774542280000061
Figure BDA0001774542280000061

以上所述，仅为本发明的具体实施方式，本说明书中所公开的任一特征，除非特别叙述，均可被其他等效或具有类似目的的替代特征加以替换；所公开的所有特征、或所有方法或过程中的步骤，除了互相排斥的特征和/或步骤以外，均可以任何方式组合。The above descriptions are only specific embodiments of the present invention, and any feature disclosed in this specification, unless otherwise stated, can be replaced by other equivalent or alternative features with similar purposes; all the disclosed features, or All steps in a method or process, except mutually exclusive features and/or steps, may be combined in any way.

Claims (4)

1.一种K近邻变换真假目标特征提取方法，其特征在于，包括下列步骤：1. a K-nearest neighbor transform true and false target feature extraction method, is characterized in that, comprises the following steps: 步骤1：输入关于雷达目标一维距离像的训练样本集，用xij表示训练样本，其中下标i为类别区分符、下标j为训练样本区分符，且1≤i≤g，1≤j≤Ni，g表示类别数量，Ni表示对应类别的样本数；Step 1: Input the training sample set about the one-dimensional range image of the radar target, and use x ij to represent the training sample, where the subscript i is the category specifier, and the subscript j is the training sample specifier, and 1≤i≤g, 1≤ j≤N i , g represents the number of categories, and N i represents the number of samples of the corresponding category; 步骤2：计算K近邻变换矩阵A的估计值
Figure FDA0001774542270000011
Figure FDA0001774542270000012
Step 2: Calculate the estimated value of the K-nearest neighbor transformation matrix A
Figure FDA0001774542270000011
Figure FDA0001774542270000012
 其中，样本矩阵
Figure FDA0001774542270000013
Among them, the sample matrix
Figure FDA0001774542270000013
 同类K近邻规则的约束系数矩阵
Figure FDA0001774542270000014
Constraint coefficient matrix for homogeneous K-nearest neighbor rules
Figure FDA0001774542270000014
 矩阵
Figure FDA0001774542270000015
matrix
Figure FDA0001774542270000015
 同类K近邻规则的约束系数
Figure FDA0001774542270000016
的设置为：若
Figure FDA0001774542270000017
或者
Figure FDA0001774542270000018
则
Figure FDA0001774542270000019
否则
Figure FDA00017745422700000110
其中下标k为训练样本区分符，σ2表示高斯参数，
Figure FDA00017745422700000111
表示同类中某个矢量的k1个近邻矢量的集合，k1为预设近邻数；Constraint coefficients of the same class of K-nearest neighbor rules
Figure FDA0001774542270000016
is set to: if
Figure FDA0001774542270000017
or
Figure FDA0001774542270000018
but
Figure FDA0001774542270000019
otherwise
Figure FDA00017745422700000110
where the subscript k is the training sample discriminator, σ 2 represents the Gaussian parameter,
Figure FDA00017745422700000111
Represents the set of k 1 neighbor vectors of a vector in the same class, where k 1 is the preset number of neighbors; 异类K近邻规则的约束系数矩阵
Figure FDA00017745422700000112
Constraint Coefficient Matrix of Heterogeneous K-Nearest Neighbor Rules
Figure FDA00017745422700000112
 矩阵
Figure FDA00017745422700000113
matrix
Figure FDA00017745422700000113
 其中异类K近邻规则的约束系数
Figure FDA0001774542270000021
的设置为：若
Figure FDA0001774542270000022
或者
Figure FDA0001774542270000023
l≠i，则
Figure FDA0001774542270000024
否则
Figure FDA0001774542270000025
其中，其中下标l为类别区分符，
Figure FDA0001774542270000026
表示异类中某个矢量的k2个近邻矢量的集合，k2为预设近邻数；where the constraint coefficient of the heterogeneous K-nearest neighbor rule
Figure FDA0001774542270000021
is set to: if
Figure FDA0001774542270000022
or
Figure FDA0001774542270000023
l≠i, then
Figure FDA0001774542270000024
otherwise
Figure FDA0001774542270000025
Among them, the subscript l is the category specifier,
Figure FDA0001774542270000026
Represents the set of k 2 nearest neighbor vectors of a vector in heterogeneous, k 2 is the preset number of neighbors; 步骤3：输入待提取子像特征的雷达真假目标一维距离像xt，根据
Figure FDA0001774542270000027
得到一维距离像xt的特征矢量yt。Step 3: Input the one-dimensional distance image x t of the radar true and false targets of the sub-image features to be extracted, according to
Figure FDA0001774542270000027
Get a eigenvector y t with a one-dimensional distance like x t . 2.如权利要求1所述的方法，其特征在于，步骤2中，2. method as claimed in claim 1, is characterized in that, in step 2, 估计值
Figure FDA0001774542270000028
为矩阵(X(DB-WB)XT)-1(X(DC-WC)XT)的M个最大特征值对应的特征向量组成的矩阵，其中M＜n，n表示训练样本xij的维度数。estimated value
Figure FDA0001774542270000028
is a matrix composed of eigenvectors corresponding to the M largest eigenvalues of the matrix (X(D B -W B )X T ) -1 (X(D C -W C )X T ), where M<n, n represents training The number of dimensions of the sample x ij . 3.如权利要求1或2所述的方法，其特征在于，近邻数k1、k2的优选取值为k1＝20，k2＝10。3 . The method according to claim 1 or 2 , wherein the preferred values of the number of neighbors k 1 and k 2 are k 1 =20 and k 2 =10. 4 . 4.如权利要求1或2所述的方法，其特征在于，高斯参数的优选值为σ2＝6.25。4. The method according to claim 1 or 2, wherein the preferred value of the Gaussian parameter is σ 2 =6.25.
CN201810964240.1A 2018-08-23 2018-08-23 A feature extraction method of true and false target by K-nearest neighbor transform Expired - Fee Related CN108845302B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810964240.1A CN108845302B (en) 2018-08-23 2018-08-23 A feature extraction method of true and false target by K-nearest neighbor transform

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810964240.1A CN108845302B (en) 2018-08-23 2018-08-23 A feature extraction method of true and false target by K-nearest neighbor transform

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 Expired - Fee Related CN108845302B (en) 2018-08-23 2018-08-23 A feature extraction method of true and false target by K-nearest neighbor transform

Country Status (1)

Country Link
CN (1) CN108845302B (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110007286B (en) * 2019-04-22 2022-05-24 电子科技大学 A Linear Discriminant Learning Method for One-Dimensional Distance Profile Feature Extraction of True and False Targets
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 A Feature Extraction Method of Sample Distribution Structure Map in Unknown Target Discrimination

Citations (11)

* Cited by examiner, † Cited by third party
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
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
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

Family Cites Families (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
NL302562A (en) * 1963-12-24
CA2199588A1 (en) * 1997-03-10 1998-09-10 Efrem Hoffman Hierarchical data matrix pattern recognition and identification system
US7693683B2 (en) * 2004-11-25 2010-04-06 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 Recognition Method Based on Diagonal Subclass Decision 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 Radar Target Recognition Method Based on Neighborhood Feature Space Discriminant 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

Patent Citations (11)

* Cited by examiner, † Cited by third party
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
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
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

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
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) A feature extraction method of true and false target by K-nearest neighbor transform
CN110007286B (en) A Linear Discriminant Learning Method for One-Dimensional Distance Profile Feature Extraction of True and False Targets
CN108761411B (en) 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
CN107037417B (en) The true and false target of radar is one-dimensional as non-linear arest neighbors subspace representation method
CN107977642B (en) A High-Resolution Range Image Target Recognition Method Based on Kernel Adaptive Mean Discriminant Analysis
CN108073895B (en) A hyperspectral target detection method based on unmixing preprocessing
CN107678006A (en) A kind of true and false target one-dimensional range profile feature extracting method of the radar of largest interval subspace
CN110658507B (en) Multi-class average maximization true and false target feature extraction method for radar target identification
CN108845303B (en) A Nonlinear Robust Subspace True and False Target Feature Extraction Method
CN108664983A (en) A kind of scale and the adaptive SURF characteristic point matching methods of characteristic strength
CN108828574B (en) Inter-class separation enhancer space true and false target feature extraction method
CN109242010A (en) A kind of sparse study RCS sequence characteristic extracting method
CN110687514B (en) Nonlinear discrimination learning true and false target one-dimensional range profile feature extraction method
CN108594202B (en) A Feature Extraction Method for Nonlinear Projection Targets of Neighborhood Distribution Structure
CN103955711B (en) A kind of mode identification method in imaging spectral target identification analysis
CN108828533B (en) Method for extracting similar structure-preserving nonlinear projection features of similar samples
CN109946670A (en) An optical image-driven radar polarization information extraction method
CN112149061A (en) Multi-class average maximization true and false target feature extraction method
CN107678007B (en) A one-dimensional range profile feature extraction method for radar true and false targets in a compact subspace in exponential domain
CN110068799B (en) A Sparse Neighborhood Center Preserving RCS Sequence Feature Extraction Method
CN110826599B (en) Sparse representation sample distribution boundary retention feature extraction method
CN109615025A (en) Quick atural object classification method
CN108549065B (en) A near-neighbor structure preserving true and false target RCS sequence feature extraction method

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
CF01 Termination of patent right due to non-payment of annual fee
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20220603