CN114280566B - One-dimensional range profile identification method based on class label association - Google Patents
One-dimensional range profile identification method based on class label association Download PDFInfo
- Publication number
- CN114280566B CN114280566B CN202111448633.5A CN202111448633A CN114280566B CN 114280566 B CN114280566 B CN 114280566B CN 202111448633 A CN202111448633 A CN 202111448633A CN 114280566 B CN114280566 B CN 114280566B
- Authority
- CN
- China
- Prior art keywords
- class
- range profile
- class label
- vector
- dimensional range
- 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
Landscapes
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a method for identifying a one-dimensional range profile associated with a class of tags, and belongs to the technical field of radar target identification. The class label association one-dimensional range profile recognition method utilizes the training data set and the corresponding class label vector to construct the class label association subspace, and the class information is directly obtained by carrying out association projection on the one-dimensional range profile sample of the input target, so that a classifier is not required to be introduced to carry out final class decision, the problem that the recognition performance is reduced due to the adoption of the classifier in the conventional recognition method is avoided, and the recognition rate of the target is improved.
Description
Technical Field
The invention belongs to the technical field of radar target recognition, and particularly relates to a method for recognizing a one-dimensional range profile associated with a class of labels.
Background
The one-dimensional range profile acquired by the broadband radar contains information of the structure and the shape of the target, which is very beneficial to target classification, so that higher classification performance can be obtained, and compared with the two-dimensional radar profile, the one-dimensional range profile has more advantages in terms of acquisition easiness and real-time recognition difficulty, so that the one-dimensional range profile is recognized as a hotspot in the current radar target recognition field.
The conventional radar target recognition method still belongs to a classical mode recognition method, firstly, the classification characteristic of the target must be extracted, then, the classifier is adopted to make a final decision on the class of the target, however, the classifier may introduce a certain classification error, so that the recognition rate of the whole system is reduced. Therefore, the performance of the conventional target recognition method has room for further improvement.
Disclosure of Invention
The invention provides a method for identifying a one-dimensional range profile associated with a class of labels, which can be used for improving the identification performance of radar targets.
The invention adopts the technical scheme that:
a kind of label association one-dimensional distance image recognition method includes:
definition x ij The j training one-dimensional distance image representing the i-th known target is 1-g, 1-j-N i Wherein g is the number of categories, N i Training sample number for the i-th class of known targets;
definition y ij Representing a one-dimensional distance image x ij Class tag vector of (2), where y ij Is a g-dimensional column vector, the value of the i element is 1, and the values of other elements are 0;
calculating an autocorrelation matrix R of a one-dimensional range profile sample XY ＝XY T Wherein, one-dimensional range profile sample matrix
Calculating a cross-correlation matrix R between one-dimensional range profile samples and corresponding class label vectors XY ＝XY T Wherein, the class label matrix
Based on autocorrelation matrix R XX And cross-correlation matrix R XY Building an associative subspace
One-dimensional distance image x of input target to be identified t Transforming to obtain its class label vector y t ＝W T x t Wherein the vector y t Is a g-dimension column vector;
based on one-dimensional distance image x t Class label vector y of (2) t Judging the category: traversal vector y t If the current element (y t,k ) Are smaller than the other elements (i.e. y t,k ＞y t,l L=1, 2, … g, l+.k), then a one-dimensional range profile x is determined t The target class of (2) is the class corresponding to the current element, namely belongs to the kth class.
To further ensure recognition performance, the method is performed according to formula R XX ＝(XX T +λI) calculating an autocorrelation matrix R between one-dimensional range profile samples and corresponding class label vectors XX Wherein I represents an identity matrix and λ represents an adjustment factor.
The technical scheme provided by the invention has at least the following beneficial effects:
the invention constructs the class label association subspace by using the training data set and the corresponding class label vector, directly obtains the class information of the input target by carrying out association projection on the one-dimensional distance image sample of the input target, does not need to introduce a classifier to carry out final class decision, and avoids the degradation of the recognition performance caused by adopting the classifier in the conventional recognition method, thereby improving the recognition rate of the target.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in further detail below.
The embodiment of the invention provides a method for identifying a class label associated target, which is characterized in that a one-dimensional distance image sample of an input target is projected through an associated transformation subspace to directly obtain corresponding class label information, so that the reduction of identification performance caused by the introduction of a classifier is avoided, and the identification rate of the target is improved.
The method for identifying the class label associated target provided by the embodiment of the invention comprises the following steps:
the class label association identification method comprises the following steps:
let x be ij (N-dimensional column vector) is the j training one-dimensional distance image of the i-th known target, 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 N i ，
Wherein g is the category number, N i The number of training samples for the i-th class of known targets, N is the total number of training samples. X is x ij Representing the corresponding class label vector as y ij Is a g-dimensional column vector, the value of the i-th element is 1, and the values of other elements are 0. From x ij Forming a one-dimensional range profile sample matrix X:
from class label vector y ij Constituting a class label matrix Y:
calculating an autocorrelation matrix R of a one-dimensional range profile sample XX ：
R XX ＝XX T (3)
Wherein, superscript'T' represents matrix transposition, in order to avoid R XX Is a singular matrix, the adjusting factor lambda is introduced, lambda is an empirical value, and a specific value can be determined by experiments.
R XX ＝(XX T +λI) (4)
Wherein I is an identity matrix. Calculating a cross-correlation matrix R between one-dimensional range profile samples and corresponding class label vectors XY
R XY ＝XY T (5)
Based on the sample autocorrelation matrix and the sample and class label cross correlation matrix, constructing an association subspace W:
one-dimensional distance image x of input target to be identified t The following transformation is carried out to obtain a corresponding class label vector y t ：
y t ＝W T x t (7)
If the conditions are satisfied:
wherein y is t,k And y t,l Is the vector y t Is a component of the group. Then judge x t Belonging to the k-th class.
In order to verify the recognition performance of the label-like associated one-dimensional range profile recognition method provided by the embodiment of the invention, the following simulation experiment is carried out:
four point targets are set: "I" style of calligraphy, "V" style of calligraphy, "dry" style of calligraphy and "little" style of calligraphy targets. The bandwidth of the radar transmission pulse is 150MHz (the distance resolution is 1m, the radar radial sampling interval is 0.5 m), the target is set as a uniform scattering point target, the scattering point of the 'I' target is 5, and the scattering points of the other three targets are 9. In the one-dimensional range images with the target attitude angles of 0-70 degrees at intervals of 1 degree, training is carried out by taking one-dimensional range images with the target attitude angles of 0 degrees, 2 degrees, 4 degrees, 6 degrees, the first and the second angles, and the one-dimensional range images with the rest attitude angles are used as test data, and then 35 test samples are arranged in each category of targets.
The identification experiments are carried out on four targets ("|" type targets, "V" type targets, "dry" type targets and "small" type targets) within the range of the attitude angle of 0-70 degrees by using the class label association identification method and the conventional characteristic subspace-based identification method provided by the embodiment of the invention, the average identification rate of the 4 targets reaches 87%, and the average identification rate of the conventional characteristic subspace-based identification method is about 82%, so that the identification performance of the multiple targets can be improved by the identification method provided by the embodiment of the invention.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
What has been described above is merely some embodiments of the present invention. It will be apparent to those skilled in the art that various modifications and improvements can be made without departing from the spirit of the invention.
Claims (2)
1. The identification method of the class label associated one-dimensional range profile is characterized by comprising the following steps:
definition x ij The j training one-dimensional distance image representing the i-th known target is 1-g, 1-j-N i Wherein g is the number of target categories, N i Training sample number for the i-th class of known targets;
definition y ij Representing a one-dimensional distance image x ij Class tag vector of (2), where y ij Is g-dimension columnThe value of the ith element of the vector, like a label vector, is 1, and the values of other elements are 0;
calculating an autocorrelation matrix R of a one-dimensional range profile sample XX ＝(XX T +λI), wherein one-dimensional range profile sample matrix
I represents an identity matrix, and lambda represents an adjusting factor;
calculating a cross-correlation matrix R between one-dimensional range profile samples and corresponding class label vectors XY ＝XY T Wherein, the class label matrix
Based on autocorrelation matrix R XX And cross-correlation matrix R XY Building an associative subspace
One-dimensional distance image x of input target to be identified t Transforming to obtain its class label vector y t ＝W T x t ；
Based on one-dimensional distance image x t Class label vector y of (2) t Judging the category: traversal vector y t To combine the current element with the vector y t The other elements in the list are compared, if the comparison result is smaller than the comparison result, the category corresponding to the current element is one-dimensional distance image x t Is a target class of (c).
2. The method of claim 1, wherein the formula R is followed by XX ＝(XX T +λI) calculating an autocorrelation matrix R between one-dimensional range profile samples and corresponding class label vectors XX Wherein I represents an identity matrix and λ represents a preset adjustment factor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111448633.5A CN114280566B (en) | 2021-11-30 | 2021-11-30 | One-dimensional range profile identification method based on class label association |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111448633.5A CN114280566B (en) | 2021-11-30 | 2021-11-30 | One-dimensional range profile identification method based on class label association |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114280566A CN114280566A (en) | 2022-04-05 |
| CN114280566B true CN114280566B (en) | 2023-05-23 |
Family
ID=80870513
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111448633.5A Active CN114280566B (en) | 2021-11-30 | 2021-11-30 | One-dimensional range profile identification method based on class label association |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114280566B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114861809A (en) * | 2022-05-20 | 2022-08-05 | 电子科技大学 | True and false target one-dimensional range profile nonlinear class label correlation identification method |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104463247B (en) * | 2014-12-09 | 2016-08-31 | 山东大学 | The abstracting method of spectral vector cross-correlation feature in a kind of classification hyperspectral imagery |
| GB2554633B (en) * | 2016-06-24 | 2020-01-22 | Imperial College Sci Tech & Medicine | Detecting objects in video data |
| CN108761411B (en) * | 2018-04-10 | 2021-11-23 | 电子科技大学 | True and false target one-dimensional range profile feature extraction method |
| EP3581961A1 (en) * | 2018-06-13 | 2019-12-18 | Technische Universität München | Method and apparatus for ultrasound imaging with improved beamforming |
| CN108845303B (en) * | 2018-08-27 | 2022-03-15 | 电子科技大学 | Nonlinear robust subspace true and false target feature extraction method |
| CN110109095B (en) * | 2019-04-30 | 2022-10-28 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | Target feature assisted multi-source data association method |
| CN110334741B (en) * | 2019-06-06 | 2023-03-31 | 西安电子科技大学 | Radar one-dimensional range profile identification method based on cyclic neural network |
| JP7298350B2 (en) * | 2019-07-08 | 2023-06-27 | 株式会社リコー | Ranging device, information processing device, ranging method, in-vehicle device, moving object, ranging system |
| CN110471104B (en) * | 2019-08-26 | 2021-03-16 | 电子科技大学 | Post-stack seismic reflection pattern recognition method based on intelligent feature learning |
| CN112183617B (en) * | 2020-09-25 | 2022-03-29 | 电子科技大学 | RCS sequence feature extraction method for sample and class label maximum correlation subspace |
-
2021
- 2021-11-30 CN CN202111448633.5A patent/CN114280566B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN114280566A (en) | 2022-04-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107238822B (en) | Method for extracting orthogonal nonlinear subspace characteristics of true and false target one-dimensional range profile | |
| CN108761411B (en) | True and false target one-dimensional range profile feature extraction method | |
| CN107037417B (en) | The true and false target of radar is one-dimensional as non-linear arest neighbors subspace representation method | |
| CN106778837B (en) | SAR image target recognition method based on polyteny principal component analysis and tensor analysis | |
| CN107545279B (en) | Image identification method based on convolutional neural network and weighted kernel feature analysis | |
| CN108845302B (en) | K-nearest neighbor transformation true and false target feature extraction method | |
| CN114280566B (en) | One-dimensional range profile identification method based on class label association | |
| CN110781766A (en) | Grassmann manifold discriminant analysis image recognition method based on characteristic spectrum regularization | |
| CN106874932B (en) | SAR target model identification method based on rapid sparse description | |
| CN113139513B (en) | Spatial spectrum active learning hyperspectral classification method based on superpixel profile and improved PSO-ELM | |
| CN108845303B (en) | Nonlinear robust subspace true and false target feature extraction method | |
| CN114782763A (en) | Sparse principal component alignment method for multi-view high-resolution remote sensing image | |
| CN110687514B (en) | Nonlinear discrimination learning true and false target one-dimensional range profile feature extraction method | |
| CN110020674B (en) | Cross-domain self-adaptive image classification method for improving local category discrimination | |
| CN109063750B (en) | SAR target classification method based on CNN and SVM decision fusion | |
| CN106199544A (en) | The Recognition of Radar Target Using Range Profiles method of local tangent space alignment is differentiated based on core | |
| CN108428226B (en) | Distortion image quality evaluation method based on ICA sparse representation and SOM | |
| CN110221267A (en) | A kind of object library attribute differentiation neighborhood holding Learning Subspaces feature extracting method | |
| CN114241233A (en) | Nonlinear class group sparse representation true and false target one-dimensional range profile identification method | |
| CN104794497B (en) | Multicenter approximating method in a kind of classification hyperspectral imagery | |
| CN109784244B (en) | Low-resolution face accurate identification method for specified target | |
| CN116824218A (en) | Small sample sequence ISAR image robust classification method based on semi-supervised transfer learning | |
| CN107678007B (en) | Method for extracting radar true and false target one-dimensional range profile features in exponential domain compact subspace | |
| CN110780270B (en) | Target library attribute discrimination local regular learning subspace feature extraction method | |
| CN114241239A (en) | Unsupervised pre-adaptive image classification method based on manifold learning |
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 |