CN111539487A - Target classification method and system based on millimeter wave polarization radiation measurement - Google Patents

Abstract

The invention discloses a target classification method and a system based on millimeter wave polarization radiation measurement, wherein horizontal polarization brightness temperature, vertical polarization brightness temperature and physical temperature of N targets are respectively measured, incidence angles are recorded, N groups of data are obtained in total, equivalent dielectric constants of the N groups of data are respectively calculated, the equivalent dielectric constants are clustered, and a clustering result is a classification result of the N targets. The method overcomes the defect that the classification is carried out only by depending on the brightness temperature of the target in the prior art, and improves the capability of detecting and identifying various targets in the observed scene by utilizing the millimeter wave radiation measurement technology.

Description

Target classification method and system based on millimeter wave polarization radiation measurement

Technical Field

The invention belongs to the technical field of passive microwave remote sensing and detection, and particularly relates to a target classification method and system based on millimeter wave polarized radiometry.

Background

All objects with physical temperatures above 0K spontaneously radiate electromagnetic energy into space in the form of electromagnetic waves, a phenomenon known as thermal radiation. Millimeter wave radiometers achieve remote sensing and detection of targets by receiving thermal radiation signals in the millimeter wave frequency band, and this technology is generally called passive millimeter wave radiometry. Millimeter wave radiometry has the advantages of all-time, quasi-all-weather operation, invisibility and penetration to a certain extent through the atmosphere, clothes, smoke, etc., and thus has been applied to fields such as remote sensing, security inspection, target detection, astronomy, etc. Polarization is a type of characteristic of electromagnetic radiation.

Because the polarization of electromagnetic radiation is related to factors such as material composition, shape and surface characteristics of the target, the millimeter wave radiation polarization imaging contains abundant target information. The targets are classified based on the polarized bright temperature data obtained by measurement, and the method is concerned in the fields of human body security inspection, military target detection and the like. The current method classifies targets directly based on brightness temperature data or after simple operations of addition, subtraction, multiplication and division are carried out on the brightness temperature data. However, the target brightness temperature is related to various factors such as the ambient brightness temperature, the target temperature itself, and the observation angle, and is not a physical quantity related only to the target material information. Thus, two different types of targets may have the same light temperature. How to extract physical quantities directly representing target material information from the brightness temperature data so as to classify targets still needs to be researched.

Disclosure of Invention

Aiming at the defects of the prior art, the invention aims to provide a target classification method based on polarized millimeter wave radiation measurement, and aims to overcome the defect that the classification is carried out only by depending on the brightness temperature of a target in the prior art.

To achieve the above object, according to one aspect of the present invention, there is provided a target classification method based on millimeter wave polarized radiometry, comprising the steps of:

s1, measuring horizontal polarization brightness temperature T of N targets respectively_hiVertical polarization brightness temperature T_viPhysical temperature T_objiRecording the angle of incidence θ_iObtaining N groups of data, and recording as { (T)_h1,T_v1,T_obj1,θ₁)；(T_h2,T_v2,T_obj2,θ₂)；…；(T_hN,T_vN,T_objN,θ_N) I is more than or equal to 1 and less than or equal to N, and N is a positive integer;

s2, calculating the equivalent dielectric constant of the N groups of data respectively, and recording as₁；₂；…；_N}；

S3. paired check₁；₂；…；_NCarry out clustering, clustering result { c }₁；c₂；…；c_NI.e. the classification results of the N targets.

Further, for the ith data (T) in S2_hi,T_vi,T_obji,θ_i) Equivalent dielectric constant of_iThe calculation method comprises the following steps:

s21, calculating PD_oP＝(T_vi-T_hi)/(2T_obji-T_hi-T_vi) If PDoP < P₀Then, then_i＝₀(ii) a Otherwise, go to S22; wherein P is₀Is a preset threshold value;

s22, calculating LPR ═ T_h-T_obj)/(T_v-T_obj) If LPR < L₀According to the equation

Obtaining_i(ii) a Otherwise according to the equation

Obtaining_i(ii) a Wherein L is₀Is a preset threshold.

Further, in S3_iAll targets 10000 are metals, c_i0; for other targets, clustering into n (n ≧ 1) classes using a clustering algorithm, c_i∈{1,2,…,n}。

According to another aspect of the present invention, there is provided an object classification system based on millimeter wave polarized radiometry, comprising:

a data acquisition module for measuring the horizontal polarization brightness temperature T of the N targets respectively_hiVertical polarization brightness temperature T_viPhysical temperature T_objiRecording the angle of incidence θ_iObtaining N groups of data, and recording as { (T)_h1,T_v1,T_obj1,θ₁)；(T_h2,T_v2,T_obj2,θ₂)；…；(T_hN,T_vN,T_objN,θ_N)}，1≤i≤N；

An equivalent dielectric constant obtaining module for respectively calculating the equivalent dielectric constants of the N groups of data and recording as₁；₂；…；_N}；

Sorting module of butt₁；₂；…；_NCarry out clustering, clustering result { c }₁；c₂；…；c_NI.e. the classification results of the N targets.

Further, the equivalent dielectric constant_iThe calculation method comprises the following steps:

s21, calculating PDoP ═ T_vi-T_hi)/(2T_obji-T_hi-T_vi) If PDoP < P₀Then, then_i＝₀(ii) a Otherwise, go to S22; wherein P is₀Is a preset threshold value;

s22, calculating LPR ═ T_h-T_obj)/(T_v-T_obj) If LPR < L₀According to the equation

Obtaining_i(ii) a Otherwise according to the equation

Obtaining_i(ii) a Wherein L is₀Is a preset threshold.

Further, in the present invention,_iall targets 10000 are metals, c_i0; for other targets, clustering into n (n ≧ 1) classes using a clustering algorithm, c_i∈{1,2,…,n}。

Compared with the prior art, the equivalent dielectric constant is obtained only through the brightness temperature, the dielectric constant of a substance is complex in a millimeter wave frequency band, and the complex dielectric constant is difficult to solve through the brightness temperature. The method utilizes the measured millimeter wave radiation polarization brightness temperature, physical temperature and incident angle of a plurality of targets to obtain the equivalent dielectric constant through data fitting, and then classifies the targets based on the equivalent dielectric constant. The method solves the defect that the classification is carried out only by depending on the brightness temperature of the target in the prior art, is not influenced by the incident angle, the brightness temperature of the environment and the physical temperature of the target, and improves the capability of detecting and identifying various targets in the observed scene by using the millimeter wave radiometric technology.

Drawings

Fig. 1 is a measurement schematic diagram of a target classification method based on millimeter wave polarized radiation measurement provided by the invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

The invention provides a target classification method based on millimeter wave polarization radiation measurement, which comprises the following steps:

(1) measuring horizontal polarization brightness temperature T of 12 targets_hVertical polarization brightness temperature T_vPhysical temperature T_objThe incident angle θ is recorded, and the measurement results are:

(2) for 12 sets of measured data, the equivalent dielectric constants were calculated:

s21, calculating PDoP ═ T_vi-T_hi)/(2T_obji-T_hi-T_vi) If PDoP < P₀Then, then_i＝₀(ii) a Otherwise, go to S22; in this example P₀＝0.05；₀＝10000；

S22, calculating LPR ═ T_h-T_obj)/(T_v-T_obj) If LPR < L₀According to the equation

Obtaining_i(ii) a Otherwise according to the equation

Obtaining_i(ii) a In this example L₀＝50。

The calculation result is as follows:

(3) check and check₁；₂；…；_NAnd then clustering is carried out. The clustering result is { 1; 1; 2; 2; 0; 0; 1; 1; 2; 2; 0; 0}, namely 1 st, 2 nd, 7 th and 8 th targets are the same type of materials; 3, 4, 9, 10 targets are the same class of materials; the 5 th, 6 th, 11 th, and 12 th targets are the same class of materials and are metal classes.

It will be understood by those skilled in the art that the foregoing is only a preferred embodiment of the present invention, and is not intended to limit the invention, and that any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims (6)

1. A target classification method based on millimeter wave polarized radiation measurement is characterized by comprising the following steps:

s1, measuring horizontal polarization brightness temperature T of N targets respectively_hiVertical polarization brightness temperature T_viPhysical temperature T_objiRecording the angle of incidence θ_iObtaining N groups of data, and recording as { (T)_h1,T_v1,T_obj1,θ₁)；(T_h2,T_v2,T_obj2,θ₂)；…；(T_hN,T_vN,T_objN,θ_N) I is more than or equal to 1 and less than or equal to N, and N is a positive integer;

s2, calculating the equivalent dielectric constant of the N groups of data respectively, and recording as₁；₂；…；_N}；

S3. paired check₁；₂；…；_NCarry out clustering, clustering result { c }₁；c₂；…；c_NI.e. the classification results of the N targets.

2. Method according to claim 1, characterized in that for the ith data (T) in S2_hi,T_vi,T_obji,θ_i) Equivalent dielectric constant of_iThe calculation method comprises the following steps:

s21, calculating PDoP ═ T_vi-T_hi)/(2T_obji-T_hi-T_vi) If PDoP < P₀Then, then_i＝₀(ii) a Otherwise, go to S22; wherein P is₀Is a preset threshold value;

s22, calculating LPR ═ T_h-T_obj)/(T_v-T_obj) If LPR < L₀According to the equation

Obtaining_i(ii) a Otherwise according to the equation

Obtaining_i(ii) a Wherein L is₀Is a preset threshold.

3. The method of claim 1, wherein in S3_iAll targets 10000 are metals, c_i0; for other targets, clustering into n (n ≧ 1) classes using a clustering algorithm, c_i∈{1,2,…,n}。

4. A target classification system based on millimeter wave polarized radiometry, comprising:

a data acquisition module for measuring the horizontal polarization brightness temperature T of the N targets respectively_hiVertical polarization brightness temperature T_viPhysical temperature T_objiRecording the angle of incidence θ_iObtaining N groups of data, and recording as { (T)_h1,T_v1,T_obj1,θ₁)；(T_h2,T_v2,T_obj2,θ₂)；…；(T_hN,T_vN,T_objN,θ_N)}，1≤i≤N；

An equivalent dielectric constant obtaining module for obtaining the equivalent dielectric constant of the semiconductor device,calculating the equivalent dielectric constant of the data of N units, and recording as₁；₂；…；_N}；

Sorting module of butt₁；₂；…；_NCarry out clustering, clustering result { c }₁；c₂；…；c_NI.e. the classification results of the N targets.

5. The system of claim 4, wherein the equivalent dielectric constant is_iThe calculation method comprises the following steps:

s21, calculating PDoP ═ T_vi-T_hi)/(2T_obji-T_hi-T_vi) If PDoP < P₀Then, then_i＝₀(ii) a If not, then,

execution of S22; wherein P is₀Is a preset threshold value;

s22, calculating LPR ═ T_h-T_obj)/(T_v-T_obj) If LPR < L₀According to the equation

Obtaining_i(ii) a Otherwise according to the equation

Obtaining_i(ii) a Wherein L is₀Is a preset threshold.

6. The system of claim 4,_iall targets 10000 are metals, c_i0; for other targets, clustering into n (n ≧ 1) classes using a clustering algorithm, c_i＝{1,2,…,n}。

Images