A method for identifying similar but different targets at room temperature based on multi-band spectral extreme value features

Abstract

This invention belongs to the field of optoelectronic information acquisition and processing, and provides a method for identifying similar but different targets at room temperature based on multi-band spectral extreme value characteristics. Starting from the fundamental infrared physics knowledge that targets scatter / radiate differently, it utilizes the material, scattering, and radiation characteristics of true and false targets in different bands, i.e., identifying the extreme points (absorption peaks) corresponding to different bands, to distinguish between true and false targets. Since targets exhibit different characteristics in different bands, the number of absorption peaks also differs. Short-wave bands primarily reflect the target's scattering characteristics, while medium and long-wave bands reflect its radiation characteristics. Therefore, when statistically analyzing spectral extreme points, the bands are divided into short-wave, medium-wave, and long-wave bands, and adjacent extreme spectral points of scattering and radiation differ between true and false targets. This invention uses a single-band discrimination formula to sequentially determine the number and amplitude of extreme points in the short-wave, medium-wave, and long-wave bands for true and false targets until a true target is found. If a true target cannot be identified in any of the three bands, the target is considered a false target.

Description

Technical Field

[0001] This invention belongs to the field of optoelectronic information acquisition and processing, and more specifically, relates to a method for identifying homomorphic but heterogeneous targets at room temperature based on multi-band spectral extreme value features. Background Technology

[0002] Decoy targets are simulated objects set up and constructed to deceive and mislead others, displaying signs of target exposure. Using them to simulate targets and set up fake positions can distract the enemy, lure them into reconnaissance and attack, and cover concealed real targets; it can also disperse enemy firepower, reducing losses to real targets. These decoy targets deliberately expose their own positions, inducing the enemy to waste a large number of expensive missiles, while simultaneously protecting real targets, and may also induce the enemy to reveal the location information of their own launchers.

[0003] Decoys generally possess the following characteristics: ① Low cost; ② Realistic appearance, including geometric dimensions and visible light color, similar to real targets; ③ Radar and infrared signal strength often exceeds that of real targets; ④ Inflatable decoys are lightweight and easy to transport; ⑤ Quick deployment and dismantling, and convenient maintenance. Real and decoy targets have the same shape but use different materials. Decoys are usually at room temperature and are difficult to distinguish from real targets at room temperature when stationary. However, current decoys only provide camouflage in certain specific wavelengths and cannot achieve full-spectrum camouflage. Summary of the Invention

[0004] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a method for identifying similar but different targets at room temperature based on multi-band spectral extreme value features, aiming to solve the problem that it is difficult to distinguish similar but different targets in the existing technology.

[0005] To achieve the above objectives, in a first aspect, the present invention provides a method for identifying homomorphic but heterogeneous targets at room temperature based on multi-band spectral extreme value features, comprising the following steps:

[0006] Acquire spectral data of two targets during the daytime and / or nighttime periods. The daytime spectrum is mainly composed of scattering spectra, while the nighttime spectrum is mainly composed of radiation spectra. The two targets have the same shape but different compositions; one is a real target and the other is a decoy target. The spectral data covers three wavebands: shortwave, medium wave, and long wave.

[0007] Wavelet decomposition is performed on the spectral data to determine multiple extreme points in the high-frequency component; the extreme points include maxima and minima.

[0008] Based on a predetermined noise spectrum range, extreme points in the two target spectral data that fall within the noise spectrum range are filtered out, and extreme points with the same wavelength as the two target extreme points are also filtered out, resulting in multiple effective extreme points after filtering of the two targets.

[0009] For spectral data during the daytime, the effective extreme points of the shortwave, mediumwave and longwave bands are judged in sequence. The target with fewer effective extreme points among two targets in a certain band is the real target, or the target with a relatively smaller sum of amplitudes of all effective extreme points in a certain band is the real target, until the data of a certain band can distinguish between real and false targets.

[0010] For spectral data during nighttime, the effective extreme points of the long-wave, medium-wave and short-wave bands are used to make judgments in sequence. The target with more effective extreme points in a certain band is the real target, or the target with a relatively larger sum of amplitudes of all effective extreme points in a certain band is the real target, until the data of a certain band can distinguish between real and false targets.

[0011] One possible implementation also includes the following steps:

[0012] Acquire spectral data of a blackbody at different ambient temperatures; the high-frequency signals of the blackbody spectral data are caused by noise from the spectral detection equipment;

[0013] Wavelet decomposition was performed on multiple spectral data of the blackbody at each room temperature to obtain the high-frequency components of the multiple spectral data;

[0014] The mean and standard deviation of multiple spectral data at each room temperature are calculated and used as the mean and standard deviation of the noise spectrum at each room temperature. The range of the noise spectrum is then determined based on the mean and standard deviation.

[0015] In one possible implementation, wavelet decomposition is performed on the spectral data to determine multiple extreme points in the high-frequency component, specifically as follows:

[0016] When the DN value of the remote sensing image pixel brightness in a certain band after wavelet decomposition is greater than or equal to the DN value of the remote sensing image pixel brightness in other bands in its neighborhood, that band is taken as the band corresponding to the maximum point.

[0017] When the DN value of a certain band after wavelet decomposition is less than or equal to the DN values ​​of other bands in its neighborhood, that band is taken as the band corresponding to the minimum point.

[0018] In one possible implementation, during the daytime, the single-band discrimination formula is as follows:

[0019]

[0020] In the above formula, λ TrueTar For the band corresponding to the effective extreme point of the true target, λ FakeTar For the effective extreme points of the false target, corresponding to multiple bands, Num(λ) TrueTar ) represents the number of valid extreme points of the true objective, Num(λ) FakeTar ) represents the number of valid extreme points of the false objective, True(λ) TrueTar) represents the spectral DN value corresponding to the effective extreme point of the true target, Fake(λ) FakeTar ) represents the spectral DN value corresponding to the effective extreme point of the false target, Sum(True(λ) TrueTar Sum(Fake(λ)) represents the magnitude of all valid extreme points of the true objective. FakeTar )) represents the amplitude of all extreme points of the false target.

[0021] In one possible implementation, during the nighttime period, the single-band discrimination formula is as follows:

[0022]

[0023] Secondly, the present invention provides a device for identifying homomorphic but heterogeneous targets at room temperature based on multi-band spectral extreme value features, comprising:

[0024] The spectral data acquisition unit is used to acquire spectral data of two targets during the daytime and / or nighttime periods. The spectrum during the daytime period is mainly composed of scattering spectra, while the spectrum during the nighttime period is mainly composed of radiation spectra. The two targets have the same shape but different compositions, one being a real target and the other a decoy target. The spectral data covers three wavebands: shortwave, medium wave, and long wave.

[0025] The wavelet decomposition unit is used to perform wavelet decomposition on spectral data to determine multiple extreme points in the high-frequency part; the extreme points include maxima and minima.

[0026] An extreme point filtering unit is used to filter out extreme points in two target spectral data that are within the noise spectrum range based on a predetermined noise spectrum range, and to filter out extreme points with the same wavelength in the two target extreme points, so as to obtain multiple effective extreme points after filtering the two targets.

[0027] The target recognition unit is used to judge the true target based on the effective extreme points of the shortwave, mediumwave, and longwave bands for spectral data during the daytime. The true target is the one with fewer effective extreme points among two targets in a certain band, or the true target is the one with a relatively smaller sum of amplitudes of all effective extreme points in a certain band, until the data of a certain band can distinguish between true and false targets. Similarly, for spectral data during the nighttime, the unit judges the true target based on the effective extreme points of the longwave, mediumwave, and shortwave bands for spectral data. The true target is the one with more effective extreme points among two targets in a certain band, or the true target is the one with a relatively larger sum of amplitudes of all effective extreme points in a certain band, until the data of a certain band can distinguish between true and false targets.

[0028] In one possible implementation, the device further includes:

[0029] The noise spectrum determination unit is used to acquire spectral data of a blackbody at different ambient temperatures; the high-frequency signal of the blackbody spectral data is caused by noise from the spectral detection device; wavelet decomposition is performed on multiple spectral data at each ambient temperature of the blackbody to obtain the high-frequency part of multiple spectral data; and the mean and standard deviation of multiple spectral data at each ambient temperature are calculated and used as the mean and standard deviation of the noise spectrum at each ambient temperature, and the range of the noise spectrum is determined based on the mean and standard deviation.

[0030] In one possible implementation, the wavelet decomposition unit performs wavelet decomposition on the spectral data to determine multiple extreme points in the high-frequency part. Specifically, when the DN value of the remote sensing image pixel brightness in a certain band after wavelet decomposition is greater than or equal to the DN value of the remote sensing image pixel brightness in other bands in its neighborhood, the band is taken as the band corresponding to the maximum value point; when the DN value of a certain band after wavelet decomposition is less than or equal to the DN value of other bands in its neighborhood, the band is taken as the band corresponding to the minimum value point.

[0031] Thirdly, this application provides an electronic device, comprising: at least one memory for storing a program; and at least one processor for executing the program stored in the memory, wherein when the program stored in the memory is executed, the processor is configured to execute the method described in the first aspect or any possible implementation thereof.

[0032] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when run on a processor, causes the processor to perform the method described in the first aspect or any possible implementation thereof.

[0033] Fifthly, this application provides a computer program product that, when run on a processor, causes the processor to perform the method described in the first aspect or any possible implementation thereof.

[0034] In summary, the technical solutions conceived by this invention have the following beneficial effects compared with the prior art:

[0035] This invention provides a method for identifying similar-shaped but different-material targets at room temperature based on multi-band spectral extreme value features. Wavelet decomposition is used to obtain high-frequency spectral information, and effective extreme points are obtained through the device noise spectrum. These extreme points reflect the scattering and radiation characteristics of the target material in different bands. Since real and false targets have different scattering and radiation characteristics in different bands, the number and amplitude of their extreme points also differ, thus enabling effective identification of similar-shaped but different-material targets at room temperature. Attached Figure Description

[0036] Figure 1This is a flowchart of a method for identifying isomorphic but heterogeneous targets at room temperature based on multi-band spectral extreme value features, provided in an embodiment of the present invention.

[0037] Figure 2 This is a flowchart of the multi-band spectral extreme point real and false target identification process provided in the embodiments of the present invention;

[0038] Figure 3 This is a comparison chart of blackbody calibration spectrum curves at different temperatures provided in the embodiments of the present invention;

[0039] Figure 4 This is a high-frequency curve of a 30°C blackbody after wavelet decomposition, provided in an embodiment of the present invention.

[0040] Figure 5 This is a high-frequency curve of a 50°C blackbody after wavelet decomposition, provided in an embodiment of the present invention.

[0041] Figure 6 This is a high-frequency curve of an 80° blackbody after wavelet decomposition, provided in an embodiment of the present invention.

[0042] Figure 7 This is a mean curve of the high-frequency component obtained after wavelet decomposition according to an embodiment of the present invention;

[0043] Figure 8 This is a standard deviation curve of the high-frequency component obtained after wavelet decomposition according to an embodiment of the present invention;

[0044] Figure 9 This is a schematic diagram of the device noise spectrum range provided in an embodiment of the present invention;

[0045] Figure 10 This is a comparison chart of the spectra of real and false targets during the daytime provided in an embodiment of the present invention;

[0046] Figure 11 This is a comparison chart of the spectra of real and false targets during nighttime, provided in an embodiment of the present invention.

[0047] Figure 12 This is the wavelet decomposition and noise spectrum of real and false targets during the day provided in the embodiments of the present invention;

[0048] Figure 13 This is a wavelet decomposition maximum analysis diagram of real and false targets during the day provided in an embodiment of the present invention;

[0049] Figure 14 This is a wavelet decomposition minimum analysis diagram of real and false targets during the day provided in an embodiment of the present invention;

[0050] Figure 15 This is the wavelet decomposition and noise spectrum of real and false targets at night provided in an embodiment of the present invention;

[0051] Figure 16 This is a wavelet decomposition maximum point analysis diagram of true and false targets at night provided in an embodiment of the present invention;

[0052] Figure 17 This is an analysis diagram of the minimum points of wavelet decomposition of true and false targets at night provided in an embodiment of the present invention;

[0053] Figure 18 This is an architectural diagram of a device for identifying homomorphic but heterogeneous targets at room temperature based on multi-band spectral extreme value features, provided in an embodiment of the present invention. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0055] In this invention, the term "and / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. In this invention, the symbol " / " indicates that the related objects are in an "or" relationship; for example, A / B means A or B.

[0056] The terms "first" and "second," etc., used in the specification and claims of this invention are used to distinguish different objects, rather than to describe a specific order of objects.

[0057] In the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more, for example, multiple processing units means two or more processing units, multiple elements means two or more elements, etc.

[0058] First, the technical terms involved in the embodiments of this application will be introduced.

[0059] (1) Target at room temperature

[0060] The room temperature mentioned in this application refers to the chemical industry, where the room temperature ranges from -20℃ to 200℃.

[0061] (2) Extreme points

[0062] In this application, the extreme point refers to a spectral point.

[0063] Next, the technical solutions provided in the embodiments of this application will be described.

[0064] This invention proposes a new method to identify true and false targets based on the differences in their internal structure and material composition, as well as the fundamental differences in the exchange of matter and energy between the true and false targets and between the true and false targets and their external environment, resulting in different infrared broadband spectral characteristics.

[0065] Figure 1 This is a flowchart of a method for identifying homomorphic but heterogeneous targets at room temperature based on multi-band spectral extreme value features, provided in an embodiment of the present invention; as shown below. Figure 1 As shown, it includes the following steps:

[0066] S101, acquire spectral data of two targets during the daytime and / or nighttime periods. The spectrum during the daytime period is mainly composed of scattering spectrum, and the spectrum during the nighttime period is mainly composed of radiation spectrum. The two targets have the same shape but different composition, one is a real target and the other is a fake target. The spectral data covers three wavebands: shortwave, medium wave and long wave.

[0067] S102, perform wavelet decomposition on the spectral data to determine multiple extreme points in the high-frequency part; the extreme points include maxima and minima;

[0068] S103, based on a predetermined noise spectrum range, filter out the extreme points in the two target spectral data that are within the noise spectrum range, and filter out the extreme points with the same wavelength of the two target extreme points, to obtain multiple effective extreme points of the two targets after filtering;

[0069] S104. For spectral data during the daytime, the effective extreme points of the shortwave, mediumwave and longwave bands are judged in sequence. The target with fewer effective extreme points in a certain band is the true target, or the target with a relatively smaller sum of amplitudes of all effective extreme points in a certain band is the true target, until the data of a certain band can distinguish between true and false targets.

[0070] S105. For spectral data during the nighttime period, the effective extreme points of the long-wave, medium-wave and short-wave bands are judged in sequence. The target with more effective extreme points in a certain band is the real target, or the target with a relatively larger sum of amplitudes of all effective extreme points in a certain band is the real target, until the data of a certain band can distinguish between real and false targets.

[0071] Specifically, the method provided in the embodiments of the present invention is analyzed as follows:

[0072] 1. Analyze the differences in reflection spectra by utilizing the different physical properties of targets with the same shape but different textures.

[0073] True targets are generally made of metal, with a smooth surface and a special paint coating, and their reflection spectrum exhibits quasi-specular reflection.

[0074] The dummy target is usually a polymer chemical material with a rough surface and a diffuse reflection spectrum.

[0075] Therefore, the scattering spectra of real and fake targets are completely different.

[0076] 2. Analyze the differences in radiation spectra by utilizing the different internal structures of targets with the same shape but different properties.

[0077] The internal structure of a real target is complex, and it also has an internal heat source. The exchange of matter and energy between its internal structures is complex, as is the exchange of matter and energy between its internal structures and the surrounding environment.

[0078] The internal structure of an inflatable decoy is simple, consisting of air. There is no complex exchange of matter and energy within it, nor is there any complex exchange of matter and energy with the surrounding external environment.

[0079] Therefore, the infrared radiation spectra of real and fake targets are completely different.

[0080] 3. Homogeneous but different spectra and homogeneous but different spectral types only exist in narrow bands and cannot all appear in the wide infrared band.

[0081] Because real and fake targets have different materials and heat capacities, their internal energy exchange and their energy exchange with the external environment are different. Therefore, the spectral information of real and fake targets are different in different wavelength bands.

[0082] 4. Analyze the spectral noise of the spectral measurement equipment to help remove the interference of noise spectrum on the identification of true and false target spectra.

[0083] In addition to the target spectral information, the signal received by the spectral association device also includes the detector's internal noise spectrum. This invention conducts blackbody calibration experiments on the spectral association device at different temperatures, performs wavelet decomposition on the n spectral data collected during blackbody calibration, extracts the small-scale components of the wavelet decomposition, and calculates the mean and standard deviation of the small-scale components. The region covered by the average value and standard deviation vector of the spectral noise varying with different spectral points is used as the benchmark for suppressing the noise spectrum.

[0084]

[0085] S inner (λ)=μ inner (λ)±σ inner (λ)

[0086] In the formula, λ is the wavelength, and S i For the high-frequency portion of the measured data after wavelet decomposition, μ inner σ is the mean of the noise spectrum. inner S represents the standard deviation of the noise spectrum. inner This refers to the noise spectrum range of the equipment.

[0087] 5. Use wavelet decomposition to obtain the changes in the measured reflection spectrum and radiation spectrum.

[0088] Wavelet transform was used to decompose the measured data, obtaining the small-scale components after spectral wavelet decomposition, and extracting their maxima and minima. The amplitude and location of the extreme points reflect the intensity of the changes in the reflection and radiation spectra of the target in different wavebands. The spectral bands corresponding to all maxima and minima of the true and false targets were calculated according to the following formula.

[0089]

[0090] In the above formula, λ0 represents the band, and λ Maximum For the band corresponding to the maximum point, λ Minimum Let S(λ) be the spectral DN value corresponding to the minimum point. S(λ) is the neighborhood of S(λ0), which is the region to the left and right of λ0.

[0091] The method for finding effective extreme points in this application is as follows: 1. First, find all extreme points; 2. Filter out extreme points in the same band as real and false targets; 3. Filter out extreme points in the noise spectrum.

[0092] Due to limitations in the manufacturing processes of existing optical components, when detecting the spectrum of a target at long distances, all sensing units in the spectral module of the image-spectrum association detection system can receive spectral energy. Therefore, the target spectral sequence obtained by the spectral measurement module may contain a mixed spectrum of the target and the background. Because the background spectrum is relatively stable, after statistically analyzing the extreme points of true and false targets, spectral points with the same extreme points between true and false targets can be filtered out, thereby obtaining more accurate extreme point information about true and false targets.

[0093] Based on the equipment noise spectrum calculated above, and using the standard deviation region of the noise spectrum as a reference standard, the extreme points of the true and false target spectral data within the noise spectrum are filtered out to obtain the effective extreme points of the true and false targets.

[0094]

[0095]

[0096] In the above formula, λ TarMax For the band corresponding to the effective maximum point of the target, λ TarMin For the band corresponding to the effective minimum point of the target, λ Maximum For the band corresponding to the original maximum point, λ Minimum For the original minimum point, the corresponding band, S inner This represents the internal noise spectrum of the equipment.

[0097] 6. Utilize the stable background spectrum to filter out identical spectral points between real and false targets.

[0098] Due to the irregularity of the target, background information may be included when measuring the target's spectrum. However, when real and false targets are in the same scene, their background spectra are stable. To obtain more realistic spectral information about the target, this invention filters out the same extreme points between real and false targets, thus obtaining the target's true spectral mechanism point data.

[0099]

[0100]

[0101] In the above formula, λ TrueTar For the true target extreme point band, λ TrueTarMax For the true target maximum point band, λ FakeTarMax This is the band representing the false target maximum point.

[0102] 7. Identify real and false targets by recursion in different wavebands.

[0103] No target can exhibit the same quality but different spectra, or the same spectrum but different quality, simultaneously across short, medium, and long bands. In other words, false and true targets will always differ in certain bands. Therefore, depending on the application scenario, we should first use the band with the best recognition capability, and then look for differences in other bands.

[0104] 8. The number of spectral extreme points and amplitude are used to identify real and false targets in a single band.

[0105] Because targets exhibit different behaviors in different spectral bands, the number of their absorption peaks also varies across different bands. Short-wave bands primarily reflect the target's scattering characteristics, while medium- and long-wave bands reflect its radiation characteristics. Therefore, when statistically analyzing spectral extreme points, the bands are divided into three categories: short-wave, medium-wave, and long-wave.

[0106] The energy exchange reaction is represented by the energy spectrum of the external environment incident on the metal or polymer target, consisting of two components: scattering and absorption. If the false target diffuses emission while the true target reflects through a standard specular surface, then the adjacent extreme points of the scattering and radiation from the true and false targets will differ.

[0107] During the daytime, due to sunlight, the number of effective extrema points of false targets at normal temperatures is greater than that of true targets in all bands, and the amplitude of the effective extrema points of false targets is greater than that of true targets. The discrimination formula for a single band is shown below:

[0108]

[0109] In the above formula, λ TrueTar For the band corresponding to the effective extreme point of the true target, λ FakeTar For the effective extreme points of the false target, corresponding to multiple bands, Num(λ) TrueTar ) represents the number of valid extreme points of the true objective, Num(λ) FakeTar ) represents the number of valid extreme points of the false objective, True(λ) TrueTar) represents the spectral DN value corresponding to the effective extreme point of the true target, Fake(λ) FakeTar ) represents the spectral DN value corresponding to the effective extreme point of the false target, Sum(True(λ) TrueTar Sum(Fake(λ)) represents the magnitude of all valid extreme points of the true objective. FakeTar )) represents the amplitude of all extreme points of the false target.

[0110] At night, due to the lack of illumination and low ambient light, the number of effective extrema points for true targets is greater than that for false targets in each band, and the amplitude of the effective extrema points for true targets is greater than that for false targets. The formula for single-band discrimination is shown below:

[0111]

[0112] In the above formula, λ TrueTar For the band corresponding to the effective extreme point of the true target, λ FakeTar For the effective extreme points of the false target, corresponding to multiple bands, Num(λ) TrueTar ) represents the number of valid extreme points of the true objective, Num(λ) FakeTar ) represents the number of valid extreme points of the false objective, True(λ) TrueTar ) represents the spectral DN value corresponding to the effective extreme point of the true target, Fake(λ) FakeTar ) represents the spectral DN value corresponding to the effective extreme point of the false target, Sum(True(λ) TrueTar Sum(Fake(λ)) represents the magnitude of all valid extreme points of the true objective. FakeTar )) represents the amplitude of all extreme points of the false target.

[0113] Understandably, when a beam of infrared light with continuous wavelengths passes through a substance, if the vibrational or rotational frequency of a certain group in the substance's molecules matches the frequency of the infrared light, the molecule absorbs energy and transitions from its original ground state vibrational (rotational) energy level to a higher energy level. After absorbing infrared radiation, the molecule undergoes transitions between vibrational and rotational energy levels, and light at that wavelength is absorbed by the substance. Recording the absorption of infrared light by molecules with an instrument yields an infrared spectrum. Infrared spectra typically use wavelength (λ) or wavenumber (σ) as the abscissa to represent the location of spectral features, and DN value or intensity (A) as the ordinate to represent spectral intensity.

[0114] Emissivity is a physical quantity that characterizes the basic infrared radiation characteristics of an object. It is the ratio of the radiance of the object to that of a standard blackbody at the same temperature. Reflectivity is the percentage of the total radiant energy reflected by the object. Emissivity and reflectivity characterize the selectivity of an object's absorption and reflection features. They are determined by the atomic and molecular structure of the substance and are key characteristic quantities that can effectively distinguish various moving targets.

[0115] When an external electromagnetic wave irradiates a molecule, if the energy of the irradiating electromagnetic wave is equal to the energy difference between the two energy levels of the molecule, the electromagnetic wave of that frequency will be absorbed by the molecule, causing a transition in the corresponding energy level of the molecule. Macroscopically, this manifests as a decrease in the intensity of transmitted light. The equality of electromagnetic wave energy with the energy difference between the two energy levels of the molecule is one of the necessary conditions for a substance to produce an infrared absorption spectrum, which determines the position of the absorption peak.

[0116] Real / false targets, interference, and backgrounds differ in material, temperature, emissivity, reflectivity, and heat capacity. Real targets are composed of different materials and temperatures (from the inside out), and their thermal infrared images exhibit texture. False targets, composed of internal air and external thin-film chemical materials, are homogeneous and at the same temperature (except for localized heat sources), and their thermal infrared images lack texture information.

[0117] The positions of adjacent extreme spectral points corresponding to reflection and absorption differ between real and false targets; that is, the pairs of adjacent extreme points are different. Real targets are intended for actual use; their surface material is machine paint and is very smooth, exhibiting quasi-specular reflection in their reflection spectrum. False targets, on the other hand, have surfaces made of polymer chemical materials, which are rougher than real targets, and their reflection spectrum is primarily diffuse reflection. This invention is based on the fundamental infrared physics principle of different target scattering / radiation patterns. It utilizes the material composition, scattering, and radiation characteristics of real and false targets in different wavelength bands—that is, the extreme points (absorption peaks) corresponding to different wavelength bands—to identify real and false targets.

[0118] The flowchart of the multi-band spectral extreme point real and false target identification method is as follows: Figure 2 As shown, the image association device obtains infrared images and spectral information of true and false targets; wavelet decomposition is performed on the true and false spectra, and the high-frequency information after wavelet decomposition is extracted, which is the radiation and scattering information of the target in different bands; the maximum and minimum points of the true and false target spectra are calculated, and the same extreme points in the true and false target spectra are filtered out, which are the same background information in the true and false targets; the image association device is calibrated to obtain the device noise spectrum, and the extreme points of the noise spectrum with one standard deviation are filtered out, and the remaining extreme points are the effective extreme points of the target; the number of extreme points of the true and false targets in short, medium and long waves are counted respectively, and the bands are fused to obtain the recognition result.

[0119] In addition to the target spectral information, the signals received by the sensors of the spectral association device also include the noise spectrum of the detector itself.

[0120] This invention relates to blackbody calibration experiments conducted at different temperatures on a spectrum correlation device. A blackbody is an idealized object that absorbs all incoming radiation without any reflection or transmission. The process of a blackbody absorbing and emitting electromagnetic waves follows the spectrum, which follows the Planck distribution. Therefore, the high-frequency signals such as "spicules" that appear when the detection device acquires blackbody data are caused by the noise of the detector itself.

[0121] The spectral data used in this invention are broadband infrared data of both real and false targets collected at an altitude of 3 km, with the detector operating at an ambient temperature of approximately -1°C. Therefore, 600 spectral curves were collected for ambient temperatures of -1°C and blackbody temperatures of 30°C, 50°C, and 80°C. Based on this data, the equipment noise spectrum during data acquisition was analyzed, and the calibration spectral curves are shown below. Figure 3 As shown.

[0122] To obtain the noise spectrum of the refrigeration equipment, wavelet transform is required on the blackbody measured at three temperatures. Wavelet transform (WT) is a novel transform analysis method. Its main characteristics are its ability to fully highlight certain features of the problem through transformation, enabling localized analysis of time (space) and frequency. It refines the signal (function) step-by-step through scaling and translation operations, ultimately decomposing it into high-frequency and low-frequency signals. The high-frequency signals represent the detailed parts of the signal, while the low-frequency signals represent the approximate contour of the signal.

[0123] This invention employs Dobermann wavelet (DB wavelet function) for wavelet decomposition. The Dobermann wavelet is primarily used in discrete wavelet transforms. Classification of the Dobermann wavelet is based on the value of the vanishing momentum (A). The smoothness of both the adjustment function (low-pass filtering) and the wavelet function (high-pass filtering) increases with the value of the vanishing momentum (A). This analysis method uses A=4 to perform wavelet decomposition on blackbody calibration data at three temperatures.

[0124] The wavelet decomposition results of the calibration curves at ambient temperature -1℃ and blackbody temperatures of 30℃, 50℃, and 80℃ are as follows: Figures 4-6 As shown.

[0125] A blackbody is an idealized object that absorbs all-wave radiation without any reflection or transmission. The absorption coefficient of a blackbody for electromagnetic waves of any wavelength is 1, and its transmission coefficient is 0. The absorption and emission of electromagnetic waves by a blackbody follows a spectrum, which is distributed according to the Planck distribution. Therefore, the high-frequency signals such as "glitch" that appear when a detection device acquires blackbody data are caused by the noise of the detector itself.

[0126] The high-frequency components of the blackbody spectra after wavelet decomposition at 30℃, 50℃, and 80℃ have been calculated above. Each temperature has 600 data points. Calculate the mean and standard deviation for each of the three temperatures using the following formulas:

[0127]

[0128]

[0129] S inner (λ)=μ inner (λ)+σ inner (λ)

[0130] In the formula, λ is the wavelength, μ inner S is the mean of the noise spectrum. i For the high-frequency portion of the measured data after wavelet decomposition, σ inner S represents the standard deviation of the noise spectrum. inner This is the noise spectrum of the equipment.

[0131] At an ambient temperature of -1℃ and blackbody temperatures of 30℃, 50℃, and 80℃, the average noise level of the equipment is as follows: Figure 7 As shown, the standard deviation curve of the noise spectrum Figure 8 As shown, one standard deviation of the noise spectrum is taken as the noise reference standard, and its noise spectrum range is as follows. Figure 9 As shown:

[0132] Specifically, this invention identifies the true and false target spectral data collected from field tests. Data collection occurred during both daytime and nighttime periods. Due to the alternating day and night temperatures of the target and background in the detection scenario, the target spectra also exhibit a reversal between daytime and nighttime. During the day, due to sunlight, the target spectrum is dominated by the scattering spectrum, with the radiation spectrum being less prominent; this difference is mainly observed in the shortwave band. At night, due to the absence of sunlight and weak ambient light, the target spectrum is dominated by the radiation spectrum, with the scattering spectrum being less prominent; this difference is mainly observed in the mid- and longwave bands. Therefore, the data is divided into daytime and nighttime periods for separate processing and analysis. This invention uses a Boeing 737 commercial airliner as the true target and an inflatable dummy Boeing 737 as the false target for analysis. A comparison of true and false targets during the daytime period is shown below. Figure 10 As shown, the real and fake target pairs during the nighttime period are as follows: Figure 11 As shown.

[0133] Similar to the method used to calculate the equipment noise spectrum, the Dobese wavelet (DB wavelet function) is employed to decompose the measured data into wavelets. The maxima and minima of the high-frequency components after wavelet decomposition are statistically analyzed; these extreme points reflect the target's response intensity in different wavebands. The spectral bands corresponding to all maxima and minima of both real and false targets are then calculated using the following formula.

[0134]

[0135] In the above formula, λ Maximum For the band corresponding to the maximum point, λ Minimum Let S(λ) be the band corresponding to the minimum point, λ be the band, and S(λ) be the spectral DN value corresponding to λ.

[0136] Due to limitations in the manufacturing processes of existing optical components, when detecting the spectrum of a target at long distances, all sensing units in the spectral module of the image-spectrum correlation detection system can receive spectral energy. Therefore, the target spectral sequence obtained by the spectral measurement module may contain a mixed spectrum of the target and the background. Because the background spectrum is relatively stable, after statistically analyzing the extreme points of true and false targets, spectral points with the same extreme points between true and false targets can be filtered out, thereby obtaining more accurate extreme point information about true and false targets.

[0137] Based on the equipment noise spectrum calculated above, and using one standard deviation of the noise spectrum as a reference standard, the extreme points of the true and false target spectral data within the noise spectrum are filtered out to obtain the effective extreme points of the true and false targets.

[0138]

[0139]

[0140] In the above formula, λ TarMax For the band corresponding to the effective maximum point of the target, λ TarMin For the band corresponding to the effective minimum point of the target, λ Maximum For the band corresponding to the original maximum point, λ Minimum For the original minimum point, the corresponding band, S inner This is the internal noise spectrum of the equipment. A schematic diagram showing all the maximum points and noise spectrum regions of both real and false targets during the day is shown below. Figure 12 As shown, the effective maxima points of the real and false targets during the day are as follows: Figure 13 As shown, the effective minimum points of the real and false targets during the day are as follows: Figure 14 As shown; a schematic diagram of all the maxima and noise spectrum regions of real and false targets at night is shown below. Figure 15 As shown, the effective maxima of real and false targets at night are as follows: Figure 16 As shown, the effective minimum points of true and false targets at night are as follows: Figure 17 As shown.

[0141] Because targets exhibit different behaviors in different spectral bands, the number of their absorption peaks also varies across different bands. Short-wave bands primarily reflect the target's scattering characteristics, while medium- and long-wave bands reflect its radiation characteristics. Therefore, when statistically analyzing spectral extreme points, the bands are divided into three categories: short-wave (2-3 μm), medium-wave (3-8 μm), and long-wave (8-14 μm).

[0142] Based on the principle of diffuse emission from a false target and standard specular reflection from a true target, a false target has more effective extreme points in each band during the day than a true target, or the amplitude of the effective extreme points of the false target is greater than that of the true target. The formula for determining a single band is as follows:

[0143]

[0144] In the above formula, λ TrueTarFor the band corresponding to the effective extreme point of the true target, λ FakeTar For the effective extreme points of the false target, corresponding to multiple bands, Num(λ) TrueTar ) represents the number of valid extreme points of the true objective, Num(λ) FakeTar ) represents the number of valid extreme points of the false objective, True(λ) TrueTar ) represents the spectral DN value corresponding to the effective extreme point of the true target, Fake(λ) FakeTar ) represents the spectral DN value corresponding to the effective extreme point of the false target, Sum(True(λ) TrueTar Sum(Fake(λ)) represents the magnitude of all valid extreme points of the true objective. FakeTar )) represents the amplitude of all extreme points of the false target.

[0145] Using the above formula, the effective extreme values ​​of the shortwave, mediumwave, and longwave spectra of true and false targets during the daytime were statistically analyzed, and the recognition rates for different bands are shown in Table 1.

[0146] Table 1. Results of spectral extreme point identification during daytime.

[0147]

[0148]

[0149] At night, due to the lack of illumination and low ambient light, the number of effective extrema points of the true target in each band is greater than that of the false target, or the amplitude of the effective extrema points of the true target is greater than that of the false target. The formula for single-band discrimination is shown below:

[0150]

[0151] In the above formula, λ TrueTar For the band corresponding to the effective extreme point of the true target, λ FakeTar For the effective extreme points of the false target, corresponding to multiple bands, Num(λ) TrueTar ) represents the number of valid extreme points of the true objective, Num(λ) FakeTar ) represents the number of valid extreme points of the false objective, True(λ) TrueTar ) represents the spectral DN value corresponding to the effective extreme point of the true target, Fake(λ) FakeTar ) represents the spectral DN value corresponding to the effective extreme point of the false target, Sum(True(λ) TrueTar Sum(Fake(λ)) represents the magnitude of all valid extreme points of the true objective. FakeTar )) represents the amplitude of all extreme points of the false target.

[0152] Using the above formula, the effective extreme values ​​of the shortwave, mediumwave, and longwave spectra of true and false targets at night were statistically analyzed, and the recognition rates for different bands are shown in Table 2.

[0153] Table 2. Results of spectral extreme point identification during nighttime hours.

[0154] Band range Recognition rate Shortwave (2-3µm) 69.10％ Medium wave (3-8um) 78.32％ Long wavelength (8-14µm) 74.57％

[0155] Since no target can simultaneously exhibit homogeneous but different spectra across short, medium, and long bands—meaning that during the daytime, the number of extreme points for false targets cannot exceed that for true targets across all bands; and during the nighttime, the number of extreme points for true targets cannot exceed that for false targets across all bands—this invention proposes a multi-band recursive fusion method for identifying extreme points of true and false targets. Based on a single-band discrimination formula, the number and amplitude of extreme points for short, medium, and long waves of true and false targets are sequentially determined until a true target is found. If a true target cannot be identified across all three bands, the target is considered a false target.

[0156] The real and fake target recognition rates after multi-band fusion are shown in Table 3:

[0157] Table 3. Multi-band recursive fusion identification results

[0158] Time period band Recognition rate daytime Full band (2-14um) 93.03％ night Full band (2-14um) 91.05％

[0159] Figure 18 This is an architectural diagram of a device for identifying homomorphic but heterogeneous targets at room temperature based on multi-band spectral extreme value features, as provided in an embodiment of the present invention. Figure 18 As shown, it includes:

[0160] The spectral data acquisition unit 1810 is used to acquire spectral data of two targets during the daytime and / or nighttime periods. The spectrum during the daytime period is mainly composed of scattering spectrum, while the spectrum during the nighttime period is mainly composed of radiation spectrum. The two targets have the same shape but different composition, one being a real target and the other a decoy target. The spectral data covers three wavebands: shortwave, medium wave, and long wave.

[0161] Wavelet decomposition unit 1820 is used to perform wavelet decomposition on spectral data to determine multiple extreme points in the high-frequency part; the extreme points include maxima and minima.

[0162] The extreme point filtering unit 1830 is used to filter out extreme points in two target spectral data that are within the noise spectrum range based on a predetermined noise spectrum range, and to filter out extreme points with the same wavelength in the two target extreme points, so as to obtain multiple effective extreme points after filtering the two targets.

[0163] The target recognition unit 1840 is used to judge the true target based on the effective extreme points of the shortwave, mediumwave and longwave bands for spectral data during the daytime. The true target is the one with fewer effective extreme points among two targets in a certain band, or the true target is the one with a relatively smaller sum of amplitudes of all effective extreme points in a certain band, until the data of a certain band can distinguish between true and false targets. Similarly, for spectral data during the nighttime, the true target is judged based on the effective extreme points of the longwave, mediumwave and shortwave bands for spectral data. The true target is the one with more effective extreme points among two targets in a certain band, or the true target is the one with a relatively larger sum of amplitudes of all effective extreme points in a certain band, until the data of a certain band can distinguish between true and false targets.

[0164] The noise spectrum determination unit 1850 is used to acquire spectral data of a blackbody at different ambient temperatures; the high-frequency signal of the blackbody spectral data is caused by noise from the spectral detection device; wavelet decomposition is performed on multiple spectral data at each ambient temperature to obtain the high-frequency part of multiple spectral data; and the mean and standard deviation of multiple spectral data at each ambient temperature are calculated and used as the mean and standard deviation of the noise spectrum at each ambient temperature, and the range of the noise spectrum is determined based on the mean and standard deviation.

[0165] It should be understood that the above-described device is used to execute the methods in the above embodiments. The implementation principle and technical effect of the corresponding program modules in the device are similar to those described in the above methods. The working process of the device can be referred to the corresponding process in the above methods, and will not be repeated here.

[0166] Based on the methods described in the above embodiments, this application provides an electronic device. The device may include at least one memory for storing a program and at least one processor for executing the program stored in the memory. When the program stored in the memory is executed, the processor performs the methods described in the above embodiments.

[0167] Based on the methods in the above embodiments, this application provides a computer-readable storage medium storing a computer program that, when run on a processor, causes the processor to execute the methods in the above embodiments.

[0168] Based on the methods in the above embodiments, this application provides a computer program product that, when run on a processor, causes the processor to execute the methods in the above embodiments.

[0169] It is understood that the processor in the embodiments of this application may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. A general-purpose processor may be a microprocessor or any conventional processor.

[0170] The method steps in the embodiments of this application can be implemented in hardware or by a processor executing software instructions. The software instructions can consist of corresponding software modules, which can be stored in random access memory (RAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, portable hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can reside in an ASIC.

[0171] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0172] It is understood that the various numerical designations used in the embodiments of this application are merely for descriptive convenience and are not intended to limit the scope of the embodiments of this application.

[0173] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for identifying homomorphic but heterogeneous targets at room temperature based on multi-band spectral extreme value features, characterized in that, Includes the following steps: Acquire spectral data of two targets during the daytime and / or nighttime periods. The daytime spectrum is mainly composed of scattering spectra, while the nighttime spectrum is mainly composed of radiation spectra. The two targets have the same shape but different composition; one is a real target and the other is a fake target. The spectral data covers three bands: shortwave, medium wave, and long wave. Wavelet decomposition is performed on the spectral data to determine multiple extreme points in the high-frequency component; the extreme points include maxima and minima. Based on a predetermined noise spectrum range, extreme points in the two target spectral data that fall within the noise spectrum range are filtered out, and extreme points with the same wavelength as the two target extreme points are also filtered out, resulting in multiple effective extreme points after filtering of the two targets. For spectral data during the daytime, the effective extreme points of the shortwave, mediumwave and longwave bands are judged in sequence. The target with fewer effective extreme points among two targets in a certain band is the real target, or the target with a relatively smaller sum of amplitudes of all effective extreme points in a certain band is the real target, until the data of a certain band can distinguish between real and false targets. For spectral data during nighttime, the effective extreme points of the long-wave, medium-wave and short-wave bands are used to make judgments in sequence. The target with more effective extreme points in a certain band is the real target, or the target with a relatively larger sum of amplitudes of all effective extreme points in a certain band is the real target, until the data of a certain band can distinguish between real and false targets.

2. The method according to claim 1, characterized in that, It also includes the following steps: Acquire spectral data of a blackbody at different ambient temperatures; the high-frequency signals of the blackbody spectral data are caused by noise from the spectral detection equipment; Wavelet decomposition was performed on multiple spectral data of the blackbody at each room temperature to obtain the high-frequency components of the multiple spectral data; The mean and standard deviation of multiple spectral data at each room temperature are calculated and used as the mean and standard deviation of the noise spectrum at each room temperature. The range of the noise spectrum is then determined based on the mean and standard deviation.

3. The method according to claim 1, characterized in that, Wavelet decomposition is performed on the spectral data to determine multiple extreme points in the high-frequency component, specifically: When the DN value of the remote sensing image pixel brightness in a certain band after wavelet decomposition is greater than or equal to the DN value of the remote sensing image pixel brightness in other bands in its neighborhood, that band is taken as the band corresponding to the maximum point. When the DN value of a certain band after wavelet decomposition is less than or equal to the DN values ​​of other bands in its neighborhood, that band is taken as the band corresponding to the minimum point.

4. The method according to claim 1, characterized in that, During the daytime, the formula for identifying a single band is as follows: In the above formula, λ TrueTar For the band corresponding to the effective extreme point of the true target, λ FakeTar For the effective extreme points of the false target, corresponding to multiple bands, Num(λ) TrueTar ) represents the number of valid extreme points of the true objective, Num(λ) FakeTar ) represents the number of valid extreme points of the false objective, True(λ) TrueTar ) represents the spectral DN value corresponding to the effective extreme point of the true target, Fake(λ) FakeTar ) represents the spectral DN value corresponding to the effective extreme point of the false target, Sum(True(λ) TrueTar Sum(Fake(λ)) represents the magnitude of all valid extreme points of the true objective. FakeTar )) represents the amplitude of all extreme points of the false target.

5. The method according to claim 4, characterized in that, During nighttime, the formula for determining a single band is as follows:

6. A device for identifying homomorphic but heterogeneous targets at room temperature based on multi-band spectral extreme value features, characterized in that, include: The spectral data acquisition unit is used to acquire spectral data of two targets during the daytime and / or nighttime periods. The spectrum during the daytime period is mainly composed of scattering spectra, while the spectrum during the nighttime period is mainly composed of radiation spectra. The two targets have the same shape but different composition; one is a real target and the other is a fake target. The spectral data covers three bands: shortwave, medium wave, and long wave. The wavelet decomposition unit is used to perform wavelet decomposition on spectral data to determine multiple extreme points in the high-frequency part; the extreme points include maxima and minima. An extreme point filtering unit is used to filter out extreme points in two target spectral data that are within the noise spectrum range based on a predetermined noise spectrum range, and to filter out extreme points with the same wavelength in the two target extreme points, so as to obtain multiple effective extreme points after filtering the two targets. The target recognition unit is used to judge the true target based on the effective extreme points of the shortwave, mediumwave, and longwave bands for spectral data during the daytime. The true target is the one with fewer effective extreme points among two targets in a certain band, or the true target is the one with a relatively smaller sum of amplitudes of all effective extreme points in a certain band, until the data of a certain band can distinguish between true and false targets. Similarly, for spectral data during the nighttime, the unit judges the true target based on the effective extreme points of the longwave, mediumwave, and shortwave bands for spectral data. The true target is the one with more effective extreme points among two targets in a certain band, or the true target is the one with a relatively larger sum of amplitudes of all effective extreme points in a certain band, until the data of a certain band can distinguish between true and false targets.

7. The apparatus according to claim 6, characterized in that, Also includes: The noise spectrum determination unit is used to acquire spectral data of the blackbody at different ambient temperatures; The high-frequency signal in the blackbody spectral data is caused by noise from the spectral detection equipment; Wavelet decomposition was performed on multiple spectral data of the blackbody at each room temperature to obtain the high-frequency components of the multiple spectral data; The mean and standard deviation of multiple spectral data at each room temperature are calculated and used as the mean and standard deviation of the noise spectrum at each room temperature. The range of the noise spectrum is then determined based on the mean and standard deviation.

8. The apparatus according to claim 6, characterized in that, The wavelet decomposition unit performs wavelet decomposition on the spectral data to determine multiple extreme points in the high-frequency part. Specifically, when the DN value of the remote sensing image pixel brightness in a certain band after wavelet decomposition is greater than or equal to the DN value of the remote sensing image pixel brightness in other bands in its neighborhood, the band is taken as the band corresponding to the maximum value point; when the DN value of a certain band after wavelet decomposition is less than or equal to the DN value of other bands in its neighborhood, the band is taken as the band corresponding to the minimum value point.

9. An electronic device, characterized in that, include: At least one memory for storing programs; At least one processor is configured to execute a program stored in the memory, wherein when the program stored in the memory is executed, the processor is configured to perform the method as described in any one of claims 1-5.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is run on the processor, it causes the processor to perform the method as described in any one of claims 1-5.

Images