Satellite-borne near-infrared two-dimensional temperature field imager and gravitational wave three-dimensional structure detection method

Abstract

The application discloses a spaceborne near-infrared two-dimensional temperature field imager and a method for detecting the three-dimensional structure of gravitational waves, wherein the imager uses the near-infrared radiation of OH (3-1) band airglow as a light source, acquires the horizontal two-dimensional spatial distribution of the radiation intensity of two spectral lines P1(2) and P1(4) in an imaging mode, and inverses the spatial distribution of the horizontal two-dimensional temperature field on the height layer through the radiation intensity ratio, compared with the existing limb observation spaceborne instrument which can only provide vertical profile information, the application can provide horizontal two-dimensional distribution information, and has important value for the fine structure research of airglow radiation, atmospheric temperature and density and gravitational wave disturbance.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of satellite remote sensing detection of middle and high atmosphere, and particularly relates to a satellite-borne near-infrared two-dimensional temperature field imager and a gravity wave three-dimensional structure detection method. BACKGROUND

[0002] In recent years, the study of the middle and high atmosphere (the atmosphere 50-300 km above the earth) has become a research hotspot in the field of atmosphere and space. As a key dynamic phenomenon in the middle and high atmosphere, the propagation process of gravity waves deeply affects the energy transport, atmospheric composition distribution and evolution of thermodynamic structure, and has important scientific and application values for studying atmospheric coupling mechanism, monitoring and guaranteeing space environment.

[0003] The satellite remote sensing technology for detecting parameters (such as airglow radiation, temperature, etc.) of the middle and high atmosphere can obtain observation data with global range and high spatiotemporal resolution, and has been widely used in scientific research and business application in this field. However, due to the extremely thin middle and high atmosphere tracer gas and the very weak passive light source signal, the current satellite remote sensing technology can only use the limb observation method to realize effective detection by increasing the atmospheric radiation or absorption path length to enhance the signal accumulation intensity. For example, the satellite-borne instruments such as SABER of TIMED satellite, MLS of AURA satellite and OSIRIS of Odin satellite all use similar limb high spectral detection technology. Although this kind of technology can effectively obtain the vertical distribution of atmospheric parameters by the "onion peeling" inversion method, it cannot obtain the horizontal two-dimensional distribution information of atmospheric parameters at a certain height layer. Since the disturbance of gravity waves to the atmosphere has a three-dimensional spatial structure, the vertical profile obtained by the current satellite remote sensing limb observation can only obtain the two-dimensional information of gravity wave parameters in the vertical direction, and the integral path is long, which greatly limits the research on gravity wave propagation information and small-scale disturbance characteristics. SUMMARY

[0004] The present application aims to provide a satellite-borne near-infrared two-dimensional temperature field imager and a gravity wave three-dimensional structure detection method, to obtain the horizontal two-dimensional spatial distribution of the radiation intensity of two spectral lines in an imaging manner, and to invert the spatial distribution of the two-dimensional temperature field at the height layer by the ratio thereof. Meanwhile, the present application provides a gravity wave three-dimensional structure detection method, which uses two satellite-borne near-infrared two-dimensional temperature field imagers to form a front and rear dual field of view cooperative observation, to form a binocular stereo vision imaging detection mode, and to obtain the change amount of the airglow layer centroid height caused by the gravity wave disturbance by means of a geometric reconstruction algorithm, so as to obtain the three-dimensional structure of the gravity wave, thereby solving the problems in the background technology.

[0005] Technical solution: The star-borne near-infrared two-dimensional temperature field imager provided by the application comprises a front light shield, the inner wall of the front light shield is provided with a plurality of light shielding ring groups, the plurality of light shielding ring groups are connected with a front zero-distortion wide-angle lens through a platform connecting fixing frame, the front zero-distortion wide-angle lens is arranged at the end of the front light shield and is used for converging incident light; a double-cement field mirror is arranged at a position 0.5 times the focal length of the front zero-distortion wide-angle lens and is used for converging a divergent light beam; a light splitting prism bin is arranged at a certain distance apart from the double-cement field mirror, a beam splitter in the light splitting prism bin is placed at 45 degrees to the optical axis and is coated with a half-reflective and half-transmissive film on the surface and is used for splitting the light beam into a transmission parallel channel and a reflection vertical channel; the parallel channel is sequentially and at a certain distance apart provided with a first double-cement collimating lens, a first electric filter wheel and a first narrow-band filter thereof, a first imaging lens, a first high-refractive-index thermal compensation lens, a first InGaAs sensor chip and a first scientific-grade camera, wherein the first narrow-band filter is used for acquiring an OH(3-1) band P1(2) spectral line airglow radiation image; the vertical channel is sequentially and at a certain distance apart provided with a second double-cement collimating lens, a second electric filter wheel and a second narrow-band filter thereof, a second imaging lens, a second high-refractive-index thermal compensation lens, a second InGaAs sensor chip and a second scientific-grade camera, wherein the second narrow-band filter is used for acquiring an OH(3-1) band P1(4) spectral line airglow radiation image; all the optical elements in the imager are fixed in a light shielding lens barrel group and are combined into a firm whole through the light shielding lens barrel group.

[0006] Further, all the optical elements and interval distances of the parallel channel and the vertical channel are the same except that the central wavelength and the bandwidth of the first narrow-band filter and the second narrow-band filter are different; the first high-refractive-index thermal compensation lens and the second high-refractive-index thermal compensation lens have a negative thermal expansion coefficient and are used for offsetting the defocus of the optical lens caused by temperature fluctuation; the first InGaAs sensor chip and the second InGaAs sensor chip are respectively located on the focal planes of the first imaging lens and the second imaging lens.

[0007] Further, the imager uses airglow radiation as a light source, utilizes the characteristic relationship that the intensity ratio of the spectral lines of the airglow radiation vibration and rotation energy levels changes regularly with temperature, and inversely calculates the temperature through the intensity of the effective spectral line.

[0008] Further, the relationship between the radiation intensity ratio R of the P1(2) spectral line and the P1(4) spectral line and the rotational temperature T r is used for inversely calculating the two-dimensional temperature field, and the formula is as follows:

[0009] .

[0010] Further, the bandwidth of the first narrow-band filter and the second narrow-band filter is 1-3 nm; the first motorized filter wheel and the second motorized filter wheel further comprise filters for obtaining atmospheric background infrared radiation images.

[0011] Further, the thickness of the first high-refractive-index thermal compensation lens and the second high-refractive-index thermal compensation lens is determined through precise calculation, for offsetting the positive thermal expansion coefficient optical lens defocus caused by temperature fluctuation.

[0012] The gravity wave stereoscopic structure detection method comprises the following steps:

[0013] (1) obtaining and preprocessing airglow image data through the front and rear double-view field cooperative observation based on the star-borne near-infrared two-dimensional temperature field imager;

[0014] (2) the stereoscopic geometric reconstruction of the OH airglow layer centroid height comprises: calculating the matching point epipolar line, calculating the airglow radiation brightness similarity through the normalized cross-correlation matching algorithm, and calculating the airglow layer centroid height through triangular inversion;

[0015] (3) visualizing the result of the three-dimensional inversion.

[0016] Further, the step (1) obtains the airglow image data as follows: the first star-borne near-infrared two-dimensional temperature field imager and the second star-borne near-infrared two-dimensional temperature field imager are symmetrically fixed on the instrument connecting rod and installed on the satellite platform through the load fixing table, the first observation field of view and the second observation field of view are both downward to the earth observation, the field of view angle is 80-100°, and there is 5-10° overlap in the middle for joint calibration; the two imagers keep time synchronization shooting, a certain distance is flown after the first detection, and the second detection is carried out, so that the projection area of the second observation field of view of the first detection and the projection area of the first observation field of view of the second detection form an overlap area of at least 50%.

[0017] Further, the preprocessing is as follows: dark noise is removed to eliminate the background signal caused by equipment noise or environmental factors, discrete bright spots are removed to remove isolated bright spots or artifacts in the image; distorted images are removed by screening images that do not meet the standard to ensure that only high-quality data is used for processing; flat field coefficient correction is performed to correct the brightness unevenness caused by the non-uniformity of the optical system; the inclination of the annular stripe is corrected; the pillow-shaped distortion is corrected by adjusting the radial distortion parameter through the inverse mapping algorithm to restore the true structure of the image; the background spectrum is removed to remove the interference of the background signal.

[0018] Further, the formula for calculating the airglow radiation brightness similarity through the normalized cross-correlation matching algorithm is as follows:

[0019] ;

[0020] wherein I 1 and I 2 are the intensities of the matching points P1 and P2, respectively, and a square window of dimension (2 p +1, 2 p +1) is defined around the points P1 and P2, and are the local windows (2 p +1) 2 the mean of the values on the pixels, ( x , y ) are the pixel positions on the reference image, (d x , d y ) are the displacements of the matching points. k 、 l are the increments of the pixel coordinates, ranging from p to p , used to iterate over all the pixels within the window around the matching points; after computing the NCC values within the range using the sliding window, the peak position of the NCC values is found; the position of the peak corresponds to the region of the image block that is most similar to the reference image block, thus locking the matching point.

[0021] Advantages: compared with the prior art, the present application has the following obvious advantages: (1) the present application provides a spaceborne near-infrared two-dimensional temperature field imager, which uses the near-infrared radiation of OH (3-1) band airglow as a light source, obtains the horizontal two-dimensional spatial distribution of the radiation intensity of P1(2) and P1(4) two spectral lines in an imaging manner, and inverses the spatial distribution of the horizontal two-dimensional temperature field on the height layer through the radiation intensity ratio, compared with the existing spaceborne instrument which can only provide vertical profile information, the present application can provide horizontal two-dimensional distribution information, which has important value for the fine structure research of airglow radiation, atmospheric temperature and density, and gravity wave disturbance. (2) the present application provides a detection method for the three-dimensional structure of gravity waves, with the help of two spaceborne near-infrared two-dimensional temperature field imagers, a binocular stereo vision imaging detection mode is constructed, and with the help of a geometric reconstruction algorithm, the change amount of the airglow layer centroid height caused by gravity wave disturbance can be inversely calculated, so that the three-dimensional structure of the gravity wave is obtained, compared with two-dimensional vertical profile detection or two-dimensional horizontal distribution detection, the detection of three-dimensional structure provides a new idea for the research of gravity waves. (3) the spaceborne near-infrared two-dimensional temperature field imager provided by the present application adopts a split light path design, and the parallel channel and the vertical channel are completely synchronous in the time and space distribution level, compared with the traditional time sequence shooting, the error caused by the time and space displacement of the satellite in orbit is greatly reduced. (4) the spaceborne near-infrared two-dimensional temperature field imager of the present application uses the airglow radiation of OH (3-1) band as a light source, the spectral lines of this band are mainly distributed in the interval of 1500-1560nm, compared with the traditional spaceborne instrument which uses the near-infrared OH (8-3), (6-2) band (760-1000nm band), the radiation intensity is about 50 times higher, so that the exposure time of each detection is greatly reduced, and the image motion problem of the ground remote sensing observation caused by the fast flight of the satellite is greatly improved. (5) the spaceborne near-infrared two-dimensional temperature field imager of the present application proposes to add a high refractive index thermal compensation lens with a negative thermal expansion coefficient between the imaging lens and the InGaAs sensor chip, the thickness of the compensation lens can be determined through calculation, which can offset the defocusing of the optical lens with a positive thermal expansion coefficient caused by temperature fluctuation, and to a certain extent, the clear imaging of the instrument is ensured. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 is the main body optical structure schematic diagram of the spaceborne near-infrared two-dimensional temperature field imager of the present application;

[0023] Figure 2 is the OH (3-1) band airglow characteristic spectral line distribution diagram adopted by the present application;

[0024] Figure 3 is the OH (3-1) band P1(2) and P1(4) spectral line radiation intensity ratio R and rotational temperature T rQuantitative relationship curve of the quantitative relationship between the two parameters;

[0025] Figure 4 Flow chart for two-dimensional temperature field detection of the satellite-borne near-infrared two-dimensional temperature field imager of the present application;

[0026] Figure 5 Schematic diagram of the detection mode of the single satellite-borne near-infrared two-dimensional temperature field imager of the present application;

[0027] Figure 6 Schematic diagram of the main structure of the dual-view satellite-borne near-infrared two-dimensional temperature field imager of the present application;

[0028] Figure 7 Schematic diagram of the binocular stereo vision imaging detection mode constructed by the dual-view satellite-borne near-infrared two-dimensional temperature field imager of the present application;

[0029] Figure 8 Flow chart for gravity wave stereo structure inversion based on different overlapping areas of the present application;

[0030] Figure 9 Schematic diagram of the dual-view epipolar geometry model of the present application;

[0031] Figure 10 Schematic diagram of the normalized cross-correlation matching based on overlapping areas of the present application;

[0032] Figure 11 Schematic diagram of the gravity wave stereo structure inversion instance of the present application;

[0033] In the figure: 1, front light shield, 2, light shield ring group, 3, platform connection fixing frame, 4, front zero distortion wide-angle lens, 5, light shield lens barrel group, 6, double-cement field mirror, 7, light splitting prism bin, 8, beam splitter, 9, first double-cement collimating lens, 10, first narrow-band filter, 11, first electric filter wheel, 12, first imaging lens, 13, first high refractive index thermal compensation lens, 14, first InGaAs sensor chip, 15, first scientific grade camera, 16, second double-cement collimating lens, 17, second narrow-band filter, 18, second electric filter wheel, 19, second imaging lens, 20, second high refractive index thermal compensation lens, 21, second InGaAs sensor chip, 22, second scientific grade camera, 23, load fixing table, 24, instrument connecting rod, 25, first satellite-borne near-infrared two-dimensional temperature field imager, 26, second satellite-borne near-infrared two-dimensional temperature field imager, 27, first observation field of view, 28, second observation field of view. DETAILED DESCRIPTION

[0034] The technical solutions of the present application will be further described below in combination with the drawings.

[0035] AsFigure 1 As shown, the embodiment of the present application provides a spaceborne near-infrared two-dimensional temperature field imager, a front light shield 1 is arranged at the front end, the inner wall of which is assembled with 7-10 light shielding ring groups 2 with decreasing inner diameters from front to back, and the funnel-shaped light collecting angle composed of the light shielding ring groups 2 is the ground observation field angle, and the light outside the field angle will be shielded and reflected to the outside by the light shielding ring groups 2. A front zero-distortion wide-angle lens 4 is arranged close to the end of the front light shield 1, and the two are fastened and connected by using a platform connecting fixing frame 3, and the platform connecting fixing frame 3 is also responsible for fastening and connecting the whole instrument with the satellite platform. A double-cement field mirror 6 is arranged at 0.5 times the focal length after the front zero-distortion wide-angle lens 4, for converging the divergent light beam. Then a beam splitter 8 fastened in a light splitting prism bin 7 is arranged at a certain distance, the beam splitter 8 is placed at 45° with the optical axis, and the surface thereof is coated with a half-reflective and half-transmissive film at a transmittance of 1:1, so as to divide the light beam from the front into two paths, resulting in that the optical system behind the beam splitter 8 is divided into two parts, wherein the light beam transmitted by the beam splitter 8 propagates along the original light path direction, and the subsequent optical elements thereof form a parallel channel, and the light beam reflected by the beam splitter 8 propagates in a direction at 90° with the original light path, and the subsequent optical elements thereof form a vertical channel. In the parallel channel, a first double-cement collimating lens 9, a first electric filter wheel 11 provided with a first narrow-band filter 10, a first imaging lens 12, a first high refractive index thermal compensation lens 13, and a first scientific grade camera 15 provided with a first InGaAs sensor chip 14 are arranged at a certain distance in sequence, wherein the first InGaAs sensor chip 14 is located on the focal plane of the first imaging lens 12. In the vertical channel, a second double-cement collimating lens 16, a second electric filter wheel 18 provided with a second narrow-band filter 17, a second imaging lens 19, a second high refractive index thermal compensation lens 20, and a second scientific grade camera 22 provided with a second InGaAs sensor chip 21 are arranged at a certain distance in sequence, wherein the second InGaAs sensor chip 21 is located on the focal plane of the second imaging lens 19. Except that the optical parameters of the first narrow-band filter 10 and the second narrow-band filter 17 are different, all the other optical elements and interval distances in the parallel channel and the vertical channel are the same. In addition, all the optical elements in the spaceborne near-infrared two-dimensional temperature field imager are fixed in a light shielding lens barrel group 5, and all the components are combined into a firm whole by the light shielding lens barrel group 5.

[0036] As Figure 1As shown, the imaging light path principle of the spaceborne near-infrared two-dimensional temperature field imager of the application is as follows: Since the instrument is generally located on the polar orbit satellite orbit 400-900 km above the earth surface, and the light source detected by the instrument comes from the airglow layer radiation about 85-92 km above the earth surface, therefore, the light radiated from different positions on the remote airglow layer will enter the front light shield 1 from different directions as approximately parallel light, first pass through the cooperation of the front zero-distortion wide-angle lens 4 and the double-cement field lens 6, converge to the rear focal length of the double-cement field lens 6, and the position of this imaging surface needs to be located at the focal length of the first double-cement collimating lens 9 and the second double-cement collimating lens 16. Therefore, in the parallel channel, after the light is transmitted through the beam splitter 8, it is converged by the first double-cement collimating lens 9 into a parallel light beam, enters the first narrow-band optical filter 10 and is filtered into the required infrared light, and then is converged by the first imaging lens 12 to form an airglow infrared radiation image on the first InGaAs sensor chip 14 located at the focal plane position of the first imaging lens 12. Similarly, in the vertical channel, the reflected light forms an airglow infrared radiation image on the second InGaAs sensor chip 21 in the same way.

[0037] Since the temperature environment inside the satellite platform occasionally fluctuates greatly during on-orbit operation, the fluctuation of the temperature causes the change of the refractive index of the optical lens, which makes the final convergence point of the light deviate from the focal plane and causes unclear imaging. Since the thermal expansion coefficients of the optical lenses in the instrument are all positive coefficients, the first high-refractive-index thermal compensation lens 13 and the second high-refractive-index thermal compensation lens 20 with negative thermal expansion coefficients are arranged between the first imaging lens 12 and the first InGaAs sensor chip 14 and between the second imaging lens 19 and the second InGaAs sensor chip 21 respectively, and the thickness of the two compensation lenses is determined through precise calculation, which can exactly offset the defocusing caused by the temperature fluctuation and ensure the final clear imaging of the instrument.

[0038] The basic principle of the spaceborne near-infrared two-dimensional temperature field imager of the application for detecting temperature is to use airglow radiation (O2 molecules, OH molecules, etc.) as a light source, utilize the characteristic relationship that the intensity ratio of the vibration and rotation energy level spectral lines of the airglow radiation changes regularly with temperature, and inversely calculate the temperature by effectively detecting the intensity of the spectral lines. Figure 2As shown, specifically, the instrument adopts the air glow radiation of OH (3-1) band, and the spectral lines of the band are mainly distributed in the range of 1500-1560 nm. Compared with the OH (8-3), (6-2) bands (760-1000 nm), the radiation intensity is about 50 times higher, and the P1(2): 1523.7 nm and P1(4): 1542.8 nm characteristic spectral lines are selected as the target light source. The spectral lines are far apart from each other, the radiation intensity ratio changes rapidly, and are suitable for temperature inversion. In addition, the non-spectral line region near 1521.4 nm is selected as the atmospheric background radiation detection spectrum. As shown in Figure 3 By using the existing Boltzmann distribution, Einstein transition probability and comprehensive analysis of spectral parameters, the P1(2) and P1(4) spectral line radiation intensity ratio R and the quantitative relationship with the rotational temperature T r is as follows:

[0039] (1);

[0040] The formula and the curve provide a theoretical basis for detecting the OH (3-1) band spectral line radiation intensity ratio to invert the temperature.

[0041] As shown in Figure 1 and Figure 2 , the center wavelength of the first narrow-band filter 10 is set to 1523.7 nm, and the bandwidth is 1-3 nm. The P1(2) spectral line air glow infrared radiation image can be obtained in the parallel channel. Similarly, the center wavelength of the second narrow-band filter 17 is set to 1542.8 nm, and the bandwidth is 1-3 nm. The P1(4) spectral line air infrared radiation image can be obtained in the vertical channel. However, due to the existence of background infrared radiation in the atmosphere, the detected image by the first narrow-band filter 10 and the second narrow-band filter 17 is the result of the combined action of the spectral line radiation and the atmospheric background infrared radiation in the spectrum. Therefore, a filter for detecting atmospheric background infrared radiation is arranged in the first motorized filter wheel 11 and the second motorized filter wheel 18, and the center wavelength is set to 1537.0 nm, and the bandwidth is 1-3 nm. After obtaining the P1(2) spectral line air glow infrared radiation image, the filter is switched, and the atmospheric background infrared radiation image is obtained.

[0042] The specific processing flow is as shown in Figure 4As shown in the figure, taking the parallel channel as an example, first, the P1(2) spectral band infrared radiation image is detected by using the filter with a center wavelength of 1523.7 nm, then the first motorized filter wheel 11 is rapidly switched to the filter with a center wavelength of 1537.0 nm, the atmospheric background infrared radiation image is detected under the same exposure parameters, then the two images are subtracted, and the remaining value becomes the gas glow P1(2) spectral line radiation image, which can be considered that the current image is entirely contributed by the gas glow P1(2) spectral line radiation. Similarly, the vertical channel can obtain the gas glow P1(4) spectral line radiation image. At this time, the P1(2) spectral line and P1(4) spectral line intensity ratio value of each corresponding pixel of the gas glow P1(2) spectral line radiation image and the gas glow P1(4) spectral line radiation image is obtained by division, and then the temperature T of each pixel in the image is calculated by using the relationship given by formula (1), to form a T horizontal two-dimensional distribution image, that is, a two-dimensional temperature field. It should be pointed out that, since the beam splitter 8 is used, the parallel channel and the vertical channel are completely synchronous in the spatial and temporal distribution level, and compared with the traditional time sequence shooting, the error caused by the spatial and temporal displacement of the satellite in orbit is greatly reduced. R r r

[0043] When the satellite carries a single satellite-borne near-infrared two-dimensional temperature field imager of the present application, the detection mode is as shown in Figure 5 OH gas glow layer radiation is relatively stable, its height is generally located at 85~90km above the earth, and its thickness is generally 3~5km, when the gravity wave activity occurs at this height, it will cause the periodic change of the state of the atmospheric temperature, density, wind speed, composition and other parameters of the gas glow layer, and the photochemical reaction of the gas glow radiation is extremely sensitive to the change of the state of the atmospheric parameters of the layer, therefore, the gas glow radiation intensity will also present periodic change. When the gas glow radiation enters the field of view angle of the satellite-borne near-infrared two-dimensional temperature field imager, through the optical path principle shown in Figure 1 , the gas glow spectral line radiation image will be formed, and the two-dimensional temperature field is obtained by the process method in Figure 4 . At this time, since the whole system belongs to the traditional passive remote sensing imaging system, the received radiation intensity is the cumulative value of the whole layer radiation of the gas glow layer, and the temperature inverted is the average temperature of the gas glow layer, therefore, it does not have the ability of height detection and resolution, when the gravity wave activity causes the periodic disturbance of the gas glow, only the period, wavelength, propagation direction and other information of the gravity wave can be obtained.

[0044] ​​​In fact, the gravity wave activity can cause obvious disturbance of the atmospheric parameter state and the centroid height of the airglow layer in three-dimensional space, and therefore the present application proposes to use two space-borne near-infrared two-dimensional temperature field imagers to form a double field of view, and to use the binocular stereoscopic imaging principle of observing the same region twice before and after the flight to invert the change of the centroid height of the airglow layer caused by the gravity wave, so as to obtain the stereoscopic structure of the gravity wave.

[0045] As shown in Figure 6 , the center of the load fixing table 23 at the top is fixed with an instrument connecting rod 24, and the first space-borne near-infrared two-dimensional temperature field imager 25 and the second space-borne near-infrared two-dimensional temperature field imager 26 are symmetrically fixed on the instrument connecting rod 24, and the first observation field of view 27 and the second observation field of view 28 are both downward to the ground observation, generally (80°~100°)×(80°~100°), one before and one after, and the middle part has an overlap of 5~10° for joint calibration. The two cameras keep time synchronization shooting.

[0046] As shown in Figure 7 , since the first observation field of view 27, i.e., the observation field of view a, and the second observation field of view 28, i.e., the observation field of view b, have a certain inclination angle, the projection of the airglow layer is two symmetrical trapezoids. After the satellite completes the first detection, it flies a certain distance to implement the second detection, so that the projection area of the observation field of view b in the first detection and the projection area of the observation field of view a in the second detection overlap, and the overlap area accounts for at least 50% of a single projection area. At this time, due to the different observation angles of the observation field of view b in the first detection and the observation field of view a in the second detection, binocular stereoscopic imaging is formed in the overlapping area of the two observations. And since the satellite flies at a very high speed, the interval between the two shootings is generally within 30 seconds, and the minimum period of the gravity wave is generally about 10~15 minutes, so it can be considered that the structure of the gravity wave is basically unchanged during the two shootings, and the stereoscopic structure inversion calculation can be carried out.

[0047] As shown in Figure 8 and Figure 9 , the basic principle of stereoscopic structure inversion is to obtain the depth information of the objects in the image by calculating the parallax of each pixel point combined with the different viewing angles of the cameras. In application, first, the matching points of the same objects in the two images are found through the stereoscopic matching algorithm (normalized cross-correlation algorithm). Then, the accurate position of each point in the three-dimensional space is recovered by calculating the parallax of these matching points combined with the internal and external parameters of the cameras. Finally, the three-dimensional distribution of the gravity wave in the observation area is obtained through the three-dimensional reconstruction of these points, including the height, speed and propagation direction of the wave. Specifically, the binocular vision inversion algorithm process based on the overlapping area of different fields of view is as follows:

[0048] Step 1, preprocessing of the airglow image data, the purpose is to eliminate the optical non-uniformity caused by equipment, to ensure that the image can accurately reflect the atmospheric signal. In order to achieve this goal, image preprocessing usually includes the following key links:

[0049] S1, dark noise removal, to eliminate the background signal caused by equipment noise or environmental factors, which helps to ensure that the image mainly reflects the target signal; For discrete bright spots, remove isolated bright spots or artifacts in the image; Remove distorted images, by screening images that do not meet the standard, ensure that only high-quality data is used for processing.

[0050] S2, flat field coefficient correction, correct the brightness unevenness caused by optical system non-uniformity; Correct the ring stripe inclination; Correct the pillow distortion, restore the true structure of the image by using a specific geometric transformation model.

[0051] S3, remove the background spectrum, remove the interference of the background signal.

[0052] Step 2, three-dimensional geometric reconstruction of the OH airglow layer centroid height, including the following steps:

[0053] (21) In the stereo vision system, the epipolar line calculation is the core of the matching process. By calculating the epipolar line geometry, the search space can be limited when matching points, thereby improving the matching efficiency and accuracy. The epipolar plane is a plane composed of a three-dimensional point, the optical centers of the two cameras, and the projection points of the three-dimensional point in the two cameras.

[0054] As shown in Figure 9 , the planes IM1 and IM2 are the image planes of the optical centers C1 and C2 of the camera lenses, respectively. In order to calculate the three-dimensional position of point P, points P1 and P2 in two images IM1 and IM2 correspond to the projection of the same physical point P. The key is to establish the correspondence between the point pairs, and then use the equations of straight lines C1P1 and C2P2 to obtain the coordinates of the intersection point P. The line C1C2 is called the baseline. Point P2 is located on the line D2 called "epipolar line P3P4", which is the intersection of the epipolar plane and the image plane. The so-called "epipolar constraint" means that the five points P, C1, C2, P1 and P2 define a plane called "epipolar plane".

[0055] The epipolar line is calculated using the fundamental matrix, which maps point P1 in C1 to epipolar line D2 in C2, the formula is:

[0056] (2);

[0057] Where Fis the fundamental matrix, P1 is the projection point in IM1, and D2 is the epipolar line in IM2. The fundamental matrix can be calculated using the camera's intrinsic and extrinsic parameters. Specifically, the geometric relationship between the two cameras can be derived using the camera's calibration parameters, parallax, and baseline distance. Considering that the airglow layer is approximately 7 kilometers thick (85-92 km), we can restrict the search range for matching points to this altitude range. Since the atmospheric signal is relatively thin, this restriction effectively reduces the search space and computational complexity.

[0058] (22) Figure 10 As shown in Figure 2, normalized cross-correlation matching of airglow radiance is performed. This method finds matching points by comparing corresponding pixels in the two images. The goal of normalized cross-correlation is to measure the similarity between two image regions. Specifically, it calculates the relative relationship between the pixel values ​​of two regions, rather than simply the difference in their absolute values.

[0059] (3);

[0060] in I 1 and I 2 are the strengths of matching points P1 and P2, and the dimension (2 p +1,2 p +1) is defined around points P1 and P2, and For the local window (2 p +1) 2 The mean of the pixel values, ( x , y ) is the pixel position on the reference image, (d x , d y ) is the displacement of the matching point. k 、 l The increment of pixel coordinates is in the range of - p arrive p , used to traverse all pixels within a window around the matching point. After calculating the NCC value within the sliding window, the peak position of the NCC value is found. The image block corresponding to the peak position is the area most similar to the reference image block, thus locking the matching point.

[0061] The window size selection directly affects the matching accuracy and computational efficiency. Larger window helps to reduce the noise influence, but increases the computation; smaller window can improve the matching precision, but may be disturbed by noise. Smaller window may produce mismatch in the environment with more noise, so it is necessary to apply filter or interpolation technology to improve the matching accuracy. Through the sub-pixel accuracy interpolation technology, the accuracy of the matching point can be further improved. For example, the bilinear interpolation or cubic interpolation method can be used to estimate the sub-pixel position of the matching point.

[0062] (23) Triangular inversion of the OH layer centroid height. Before calculating the height, it is necessary to determine the parallax of the OH layer radiation point. Through the matching step found in the previous step, the position difference of the matching points in the two images is calculated to obtain the parallax information. The accuracy of parallax positioning directly affects the accuracy of the final OH layer centroid height calculation. Parallax is the horizontal offset of the corresponding points in the two images. By calculating the parallax of each pair of matching points in the image, a parallax map is generated. The larger the parallax, the closer the object is to the camera, and vice versa. The parallax is used to perform triangulation to recover the three-dimensional coordinates using the internal and external parameters of the camera (such as focal length, baseline length, etc.) and the parallax of the matching points. Triangulation uses the camera projection model to map two-dimensional image points to three-dimensional space. In the reconstruction process, errors may be encountered, such as camera calibration errors, matching errors, etc. In order to reduce the influence of errors, an optimization algorithm is used to further improve the reconstruction accuracy.

[0063] Based on the baseline length of the two cameras B , the camera focal length f , and the parallax of the corresponding points in the images taken by the two cameras d , the height of the OH layer can be calculated Z :

[0064] (4);

[0065] This is the core basis for constructing the three-dimensional map of the OH layer centroid height.

[0066] In order to quantify the intensity of the gravity wave at the height of the OH layer, the kinetic energy density of the gravity wave is calculated:

[0067] (5);

[0068] wherein, W is the kinetic energy density of the gravity wave, p is the air density at 87 km height, a is the amplitude of the gravity wave, π is the circumference ratio, T is the period of the gravity wave. The amplitude aThe vertical fluctuation amplitude of the three-dimensional surface map is extracted. The calculation results are compared with the results of other observation methods to verify the effectiveness of the stereoscopic imaging method.

[0069] After the three-dimensional reconstruction is completed, the results need to be displayed through visualization technology:

[0070] As shown in Figure 11 , from the projection area of the first observation result and the projection area of the second observation result, by comparing the spatial coverage of the two, the overlapping area of the two observations can be clearly circled. This overlapping area is the core range of subsequent analysis. With the stereoscopic matching technology described above, the observation data in this area is processed, the feature points at corresponding positions in different fields of view are identified, the matching calculation is combined with the spatial geometric relationship, and then the three-dimensional reconstruction is completed. Finally, the three-dimensional distribution results of the centroid height of the airglow layer and the atmospheric temperature can be obtained. From Figure 11 the three-dimensional distribution map of the centroid height of the airglow layer reconstructed from Figure 11 , it can be seen that the centroid height distribution is concentrated in the height interval of 85-95 km, and with the propagation of gravity waves, the height shows regular periodic fluctuations. This fluctuation clearly reflects the dynamic changes of the airglow layer under the action of gravity waves. From Figure 11 the three-dimensional distribution map of the atmospheric temperature of the airglow layer reconstructed from Figure 11 , it can be seen that the atmospheric temperature distribution is roughly in the range of 180~210K, and shows a gradually decreasing trend with the increase of height. The temperature difference from low to high reaches about 30K, forming a significant vertical temperature gradient.

Claims

1. A method for detecting three-dimensional structure of gravity waves, characterized in that: The following steps are involved: (1) Two airglow image data were obtained and preprocessed by forming a front and rear dual field of view collaborative observation based on the satellite-borne near-infrared two-dimensional temperature field imager; the first satellite-borne near-infrared two-dimensional temperature field imager and the second satellite-borne near-infrared two-dimensional temperature field imager were symmetrically fixed on the instrument connecting rod and installed on the satellite platform through the payload fixing table. The first observation field of view and the second observation field of view both observed downward to the ground, with a field of view angle of 80°~100°, and an overlap of 5~10° in the middle for joint calibration; the two imagers kept the time synchronization shooting, and after the first detection, the satellite flew a certain distance and carried out the second detection, so that the projection area of ​​the second observation field of view of the first detection formed an overlap area of ​​at least 50% with the projection area of ​​the first observation field of view of the second detection, and the time interval between the two shots was 30 Within seconds; wherein, the first satellite-borne near-infrared two-dimensional temperature field imager and the second satellite-borne near-infrared two-dimensional temperature field imager have the same structure, including: a front light shield (1), the inner wall of which is provided with a plurality of light shielding ring groups (2), the plurality of light shielding ring groups (2) are connected to the front zero-distortion wide-angle lens (4) through a platform connection fixing frame (3), the front zero-distortion wide-angle lens (4) is arranged at the end of the front light shield (1), and is used to converge the incident light; a double-glued field mirror (6) is provided at 0.5 times the focal length behind the front zero-distortion wide-angle lens (4) for gathering the divergent light beam; a beam splitter chamber (7) is provided at a certain distance between the double-glued field mirror (6), and the beam splitter (8) in the beam splitter chamber (7) is placed at 45 degrees to the optical axis, and is coated with a semi-reflective and semi-transparent film on the surface, and is used to divide the light beam into a transmitted parallel channel and a reflected vertical channel; in the parallel channel, a first double-glued collimating lens (9), a first electric filter wheel (11) and its first narrow A filter (10), a first imaging lens (12), a first high-refractive-index thermal compensation lens (13), a first InGaAs sensor chip (14) and a first scientific-grade camera (15), wherein the first narrow-band filter (10) is used to obtain an airglow radiation image of the OH (3-1) spectral band P1 (2) spectral line; a second double-glued collimating lens (16), a second electric filter wheel (18) and its second narrow-band filter (17), a second imaging lens (19), a second high-refractive-index thermal compensation lens (20), a second InGaAs sensor chip (21) and a second scientific-grade camera (22) are sequentially arranged at a certain distance in the vertical channel, wherein the second narrow-band filter (17) is used to obtain an airglow radiation image of the OH (3-1) spectral band P1 (4) spectral line; all optical elements in the imager are fixed in a light-shielding lens barrel group (5), and all components are combined into a solid whole through the light-shielding lens barrel group (5); (2) The three-dimensional geometric reconstruction of the OH airglow layer centroid height includes: calculating the epipolar lines of the matching points, calculating the airglow radiance similarity using the normalized cross-correlation matching algorithm, and calculating the airglow layer centroid height using triangular inversion; wherein, the epipolar lines of the matching points are calculated based on two airglow images, the airglow radiance similarity is calculated using the normalized cross-correlation matching algorithm, the matching points are found by comparing the corresponding pixels in the two airglow images, and the airglow layer centroid height is calculated using triangular inversion using the focal length, baseline length, and parallax of the matching points of the two satellite-borne near-infrared two-dimensional temperature field imagers; (3) Visualize the results of three-dimensional inversion.

2. The method for detecting three-dimensional structure of gravity waves according to claim 1, characterized in that: The preprocessing in step (1) is as follows: dark noise removal is performed to eliminate background signals caused by equipment noise or environmental factors, and discrete bright spots are removed to remove isolated bright spots or artifacts in the image; distorted images are removed, and images that do not meet the standards are screened to ensure that only high-quality data are used for processing; Flat field coefficient correction, correcting the uneven brightness caused by the unevenness of the optical system; Corrected the tilt of the ring stripes; Correct pincushion distortion by adjusting radial distortion parameters using the inverse mapping algorithm to restore the true structure of the image; remove the background spectrum to eliminate the interference of background signals.

3. The method for detecting three-dimensional structure of gravity waves according to claim 1, characterized in that: Step (2) The formula for calculating the airglow radiance similarity using the normalized cross-correlation matching algorithm is as follows: ； Among them, I 1 and I 2 are the strengths of matching points P1 and P2, and the dimension (2 p +1,2 p +1) is defined around points P1 and P2, and For the local window (2 p +1) 2 The mean of the pixel values, ( x , y ) is the pixel position on the reference image, (d x , d y ) is the displacement of the matching point; k 、 l The increment of pixel coordinates is in the range of - p arrive p , used to traverse all pixels in the window around the matching point; after using the sliding window to calculate the NCC value within the range, find the peak position of the NCC value; the image block corresponding to the peak position is the area most similar to the reference image block, thereby locking the matching point.

4. The method for detecting three-dimensional structure of gravity waves according to claim 1, wherein: Except for the central wavelengths and bandwidths of the first narrowband filter (10) and the second narrowband filter (17), all optical elements and spacing distances of the parallel channel and the perpendicular channel are the same; wherein the first high-refractive index thermal compensation lens (13) and the second high-refractive index thermal compensation lens (20) have negative thermal expansion coefficients and are used to offset the defocusing of the optical lens caused by temperature fluctuations; the first InGaAs sensor chip (14) and the second InGaAs sensor chip (21) are respectively located on the focal planes of the first imaging lens (12) and the second imaging lens (19).

5. The method for detecting three-dimensional structure of gravity waves according to claim 1, characterized in that: The imager uses airglow radiation as a light source, and utilizes the characteristic relationship that the ratio of the intensity of the airglow radiation vibration and rotational energy level spectral lines changes regularly with temperature, and inverts the temperature by effectively detecting the intensity of the spectral lines.

6. The method for detecting three-dimensional structure of gravity waves according to claim 1, characterized in that: The radiation intensity ratio R of the P1(2) spectral line to the P1(4) spectral line and the rotation temperature T r The relationship is used to invert the two-dimensional temperature field, and the formula is as follows: 。 7. The method for detecting three-dimensional structure of gravity waves according to claim 1, characterized in that: The bandwidth of the first narrowband filter (10) and the second narrowband filter (17) is 1-3 nm; the first electric filter wheel (11) and the second electric filter wheel (18) further include filters for acquiring an atmospheric background infrared radiation image.

8. The method for detecting three-dimensional structure of gravity waves according to claim 1, characterized in that: The thicknesses of the first high-refractive-index thermal compensation lens (13) and the second high-refractive-index thermal compensation lens (20) are determined by calculation and are used to offset the defocusing of the positive thermal expansion coefficient optical lens caused by temperature fluctuations.

Images