CN102645279A - Interference imaging spectrometer hyperspectral data simulation method for lunar-surface minerals - Google Patents

Abstract

The invention discloses a method for simulating hyperspectral data of an interference imaging spectrometer for lunar surface minerals, characterized in that the method for simulating hyperspectral data of lunar surface minerals includes the following steps: S1: using a Gaussian spectral response model to analyze known lunar minerals Resampling of the sample spectral curve; S2: construct pure mixed spectral image data; S3: calculate the signal-to-noise ratio of the real data of the interferometric imaging spectrometer; S4: refer to the noise distribution of the real interferometric imaging spectrometer data, and add noise to the simulated data. The hyperspectral data of the interferometric imaging spectrometer simulated by the present invention can be used for verification and accuracy evaluation of algorithms such as real interferometric imaging spectrometer data endmember extraction, mixed pixel decomposition, and mineral abundance inversion, making up for lunar surface sampling data and Insufficient validation data.

Description

Simulation Method of Hyperspectral Data of Interferometric Imaging Spectrometer for Lunar Surface Minerals

technical field

The invention relates to the technical field of hyperspectral data simulation for deep space exploration, in particular to a method for simulating hyperspectral data of an interference imaging spectrometer for lunar surface minerals.

Background technique

Since the spectrum of a substance is closely related to its properties, sunlight is diffusely reflected after it hits the lunar surface, and different substances will present different reflection spectra. Imaging spectrometers use this principle to combine different reflection spectra with known minerals By comparing the typical multi-spectral sequence images, the type and content information of the detected target minerals can be obtained.

The hyperspectral remote sensing detection of the lunar surface is carried out through the Interference Imaging Spectrometer (IIM) detection equipment. By processing the data of the observed elements, minerals and rocks, we can understand their types, contents and distribution on the lunar surface, and use The results of the detection can draw the distribution map of the whole moon of each element, discover the resource-enriched area on the surface of the moon, and provide data on the distribution of resources for the development and utilization of the moon

In 2007, China launched the Chang'e-1 lunar exploration satellite, which opened the prelude to China's independent exploration of the moon. The Interferometric Imaging Spectrometer (IIM) carried on the Chang'e-1 satellite has acquired hyperspectral data covering 84% of the entire lunar surface, with 32 continuous bands covering the spectral range of 480-960nm, with a spectral resolution of 325.5cm ^-1 , which is planned to be used in Identification and abundance distribution mapping of lunar regolith minerals.

To simulate the hyperspectral data of lunar surface minerals, it is first necessary to understand the basic types and distribution of lunar surface minerals. According to the existing exploration knowledge of the moon (Ouyang Ziyuan, 2005), the mineral types on the lunar surface are relatively simple, mainly composed of plagioclase, pyroxene, olivine, and ilmenite. Among them, plagioclase is the most abundant, up to 70% in the highlands; pyroxene is the lunar surface mineral whose content is second only to plagioclase, and it is distributed in the lunar maria and highlands, up to about 60%; ilmenite Mainly distributed in mare basalts; the abundance range of olivine varies greatly in different regions.

Remote sensing image simulation technology is a technology to obtain simulated images under specific conditions through mathematical and physical calculations on the basis of remote sensing theoretical models and prior knowledge of remote sensing. Hyperspectral data simulation technology is mainly divided into two types according to the difference of ground object mixture model: spectral linear hybrid simulation technology and spectral nonlinear hybrid simulation technology. Spectral linear mixing is mainly aimed at surface mixed surface types, while spectral nonlinear mixing technology is mainly aimed at compact mixed surface types.

Normally, the mixed spectrum of a target can be considered as a linear mixture of the spectra of its end member components (Tong Qingxi, 2006):

$\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

0≤c _i ≤1 (3)

Where N is the number of endmembers, m is the L-dimensional mixed spectrum vector (L is the number of image bands), e _i is the endmember vector, c _i is the proportion of endmember e _i in the mixed spectrum, and n is the error term. The linear mixed model is suitable for spectral decomposition of planar mixed terrain.

The mixture of rocks and minerals belongs to the compact mixture. For the compact mixture of ground features, the incident solar radiation interacts with different ground features many times, resulting in the nonlinear characteristics of the mixed spectrum. The spectral mixing model is a model that describes the relationship between the spectral reflectance of a mixed material after mixing different substances and the spectral reflectance of a single material before mixing. The most widely used nonlinear mixing model is the Hapke spectral mixing model named by scientist Bruce Hapke. Based on a rigorous set of radiative transfer theories.

Evaluation of the accuracy of remote sensing data inversion results is one of the key steps in remote sensing mapping. However, since the sampling of lunar samples by humans was limited to the limited locations of the US Apollo program and the former Soviet Union’s Luna program in the 1950s-1970s, the US and the former Soviet Union’s lunar landing sites were concentrated in the low-latitude areas on the front of the moon, and the collected lunar samples were from There are significant limitations in terms of quantity, geographical distribution, and representativeness of petrochemical composition. In addition, the spatial resolution of the data acquired by the IIM is 200 meters, and the spatial resolution of the sampling points of the lunar samples and the IIM data is very different, which also makes it difficult to use the sampling point analysis data to invert the method and results of the IIM data. Accuracy evaluation.

Due to the advantages of manual controllability and known parameters, simulated data is an effective way to evaluate the accuracy of inversion methods such as IIM data endmember extraction and mineral abundance, and helps to determine the mineral inversion for real IIM data. methods and processes. The mining of IIM data is in the ascendant, and the lunar surface mineral simulation of IIM data has not yet been carried out.

Contents of the invention

(1) Technical problems to be solved

The technical problem to be solved by the present invention is a method for simulating hyperspectral data of interferometric imaging spectrometers for lunar surface minerals, which is used for the verification and accuracy of algorithms such as endmember extraction, mixed pixel decomposition, and mineral abundance inversion of real interferometric imaging spectrometer data. Evaluation work to make up for the lack of lunar surface sampling data and verification data.

(2) Technical solution

In order to achieve the above object, the present invention provides a method for simulating hyperspectral data of an interference imaging spectrometer of lunar surface minerals, the method for simulating hyperspectral data of lunar surface minerals includes the following steps: S1: Utilize a Gaussian spectral response model to analyze known lunar minerals Resampling of the sample spectral curve; S2: construct pure mixed spectral image data; S3: calculate the signal-to-noise ratio of the real data of the interferometric imaging spectrometer; S4: refer to the noise distribution of the real interferometric imaging spectrometer data, and add noise to the simulated data.

其中μ为中心波长，

其中FWHM为响应波谱曲线的半波宽。Preferably, it is characterized in that the Gaussian spectral response model in S1 is:

where μ is the central wavelength,

Where FWHM is the half-wave width of the response spectrum curve.

Preferably, said S2 includes: S21: Simulate the lunar maria, transition zone and highlands of the moon according to the different mixing ratios of plagioclase and ilmenite, and calculate the background mineral image data through the Hapke nonlinear mixed model, so The mixing ratios of plagioclase and ilmenite in the mare, transition zone and highland are: 60%:40%, 80%:20%, 90%:10%; S22: Main minerals used on the lunar surface: plagioclase, The end members of clinopyxene, olivine and ilmenite are mixed with the background mineral image data described in S21, and the lunar surface mineral mixed pixel is calculated through the Hapke nonlinear mixing model; The surface mineral mixed pixels are arranged in an orderly manner to form pure mixed spectral image data.

Preferably, said S3 includes the following steps: S31: Calculate the noise variance after the image is divided into blocks;

S32: noise optimal estimation; S33: calculating the signal-to-noise ratio of the interferometric imaging spectrometer data.

Preferably, in calculating the noise variance after the image is divided into blocks in S31, the image is divided into squares with 6 rows and 8 columns.

(3) Beneficial effects

The hyperspectral data of lunar surface minerals simulated for the Chang'e-1 interferometric imaging spectrometer has the same spectral range, central wavelength, spectral resolution, number of bands and noise distribution with the same trend as the real interferometric imaging spectrometer data, and the mineral mixing ratio and mineral spectral curves are known. The hyperspectral simulation data of lunar surface minerals by the interferometric imaging spectrometer can be used for the verification and accuracy evaluation of algorithms such as end member extraction, mixed pixel decomposition and mineral abundance inversion of real interferometric imaging spectrometer data, making up for lunar surface sampling data and verification data. lack of.

Description of drawings

Fig. 1 is the hyperspectral data simulation flowchart of the interference imaging spectrometer of lunar surface mineral of the present invention;

Fig. 2 is the reflectance spectrum of the moon return sample mineral in the prior art;

Fig. 3 is the moon return sample mineral spectrum sampled to the spectral resolution of the interference imaging spectrometer in the present invention;

Fig. 4 is the schematic diagram of the hyperspectral data of the lunar surface mineral interference imaging spectrometer simulated by the present invention

Fig. 5 is the calculation result of the signal-to-noise ratio of each band of real data of the interference imaging spectrometer of the present invention;

Fig. 6 is the effect diagram after noise is added to the simulated data of the interference imaging spectrometer of the present invention;

Fig. 7 is a diagram of spectrum changes after noise is added to the simulated data of the interference imaging spectrometer of the present invention.

Detailed ways

The specific implementation manners of the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

Fig. 1 is the flow chart of the hyperspectral data simulation of the interference imaging spectrometer of the lunar surface minerals of the present invention, as shown in Fig. 1, the hyperspectral data simulation process of the interference imaging spectrometer of the lunar surface minerals of the present invention is as follows:

S1: Resampling known spectral curves of lunar mineral samples using a Gaussian spectral response model

First, collect the spectra of the samples returned from the moon by the American Apollo and the former Soviet Union Luna programs measured by the RELAB laboratory of Brown University. The main minerals include plagioclase (No.: LS-CMP-011), clinopyroxene (No.: LS-CMP-009 ), olivine (No.: LS-CMP-005), ilmenite (No.: PI-CMP-006), the spectral curves of these minerals are shown in Figure 2.

Using the Gaussian spectral response model for spectral resampling, resampling the mineral spectrum measured by the RELAB laboratory to the spectral resolution and central wavelength of each band of the interference imaging spectrometer (see Table 1 for the technical parameters of the interference imaging spectrometer), the resampled spectrum is as follows Figure 3 shows. Wherein, the Gaussian response function is formula (4), and in order to ensure that the spectral response at the central wavelength is 1, the Gaussian model of the present invention is simplified into formula (5).

$\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\sqrt{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}}\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span>\frac{{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&mu;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; 2</span>}{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

$\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span>\frac{{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&mu;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; 2</span>}{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; )</span>$

$\mathrm{<span\; class="notranslate">\; \&sigma;</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; wxya</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}\sqrt{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; ln</span>}\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

Among them, μ is the central wavelength, FWHM is the half-wave width (Full Widthat Half Maximum) of the response spectrum curve, μ and FWHM can be read from the data header file of the interferometric imaging spectrometer, the integration interval is selected ±3σ, and the responsivity is 99.74%, which can ensure the information No distortion.

Table 1 Main index parameters of the interference imaging spectrometer

Imaging width 25.6km ground resolution 200m (sub-satellite point) imaging area 75°N-75°S (when the sun altitude angle is less than 15°) spectral range 480-960nm Spectral resolution 325.5cm ^-1

number of bands 32 Quantitative level 12bit MTF ≥0.2 (black and white quick view)

S2: Constructing pure hybrid spectral image data

Plagioclase, clinopyroxene and olivine are the most common and the most abundant main minerals in the lunar crust and mantle rocks. One of the remarkable features of the lunar sea basalt is that it contains a large amount of ilmenite, which is the origin of the earth's basalt ilmenite. More than 2 times. In this simulation scheme, plagioclase, clinopyroxene, olivine and ilmenite are selected as the main minerals of the lunar surface for mixing. The main landform units of the moon are highlands and maria. The main feature of the mineral distribution in the highlands and mare is that plagioclase is widely distributed in the highlands, with the highest content. Although the content of ilmenite in the mare is not the highest, it is the background mineral of the mare basalt, which is widely distributed.

S21: According to the different mixing ratios of plagioclase and ilmenite, respectively simulate the lunar maria, transition zone and highlands, and calculate the background mineral image data through the Hapke nonlinear mixture model. The lunar maria, transitional zone and highlands The mixing ratio of plagioclase and ilmenite is: 60%: 40%, 80%: 20%, 90%: 10%;

Fig. 4 is the schematic diagram of the hyperspectral data of the interferometric imaging spectrometer of the simulated lunar surface minerals of the present invention, as shown in Fig. 4, specifically, in S21, the simulated data are divided into lunar maria and transition zone according to the ratio of plagioclase feldspar and ilmenite Simulations were carried out in three regions, the Highland and the Highland. The proportions of plagioclase and ilmenite in the three regions were mixed according to 60%:40%, 80%:20%, and 90%:10%, respectively. The background was calculated using the Hapke nonlinear mixture model Mineral data, the specific calculation steps are as follows:

S211: Convert mineral reflectance spectrum to single scattering albedo:

Mineral reflectance is relative reflectance, firstly, the intermediate value γ needs to be obtained from relative reflectance:

$\mathrm{<span\; class="notranslate">\; \&gamma;</span>}<span\; class="notranslate">\; =</span>\frac{{<span\; class="notranslate">\; [</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&mu;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 4</span>}{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}\mathrm{<span\; class="notranslate">\; \&mu;\&tau;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>}^{\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&mu;</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; \&tau;</span>}}{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 4</span>}{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}\mathrm{<span\; class="notranslate">\; \&mu;\&tau;</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 7</span>}<span\; class="notranslate">\; )</span>$

Wherein, τ is the relative reflectance, μ ₀ =cosi, μ=cose, i and e are the incident angle and the outgoing angle of light, respectively.

The single scattering albedo is calculated by formula (8):

ω＝1-γ ² (8)

S212: Linearly mix the single scattering albedo of each mineral:

The single-scattering albedo of the mixed terrain can be considered as a linear mixture, so the single-scattering albedo of the four selected minerals is linearly mixed according to the mixing scheme set above to obtain the single-scattering albedo of the mixed minerals;

S213: converting the mixed single scattering albedo into a mixed spectral reflectance;

To convert the single scattering albedo of mixed minerals obtained after mixing into the spectral reflectance of mixed minerals, first calculate the intermediate value γ from the mixed single scattering albedo according to formula (9):

$\mathrm{<span\; class="notranslate">\; \&gamma;</span>}<span\; class="notranslate">\; =</span>{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&omega;</span>}<span\; class="notranslate">\; )</span>}^{\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 9</span>}<span\; class="notranslate">\; )</span>$

Then the relative reflectance of the mixed minerals is calculated by formula (10):

$\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&gamma;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; r</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; sample</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; r</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; dard</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&gamma;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&gamma;</span>}{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&gamma;\&mu;</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 10</span>}<span\; class="notranslate">\; )</span>$

Where sample is the reflectance of the sample, and standard is the standard reflectance.

S214: Combining the obtained mixed spectral reflectance into background mineral image data:

The mixed spectral reflectance of plagioclase and ilmenite calculated according to different mixing ratios constitutes each pixel of the mixed spectral image, and the background mineral image data is combined from each pixel of the mixed spectral image.

S22: Use the main minerals on the lunar surface: plagioclase, clinopyroxene, olivine, and ilmenite to mix with the background mineral image data described in S21, and calculate the mixed pixels of the lunar surface minerals through the Hapke nonlinear mixing model ;

In S22, the four mineral endmembers (reflectances of the four pure minerals) are mixed according to the ratio shown on the left edge of Figure 4 and the background mineral image data, and the lunar surface mineral mixed pixel is calculated through the Hapke nonlinear mixing model , forming a mixed mineral image of the lunar surface. For ease of viewing, each pixel consists of 6 rows×8 columns of identical pixels. The calculation process through the Hapke nonlinear mixed model is the same as the above-mentioned calculation process, and will not be repeated here. The mixed spectral reflectance obtained at the end of the calculation constitutes the mixed pixel of lunar surface minerals in this step.

S23: Orderly arrange the mixed pixels of lunar surface minerals in S22 according to the mixing scheme in Fig. 4 to form pure mixed spectral image data.

S3: Calculate the signal-to-noise ratio of the real data of the interferometric imaging spectrometer;

The main operation steps are as follows:

S31: Calculate the noise variance after the image is divided into blocks:

计算得出，其中

为x_i，j，k的估计值，其中， ${\hat{x}}_{i,j,k}={\mathrm{ax}}_{i,j,k-1}+{\mathrm{bx}}_{i,j,k+1}+{\mathrm{cx}}_{p,k}+d,$ $x_{p,k}=\left\{\begin{array}{cc}{x}_{i-1,j,k}& i>1\\ x_{i,j-1,k}& i=1,j>1\end{array}\right.,$ 在 ${\hat{x}}_{i,j,k}={\mathrm{ax}}_{i,j,k-1}+{\mathrm{bx}}_{i,j,k+1}+{\mathrm{cx}}_{p,k}+d$ 中a，b，c，d的值，是通过最小二乘法约束

(i，j)≠(1，1)中的S²为最小值时计算得出的。Divide the image into blocks of w columns and h rows (w=8, h=8 in this scheme), and calculate the noise variance of each block. First calculate the residual of a single pixel, let x _{i, j, k} represent the pixel value of the k-th band in row i, column j, and the residual of a single pixel by the formula:

calculated, where

is the estimated value of xi _{, j, k} , where, ${\hat{x}}_{i,j,k}={\mathrm{ax}}_{i,j,k-1}+{\mathrm{bx}}_{i,j,k+1}+{\mathrm{cx}}_{p,k}+d,$ $x_{p,k}=\left\{\begin{array}{cc}{x}_{i-1,j,k}& i>1\\ x_{i,j-1,k}& i=1,j>1\end{array}\right.,$ exist ${\hat{x}}_{i,j,k}={\mathrm{ax}}_{i,j,k-1}+{\mathrm{bx}}_{i,j,k+1}+{\mathrm{cx}}_{p,k}+d$ The values of a, b, c, and d are constrained by the least squares method

(i, j) ≠ (1, 1) S ² is calculated when it is the minimum value.

M＝ω×h-1计算出每个方格的噪声方差。Finally according to the formula

M=ω×h-1 calculates the noise variance of each square.

S32: Noise optimal estimation

Arrange the noise variance of all squares in each band from small to large, remove the smallest 15% and largest 15% noise variance, and further calculate the signal-to-noise ratio for the remaining 70% noise variance.

S33: Calculate the signal-to-noise ratio

The average value of all pixels in each divided grid is taken as the signal value, the noise variance of each grid represents the noise value, and the ratio of the signal value to the noise value is the signal-to-noise ratio. The signal-to-noise ratio of the entire image is the mean value of the signal-to-noise ratio of the remaining squares after noise optimal estimation.

Fig. 5 is the calculation result of the signal-to-noise ratio of each band of the real data of the interference imaging spectrometer of the present invention. As shown in Fig. 5, the area near the Copernicus impact crater of the real data of the interference imaging spectrometer 2874 bands is the research object, calculated according to the above method Signal-to-noise ratio of interferometric imaging spectrometer data.

S4: Refer to the noise distribution of the real interferometric imaging spectrometer data, and add noise to the simulated data

Due to the influence of noise in the process of instruments and equipment, radiation transmission, photoelectric conversion, etc., it is proposed to add noise effects to the simulated data that are consistent with the noise level distribution of each band in the real data.

The noise of the interference imaging spectrometer is additive noise, and the noise distribution of each band is shown in Figure 5. According to the real noise distribution of the IIM data, random noise with the same noise distribution trend is added to the simulated data of the interferometric imaging spectrometer. The effect after adding is shown in Figure 6 and Figure 7.

The above embodiments are only used to illustrate the present invention, but not to limit the present invention. Those of ordinary skill in the relevant technical field can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, all Equivalent technical solutions also belong to the category of the present invention, and the scope of patent protection of the present invention should be defined by the claims.

Claims (5)

1. the inteference imaging spectrometer high-spectral data analogy method of menology mineral is characterized in that, said menology mineral high-spectral data analogy method may further comprise the steps:

S1: utilize Gauss's spectrum response model that known lunar mineral sample spectra curve is resampled;

S2: make up pure mixed spectra view data;

S3: calculate inteference imaging spectrometer True Data signal to noise ratio (S/N ratio);

S4: distribute with reference to true inteference imaging spectrometer data noise, simulated data is added noise.

2. the inteference imaging spectrometer high-spectral data analogy method of menology mineral as claimed in claim 1; It is characterized in that; Gauss's spectrum response model among the said S1 is:

wherein μ is centre wavelength,

wherein FWHM is wide for the half-wave of response wave spectrum curve.

3. the inteference imaging spectrometer high-spectral data analogy method of menology mineral as claimed in claim 1 is characterized in that said S2 comprises:

S21: respectively lunar maria, zone of transition and the highland of the moon are simulated according to the different of plagioclase and ilmenite blending ratio; Calculate background mineral view data through the non-linear mixture model of Hapke, the plagioclase on said lunar maria, zone of transition and highland and ilmenite blending ratio are: 60%: 40%, 80%: 20%, 90%: 10%;

S22: utilize the menology essential mineral: the end member of plagioclase, clinopyroxene, peridot and ilmenite mixes with the background mineral view data described in the S21, and calculates menology mineral mixed pixel through the non-linear mixture model of Hapke;

S23: the menology mineral mixed pixel among the said S22 is arranged in order, to form pure mixed spectra view data.

4. the inteference imaging spectrometer high-spectral data analogy method of menology mineral as claimed in claim 1 is characterized in that said S3 may further comprise the steps:

S31: calculating noise variance behind the image block;

S32: noise optimal estimation;

S33: the signal to noise ratio (S/N ratio) of calculating the inteference imaging spectrometer data.

5. the inteference imaging spectrometer high-spectral data analogy method of menology mineral as claimed in claim 4 is characterized in that, in the calculating noise variance, said image is divided into the square of 6 row, 8 row behind the image block of S31.

Images