A hematite abundance extraction method for Gaofen-5 satellite hyperspectral data

Abstract

The present invention belongs to the field of hyperspectral remote sensing technology, and specifically relates to a method for extracting hematite abundance from Gaofen-5 satellite hyperspectral data. The method comprises the following steps: obtaining ground reflectance image data from the Gaofen-5 satellite; extracting band data that preserves spectral extraction characteristics of hematite to obtain a resampled Gaofen-5 satellite hyperspectral data image; resampling a hematite sample standard spectral curve based on the resampled Gaofen-5 satellite hyperspectral data image; obtaining a result image pixel value H1 using a spectral matching method; obtaining a result image pixel value H2 by subtracting the pixel value of the longest band image from the pixel value of the shortest band image in the resampled Gaofen-5 hyperspectral data; and when the image pixel value H2 is less than 0, subtracting H2 from H1 to obtain the hematite abundance value H within the Gaofen-5 satellite hyperspectral data image pixel. The present invention can enhance the accuracy and efficiency of hematite information extraction from Gaofen-5 satellite hyperspectral data.

Description

Technical Field

[0001] The present invention belongs to the technical field of hyperspectral remote sensing, and in particular relates to a hematite abundance extraction method applicable to hyperspectral data of the Gaofen-5 satellite. Background Art

[0002] The Gaofen-5 satellite's hyperspectral data contains a total of 330 bands, resulting in a large amount of data. Processing all of this data requires a long time. Hyperspectral hematite extraction methods primarily include spectral matching, spectral angles, and binary encoding. However, due to the large number of bands, these methods are subject to certain errors when extracting information from the Gaofen-5 satellite data. They are also susceptible to interference factors and band influences. These interference factors and irrelevant bands slow data processing and can lead to misinterpretations of extraction results. For example, riverbeds and mudflats can be mistaken for hematite during spectral matching.

[0003] Most existing hematite extraction methods utilize absorption peaks, but hematite has relatively few and less pronounced absorption peaks. Hematite's characteristic wavelengths primarily lie within the visible light range, limiting the number of usable bands. Given the numerous bands of the Gaofen-5 satellite, extracting hematite data requires streamlining the number of bands, quickly and efficiently removing the influence of various interfering factors and irrelevant bands, and improving band utilization efficiency to adapt to current developments and promote the application and dissemination of Gaofen-5 satellite data.

[0004] The Gaofen-5 satellite is the latest hyperspectral satellite launched by China and has its own unique band data type. Based on its characteristics, this patent urgently needs to develop a hematite extraction method suitable for the Gaofen-5 satellite hyperspectral. Summary of the Invention

[0005] The object of the present invention is to provide a hematite abundance extraction method suitable for Gaofen-5 satellite hyperspectral data. The method is based on the Gaofen-5 satellite hyperspectral satellite data, a new hematite extraction method, streamlines the number of processed bands, and retains the hematite spectral characteristics corresponding to the image pixels, thereby overcoming the shortcomings of the original spectral matching method. According to the spectral characteristics of the interference factors, the influence of the interference factors is removed, and the accuracy and efficiency of hematite information extraction from the Gaofen-5 satellite hyperspectral data are enhanced.

[0006] The technical solution for achieving the purpose of the present invention is as follows:

[0007] A method for extracting hematite abundance from Gaofen-5 satellite hyperspectral data, comprising the following steps:

[0008] Step 1: Obtain ground reflectance image data from the Gaofen-5 satellite hyperspectral data;

[0009] Step 2: extract the band data that maintains the spectral extraction characteristics of hematite, and obtain the resampled Gaofen-5 satellite hyperspectral data image;

[0010] Step 3: resample the standard spectral curve of the hematite sample based on the resampled Gaofen-5 satellite hyperspectral data image;

[0011] Step 4: Use the spectral matching method to match the resampled spectral curve with the resampled hyperspectral data to obtain the resulting image pixel value H1;

[0012] Step 5: Subtract the pixel value of the longest band image from the shortest band image in the resampled hyperspectral image data to obtain the result image pixel value H2;

[0013] Step 6: When the image pixel value H2 is less than 0, H2 is subtracted from H1 to obtain the abundance value H of hematite in the Gaofen-5 satellite hyperspectral data image pixel, and the relative content of hematite in the pixel is obtained.

[0014] The step 1 specifically includes: obtaining high-spectral image data from the Gaofen-5 satellite, preprocessing the data, performing atmospheric correction on the image using an atmospheric correction method, and obtaining satellite ground reflectance image data.

[0015] The wavelength bands for maintaining the spectral extraction characteristics of hematite in step 2 include: 488.74nm, 505.86nm, 540.14nm, 570.03nm, 608.6nm, 672.8nm, 681.39nm, 689.87nm, 741.1nm, 762.79nm, 818.1nm, 856.6nm, 882.27nm, 925.05nm, 959.27nm and 984.94nm.

[0016] The resampled hematite sample standard spectral curve in step 3 is consistent with the spectral range and wavelength spacing of the resampled Gaofen-5 satellite hyperspectral data image, and the number of bands of the Gaofen-5 satellite hyperspectral data image and the hematite standard curve is 16.

[0017] The step 4 is specifically as follows: using a spectral matching method, a spectral matching operation is performed on the resampled hematite sample standard spectral curve and the resampled Gaofen-5 satellite hyperspectral data image, and whether the image pixel contains the target mineral, i.e., hematite, is determined based on the degree of similarity, and the resulting image pixel value H1 is obtained.

[0018] The step 5 is specifically as follows: obtaining the pixel value of the 488.74 nm band image minus the pixel value of the 984.94 nm band image, and recording it as the result image pixel value H2.

[0019] The step 6 is specifically as follows: when the image pixel value H2 is less than 0, the difference between H1 and H2 is recorded as the abundance value H of hematite in the Gaofen-5 satellite hyperspectral data image pixel, and the relative content of hematite in the pixel is obtained.

[0020] The beneficial technical effects of the present invention are:

[0021] 1. The present invention provides a hematite abundance extraction method suitable for Gaofen-5 satellite hyperspectral data. According to the characteristics of Gaofen-5 satellite hyperspectral data, a few bands are extracted and a series of calculations are performed. The method requires a small amount of data and has a fast speed (about 20 times faster than the original speed). It also effectively removes the influence of interfering objects (such as riverbeds, clouds, water bodies, ice surfaces, shadows, etc.), thereby improving the accuracy of information extraction.

[0022] 2. The present invention provides a hematite abundance extraction method suitable for the Gaofen-5 satellite hyperspectral data, which has a good effect and significance for the rapid and efficient processing of hematite information extracted using the Gaofen-5 satellite hyperspectral satellite data. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 This is a diagram showing the abundance of minerals in Example 1 of the present invention. DETAILED DESCRIPTION

[0024] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0025] The present invention provides a method for extracting hematite abundance from Gaofen-5 satellite hyperspectral data, which specifically includes the following steps:

[0026] Step 1: Obtain ground reflectivity image data from the Gaofen-5 satellite

[0027] Obtain hyperspectral image data from the Gaofen-5 satellite, preprocess the data, perform atmospheric correction on the image using atmospheric correction methods, and obtain satellite ground reflectance image data;

[0028] Step 2: Extract the band data that maintains the spectral extraction characteristics of hematite and obtain the resampled Gaofen-5 satellite hyperspectral data image

[0029] The image data was resampled to extract the Gaofen-5 satellite hyperspectral data images with a total of 16 bands near 488.74nm, 505.86nm, 540.14nm, 570.03nm, 608.6nm, 672.8nm, 681.39nm, 689.87nm, 741.1nm, 762.79nm, 818.1nm, 856.6nm, 882.27nm, 925.05nm, 959.27nm, and 984.94nm. Extracting these 16 bands can reduce the number of processed bands, reduce interference from other bands, maintain the spectral extraction characteristics of hematite, and improve the speed and accuracy of data processing.

[0030] Step 3: Resample the resampled hematite sample standard spectral curve based on the resampled Gaofen-5 satellite hyperspectral data image for subsequent operations

[0031] Using the mineral spectral library in the remote sensing data processing software ENVI, the hematite sample standard spectral curve in the spectral library was resampled based on the resampled Gaofen-5 satellite hyperspectral data image to obtain a resampled hematite sample standard spectral curve. The spectral range and wavelength spacing of the resampled hematite sample standard spectral curve were made consistent with the resampled Gaofen-5 satellite hyperspectral data image. The number of bands in the Gaofen-5 satellite hyperspectral data image and the hematite standard curve is 16.

[0032] Step 4: Use the spectral matching method to match the resampled spectral curve with the resampled hyperspectral image data to obtain the resulting image pixel value H1

[0033] Using the spectral matching method, the resampled hematite sample standard spectral curve is matched with the resampled Gaofen-5 satellite hyperspectral data image to determine whether the image pixel contains the target mineral, i.e. hematite, based on the similarity, and the resulting image pixel value H1 is obtained;

[0034] Step 5: Subtract the pixel value of the longest band image from the pixel value of the shortest band image in the resampled GF-5 hyperspectral data to obtain the resulting image pixel value H2

[0035] Obtain the pixel value of the 488.74nm band image minus the pixel value of the 984.94nm band image, and record it as the result image pixel value H2;

[0036] Step 6: When the image pixel value H2 is less than 0, the difference between H1 and H2 is recorded as the abundance value H of hematite in the Gaofen-5 satellite hyperspectral data image pixel to obtain the hematite content in the pixel.

[0037] The value of H represents the abundance of hematite in the pixel of the Gaofen-5 satellite hyperspectral data image. The larger the H value, the greater the abundance of hematite in the pixel, that is, the higher the relative content.

[0038] Example:

[0039] Taking the northern part of Kubai, Tarim, Xinjiang as an example, the above method is used to process the hyperspectral data images of Gaofen-5 to obtain the abundance results of hematite and several other minerals. If only the results with abundance greater than 0.15 are retained, the results will be as follows: Figure 1 As shown, through Figure 1 From the image, we can see that the distribution of various minerals is clear, and the interference of clouds and noise strips is effectively removed.

[0040] The present invention has been described in detail above with reference to the accompanying drawings and embodiments. However, the present invention is not limited to the above embodiments. Various modifications can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention. Any content not described in detail in the present invention may be adapted from existing technologies.

Claims

1. A method for extracting hematite abundance from Gaofen-5 satellite hyperspectral data, characterized in that: The method comprises the following steps: Step 1: Obtain ground reflectance image data from the Gaofen-5 satellite hyperspectral data; Step 2: extracting the band data that maintains the spectral extraction characteristics of hematite to obtain the resampled Gaofen-5 satellite hyperspectral data image; Step 3: Resample the standard spectral curve of the hematite sample based on the resampled Gaofen-5 satellite hyperspectral data image for subsequent operations; Step 4: Using the spectral matching method, the resampled spectral curve is matched with the resampled hyperspectral image data to obtain the result image pixel value H1; Step 5: Subtract the pixel value of the longest band image from the shortest band in the resampled hyperspectral data to obtain the result image pixel value H2; Step 6: When the image pixel value H2 is less than 0, H2 is subtracted from H1 to obtain the abundance value H of hematite in the Gaofen-5 satellite hyperspectral data image pixel, and the relative content of hematite in the pixel is obtained.

2. The hematite abundance extraction method suitable for Gaofen-5 satellite hyperspectral data according to claim 1, characterized in that: The step 1 specifically includes: obtaining high-spectral image data from the Gaofen-5 satellite, preprocessing the data, performing atmospheric correction on the image using an atmospheric correction method, and obtaining satellite ground reflectance image data.

3. The hematite abundance extraction method applicable to the Gaofen-5 satellite hyperspectral data according to claim 1, characterized in that: The wavelength bands for maintaining the spectral extraction characteristics of hematite in step 2 include: 488.74nm, 505.86nm, 540.14nm, 570.03nm, 608.6nm, 672.8nm, 681.39nm, 689.87nm, 741.1nm, 762.79nm, 818.1nm, 856.6nm, 882.27nm, 925.05nm, 959.27nm and 984.94nm.

4. The method for extracting hematite abundance from the Gaofen-5 satellite hyperspectral data according to claim 3, characterized in that: The resampled hematite sample standard spectral curve in step 3 is consistent with the spectral range and wavelength spacing of the resampled Gaofen-5 satellite hyperspectral data image, and the number of bands of the Gaofen-5 satellite hyperspectral data image and the hematite standard curve is 16.

5. The method for extracting hematite abundance from the Gaofen-5 satellite hyperspectral data according to claim 4, characterized in that: The step 4 is specifically as follows: using a spectral matching method, a spectral matching operation is performed on the resampled hematite sample standard spectral curve and the resampled Gaofen-5 satellite hyperspectral data image, and whether the image pixel contains the target mineral, i.e., hematite, is determined based on the degree of similarity, and the resulting image pixel value H1 is obtained.

6. The method for extracting hematite abundance from the Gaofen-5 satellite hyperspectral data according to claim 5, characterized in that: The step 5 is specifically as follows: obtaining the pixel value of the 488.74 nm band image minus the pixel value of the 984.94 nm band image, and recording it as the result image pixel value H2.

7. The method for extracting hematite abundance from the Gaofen-5 satellite hyperspectral data according to claim 6, characterized in that: The step 6 is specifically as follows: when the image pixel value H2 is less than 0, the difference between H1 and H2 is recorded as the abundance value H of hematite in the Gaofen-5 satellite hyperspectral data image pixel, and the hematite content in the pixel is obtained.

Images