Remote sensing prospecting method related to light granite in high-cold and high-altitude area

Abstract

The invention provides a remote sensing prospecting method related to light granite in a high-cold and high-altitude area, and belongs to the technical field of geological remote sensing, and the method comprises the steps: obtaining remote sensing image data of a high-cold and high-altitude research area; geological data of the high-cold and high-altitude research area are collected and analyzed; according to the collected geological data, constructing a remote sensing prospecting model based on light granite in the high-cold and high-altitude research region; identifying and delineating light granite by combining an image enhancement technology and a machine learning computer automatic extraction method; determining mineralization alteration abnormity through the geological environment data, and extracting alteration abnormity information by using GF-5 satellite-borne hyperspectral data; and delineating a prospecting prediction area according to the combination of the position of the light granite and the surrounding alteration abnormal information. Based on the process, the rare metal mineralization favorable area of the high-cold and high-altitude research area is determined. The problems that an existing prospecting method is high in cost, long in consumed time, low in efficiency and the like are solved.

Description

Technical Field

[0001] The invention relates to the field of geological remote sensing technology, in particular to a remote sensing prospecting method related to light-colored granite in high-cold and high-altitude areas. Background Art

[0002] In high-altitude and cold regions, the acquisition of conventional geological information by traditional geological means is inefficient and costly, and can only be sampled locally, which brings a series of challenges to regional large-scale mineral exploration. In high-altitude and cold regions with special landscapes, the ecological environment is relatively fragile, and the investment and long-term use of large-scale traditional prospecting and exploration techniques will inevitably increase the pressure on the ecological environment.

[0003] In recent years, the use of remote sensing to conduct mineral exploration has become a conventional means and hot technology for mineralization prediction. Rare metals have become an important strategic mineral resource for the country, and their mineralization is closely related to light-colored granite. However, the geological work in high-cold and high-altitude areas is relatively low, and it is urgent to carry out rapid positioning and prospecting research on light-colored granite in high-cold and high-altitude landscape areas. However, traditional geological research methods have the disadvantages of high cost, long time consumption and low efficiency. There is no effective method for prospecting granite deposits in high-cold and high-altitude areas. Summary of the invention

[0004] In order to overcome the deficiencies of the prior art, the purpose of the present invention is to provide a remote sensing prospecting method for light-colored granite in high-cold and high-altitude areas, which solves the problems of high cost, long time consumption and low efficiency of current prospecting methods.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] A remote sensing prospecting method for light-colored granite in high-cold and high-altitude areas comprises the following steps:

[0007] Acquire remote sensing image data of high-altitude and cold research areas;

[0008] Collect and analyze geological data in high-altitude and cold study areas;

[0009] According to the collected geological data, a remote sensing prospecting model based on light-colored granite in the high-altitude and cold research area is constructed;

[0010] Combine image enhancement technology and machine learning computer automatic extraction method to identify and delineate light-colored granite;

[0011] Determine the mineralization alteration anomaly through the geological environment data, and extract the alteration anomaly information by using the GF-5 satellite-borne hyperspectral data;

[0012] The prospecting prediction area is delineated based on the location of the light-colored granite and the abnormal alteration information around it.

[0013] Preferably, the acquiring of remote sensing image data of the high-cold and high-altitude research area includes: acquiring multispectral image data and hyperspectral image data of the high-cold and high-altitude research area through a satellite detector, and the spectral types of the remote sensing image data are diverse.

[0014] Preferably, the remote sensing image data is selected from one of Landsat-8, Sentinel-2 and ASTER data, combined with GF-5 hyperspectral data, to ensure that the remote sensing image data has high spectral resolution and medium-high spatial resolution.

[0015] Preferably, the collection and analysis of geological data of the high-cold and high-altitude research area includes: collecting and analyzing the mineralization background, geological structure, stratigraphic distribution, rock type, and typical mineral deposit information of the high-cold and high-altitude research area based on the geological mineralization background and mineralization rock type of the high-cold and high-altitude research area.

[0016] Preferably, before combining image enhancement technology and machine learning computer automatic extraction method to identify and delineate light-colored granite, it also includes: pre-processing the remote sensing image data by means of denoising and atmospheric correction.

[0017] Preferably, the method of combining image enhancement technology and machine learning computer automatic extraction method to identify and delineate light-colored granite includes:

[0018] The image enhancement processing results are obtained by synthesizing the bands of the pre-processed remote sensing image data through four image enhancement processing techniques: optimal index factor, band ratio method, principal component analysis and minimum noise fraction.

[0019] Inputting the image enhancement processing results and geological environment data as training samples into the light-colored granite prospecting model for model training;

[0020] Applying Bayesian optimization to the light-colored granite prospecting model to optimize the model parameters, and then classifying the lithology of the light-colored granite prospecting model by using a support vector machine and a random forest algorithm;

[0021] Inputting the remote sensing image data into the light-colored granite prospecting model, performing feature classification according to the band spectrum information and remote sensing image texture structure, and obtaining feature classification results;

[0022] The characteristic classification result is compared with the spectral information and texture of the light-colored granite to identify and delineate the distribution range of the light-colored granite in the remote sensing image data.

[0023] Preferably, the mineralization alteration anomaly is determined by the geological environment data, and the alteration anomaly information is extracted by using the GF-5 satellite-borne hyperspectral data, including:

[0024] According to the hydrothermal mineralization and alteration type and mineral combination information of the geological environment data, the ore-forming / controlling conditions in the remote sensing image data are determined, and the mineralization and alteration mineral information is extracted using GF-5 hyperspectral satellite remote sensing data.

[0025] Preferably, the prospecting prediction area is delineated based on the location of the light-colored granite and the abnormal alteration information around it, including: quickly locating the favorable rare metal mineralization areas in the high-cold and high-altitude research area by combining the distribution position of the light-colored granite in the remote sensing image data with the extraction position of the abnormal alteration information around it, and conducting field inspections on the determined favorable mineralization areas.

[0026] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0027] (1) The present invention provides a remote sensing prospecting method related to light-colored granite in high-altitude and cold regions. The method mainly adopts medium-high resolution (10-15m) public data with higher resolution, which saves economic costs in research and investigation and improves prospecting efficiency.

[0028] (2) The present invention utilizes image enhancement processing technology and machine learning computer automatic extraction methods for prospecting, performs lithology classification based on a more efficient machine learning computer automatic extraction method, and utilizes GF-5 hyperspectral imaging data to assist in alteration information, thereby achieving double improvements in efficiency and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0030] Figure 1 This is a flow chart of a remote sensing prospecting method for light-colored granite in high-cold and high-altitude areas of the present invention;

[0031] Figure 2 A color composite image of the 1-2-11 combination band in the Sentinel-2 image provided in the first embodiment of the present invention;

[0032] Figure 3 A color composite image of the 3-9-11 / 12 combined bands in the Sentinel-2 image provided in the first embodiment of the present invention;

[0033] Figure 4 A color composite image of a 4-3-2 combined band after combining Sentinel-2 with the PCA method provided in the first embodiment of the present invention;

[0034] Figure 5 This is a color composite image obtained by combining Sentinel-2 and MNF provided in the first embodiment of the present invention, wherein: Figure 5 a is a color composite image of the 3-2-1 combination band, Figure 5 b is the color composite image of the 4-3-2 combination band;

[0035] Figure 6 A schematic diagram of delineating light-colored granite provided in Example 1 of the present invention;

[0036] Figure 7 A detailed distribution map of remote sensing altered minerals in Jiaoxi area extracted using GF-5 hyperspectral image data provided in Example 1 of the present invention;

[0037] Figure 8 This is a schematic diagram for delineating favorable areas for mineral prospecting provided in Example 1 of the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] In order to make the purpose, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0040] Embodiment 1

[0041] like Figure 1 As shown, the present invention provides a remote sensing prospecting method related to light-colored granite in high-cold and high-altitude areas, comprising the following steps:

[0042] Step 100: Acquire remote sensing image data of the high-cold and high-altitude research area;

[0043] Step 200: Collect and analyze geological data of the high-cold and high-altitude research area;

[0044] Step 300: constructing a remote sensing prospecting model based on light-colored granite in the high-cold and high-altitude research area according to the collected geological data;

[0045] Step 400: Identify and delineate light-colored granite by combining image enhancement technology and machine learning computer automatic extraction method;

[0046] Step 500: Determine the mineralization alteration anomaly through the geological environment data, and extract the alteration anomaly information using the GF-5 satellite-borne hyperspectral data;

[0047] Step 600: Delineate the prospecting prediction area based on the location of the light-colored granite and the abnormal alteration information around it.

[0048] Among them, in the above step 100, it specifically includes: obtaining multispectral image data and hyperspectral image data of the high-cold and high-altitude research area through a satellite detector, and the spectral types of the remote sensing image data are diverse. For the remote sensing image data, one of Landsat-8, Sentinel-2 and ASTER data is selected, combined with GF-5 hyperspectral data, to ensure that the remote sensing image data has high spectral resolution and medium-high spatial resolution.

[0049] In the above step 200, specifically including: according to the geological mineralization background and mineralization rock type of the high-altitude and high-cold research area, collecting and analyzing the mineralization background, geological structure, stratigraphic distribution, rock type, typical mineral deposits and other information of the high-altitude and high-cold research area.

[0050] In step 400, before the light-colored granite is identified and delineated by combining the image enhancement technology and the machine learning computer automatic extraction method, the remote sensing image data is preprocessed by means of denoising and atmospheric correction, specifically:

[0051] For the later remote sensing image data processing, the remote sensing image data are first preprocessed, including applying the Fast Line-of-Sight Atmospheric Analysis Using Spectral Hypercube (FLAASH) method to the Landsat-8 data for radiometric calibration and atmospheric correction of the OLI data; resampling the ASTER L1T data to a spatial resolution of 15 meters, performing radiometric calibration, and using the FLAASH (Fast Line-of-Sight Atmospheric Analysis of Hypercubes) module for atmospheric correction; resampling the Sentinel-2A data to a resolution of 15 meters and performing band synthesis.

[0052] However, in step 400, it further specifically includes:

[0053] S401, performing band synthesis on the pre-processed remote sensing image data by using four image enhancement processing technologies, namely, optimal index factor, band ratio method, principal component analysis and minimum noise fraction, to obtain an image enhancement processing result;

[0054] The multispectral remote sensing data of the area to be studied are used to synthesize the bands of the preprocessed remote sensing images using four image enhancement processing techniques to determine the best band combination, including optimal index factor (OIF), band ratio method (RBD), principal component and independent component analysis (PCA), and minimum noise fraction (MNF).

[0055] Table 1 Sentinel-2 data OIF table

[0056] Band Combination OIF Value 1-2-11 3496.053994 1-2-12 3417.273003 1-11-12 3210.773413 1-3-11 3194.951801 1-3-12 3124.357106 2-11-12 3005.472004

[0057] Referring to Table 1, the OIF calculated for the Sentinel-2 data in the region shows that the combination of bands 1-2-11 in the Sentinel-2 image is the best choice for false color synthesis that contains the most information and shows the best lithology. Figure 2 As shown, in the color composite image with the Sentinel-2 image band combination 1-2-11, the light-colored granite appears light purple and the Quaternary snow-capped mountains appear in yellow tones.

[0058] The results obtained using the band ratio method (RBD) are as follows Figure 3 As shown in the figure, it can be found that the color composite image with RGB of Sentinel-23-9-11 / 12 has a better effect. Figure 3 , the light blue area is light granite, snow and lakes are white, and the Shiquan River Group is darker blue;

[0059] The results obtained using principal component analysis and independent component analysis (PCA) are as follows Figure 4 As shown in the figure, in the application process of principal component analysis (PCA), it is usually found that the first four generated principal components contain most of the information, and sometimes this proportion even exceeds 99%. This means that these main components capture most of the variability in the data set. The other principal components that follow contain relatively less information. Even if they carry some specific rock type information, they may be interfered with and affected by noise to a large extent, resulting in this information not being clear and reliable enough. Therefore, in practical applications, the first few principal components are usually focused on and utilized, while those principal components with lower information content and that may be contaminated by noise are ignored. Therefore, only the color synthesis of the first four bands is shown, and the image with better effect is the color composite image of the 4-3-2 combined band after combining Sentinel-2 with PCA, refer to Figure 4 The lavender area is the location of light-colored granite, the Shiquan River Group is the brown-yellow area, and the snow has a green hue.

[0060] In addition, in this example, MNF is used on Sentinel-2 data to significantly enhance the lithology identification in the study area. The image effect obtained by combining the 4-3-2 band of Sentinel-2 data and the 2-3-1 band of Sentinel-2 data is better, and the location of light-colored granite is obvious. Figure 5 , using the 4-3-2 bands of Sentinel-2 data for combination, the rose-red area is light-colored granite, such as Figure 5 As shown in a. The 3-2-1 band of Sentinel-2 data is combined, and the dark blue area is light-colored granite, such as Figure 5 As shown in b.

[0061] S402, inputting the image enhancement processing result and geological environment data as training samples into the light-colored granite prospecting model for model training;

[0062] S403, applying Bayesian optimization to the light-colored granite prospecting model to optimize the model parameters, and then performing lithology classification on the light-colored granite prospecting model by using a support vector machine and a random forest algorithm;

[0063] S404, inputting the remote sensing image data into the light-colored granite prospecting model, performing feature classification according to the band spectrum information and remote sensing image texture structure, and obtaining feature classification results;

[0064] S405: Compare the feature classification result with the spectral information and texture of the light-colored granite to identify and delineate the distribution range of the light-colored granite in the remote sensing image data.

[0065] In the above process, refer to Figure 6 The delineation results provided are based on the results of image enhancement processing and the automatic classification of lithology, combined with geological maps and other information, to screen out interference information in the area and then delineate the light-colored granite.

[0066] In the above step 500, the mineralization alteration anomaly is determined by the geological environment data, and the alteration anomaly information is extracted by using the GF-5 satellite-borne hyperspectral data, including:

[0067] According to the hydrothermal mineralization and alteration type and mineral combination information of the geological environment data, the ore-forming / controlling conditions in the remote sensing image data are determined, and the mineralization and alteration mineral information is extracted using GF-5 hyperspectral satellite remote sensing data.

[0068] The altered mineral information was extracted by using the spectral angle matching method through the preprocessed GF-5 hyperspectral image data.

[0069] In this embodiment, refer to Figure 7The alteration mineral anomaly extraction results are interpreted by using three parts: minimum noise fraction (MNF) transformation, PPI calculation, N-dimensional visualization and end member selection to identify the alteration information of muscovite, sodium mica, polysilicon muscovite, magnesian chlorite, ferrochlore and siderite in the study area.

[0070] In the above step 600, the location of the light-colored granite is combined with the abnormal alteration information around it to delineate the prospecting prediction area, such as Figure 8 As shown, specifically including:

[0071] Based on the research on remote sensing prospecting models of typical mineral deposits, through the comprehensive application of multi-source remote sensing data, multi-scale remote sensing anomaly screening and comprehensive research, the distribution position of the light-colored granite in the remote sensing image data and the extraction position of the surrounding alteration anomaly information are used to quickly locate the favorable areas for rare metal mineralization in the high-cold and high-altitude research areas; finally, the determined prospecting target areas and favorable areas for mineralization need to be tested in the field to further verify the feasibility of the method provided in this embodiment.

[0072] Therefore, the above-mentioned remote sensing prospecting method related to light-colored granite in high-altitude and cold areas is adopted. The image processing method is combined with the machine learning algorithm to objectively and accurately locate the position of the light-colored granite. At the same time, on this basis, hyperspectral images are used to extract the related alteration of light-colored granite, and the favorable prospecting areas are determined according to their distribution patterns, which solves the problems of high cost, time-consuming and low efficiency of current prospecting methods.

[0073] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A remote sensing prospecting method for light-colored granite in high-altitude and cold regions, characterized in that: The following steps are involved: Acquire remote sensing image data of high-altitude and cold research areas; Collect and analyze geological data in high-altitude and cold study areas; According to the collected geological data, a remote sensing prospecting model based on light-colored granite in the high-altitude and cold research area is constructed; Combine image enhancement technology and machine learning computer automatic extraction method to identify and delineate light-colored granite; Determine the mineralization alteration anomaly through the geological environment data, and extract the alteration anomaly information by using the GF-5 satellite-borne hyperspectral data; The prospecting prediction area is delineated based on the location of the light-colored granite and the abnormal alteration information around it.

2. The remote sensing prospecting method for light-colored granite in high-cold and high-altitude areas according to claim 1 is characterized in that: The method of obtaining remote sensing image data of the high-cold and high-altitude research area includes: obtaining multispectral image data and hyperspectral image data of the high-cold and high-altitude research area through satellite detectors, and the spectral types of the remote sensing image data are diverse.

3. The remote sensing prospecting method for light-colored granite in high-cold and high-altitude areas according to claim 2 is characterized in that: The remote sensing image data is selected from one of Landsat-8, Sentinel-2 and ASTER data, combined with GF-5 hyperspectral data, to ensure that the remote sensing image data has high spectral resolution and medium-high spatial resolution.

4. The remote sensing prospecting method for light-colored granite in high-cold and high-altitude areas according to claim 1 is characterized in that: The collection and analysis of geological data in the high-cold and high-altitude research area includes: collecting and analyzing the mineralization background, geological structure, stratigraphic distribution, rock type, and typical mineral deposit information of the high-cold and high-altitude research area based on the geological mineralization background and mineralization rock type of the high-cold and high-altitude research area.

5. The remote sensing prospecting method for light-colored granite in high-cold and high-altitude areas according to claim 1 is characterized in that: Before combining image enhancement technology and machine learning computer automatic extraction method to identify and delineate light-colored granite, it also includes: pre-processing the remote sensing image data by means of denoising and atmospheric correction.

6. The remote sensing prospecting method for light-colored granite in high-cold and high-altitude areas according to claim 5 is characterized in that: The method of combining image enhancement technology and machine learning computer automatic extraction to identify and delineate light-colored granite includes: The image enhancement processing results are obtained by synthesizing the bands of the pre-processed remote sensing image data through four image enhancement processing techniques: optimal index factor, band ratio method, principal component analysis and minimum noise fraction. Inputting the image enhancement processing results and geological environment data as training samples into the light-colored granite prospecting model for model training; Applying Bayesian optimization to the light-colored granite prospecting model to optimize the model parameters, and then classifying the lithology of the light-colored granite prospecting model by using a support vector machine and a random forest algorithm; Inputting the remote sensing image data into the light-colored granite prospecting model, performing feature classification according to the band spectrum information and remote sensing image texture structure, and obtaining feature classification results; The characteristic classification result is compared with the spectral information and texture of the light-colored granite to identify and delineate the distribution range of the light-colored granite in the remote sensing image data.

7. The remote sensing prospecting method for light-colored granite in high-cold and high-altitude areas according to claim 1 is characterized in that: The mineralization alteration anomaly is determined through the geological environment data, and the GF-5 satellite-borne hyperspectral data is used to extract the alteration anomaly information, including: According to the hydrothermal mineralization and alteration type and mineral combination information of the geological environment data, the ore-forming / controlling conditions in the remote sensing image data are determined, and the mineralization and alteration mineral information is extracted using GF-5 hyperspectral satellite remote sensing data.

8. The remote sensing prospecting method for light-colored granite in high-cold and high-altitude areas according to claim 1 is characterized in that: The method of delineating a prospecting prediction area based on the location of the light-colored granite in combination with the abnormal alteration information around it includes: quickly locating favorable areas for rare metal mineralization in the high-altitude and high-cold research area by combining the distribution location of the light-colored granite in the remote sensing image data with the extraction location of the abnormal alteration information around it, and conducting field inspections on the determined favorable areas for mineralization.

Images