CN-121074703-B - Mineral geological investigation system based on remote sensing image texture analysis

CN121074703BCN 121074703 BCN121074703 BCN 121074703BCN-121074703-B

Abstract

The invention relates to the technical field of remote sensing geological information processing and discloses a mineral geological investigation system based on remote sensing image texture analysis, which comprises a time sequence base library construction unit, an ecological agent texture channel processing module and a collaborative verification engine module, wherein the time sequence base library construction unit is used for generating a time sequence statistical base line for defining a normal fluctuation range, the ecological agent texture channel processing module is used for purifying normalized vegetation index textures of a current image according to the time sequence statistical base line so as to generate an ecological agent texture abnormality map, the texture channel processing module is used for generating a structural texture abnormality map, and the collaborative verification engine module is used for evaluating the information quality of two channels and dynamically selecting verification logic according to image information entropy before space coupling judgment is executed.

Inventors

LI KAI
CHEN HAOPENG
Suleiman Kanjhaz
ZHANG TAO

Assignees

江西有色地质矿产勘查开发院

Dates

Publication Date: 20260508
Application Date: 20250829

Claims (9)

1. A mineral geological survey system based on remote sensing image texture analysis, the system comprising: A time sequence base line library construction unit which acquires a plurality of history contemporaneous remote sensing images of the investigation region and calculates statistical parameters of normalized vegetation index textures of the history contemporaneous remote sensing images to generate a time sequence statistical base line defining a normal fluctuation range; the ecological agent texture channel processing module is used for calculating a normalized vegetation index diagram of the current remote sensing image, performing texture analysis on the normalized vegetation index diagram, and comparing a texture analysis result with a time sequence statistical baseline to generate an ecological agent texture anomaly diagram only comprising a region with a texture value exceeding a normal fluctuation range; The texture constructing channel processing module is used for acquiring short wave infrared band images insensitive to vegetation and carrying out texture analysis on the images so as to generate a texture constructing abnormal graph; And the cooperative verification engine module is used for receiving the ecological agent texture anomaly map and the construction texture anomaly map, calculating image information entropy values of the two maps, comparing the information entropy values with a reference interval obtained according to multi-sample statistics, and executing strong verification logic when the information entropy values of the two maps are both located in the reference interval, so that an anomaly area which is spatially coupled in the two maps is determined as a mineral exploration target area.
2. The system of claim 1, wherein the collaborative verification engine module performs the strong verification logic by a coupling index determined by a combination of spatial overlap and spatial proximity To determine whether the abnormal region is spatially coupled, the coupling index The calculation rule of (2) is as follows: , wherein, To construct the spatial overlapping area of the anomaly connected clusters in the texture anomaly map and the anomaly connected clusters in the ecological agent texture anomaly map, Is the sum of the areas of two connected clusters, For the minimum edge distance of two communicating clusters, when two communicating clusters have a spatial overlap area, Is 0; for a maximum search distance, Is set to 1.5 times the average radius of influence of a typical geochemical halo or alteration zone within the survey area, And (3) with To meet the requirements of Is not negative, and the cooperative authentication engine module couples the index The region greater than the decision threshold is determined as the spatially coupled anomalous region.
3. The mining geological survey system based on remote sensing image texture analysis of claim 1, wherein the ecological agent texture channel processing module and the structural texture channel processing module both perform texture analysis by using a gray level co-occurrence matrix algorithm, wherein the structural texture channel processing module calculates at least one texture parameter of contrast or heterogeneity of the short-wave infrared band image, and the ecological agent texture channel processing module calculates at least one texture parameter of variance or entropy of the normalized vegetation index map.
4. The mining geological survey system based on remote sensing image texture analysis according to claim 1, wherein the collaborative verification engine module performs weak verification logic, when any information entropy value of two graphs is not located in a reference interval, wherein the weak verification logic takes an abnormal region in an abnormal graph generated by a channel with the information entropy value located in the reference interval as a main target, and searches whether an edge response exists in another channel only in a bandwidth range of the main target.
5. The mining geological survey system based on texture analysis of remote sensing images according to claim 1, wherein the texture channel processing module selects a remote sensing image with a low solar altitude acquired at sunrise or sunset time of the area when the survey area is determined to be a homogeneous lithology coverage area, and generates a texture anomaly map by performing edge detection on the remote sensing image to extract shadow distribution due to weak topography relief.
6. The mineral geological exploration system based on remote sensing image texture analysis of claim 1, further comprising a terrain context modulation module, wherein the terrain context modulation module calculates a terrain humidity index map of an exploration area based on a digital elevation model, and the ecological agent texture channel processing module spatially modulates a normal fluctuation range according to the terrain humidity index map when generating a time sequence statistical baseline, enlarges a limit value of the normal fluctuation range in a region with a high terrain humidity index value, and reduces the limit value of the normal fluctuation range in a region with a low terrain humidity index value.
7. The mineral geological survey system based on remote sensing image texture analysis of claim 1, further comprising a structural permeability detection module, wherein the structural permeability detection module automatically acquires normalized difference water index images of the region at two different moments of soil near saturation and drying after a rainfall event is experienced by the survey region, generates an image representing spatial difference of soil moisture attenuation rate by subtracting the two normalized difference water index images pixel by pixel, and the collaborative verification engine module uses the region with abnormal moisture attenuation rate in the image as a third information layer to perform collaborative verification with the structural texture anomaly map and the ecological agent texture anomaly map to determine a mineral survey target region.
8. The system of claim 1, further comprising a residual information interpretation module that performs a spatial logical exclusive OR operation on the texture anomaly map and the ecological agent texture anomaly map to extract uncoupled anomaly regions that occur only in a single channel, and marks regions derived from the texture channel processing module as potential deep texture regions based on source channels of uncoupled anomaly regions, marks regions derived from the ecological agent texture channel processing module as shallow environmental anomaly regions, and generates the auxiliary decision information.
9. The mining geological survey system based on texture analysis of remote sensing images according to claim 1, wherein the time sequence base line library construction unit acquires all historical images of the same month as the current image in the past ten years of the survey area, and the ecological agent texture channel processing module automatically determines a threshold value for dividing the texture analysis result by adopting a mode based on a historical sample mean value and standard deviation statistics so as to primarily identify an abnormal region and compare the primarily identified abnormal region with a time sequence statistics base line.

Description

Mineral geological investigation system based on remote sensing image texture analysis Technical Field The invention relates to a mineral geological investigation system based on remote sensing image texture analysis, and belongs to the technical field of remote sensing geological information processing. Background In the field of mineral geological exploration service developed by using high technology, remote sensing technology has become a common technical means in early general investigation and target area screening stage due to the characteristics of macroscopicity, rapidness and economy, and is mainly applied to identifying geological structure, lithology abnormality or alteration information related to ore formation by interpreting spectrum or texture information in images. However, in most areas of the earth covered by vegetation, soil and the like, the application of the above technology faces a principle dilemma, and in order to obtain the covered bedrock information, the currently mainstream technical manner generally follows a direct recovery logic, namely, attempts to strip the influence of the surface covering through a complex image algorithm or relying on a high-cost sensor with certain penetration capability, so as to directly observe the geological information under the surface covering, and in the exploration practice, especially in the general investigation stage with strict economic requirements, the method is solidified into a technical trade-off, namely, high investigation cost is input for pursuing the substantivity of the information, which violates the economic principle of early investigation, or the reliability of an interpretation result is reduced due to unavoidable artifacts and uncertainty introduced in the stripping process when general remote sensing data are used. To get rid of this dilemma, the industry is continuously striving to develop more advanced sensors or more sophisticated algorithms, but these efforts are still essentially improved on the original direct reduction concept, failing to jump out of their inherent limitations, and the technical inertia makes the industry long neglect a key possibility, namely the surface covering itself as a exploration obstacle, the texture features of its spatial spread, in fact, a surface ecological response under the long-term and comprehensive effect of the deep geological environment, containing abundant and interpretable indirect information. The response mechanism is established mainly through two ways, namely 1, the geochemical environment of overlying soil can be changed through geological actions such as weathering leaching and the like, element anomalies with stress or selectivity on plant growth are formed, the long-term geochemical pressure can cause observable systematic differences of surface vegetation in aspects such as species constitution, growth density, health condition and the like, and 2, geological structures such as fracture, broken bands and the like related to ore formation can directly influence the hydrological conditions and physical properties of local soil, and further the types and growth vigor of vegetation are regulated and controlled, so that linear or annular ecological anomalies distributed along the structure are formed. Therefore, how to create an information processing mode, the ground surface covering is not tried to be stripped or penetrated, but the texture characteristics of the covering are firstly used as a proxy information channel mutually independent with geological structure information, and the reliable identification of hidden geological anomalies is realized under the condition of only relying on standard and general remote sensing data through collaborative verification of two channel information, so that the invention is the technical problem to be solved. Disclosure of Invention The invention provides a mineral geological investigation system based on remote sensing image texture analysis, which mainly aims to solve the dilemma that cost and reliability are difficult to be compatible due to the fact that a direct reduction logic is fixed when a vegetation coverage remote sensing image is processed in the prior art, and particularly, the problem of high reliability and low cost identification of hidden geological anomalies is realized by using general remote sensing data through an information processing mode of converting ground surface covering information from noise into proxy signals. In order to achieve the above object, the present invention provides a mineral geological survey system based on remote sensing image texture analysis, the system comprising: A time sequence base line library construction unit which acquires a plurality of history contemporaneous remote sensing images of the investigation region and calculates statistical parameters of normalized vegetation index textures of the history contemporaneous remote sensing images to generate a time sequence statis