Geological exploration method integrated with remote sensing image processing technology

Abstract

The invention relates to the technical field of image processing, in particular to a geological exploration method of a comprehensive remote sensing image processing technology, which comprises the following steps of: obtaining a plurality of suspected landslide areas in a geological remote sensing image at a real-time stage; according to the geological whitening degree, the geological position deviation degree and the gray scale difference condition of the pixel points in each suspected landslide area, the surface texture discrimination degree of each suspected landslide area is obtained; acquiring landslide deformation feature conformity of each suspected landslide area according to the distribution condition of height data of pixel points in the geological remote sensing image of each suspected landslide area in the real-time stage and the geological remote sensing image of the adjacent sampling stage; obtaining a landslide hidden danger trend index of each suspected landslide area according to the landslide deformation feature conformity and the landslide surface quality presentation degree; and exploring a landslide hidden danger area in the geological remote sensing image in the real-time stage based on the landslide hidden danger trend index. According to the invention, the accuracy of identifying the landslide hidden danger area in the remote sensing image is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and particularly to a geological exploration method integrating remote sensing image processing technology. Background Art

[0002] Landslides are the most frequent and most damaging geological natural disasters globally. Geological warning information will be sent before a landslide occurs, such as unstable deformation of the geology, resulting in uneven surface stress. If rainfall or engineering construction occurs in the landslide-prone area, a landslide disaster is likely to occur. Therefore, by using remote sensing images to explore and monitor the landslide-prone area based on landslide characteristics, the timeliness of people's response to landslides can be effectively improved, and the disaster losses caused by landslides can be minimized. In the traditional process of identifying landslide-prone areas based on optical remote sensing images, unmanned aerial vehicles are used to collect remote sensing images of the monitored mountain area, and a trained neural network model is used to identify each landslide-prone area in the remote sensing image. In the existing neural network training and identification process of mountain remote sensing images, the identification of landslide-prone areas mainly relies on the similarity matching of collapsed cracks within the area. However, due to the widely different mountain contours in the remote sensing image of the actual scene and the complex deep edge morphology inside, the matching difficulty between the existing neural network crack training set and the crack edges in the actual remote sensing image is relatively large, resulting in inaccurate identification of the landslide-prone areas in the remote sensing image. Summary of the Invention

[0003] To solve the above problems, the present invention provides a geological exploration method integrating remote sensing image processing technology, and the method includes: Obtaining geological remote sensing images of several acquisition stages, and height data of each pixel point in the geological remote sensing images of each acquisition stage; Denoting the geological remote sensing image of the last acquisition stage as the geological remote sensing image of the real-time stage; according to the difference situation of the height data of each pixel point in the geological remote sensing image of the real-time stage and its corresponding pixel points in the geological remote sensing images of other acquisition stages, obtaining the geological position offset degree of each pixel point in the geological remote sensing image of the real-time stage; clustering the geological remote sensing image of the real-time stage according to the geological position offset degree of the pixel points to obtain several suspected landslide areas in the geological remote sensing image of the real-time stage; according to the gray distribution situation of the pixel points in each suspected landslide area, obtaining the geological whitening degree of each suspected landslide area; according to the geological whitening degree, the geological position offset degree, and the gray difference situation of the pixel points in each suspected landslide area, obtaining the surface geology differentiation degree of each suspected landslide area; According to the surface texture differentiation degree of each suspected landslide area in the geological remote sensing images at different sampling stages, obtain the landslide surface texture presentation degree of each suspected landslide area; according to the distribution of the height data of the pixel points in the geological remote sensing image at the real-time stage and the geological remote sensing images at its adjacent sampling stages of each suspected landslide area, obtain the landslide deformation characteristic compliance degree of each suspected landslide area; according to the landslide deformation characteristic compliance degree and the landslide surface texture presentation degree, obtain the landslide hazard tendency index of each suspected landslide area; Based on the landslide hazard tendency index, explore the landslide hazard areas in the geological remote sensing image at the real-time stage.

[0004] Preferably, the method for obtaining the geological position offset degree of each pixel point in the geological remote sensing image at the real-time stage according to the difference in the height data of each pixel point in the geological remote sensing image at the real-time stage and its height data in the geological remote sensing images at other acquisition stages includes the following specific method: For the th pixel point in the geological remote sensing image at the real-time stage and any one of the geological remote sensing images other than the geological remote sensing image at the real-time stage, record the absolute value of the difference between the height data of the th pixel point in the geological remote sensing image at the real-time stage and the height data of the th pixel point in the any one of the geological remote sensing images other than the geological remote sensing image at the real-time stage as the height difference value of the th pixel point in the any one of the geological remote sensing images other than the geological remote sensing image at the real-time stage; take the average value of the height difference values of the th pixel point in all the geological remote sensing images other than the geological remote sensing image at the real-time stage as the geological position offset degree of the th pixel point in the geological remote sensing image at the real-time stage.

[0005] Preferably, the method for clustering the geological remote sensing image at the real-time stage according to the geological position offset degree of the pixel points to obtain several suspected landslide areas in the geological remote sensing image at the real-time stage includes the following specific method: Input the geological position offset degrees of all pixel points in the geological remote sensing image at the real-time stage into the DBSCAN density clustering algorithm to cluster all pixel points in the geological remote sensing image at the real-time stage to obtain several clustering clusters; in the geological remote sensing image at the real-time stage, record the area formed by all pixel points in each clustering cluster as a suspected landslide area.

[0006] Preferably, the method for obtaining the geological whitening degree of each suspected landslide area according to the gray distribution of the pixel points in each suspected landslide area includes the following specific method: For the A suspected landslide area, the mean value of the gray values of all pixel points in the suspected landslide area is subtracted from the minimum value of the gray values of all pixel points in the suspected landslide area, and the difference is denoted as the first difference; the maximum value of the gray values of all pixel points in the suspected landslide area is subtracted from the minimum value of the gray values of all pixel points in the suspected landslide area, and the difference is denoted as the first extreme value; the ratio of the first difference to the first extreme value is used as the geological whiteness degree of the

[0007] suspected landslide area. Preferably, the method for obtaining the surface quality differentiation degree of each suspected landslide area according to the geological whiteness degree, the geological position deviation degree, and the gray value difference of the pixel points in each suspected landslide area includes the following specific method: The absolute value of the difference between the gray value of the pixel point in the suspected landslide area and the mean value of the gray values of all pixel points in the suspected landslide area is denoted as the gray value difference of the pixel point; the mean value of the gray value differences of all pixel points in the suspected landslide area is denoted as the gray value difference of the suspected landslide area; the normalization value of the product of the gray value difference of the suspected landslide area, the geological whiteness degree of the suspected landslide area, and the mean value of the geological position deviation degrees of all pixel points in the suspected landslide area is used as the surface quality differentiation degree of the suspected landslide area.

[0008] Preferably, the method for obtaining the landslide surface quality presentation degree of each suspected landslide area according to the surface quality differentiation degree of each suspected landslide area in the geological remote sensing images at different sampling stages includes the following specific method: Taking the serial number of the sampling stage as the abscissa and the surface quality differentiation degree as the ordinate to construct a two-dimensional coordinate system, inputting the surface quality differentiation degrees of the suspected landslide area in the geological remote sensing images at all sampling stages into the two-dimensional coordinate system, and using the least squares method for curve fitting to obtain the surface quality differentiation change curve of the suspected landslide area; Using the STL trend term decomposition method to decompose the surface quality differentiation change curve of the suspected landslide area to obtain the trend term curve of the suspected landslide area; Denote the mean value of the slopes of all data points on the trend item curve of the th suspected landslide area as the mean trend item slope of the th suspected landslide area; Denote the ratio between the mean trend item slope of the th suspected landslide area and the maximum value of the mean trend item slopes of all suspected landslide areas as the trend item slope ratio; Multiply the trend item slope ratio by the surface quality differentiation degree of the th suspected landslide area in the geological remote sensing image in the real-time stage as the landslide surface quality presentation degree of the th suspected landslide area.

[0009] Preferably, the method for obtaining the landslide deformation feature compliance degree of each suspected landslide area according to the distribution of the height data of the pixel points in the geological remote sensing image of each suspected landslide area in the real-time stage and the geological remote sensing image of its adjacent sampling stage includes the following specific methods: Denote the geological remote sensing image of the previous acquisition stage of the geological remote sensing image in the real-time stage as the geological remote sensing image in the comparison stage; According to the distribution difference of the height data of the pixel points in the geological remote sensing image of the th suspected landslide area in the real-time stage and the geological remote sensing image in the comparison stage, obtain the deformation subsidence ratio of the target pixel points of the th suspected landslide area; Denote the difference between the maximum value of the height data of all pixel points in the th suspected landslide area and the height data of the target pixel points of the th suspected landslide area as the deformation height difference. Multiply the deformation subsidence ratio of the target pixel points of the th suspected landslide area by the reciprocal of the deformation height difference as the landslide deformation feature compliance degree of the th suspected landslide area.

[0010] Preferably, the method for obtaining the deformation subsidence ratio of the target pixel points of the th suspected landslide area according to the distribution difference of the height data of the pixel points in the geological remote sensing image of the th suspected landslide area in the real-time stage and the geological remote sensing image in the comparison stage includes the following specific methods: In the th suspected landslide area in the geological remote sensing image in the real-time stage, denote the pixel point with the maximum height data as the height maximum value pixel point of the th suspected landslide area; Denote the pixel point with the minimum height data as the height minimum value pixel point of the th suspected landslide area; The straight-line connection between the pixel point with the maximum height value and the pixel point with the minimum height value in the th suspected landslide area is denoted as the soil flow direction straight line of the th suspected landslide area; For the th pixel point on the soil flow direction straight line of the th suspected landslide area, use the line passing through the th pixel point and perpendicular to the soil flow direction straight line of the th suspected landslide area to divide the th suspected landslide area into an upper area and a lower area; Denote the average value of the height data of all pixel points in the lower area of the th suspected landslide area in the geological remote sensing image at the real-time stage as the first height average value; denote the average value of the height data of all pixel points in the lower area of the th suspected landslide area in the geological remote sensing image at the comparison stage as the second height average value; denote the absolute value of the difference between the first height average value and the second height average value as the convexity degree of the lower area of the th pixel point; Denote the average value of the height data of all pixel points in the upper area of the th suspected landslide area in the geological remote sensing image at the real-time stage as the third height average value; denote the average value of the height data of all pixel points in the upper area of the th suspected landslide area in the geological remote sensing image at the comparison stage as the fourth height average value; denote the absolute value of the difference between the third height average value and the fourth height average value as the concavity degree of the upper area of the th pixel point; Denote the ratio of the concavity degree of the upper area of the th pixel point to the convexity degree of the lower area of the th pixel point as the deformation convexity ratio of the th pixel point; Among all pixel points on the soil flow direction straight line of the th suspected landslide area, denote the pixel point corresponding to the maximum deformation convexity ratio as the target pixel point of the th suspected landslide area.

[0011] Preferably, the specific method for obtaining the landslide hazard tendency index for each suspected landslide area according to the landslide deformation feature conformity and the landslide surface quality presentation degree is as follows: Multiply the landslide deformation feature conformity of the th suspected landslide area by the landslide surface quality presentation degree of the th suspected landslide area as the The degree of landslide hazard reflection in a suspected landslide area; The normalized value of the ratio between the degree of landslide hazard reflection in the th suspected landslide area and the average value of the degrees of landslide hazard reflection in all suspected landslide areas is used as the landslide hazard tendency index for the

[0012] Preferably, the exploration of the landslide hazard area in the geological remote sensing image in the real - time stage based on the landslide hazard tendency index includes the following specific method: Preset two threshold parameters and . For any suspected landslide area, if the landslide hazard tendency index of the any suspected landslide area is less than the threshold parameter , mark the any suspected landslide area as a low - landslide - risk area; if the landslide hazard tendency index of the any suspected landslide area is greater than the threshold parameter , mark the any suspected landslide area as a high - landslide - risk area.

[0013] The beneficial effects of the technical solution of the present invention are as follows: The present invention clusters the geological remote sensing image in the real - time stage according to the geological position offset degree of pixel points, and obtains several suspected landslide areas in the geological remote sensing image in the real - time stage; obtains the geological whitening degree of each suspected landslide area; obtains the surface geology differentiation degree of each suspected landslide area according to the geological whitening degree, the geological position offset degree, and the gray - scale difference situation of pixel points in each suspected landslide area; obtains the landslide surface geology presentation degree of each suspected landslide area; obtains the landslide deformation feature conformity degree of each suspected landslide area according to the distribution of height data of pixel points in the geological remote sensing image of each suspected landslide area in the real - time stage and its adjacent sampling stage; obtains the landslide hazard tendency index of each suspected landslide area according to the landslide deformation feature conformity degree and the landslide surface geology presentation degree; explores the landslide hazard area in the geological remote sensing image in the real - time stage based on the landslide hazard tendency index; thereby, it can combine the performance characteristics of landslides in geological remote sensing images and stress deformation characteristics to obtain a more accurate landslide hazard area, and further improve the accuracy of identifying landslide hazard areas in remote sensing images. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0015] Figure 1 The flowchart of steps of the geological exploration method integrating remote sensing image processing technology of the present invention; Figure 2 The flowchart of characteristic relationships of the geological exploration method integrating remote sensing image processing technology of the present invention. Specific implementation manners

[0016] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation manners, structures, characteristics and their effects of the geological exploration method integrating remote sensing image processing technology according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0017] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0018] The following specifically describes the specific solution of the geological exploration method integrating remote sensing image processing technology provided by the present invention in conjunction with the accompanying drawings.

[0019] Please refer to Figure 1 , which shows the flowchart of steps of the geological exploration method integrating remote sensing image processing technology provided by an embodiment of the present invention. The method includes the following steps: Step S001: Obtain geological remote sensing images of several acquisition stages, and height data of each pixel point in the geological remote sensing images of each acquisition stage.

[0020] It should be noted that unstable mountain structures are more likely to cause landslides in harsh environments or engineering construction. When a landslide occurs, the surface soil of the geological layer deforms under unstable stress and collapses towards the bottom of the mountain, bringing a large amount of high-risk disaster losses; since landslides often have a translational manifestation, that is, the upper area collapses, and the collapsed soil accumulates and bulges in the lower area, so multiple collapse cracks will appear in the upper and middle areas of the landslide hidden danger area. If the cracks in the area are stretched more, it means that the probability of a landslide occurring in this area is higher. Therefore, in the existing neural network training and recognition process of mountain remote sensing images, the identification of landslide hidden danger areas mainly relies on the similarity matching of collapse cracks in the area; and due to the widely different mountain contours in the remote sensing images in the actual scene and the complex internal deep edge morphology, it is difficult to match the existing neural network crack training set with the crack edges in the actual remote sensing images. Therefore, in this embodiment, the characteristic manifestations of landslides in remote sensing images and the landslide stress deformation law in the actual scene are combined and analyzed to identify more accurate landslide hidden danger areas in mountain remote sensing images.

[0021] Specifically, it is first necessary to collect geological remote sensing images of several acquisition stages, as well as the height data of each pixel in the geological remote sensing images of each acquisition stage. The specific process is as follows: Each acquisition stage is one month apart. Each time, the drone conducts optical remote sensing image acquisition on the monitored mountain area, and performs median filtering denoising and defogging operations on the optical remote sensing images collected each time. After collecting for one year, several geological remote sensing images of the acquisition stages are obtained. At the same time, during each acquisition stage, the radar system of the drone is used to obtain the height data of each pixel in the geological remote sensing images of each acquisition stage. Among them, median filtering denoising and defogging operations are existing technologies, and will not be elaborated here in this embodiment.

[0022] Thus far, several geological remote sensing images of the acquisition stages, as well as the height data of each pixel in the geological remote sensing images of each acquisition stage, are obtained through the above method.

[0023] Step S002: Denote the geological remote sensing image of the last acquisition stage as the geological remote sensing image of the real-time stage; according to the difference in the height data of each pixel in the geological remote sensing image of the real-time stage and its corresponding pixel in the geological remote sensing images of other acquisition stages, obtain the geological position offset degree of each pixel in the geological remote sensing image of the real-time stage; cluster the geological remote sensing image of the real-time stage according to the geological position offset degree of the pixels to obtain several suspected landslide areas in the geological remote sensing image of the real-time stage; according to the gray-scale distribution of the pixels in each suspected landslide area, obtain the geological whiteness degree of each suspected landslide area; according to the geological whiteness degree, the geological position offset degree, and the gray-scale difference of the pixels in each suspected landslide area, obtain the surface geology differentiation degree of each suspected landslide area.

[0024] It should be noted that considering that directly clustering the geological remote sensing image may be greatly affected by the resolution of the remote sensing image, it is necessary to judge and analyze the geological height information. That is, if the height data of the pixel at a certain position in the geological remote sensing image of the real-time stage deviates greatly from that in the geological remote sensing image of the historical acquisition stage, it indicates that the pixel at that position may have a large stress deformation, and thus is more likely to have the characteristics of a landslide hazard. Moreover, the stress deformation characteristics of the pixels in the same landslide hazard area should have a certain degree of aggregation. Therefore, density clustering is further performed on the geological position offset degree reflecting the stress deformation characteristics of the pixels to achieve the division of suspected landslide areas. Since the soil quality in the landslide hazard area collapses and renovates more frequently than in the normal area, and the renovated soil has a grayer-scale appearance that tends to be brighter white, the surface geology differentiation degree of each suspected landslide area is further obtained by combining the gray-scale performance.

[0025] Specifically, denote the geological remote sensing image of the last acquisition stage as the geological remote sensing image of the real-time stage.

[0026] Preferably, in some implementation manners of the embodiments of the present invention, if the height data deviation of the pixel points at a certain position in the geological remote sensing image in the real-time stage and the geological remote sensing image in the historical acquisition stage is large, it indicates that there may be a large stress deformation at the pixel points at that position, and it is more likely to have the characteristics of landslide hazards; according to the difference situation of the height data of each pixel point in the geological remote sensing image in the real-time stage and the height data of the same pixel point in the geological remote sensing images in other acquisition stages, the specific method for obtaining the geological position offset degree of each pixel point in the geological remote sensing image in the real-time stage is as follows: For the th pixel point in the geological remote sensing image in the real-time stage, and any one geological remote sensing image other than the geological remote sensing image in the real-time stage, the absolute value of the difference between the height data of the th pixel point in the geological remote sensing image in the real-time stage and the height data of the th pixel point in the said any one geological remote sensing image is denoted as the height difference value of the th pixel point in the said any one geological remote sensing image; the average value of the height difference values of the th pixel point in all geological remote sensing images other than the geological remote sensing image in the real-time stage is used as the geological position offset degree of the th pixel point in the geological remote sensing image in the real-time stage; The specific formula is: In the formula, represents the geological position offset degree of the th pixel point in the geological remote sensing image in the real-time stage; represents the number of all geological remote sensing images other than the geological remote sensing image in the real-time stage; represents the height data of the th pixel point in the geological remote sensing image in the real-time stage; represents the height data of the th pixel point in the

[0027] th geological remote sensing image in the acquisition stage;

[0028] represents taking the absolute value. It should be noted that if the difference between the height data of the pixel points at a certain position in the geological remote sensing image in the real-time stage and the height data of the pixel points at the same position in the geological remote sensing images in other acquisition stages is larger, it indicates that the geological position offset at that position is more significant, and it indicates that the geology at that position is more likely to have a large deformation.

[0028] Preferably, in some implementation manners of the embodiments of the present invention, since the stress and deformation characteristics of pixel points in the same landslide hazard area should have a certain aggregation property, the density clustering of the geological position offset degree reflecting the stress and deformation characteristics of pixel points can be performed to complete the division of the suspected landslide area in the geological remote sensing image in the real-time stage; the specific method for clustering the geological remote sensing image in the real-time stage according to the geological position offset degree of pixel points to obtain several suspected landslide areas in the geological remote sensing image in the real-time stage is as follows: Input the geological position offset degrees of all pixel points in the geological remote sensing image in the real-time stage into the DBSCAN density clustering algorithm to cluster all pixel points in the geological remote sensing image in the real-time stage, and obtain several clustering clusters; in the geological remote sensing image in the real-time stage, the area formed by all pixel points in each clustering cluster is recorded as a suspected landslide area.

[0029] Among them, the DBSCAN density clustering algorithm is a prior art, and no more details will be described here in this embodiment.

[0030] Preferably, in some implementation manners of the embodiments of the present invention, since the surface soil renovation in the landslide hazard area is more frequent than that in the normal area, the landslide hazard area has a bright white gray-scale performance with obvious distinguishability in the geological remote sensing image; then, according to the gray-scale distribution of pixel points in each suspected landslide area, the specific method for obtaining the geological whitening degree of each suspected landslide area is as follows: For the th suspected landslide area in the geological remote sensing image in the real-time stage, the difference between the mean value of the gray-scale values of all pixel points in the th suspected landslide area and the minimum value of the gray-scale values of all pixel points in the th suspected landslide area is recorded as the first difference; the difference between the maximum value of the gray-scale values of all pixel points in the th suspected landslide area and the minimum value of the gray-scale values of all pixel points in the th suspected landslide area is recorded as the first extreme value; the ratio of the first difference to the first extreme value is used as the geological whitening degree of the th suspected landslide area. The specific formula is: In the formula, represents the geological whitening degree of the th suspected landslide area; represents the mean value of the gray-scale values of all pixel points in the th suspected landslide area; represents the minimum value of the gray-scale values of all pixel points in the th suspected landslide area; represents the maximum value of the gray-scale values of all pixel points in the The maximum gray value of all pixel points in a suspected landslide area.

[0031] Preferably, in some implementation manners of the embodiments of the present invention, due to frequent soil renovation, the internal structure of the landslide area will be more chaotic, that is, it shows a more uneven pixel gray performance; then, according to the degree of geological whitening, the degree of geological position deviation, and the gray difference of pixel points in each suspected landslide area, the specific method for obtaining the surface mass differentiation degree of each suspected landslide area is as follows: The gray value of the th pixel point in the th suspected landslide area and the mean value of the gray values of all pixel points in the th suspected landslide area, the absolute value of the difference therebetween is denoted as the gray difference value of the th pixel point; the mean value of the gray difference values of all pixel points in the th suspected landslide area is denoted as the gray difference of the th suspected landslide area; the normalized value of the product of the gray difference of the th suspected landslide area, the degree of geological whitening of the th suspected landslide area, and the mean value of the geological position deviation degrees of all pixel points in the th suspected landslide area is used as the surface mass differentiation degree of the th suspected landslide area; In the formula, represents the surface mass differentiation degree of the th suspected landslide area; represents the degree of geological whitening of the th suspected landslide area; represents the mean value of the geological position deviation degrees of all pixel points in the th suspected landslide area; represents the number of all pixel points in the th suspected landslide area; represents the gray value of the th pixel point in the th suspected landslide area; represents the mean value of the gray values of all pixel points in the th suspected landslide area; represents taking the absolute value;

[0032] Specifically, map the positions of the suspected landslide areas in the geological remote sensing images in the real-time stage to the geological remote sensing images in other sampling stages, and obtain the surface quality differentiation degrees of the suspected landslide areas in the geological remote sensing images in other sampling stages through the above method.

[0033] So far, the surface quality differentiation degrees of each suspected landslide area in the geological remote sensing images in each sampling stage are obtained through the above method.

[0034] Step S003: Obtain the landslide surface quality presentation degree of each suspected landslide area according to the surface quality differentiation degrees of each suspected landslide area in the geological remote sensing images in different sampling stages; obtain the landslide deformation feature compliance degree of each suspected landslide area according to the distribution of the height data of the pixel points in the geological remote sensing image of each suspected landslide area in the real-time stage and the geological remote sensing image of its adjacent sampling stage; obtain the landslide hazard tendency index of each suspected landslide area according to the landslide deformation feature compliance degree and the landslide surface quality presentation degree.

[0035] It should be noted that considering that the geological surface soil in the landslide hazard area is frequently renovated, while the normal area is relatively stable, the landslide surface quality presentation degree of the suspected landslide area is obtained according to the updated performance of the surface quality differentiation degree in different acquisition stages; at the same time, the more obvious the depression feature in the upper part and the convex feature in the lower part of the suspected landslide area, the more uneven the internal stress deformation of the suspected landslide area, and the higher the landslide hazard of the suspected landslide area. Therefore, the landslide hazard reflection degree of each suspected landslide area is further obtained by combining the internal stress deformation law of the suspected landslide area.

[0036] Preferably, in some implementation manners of the embodiment of the present invention, the specific method for obtaining the landslide surface quality presentation degree of each suspected landslide area according to the surface quality differentiation degrees of each suspected landslide area in the geological remote sensing images in different sampling stages is as follows: Construct a two-dimensional coordinate system with the serial number of the sampling stage as the abscissa and the surface quality differentiation degree as the ordinate, and input the surface quality differentiation degrees of the th suspected landslide area in the geological remote sensing images in all sampling stages into the two-dimensional coordinate system, and use the least squares method for curve fitting to obtain the surface quality differentiation change curve of the th suspected landslide area; Use the STL trend item decomposition method to decompose the surface quality differentiation change curve of the th suspected landslide area to obtain the trend item curve of the th suspected landslide area; Take the mean value of the slopes of all data points on the trend item curve of the th suspected landslide area as the mean value of the trend item slopes of the th suspected landslide area; Take the The ratio between the mean slope of the trend term of a suspected landslide area and the maximum value of the mean slope of the trend terms of all suspected landslide areas is denoted as the trend term slope ratio; the product of the trend term slope ratio and the surface quality differentiation degree of the th suspected landslide area in the geological remote sensing image at the real-time stage is taken as the landslide surface quality presentation degree of the th suspected landslide area; The specific formula is as follows: In the formula, represents the landslide surface quality presentation degree of the th suspected landslide area; represents the mean value of the slopes of all data points on the trend term curve of the th suspected landslide area; represents the maximum value of the mean slopes of the trend terms of all suspected landslide areas; represents the surface quality differentiation degree of the th suspected landslide area in the geological remote sensing image at the real-time stage.

[0037] It should be noted that the more the corresponding trend term curve of the suspected landslide area shows an upward trend, and at the same time, the greater the surface quality differentiation degree of the suspected landslide area in the geological remote sensing image at the real-time stage, the higher the landslide hidden danger of the suspected landslide area; the least squares method and the STL trend term decomposition method are existing technologies, and there is no need to elaborate here in this embodiment.

[0038] Preferably, in some implementation manners of the embodiment of the present invention, since there is often a deformation demarcation line in the landslide hidden danger area, a depression feature appears above the deformation demarcation line, and a convex feature appears below the deformation demarcation line; according to the distribution of the height data of the pixel points in the geological remote sensing image of each suspected landslide area at the real-time stage and the geological remote sensing image of its adjacent sampling stage, the specific method for obtaining the landslide deformation feature compliance degree of each suspected landslide area is as follows: The geological remote sensing image of the previous acquisition stage of the geological remote sensing image at the real-time stage is denoted as the geological remote sensing image of the comparison stage; In the th suspected landslide area in the geological remote sensing image at the real-time stage, the pixel point with the maximum height data is denoted as the height maximum value pixel point of the th suspected landslide area; the pixel point with the minimum height data is denoted as the height minimum value pixel point of the th suspected landslide area; the straight line connection between the height maximum value pixel point of the th suspected landslide area and the height minimum value pixel point of the th suspected landslide area is denoted as the The soil flow direction line of a suspected landslide area; For the th pixel point on the soil flow direction line of the th suspected landslide area, use the line passing through the th pixel point and perpendicular to the soil flow direction line of the th suspected landslide area to divide the th suspected landslide area into an upper area and a lower area; Take the average value of the height data of all pixel points in the lower area of the th suspected landslide area in the geological remote sensing image in the real-time stage as the first height average value; take the average value of the height data of all pixel points in the lower area of the th suspected landslide area in the geological remote sensing image in the comparison stage as the second height average value; take the absolute value of the difference between the first height average value and the second height average value as the th pixel point's lower area convexity degree; Take the average value of the height data of all pixel points in the upper area of the th suspected landslide area in the geological remote sensing image in the real-time stage as the third height average value; take the average value of the height data of all pixel points in the upper area of the th suspected landslide area in the geological remote sensing image in the comparison stage as the fourth height average value; take the absolute value of the difference between the third height average value and the fourth height average value as the th pixel point's upper area concavity degree; Take the ratio of the upper area concavity degree of the th pixel point to the lower area convexity degree of the th pixel point as the th pixel point's deformation depression and bulge ratio; Among all pixel points on the soil flow direction line of the th suspected landslide area, mark the pixel point corresponding to the maximum deformation depression and bulge ratio as the th suspected landslide area's target pixel point; take the difference between the maximum value of the height data of all pixel points in the th suspected landslide area and the height data of the th suspected landslide area's target pixel point as the deformation height difference, and take the product of the deformation depression and bulge ratio of the th suspected landslide area's target pixel point and the reciprocal of the deformation height difference as the th suspected landslide area's landslide deformation characteristic compliance degree; The specific formula is: In the formula, represents the The degree of conformity of landslide deformation characteristics in a suspected landslide area; Denote the maximum value of the height data of all pixel points in the th suspected landslide area; Denote the height data of the target pixel point in the th suspected landslide area; Denote the deformation subsidence ratio of the target pixel point in the

[0039] It should be noted that as time goes by, for the landslide hidden danger area, the concave soil part inside it will gradually accumulate towards the convex part, resulting in the upward movement of the deformation demarcation line. Therefore, the more the deformation demarcation line tends to the lowest point of height, the smaller the landslide hidden danger in the suspected landslide area.

[0040] Preferably, in some implementation manners of the embodiment of the present invention, the greater the degree of presentation of the landslide surface quality in the suspected landslide area and the higher the degree of conformity of the landslide deformation characteristics, the greater the probability of reflecting the landslide hidden danger in the suspected landslide area; the specific method for obtaining the landslide hidden danger tendency index of each suspected landslide area according to the degree of conformity of the landslide deformation characteristics and the degree of presentation of the landslide surface quality is as follows: Multiply the degree of conformity of the landslide deformation characteristics of the th suspected landslide area by the degree of presentation of the landslide surface quality of the th suspected landslide area as the degree of reflection of the landslide hidden danger of the th suspected landslide area; Take the normalization value of the ratio between the degree of reflection of the landslide hidden danger of the th suspected landslide area and the average value of the degrees of reflection of the landslide hidden danger of all suspected landslide areas as the landslide hidden danger tendency index of the th suspected landslide area.

[0041] Thus, the landslide hidden danger tendency index of each suspected landslide area is obtained through the above method.

[0042] Step S004: Explore the landslide hidden danger area in the geological remote sensing image in the real-time stage based on the landslide hidden danger tendency index.

[0043] Preferably, in some implementation manners of the embodiment of the present invention, the specific method for exploring the landslide hidden danger area in the geological remote sensing image in the real-time stage based on the landslide hidden danger tendency index is as follows: Preset two threshold parameters and , where in this embodiment, and are taken as examples for description, and this embodiment does not make specific limitations. Among them, and It depends on the specific implementation situation; For any suspected landslide area, if the landslide hazard tendency index of the any suspected landslide area is less than the threshold parameter , mark the any suspected landslide area as a low landslide risk area; if the landslide hazard tendency index of the any suspected landslide area is greater than the threshold parameter , mark the any suspected landslide area as a high landslide risk area.

[0044] Please refer to Figure 2 , which shows the characteristic relationship flowchart of the geological exploration method of the integrated remote sensing image processing technology; So far, this embodiment is completed.

[0045] The above are only the preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A geological prospecting method using integrated remote sensing image processing technology, characterized in that: The method comprises the following steps: Acquire geological remote sensing images of several acquisition stages and height data of each pixel point in the geological remote sensing images of each acquisition stage; According to the difference between the height data of each pixel point in the geological remote sensing image in the real-time stage and that in the geological remote sensing image in other acquisition stages, the geological position offset of each pixel point in the geological remote sensing image in the real-time stage is obtained; the geological remote sensing image in the real-time stage is clustered according to the geological position offset of the pixel point, and several suspected landslide areas in the geological remote sensing image in the real-time stage are obtained; according to the grayscale distribution of the pixel points in each suspected landslide area, the geological whitening degree of each suspected landslide area is obtained; according to the geological whitening degree, the geological position offset and the grayscale difference of the pixel points in each suspected landslide area, the surface texture distinction of each suspected landslide area is obtained; According to the surface texture distinction of each suspected landslide area in the geological remote sensing images at different sampling stages, the landslide surface texture presentation degree of each suspected landslide area is obtained; according to the distribution of the height data of the pixel points in the geological remote sensing images of each suspected landslide area in the real-time stage and the geological remote sensing images of the adjacent sampling stages, the landslide deformation feature conformity of each suspected landslide area is obtained; according to the landslide deformation feature conformity and the landslide surface texture presentation degree, the landslide hazard trend index of each suspected landslide area is obtained; Exploring landslide hazard areas in real-time geological remote sensing images based on landslide hazard trend indicators; The specific method for obtaining the geological whitening degree of each suspected landslide area includes: For the jth suspected landslide area in the geological remote sensing image in the real-time stage, the difference between the mean value of the grayscale values ​​of all pixels in the jth suspected landslide area and the minimum value of the grayscale values ​​of all pixels in the jth suspected landslide area is recorded as the first difference; the difference between the maximum value of the grayscale values ​​of all pixels in the jth suspected landslide area and the minimum value of the grayscale values ​​of all pixels in the jth suspected landslide area is recorded as the first extreme value; the ratio of the first difference to the first extreme value is taken as the geological whitening degree of the jth suspected landslide area; The absolute value of the difference between the grayscale value of the yth pixel in the jth suspected landslide area and the mean grayscale value of all pixels in the jth suspected landslide area is recorded as the grayscale difference value of the yth pixel; the mean grayscale difference value of all pixels in the jth suspected landslide area is recorded as the grayscale difference of the jth suspected landslide area; the normalized value of the product of the grayscale difference of the jth suspected landslide area, the geological whitening degree of the jth suspected landslide area and the mean geological position deviation of all pixels in the jth suspected landslide area is taken as the surface texture discrimination of the jth suspected landslide area.

2. The geological prospecting method based on the integrated remote sensing image processing technology according to claim 1 is characterized in that: The specific method of obtaining the geological position deviation of each pixel point in the geological remote sensing image in the real-time stage includes: For the geological remote sensing images in the real-time stage pixel points, as well as any geological remote sensing image of any acquisition stage except the geological remote sensing image of the real-time stage, The height data of the pixel point in the geological remote sensing image in real time is The absolute value of the difference between the height data of the pixel points in the geological remote sensing image at any acquisition stage is recorded as The height difference value of the pixel point in the geological remote sensing image at any acquisition stage; The average of the height difference values ​​of the pixel points in all the geological remote sensing images collected in the real-time stage is taken as the height difference of the first pixel point in the geological remote sensing image in the real-time stage. The geological position offset of each pixel.

3. The geological prospecting method based on integrated remote sensing image processing technology according to claim 1, characterized in that: The specific method of obtaining several suspected landslide areas in the geological remote sensing image in the real-time stage includes: The geological position offset of all pixels in the real-time geological remote sensing image is input into the DBSCAN density clustering algorithm, and all pixels in the real-time geological remote sensing image are clustered to obtain several clusters; in the real-time geological remote sensing image, the area composed of all pixels in each cluster is recorded as a suspected landslide area.

4. The geological prospecting method based on integrated remote sensing image processing technology according to claim 1, characterized in that: The specific method for obtaining the landslide surface quality presentation degree of each suspected landslide area includes: A two-dimensional coordinate system is constructed with the sampling stage number as the horizontal coordinate and the surface quality differentiation as the vertical coordinate. The surface texture discrimination of the suspected landslide area in the geological remote sensing images of all sampling stages is input into the two-dimensional coordinate system, and the least squares method is used for curve fitting to obtain the first The surface geology differentiation curve of the suspected landslide area; Using STL trend term decomposition method to The surface texture variation curve of the suspected landslide area is decomposed to obtain the Trend curve of suspected landslide area; The first The average slope of all data points on the trend curve of the suspected landslide area is recorded as The mean slope of the trend term in the suspected landslide area; The first The ratio between the mean value of the trend item slope of the suspected landslide area and the maximum value of the mean value of the trend item slope of all suspected landslide areas is recorded as the trend item slope ratio; The trend term slope ratio is compared with the first The product of the surface texture differentiation of the suspected landslide area is taken as the The degree of landslide surface texture in the suspected landslide area.

5. The geological prospecting method based on integrated remote sensing image processing technology according to claim 1, characterized in that: The specific methods for obtaining the conformity of landslide deformation characteristics of each suspected landslide area are as follows: The geological remote sensing image of the previous acquisition stage of the geological remote sensing image of the real-time stage is recorded as the geological remote sensing image of the comparison stage; According to The distribution difference of the height data of the pixels in the geological remote sensing images in the real-time stage and the geological remote sensing images in the comparison stage of the suspected landslide area is obtained. The deformation depression-rise ratio of the target pixel point in the suspected landslide area; The first The maximum value of the height data of all pixels in the suspected landslide area is The difference between the height data of the target pixel points in the suspected landslide area is recorded as the deformation height difference. The product of the deformation uplift ratio and the inverse of the deformation height difference of the target pixel point in the suspected landslide area is taken as the The conformity of landslide deformation characteristics in the suspected landslide area.

6. The geological prospecting method using integrated remote sensing image processing technology according to claim 5 is characterized in that: The acquisition The deformation depression-rise ratio of the target pixel point in the suspected landslide area includes the following specific methods: In the real-time geological remote sensing image In the suspected landslide area, the pixel with the maximum value of the height data is recorded as The pixel with the maximum height value in the suspected landslide area is recorded as the pixel with the minimum height data. The height minimum pixel point of the suspected landslide area; The maximum pixel point of the height of the suspected landslide area is The straight line connecting the minimum height pixels in the suspected landslide area is recorded as A straight line of soil flow in the suspected landslide area; For The soil flow direction of the suspected landslide area is pixels, using the pixels, and perpendicular to the The soil flow direction of the suspected landslide area is straight line. The suspected landslide area is divided into upper and lower areas; The first The average of the height data of all pixels in the lower area of ​​the suspected landslide area in the real-time geological remote sensing image is recorded as the first height average; The average of the height data of all pixels in the lower area of ​​the suspected landslide area in the geological remote sensing image in the comparison stage is recorded as the second height average; the absolute value of the difference between the first height average and the second height average is recorded as the first The degree of convexity of the area below the pixel point; The first The average of the height data of all pixels in the upper area of ​​the suspected landslide area in the real-time geological remote sensing image is recorded as the third height average; The average of the height data of all pixels in the upper area of ​​the suspected landslide area in the geological remote sensing image in the comparison stage is recorded as the fourth height average; the absolute value of the difference between the third height average and the fourth height average is recorded as the fourth height average. The degree of depression in the upper area of ​​each pixel; The first The degree of depression in the upper area of ​​the pixel is similar to that of the The ratio of the convexity of the lower area of ​​the pixel point is taken as the The deformation convexity ratio of each pixel; In the Among all the pixels on the soil flow direction line of the suspected landslide area, the pixel corresponding to the maximum value of the deformation collapse ratio is recorded as The target pixel points of the suspected landslide area.

7. The geological prospecting method based on integrated remote sensing image processing technology according to claim 1, characterized in that: The specific method of obtaining the landslide hazard trend index of each suspected landslide area according to the landslide deformation characteristic conformity and the landslide surface quality presentation degree is as follows: The first The conformity of the landslide deformation characteristics in the suspected landslide area is consistent with that in the The product of the landslide surface quality presentation degree of the suspected landslide area is taken as the The degree of landslide hazard reflected in each suspected landslide area; The first The normalized value of the ratio of the landslide hazard reflection degree of the suspected landslide area to the average of the landslide hazard reflection degrees of all suspected landslide areas is taken as the first Landslide hazard trend indicators in suspected landslide areas.

8. The geological prospecting method based on integrated remote sensing image processing technology according to claim 1, characterized in that: The specific method of exploring the landslide hazard area in the geological remote sensing image in the real-time stage based on the landslide hazard trend index includes: Preset two threshold parameters and For any suspected landslide area, if the landslide hazard trend index of any suspected landslide area is less than the threshold parameter , any suspected landslide area is recorded as a low landslide risk area; If the landslide hazard trend index of any suspected landslide area is greater than the threshold parameter , any suspected landslide area is recorded as a high landslide risk area.

Images