Geological exploration digital surveying and mapping method, system and device

Abstract

The invention discloses a geological exploration digital surveying and mapping method, system and device, and relates to the technical field of image processing. The method comprises the following steps: acquiring a multi-modal geological image of a target area; on the basis of the feature point matching relation between the geological images, each geological image is divided into a plurality of feature blocks, the feature blocks are used for representing a local image region obtained by region growth with a reference feature point as a seed point, and the reference feature point is a feature point successfully matched with other geological images in the geological image; respectively determining the feature weight of each reference feature point in each geological image based on the texture feature of each feature block and the information gain of each reference feature point; and based on the feature weight of each reference feature point in each geological image, fusing the multi-modal geological images of the target area to obtain a target fused geological image of the target area. According to the invention, the accuracy of digital surveying and mapping can be improved.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a geological exploration digital mapping method, system and device. Background Art

[0002] Digital surveying and mapping utilizes modern information technology, integrating global positioning systems, geographic information systems, and remote sensing technologies to transform traditional surveying and mapping operations into a process for collecting, processing, and managing digital information. In practice, complex environments often experience occlusion caused by factors such as undulating terrain and inclement weather, which can easily lead to localized distortion. Therefore, comprehensive analysis and processing of multimodal geological images is necessary to obtain accurate surveying and mapping results.

[0003] Currently, digital mapping typically uses various exploration technologies to collect geological images of the target area. Decision tree models are then used to perform weighted fusion of the features of these multimodal geological images.

[0004] However, the feature weighting method used in decision tree models relies heavily on the characteristics of single-modal geological images. This makes it difficult to achieve coordinated attention between the features of each modality during data fusion. Consequently, the fused geographic information data still exhibits distortion in some dimensions, resulting in poor accuracy in digital mapping. Summary of the Invention

[0005] The embodiments of the present invention provide a geological exploration digital mapping method, system and device, which can improve the accuracy of digital mapping.

[0006] A first aspect of an embodiment of the present invention provides a method for digital mapping of geological exploration, the method comprising: Acquire multimodal geological images of the target area; Based on the matching relationship between the feature points of each geological image, each geological image is divided into multiple feature blocks. The feature blocks are used to represent the local image area obtained by region growing with reference feature points as seed points. The reference feature points are the feature points in the geological image that are successfully matched with other geological images. Based on the texture features of each feature block and the information gain of each reference feature point, the feature weight of each reference feature point in each geological image is determined respectively; Based on the feature weights of each reference feature point in each geological image, the multimodal geological images of the target area are fused to obtain a target fused geological image of the target area.

[0007] Furthermore, the present invention also proposes to determine the feature weight of each reference feature point in each geological image based on the texture features of each feature block and the information gain of each reference feature point, which may specifically include: Fusing the matching feature blocks between the geological images to obtain a fused block image; Based on the overall information entropy of the fused block image and the conditional entropy of each feature point in the fused block image to the fused block image, the detail contribution of each feature point in the fused block image is determined respectively; Based on the detail contribution of each feature point in each feature block, the basic reference degree of each feature block is determined respectively; Based on the basic reference degree of each feature block in the geological image, the basic description index of the geological image is determined; The information gain of each reference feature point is divided by the basic description index of the corresponding geological image to obtain the feature weight of each reference feature point in each geological image.

[0008] Furthermore, the present invention also proposes to fuse the matching feature blocks between the geological images to obtain a fused block image, which may specifically include: Obtain the matching relationship of feature points between geological images; Based on the feature point matching relationship, a first geometric transformation model is constructed, where the geometric transformation model is used to represent at least one of a translation transformation relationship, a rotation transformation relationship, and a scaling transformation relationship between geological images; Based on the first geometric transformation model, resampling the matching feature blocks between the geological images to obtain geometrically aligned resampled blocks; An image fusion operation based on wavelet transform is performed on each resampled block to obtain a fused block image.

[0009] Furthermore, the present invention also proposes to determine the detail contribution of each feature point in the fused block image based on the overall information entropy of the fused block image and the conditional entropy of each feature point in the fused block image to the fused block image, which may specifically include: Determining the overall information entropy of the fused block image using the pixel features of each pixel in the fused block image, and determining the block information entropy of each feature block using the pixel features of each pixel in each feature block, wherein the pixel features include at least one of grayscale, gradient value, and gradient direction; The conditional entropy of the target feature point on the fused block image is determined by using the pixel features of the target feature point and the information entropy of each block. The target feature point is any feature point in the fused block image. The difference between the overall information entropy and the conditional entropy of the target feature point on the fused block image is divided by the overall information entropy to obtain the detail contribution of the target feature point.

[0010] Furthermore, the present invention also proposes to determine the basic reference degree of each feature block based on the detail contribution of each feature point in each feature block, which may specifically include: Obtaining a first gradient mean of pixels within a preset neighborhood of each feature point in the target feature block, and a second gradient mean of pixels within a preset neighborhood of each corresponding matching feature point in the fused block image, where the target feature block is any feature block; The basic reference degree of the target feature block is determined by using the first gradient mean values, the second gradient mean values and the detail contribution of each feature point in the target feature block.

[0011] Furthermore, the present invention also proposes to determine the basic description index of the geological image based on the basic reference degree of each feature block in the geological image, which may specifically include: The detail contribution of the feature points in each feature block is processed by averaging to obtain the detail weight of each feature block; The product of the detail weight of each feature block and the basic reference degree is averaged to obtain the basic description index of the geological image.

[0012] Furthermore, the present invention also proposes dividing each geological image into a plurality of feature blocks based on the matching relationship of feature points between the geological images, which may specifically include: Perform feature point detection on each geological image to obtain the feature points in each geological image; Match the feature points in each geological image to obtain the feature point matching relationship; Based on the feature point matching relationship, determining the reference feature points in the target geological image, where the target geological image is any geological image; Region growing is performed using each reference feature point in the target geological image as a seed point until a preset stop-growth condition is reached, thereby obtaining multiple feature blocks corresponding to the target geological image.

[0013] Furthermore, the present invention also proposes to fuse the multimodal geological images of the target area based on the feature weights of each reference feature point in each geological image to obtain a target fused geological image of the target area, which may specifically include: According to the feature weights of each reference feature point in each geological image, the interpolation method is used to assign corresponding pixel weights to each pixel point in each geological image; Constructing a second geometric transformation model based on the matching relationship of reference feature points between the geological images; Based on the second geometric transformation model, the geological images are resampled to obtain geometrically aligned resampled images; According to the pixel weight of each pixel in the resampled image, the pixel features of each resampled image are fused to obtain the target fused geological image of the target area.

[0014] A second aspect of an embodiment of the present invention provides a digital mapping system for geological exploration, the system comprising: An image acquisition module, used to acquire multimodal geological images of the target area; An image segmentation module is used to divide each geological image into multiple feature blocks based on the matching relationship between feature points between the geological images. The feature blocks are used to represent the local image area obtained by region growing using reference feature points as seed points. The reference feature points are feature points in the geological image that are successfully matched with other geological images. A weight determination module is used to determine the feature weight of each reference feature point in each geological image based on the texture features of each feature block and the information gain of each reference feature point; The image fusion module is used to fuse the multimodal geological images of the target area based on the feature weights of each reference feature point in each geological image to obtain a target fused geological image of the target area.

[0015] According to a third aspect of an embodiment of the present invention, a digital mapping device for geological exploration is provided, comprising: a processor and a memory storing computer program instructions; the processor implements any one of the above-mentioned digital mapping methods for geological exploration when executing the computer program instructions.

[0016] The present invention has the following beneficial effects: In the digital mapping method for geological exploration provided by an embodiment of the present invention, a multimodal geological image of the target area is first acquired. Then, based on the matching relationship between feature points in each image, the image is divided into feature blocks obtained by regional growth using successfully matched reference feature points as seed points. This method can more accurately locate local areas of the image and provide more detailed units for subsequent processing. Then, based on the texture features of each feature block and the information gain of each reference feature point, the feature weights of the reference feature points in each image are determined. This method, which considers the correlation between multimodal images and the feature information gain, can more scientifically measure the importance of each feature in the fusion process and avoid the limitations of relying solely on single-modal image features. Finally, the multimodal geological images are fused based on the feature weights of each reference feature point. Due to the more reasonable weight determination, the features of each modal image can be better coordinated and focused, thereby effectively reducing the distortion of the fused geographic information data in some data dimensions, making the final target fused geological image more accurate, thereby improving the accuracy of digital mapping. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. 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 any creative work.

[0018] Figure 1 A schematic diagram of a process flow of a digital mapping method for geological exploration provided by one embodiment of the present invention; Figure 2 A schematic diagram of the process of S103 provided in one embodiment of the present invention; Figure 3 A schematic structural diagram of a digital mapping system for geological exploration provided by one embodiment of the present invention; Figure 4 The present invention provides a structural diagram of a digital mapping device for geological exploration according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] To further illustrate the technical means and effectiveness of the present invention in achieving its intended objectives, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail a digital geological surveying and mapping method, system, and device according to the present invention, including its specific implementation, structure, features, and effectiveness. In the following description, references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0020] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0021] It should be noted that the acquisition, storage, use, and processing of data in the technical solution of the present invention comply with the relevant provisions of laws and regulations.

[0022] It should be noted that in the embodiments of the present invention, certain software, components, models and other existing solutions in the industry may be mentioned. They should be regarded as exemplary and their purpose is only to illustrate the feasibility of implementing the technical solution of the present invention, but it does not mean that the applicant has or will necessarily use the solution.

[0023] Digital surveying and mapping uses modern information technology to integrate global positioning systems, geographic information systems, and remote sensing technologies to transform traditional surveying and mapping operations into a process for collecting, processing, and managing digital information. In geological exploration, digital surveying and mapping technology is of great value for applications such as mineral resource exploration and geological disaster early warning. In actual work, when faced with complex geological environments, image distortion caused by factors such as undulating terrain, rock obstruction, and inclement weather is particularly prominent. This makes it difficult for single-modal geological images to fully reflect geological characteristics. Therefore, in order to obtain accurate surveying and mapping results, comprehensive analysis and processing of multimodal geological images is necessary.

[0024] Current mainstream digital mapping solutions typically employ multi-source data fusion technology, using various exploration equipment to collect multimodal geological image data from the target area. Existing technologies primarily employ decision tree models to perform weighted fusion processing on the features of these multimodal geological images. However, this fusion approach suffers from significant technical flaws. The feature weighting method employed by decision tree models primarily relies on the inherent characteristics of single-modal geological images and lacks consideration of the inter-modal correlation features. This severely impacts the accuracy and reliability of digital mapping results.

[0025] When faced with the above problems, the present invention first recognizes that the root cause of the distortion of multimodal geological image fusion lies in the lack of cross-modal feature coordination mechanism and dynamic adaptation capability of block boundaries. The traditional method adopts a global weight adjustment strategy, which is unable to establish cross-modal feature associations in local areas, resulting in spatial offset between the feature block boundaries and the real geological interface. In this regard, the present invention proposes to establish a local association network based on the successfully matched feature points, and form feature blocks adaptive to the geological structure through regional growth, wherein the matching feature points are used as seed points to guide the block boundaries to fit the actual geological interface, avoiding segmentation errors caused by artificially set thresholds. On this basis, the present invention finds that the information gain of the matching feature points and the texture features of the blocks they are located in jointly determine the fusion weight, wherein the information gain reflects the complementarity of cross-modal features, and the texture features characterize the consistency of local structure. The combination of the two can suppress the interference caused by the dominance of single-modal features and realize the collaborative optimization of multi-source features.

[0026] In this regard, Figure 1 As shown, the present invention provides a flow chart of a geological exploration digital mapping method. The geological exploration digital mapping method can be applied to a server, and the geological exploration digital mapping method can include the following steps S101 to S104: S101, acquiring a multimodal geological image of a target area; S102, based on the matching relationship between the feature points of each geological image, each geological image is divided into a plurality of feature blocks, wherein the feature blocks are used to represent a local image region obtained by region growing using a reference feature point as a seed point, where the reference feature point is a feature point in the geological image that successfully matches other geological images; S103, determining the feature weight of each reference feature point in each geological image based on the texture feature of each feature block and the information gain of each reference feature point; S104 , based on the feature weights of the reference feature points in the geological images, the multimodal geological images of the target area are fused to obtain a target fused geological image of the target area.

[0027] In this embodiment, multimodal geological images refer to multiple types of geological images obtained through different exploration technologies or sensors, which can be specifically realized by satellite remote sensing images, ground radar detection images, and electromagnetic wave imaging data. Different modal images complement each other in terms of resolution, spectral characteristics, and spatial coverage.

[0028] Feature point matching relationship refers to the geometric or texture association between key points that can correspond to each other in different geological images. Specifically, it can be achieved by using the Scale-Invariant Feature Transform (SIFT) algorithm and the accelerated robust feature algorithm to identify overlapping areas of the same geological structure in multimodal images through matching relationships.

[0029] Feature segmentation refers to the local image area obtained by expanding the successfully matched feature points through the region growing algorithm. Specifically, pixel similarity threshold and edge continuity constraint can be used as growth conditions to make the segmentation boundary fit the morphological changes of the actual geological structure.

[0030] Texture features refer to statistics that reflect the grayscale distribution patterns of local areas in geological images. They can be implemented using grayscale co-occurrence matrices and are used to quantify differences in geological properties such as rock strata orientation and fracture density.

[0031] Information gain refers to the amount of mutual information contributed by the reference feature point in the multimodal image matching process. It can be achieved by calculating the change in entropy value of the block where the feature point is located before and after fusion, reflecting the role of the reference feature point in eliminating the uncertainty of multi-source data.

[0032] Feature weight refers to the influence coefficient of feature points that is dynamically adjusted based on texture features and information gain. It can be implemented using linear weighting method or entropy weight method to balance the contribution ratio of reliable features in images of different modalities.

[0033] Image fusion refers to the integration of spatial and spectral information of multimodal geological images. Specifically, it can be achieved by wavelet transform fusion and principal component analysis fusion methods. Weight optimization can be used to suppress local distortion and enhance the continuity of geological features.

[0034] The core innovation of this invention is to construct feature blocks as fusion units through feature point matching relationships, dynamically allocate weights based on texture features and information gain, so as to achieve collaborative optimization of multimodal geological images in terms of local details and global consistency, and solve the fusion distortion problem caused by traditional single-modal feature weighting.

[0035] The working process and principle of the present invention are as follows: first, multimodal geological images of the target area are acquired. These geological images may include different types of remote sensing data, such as optical, infrared, and radar. Next, each geological image is divided into multiple feature blocks based on the matching relationships between feature points in each geological image. Feature blocks are local image regions obtained by region growing using reference feature points as seed points. Reference feature points are feature points in the geological image that have successfully matched other geological images. This division method ensures the corresponding relationship between images of different modalities.

[0036] Next, the feature weight of each reference feature point in each geological image is determined based on the texture features of each feature block and the information gain of each reference feature point. Texture features reflect the structural information of the local area, while information gain indicates the contribution of the reference feature point to the overall information. By combining these two factors, the importance of each reference feature point can be more accurately assessed.

[0037] Finally, based on the feature weights of each reference feature point in each geological image, the multimodal geological images of the target area are fused to obtain a target fused geological image of the target area. This feature weight-based fusion method can fully utilize the advantages of each modality image while suppressing noise and interference, thereby obtaining a more accurate and comprehensive fusion result.

[0038] As an example, a multimodal geological image of the target area is first acquired, for example, an optical remote sensing image, a hyperspectral image, and a synthetic aperture radar image of the target area can be acquired simultaneously.

[0039] Next, feature points are extracted and matched for each modality's geological image. The SIFT algorithm can be used to extract feature points, which are then matched using nearest neighbor matching and the Random Sample Consensus (RANSAC) algorithm to determine the matching relationships between the feature points of each geological image. Based on this matching relationship, a region growing algorithm is used to partition each geological image into blocks, using successfully matched reference feature points as seed points. The stopping condition for region growing can be set to a pixel grayscale difference threshold or a region area threshold.

[0040] For each feature block, its texture features are calculated. Texture features such as energy, contrast, and entropy can be extracted using the gray-level co-occurrence matrix. Simultaneously, the information gain of each reference feature point is calculated using information theory methods such as mutual information or relative entropy. Combining texture features and information gain, a feature weight is assigned to each reference feature point. Weight calculation can use a weighted average or nonlinear mapping function.

[0041] Finally, based on the feature weights, the multimodal geological images are fused. Methods such as weighted averaging or wavelet transform can be used for pixel-level fusion to obtain the target fused geological image.

[0042] Through this embodiment, a multimodal geological image of the target area is first obtained. Then, based on the matching relationship between the feature points of each image, the image is divided into feature blocks obtained by regional growth with the successfully matched reference feature points as seed points. This method can more accurately locate the local area of the image and provide more detailed units for subsequent processing. Then, based on the texture features of each feature block and the information gain of each reference feature point, the feature weights of the reference feature points in each image are determined respectively. This method, which takes into account the correlation between multimodal images and the feature information gain to determine the weights, can more scientifically measure the importance of each feature in the fusion process and avoid the limitations brought by relying solely on single-modal image features. Finally, the multimodal geological images are fused based on the feature weights of each reference feature point. Since the weight determination is more reasonable, the features of each modal image can be better coordinated and focused, thereby effectively reducing the distortion of the fused geographic information data in some data dimensions, making the final target fused geological image more accurate, thereby improving the accuracy of digital mapping.

[0043] In some of the above-mentioned schemes of the present invention, when determining feature weights based on the texture features of the feature blocks and the information gain of the reference feature points, due to the lack of effective evaluation of the detail information in the fused block images, the feature weight allocation cannot accurately reflect the contribution differences of different feature points in the image fusion process, affecting the detail preservation effect of the fused geological image.

[0044] In this regard, Figure 2 As shown, the present invention further proposes that S103 may specifically include the following S201 to S205: S201, fusing the matching feature blocks between the geological images to obtain a fused block image; S202, determining the detail contribution of each feature point in the fused block image based on the overall information entropy of the fused block image and the conditional entropy of each feature point in the fused block image to the fused block image; S203, determining a basic reference degree for each feature block based on the detail contribution of each feature point in each feature block; S204, determining a basic description index of the geological image based on the basic reference degree of each feature block in the geological image; S205 , dividing the information gain of each reference feature point by the basic description index of the corresponding geological image to obtain the feature weight of each reference feature point in each geological image.

[0045] In this embodiment, the fused block image is obtained by fusing matching feature blocks from different geological images. The fusion process aims to integrate information from the same feature blocks in different geological images to obtain more comprehensive and accurate geological information. For example, matching feature blocks from two geological images are fused together using wavelet transform to form a new fused block image.

[0046] Information entropy is a measure of information uncertainty. Overall information entropy is used to measure the degree of uncertainty in the information within the fused block image. It reflects the distribution of all pixel values within the fused block image. A higher information entropy indicates more complex information within the fused block image and a higher degree of uncertainty.

[0047] Conditional entropy measures the uncertainty contribution of a feature point to the information in the fused image block, given the information of other feature points in the fused image block. Conditional entropy reflects the degree of correlation between the feature point and other information in the fused image block. Assuming there are multiple feature points in the fused image block, calculate the uncertainty contribution of a feature point to the overall image information, given the information of other feature points. This is the conditional entropy of that feature point.

[0048] The detail contribution is based on the overall information entropy of the fused block image and the conditional entropy of each feature point in the fused block image. It determines the contribution of each feature point to the detail information of the fused block image. The higher the detail contribution, the more important the feature point is in describing the image details.

[0049] The basic reference degree is the fundamental reference level of a feature block, determined based on the detail contribution of each feature point within the block. This reflects the importance and reference value of the feature block within the overall geological image. For example, the basic reference degree of a feature block can be calculated by comprehensively calculating the detail contribution of all feature points within the block (e.g., averaging or weighted averaging).

[0050] Basic descriptive metrics are indicators used to describe the overall characteristics of a geological image, based on the basic reference degrees of each feature block within the image. Basic descriptive metrics integrate information from each feature block within the image and are used to subsequently calculate feature weights. For example, the basic descriptive metrics for a geological image can be obtained by combining the basic reference degrees of all feature blocks within the image (e.g., summing or weighted summing).

[0051] As an example, first, a feature point matching algorithm is used to find matching feature points between geological images; based on the matching feature points, the corresponding feature blocks are determined; and then an image fusion algorithm is used to fuse the matching feature blocks to obtain a fused block image.

[0052] The probability of occurrence of all pixel values in the fused image blocks is then calculated, and the overall information entropy of the fused image blocks is calculated using a preset information entropy formula. For each feature point, the uncertainty contribution to the overall image information, i.e., the conditional entropy, is calculated, given the information of other feature points. By combining the overall information entropy and the conditional entropy through certain combination operations, the detail contribution of each feature point is determined.

[0053] Then, for each feature block, the detail contribution of all feature points is calculated; a comprehensive calculation method (such as weighted average) is used to obtain the basic reference degree of the feature block. For a geological image, the basic reference degree of all feature blocks is calculated; a comprehensive calculation method (such as weighted summation) is used to obtain the basic description index of the geological image.

[0054] Finally, the feature weight of each reference feature point in the geological image is determined by the following formula 1: Formula 1 In formula 1, The feature weight used to characterize the rth reference feature point in the eth geological image, The information gain used to characterize the rth reference feature point in the eth geological image is: Basic descriptive indicators used to characterize the e-th geological image.

[0055] Through this embodiment, the correlation between multimodal geological images can be fully considered to achieve coordinated adjustment of feature weights. This improves the accuracy of geological image fusion, thereby improving the accuracy of digital mapping. At the same time, the solution can also effectively reduce the risk of information distortion in local areas, ensuring the integrity and reliability of the fused geological image. It should be noted that in order to ensure that the calculation results are meaningful, when performing fractional operations in the embodiment of the present invention, when the denominator is 0, a parameter adjustment factor greater than 0 needs to be added to the denominator to prevent the denominator from being 0. The value of the parameter adjustment factor is set by the implementer according to the actual situation, and this application does not impose any special restrictions.

[0056] In some of the above-mentioned solutions of the present invention, during the fusion process, the feature blocks may not be accurately aligned in space due to geometric differences between different geological images, thereby affecting the quality of the fused block image.

[0057] In this regard, the present invention further proposes that S201 may specifically include: Obtain the matching relationship of feature points between geological images; Based on the feature point matching relationship, a first geometric transformation model is constructed, where the geometric transformation model is used to represent at least one of a translation transformation relationship, a rotation transformation relationship, and a scaling transformation relationship between geological images; Based on the first geometric transformation model, resampling the matching feature blocks between the geological images to obtain geometrically aligned resampled blocks; An image fusion operation based on wavelet transform is performed on each resampled block to obtain a fused block image.

[0058] In this embodiment, the first geometric transformation model describes the spatial transformation relationship between geological images using a parameter matrix. For example, an affine transformation matrix is used to implement a combination of translation, rotation, and scaling. The resampled blocks use bilinear interpolation to adjust pixel positions to ensure smooth transitions at the block edges. The wavelet transform-based image fusion operation uses discrete wavelet decomposition to decompose the blocks into low-frequency and high-frequency subbands. For example, a three-layer decomposition of the blocks is performed using the Haar wavelet basis function. The weighted average of the subband coefficients is then used to reconstruct the fused block image.

[0059] Specifically, feature point matching relationships are determined by calculating the Hamming distance between feature descriptors, and spatial coordinate correspondences are generated for successfully matched feature points. Based on these feature point correspondences, the least squares method is used to fit the parameter matrix of the first geometric transformation model. For example, the optimal rotation and translation parameters are determined through singular value decomposition (SVD) decomposition. During the resampling and block-by-block process, the original block pixel coordinates are inversely mapped according to the first geometric transformation model to generate a new pixel array in the target coordinate system. For wavelet transform fusion, the low-frequency subband uses a mean fusion rule to preserve basic structural information, while the high-frequency subband uses a local variance fusion rule to enhance detailed textures. For example, in the third-level wavelet decomposition, the high-frequency coefficients corresponding to the block with the largest variance are selected as the fusion result. Through the dual processing of geometric alignment and frequency domain fusion, spatial misalignment and spectral interference between blocks are effectively eliminated.

[0060] As an example, first obtain the feature point matching relationship between various geological images. For example, the feature points of each geological image can be extracted using the SIFT algorithm, and the feature points can be matched using the K-Nearest Neighbors (KNN) algorithm to obtain the feature point matching relationship.

[0061] Based on the feature point matching relationship, a first geometric transformation model is constructed. Specifically, at least four pairs of matching feature points can be selected and an affine transformation matrix is constructed using these matching point pairs. The affine transformation matrix is used to represent the translation transformation relationship, rotation transformation relationship, and scaling transformation relationship between geological images.

[0062] Based on the first geometric transformation model, the matching feature blocks between the geological images are resampled. Furthermore, the feature blocks in the geological images can be mapped to the target image coordinate system using an affine transformation matrix, and the mapped pixels are resampled using bicubic interpolation to obtain geometrically aligned resampled blocks.

[0063] Finally, a wavelet transform-based image fusion operation is performed on each resampled block. Discrete wavelet transform can be used to decompose each resampled block into low-frequency sub-bands and high-frequency sub-bands. The low-frequency sub-bands are fused using the weighted average method, and the high-frequency sub-bands are fused using the maximum absolute value method. Finally, an inverse wavelet transform is performed to obtain the fused block image.

[0064] This embodiment achieves precise alignment and fusion of multimodal geological image feature blocks. This improves the geometric accuracy and detail preservation of the fused block images, paving the way for subsequent feature weight calculations based on the fused block images, and ultimately improves the overall quality and accuracy of multimodal geological image fusion.

[0065] In some of the above-mentioned schemes of the present invention, it is proposed to allocate feature weights by calculating the detail contribution of feature points. However, in this process, since the differentiated effects of feature points in different modal images on local texture information are not taken into account, there are deviations in the calculation results of the detail contribution, which in turn affects the accuracy of subsequent feature weight allocation.

[0066] In this regard, the present invention further proposes that S202 may specifically include: Determining the overall information entropy of the fused block image using the pixel features of each pixel in the fused block image, and determining the block information entropy of each feature block using the pixel features of each pixel in each feature block, wherein the pixel features include at least one of grayscale, gradient value, and gradient direction; The conditional entropy of the target feature point on the fused block image is determined by using the pixel features of the target feature point and the information entropy of each block. The target feature point is any feature point in the fused block image. The difference between the overall information entropy and the conditional entropy of the target feature point on the fused block image is divided by the overall information entropy to obtain the detail contribution of the target feature point.

[0067] In this embodiment, the pixel feature is a feature triple consisting of the grayscale, gradient value, and gradient direction of the pixel. For example, the pixel feature of the jth pixel is the feature triple .

[0068] As an example, the overall information entropy of the fused block image can be determined by the following formula 2: Formula 2 In formula 2, Used to characterize the overall information entropy of the c-th fused block image, It is used to represent the probability of occurrence of the pixel feature of the jth pixel in the cth fused block image, J is used to represent the total number of pixel feature categories in the cth fused block image, and log is used to represent the logarithmic function operation.

[0069] in, This reflects the relative frequency of that pixel feature within that image block. The nature of logarithmic operations means that when the probability of occurrence is greater, its logarithmic value is smaller; when the probability of occurrence is smaller, its logarithmic value is larger. Thus, when the probability distribution of all pixel features is more uniform, the overall information entropy is greater, indicating richer information and less uncertainty in the image. Conversely, if the probability of some pixel features is much greater than that of others, the overall information entropy is smaller, indicating that the information in the image is more concentrated and more certain.

[0070] The calculation method of the block information entropy of the feature block is the same as the calculation method of the overall information entropy of the fused block image, and can also be determined by the above formula 2.

[0071] The conditional entropy of the target feature point on the fused block image can be determined by the following formula 3: Formula 3 In formula 3, Used to characterize target feature points The conditional entropy of the c-th fused block image, Used to characterize target feature points The corresponding pixel features are in the feature block The probability of occurrence in Used to characterize feature blocks The block information entropy, V is used to characterize the target feature points The number of feature blocks corresponding to the pixel feature.

[0072] Among them, each block feature The block information is multiplied by the target feature point The probability of occurrence of the corresponding pixel feature in it indicates that when considering the target feature point In the case of Contribution to the overall conditional entropy. Then add the weighted information entropy of all feature blocks to get the The conditional entropy of the entire fused block image under the condition of .

[0073] If the conditional entropy is small, it means that at the known target feature point After that, the information in the image is relatively certain; on the contrary, if the conditional entropy is large, it means that even if the target feature points are known, , there are still many uncertainties in the image.

[0074] The detail contribution of the target feature point can be determined by the following formula 4: Formula 4 In formula 4, Used to characterize target feature points The contribution of details, Used to characterize the overall information entropy of the c-th fused block image, Used to characterize target feature points Conditional entropy of the c-th fused block image.

[0075] Among them, when the target feature points are introduced After that, the uncertainty of the fused block images is reduced, and the information entropy is reduced from Reduce to Difference This represents the degree of reduction; by dividing the difference by the original information entropy , that is, the normalized contribution , whose value is between 0 and 1.

[0076] Target feature points The detail contribution is close to 1, which means that the target feature point It contributes greatly to the image information entropy, that is, the target feature point Carries a large amount of image information; target feature points The detail contribution of is close to 0, which means that the target feature point The contribution to the image information entropy is very small, that is, the target feature point The impact on image information is not significant.

[0077] This embodiment accurately quantifies the information contribution of each feature point to the fused block image. By calculating the difference between the overall information entropy and the conditional entropy, the influence of a feature point on the overall image information can be effectively measured. Normalizing this difference by the overall information entropy makes the detail contribution comparable across different images. This information-theoretic approach objectively reflects the importance of feature points, providing a reliable basis for subsequent feature weight determination, thereby improving the accuracy and effectiveness of geological image fusion.

[0078] In some of the above-mentioned schemes of the present invention, when determining the basic reference degree of the feature block by the detail contribution of the feature point, relying solely on the detail contribution of the feature point may lead to insufficient evaluation of the stability of the feature block, and the local gradient changes are not effectively taken into consideration, affecting the accuracy of the basic reference degree, thereby reducing the quality of the fused image.

[0079] In this regard, the present invention further proposes that S203 may specifically include: Obtaining a first gradient mean of pixels within a preset neighborhood of each feature point in the target feature block, and a second gradient mean of pixels within a preset neighborhood of each corresponding matching feature point in the fused block image, where the target feature block is any feature block; The basic reference degree of the target feature block is determined by using the first gradient mean values, the second gradient mean values and the detail contribution of each feature point in the target feature block.

[0080] In this embodiment, the preset neighborhood range can be set to a 3×3 or 5×5 pixel area centered on the feature point, and the first gradient mean and the second gradient mean are obtained by calculating the mean gradient amplitude of the pixels in the neighborhood.

[0081] Specifically, when determining the basic reference degree of the target feature block, the gradient information in the neighborhood of each feature point is first extracted, and the first gradient mean and the second gradient mean are used to reflect the local texture stability in the target feature block of the original geological image and the fused block image, respectively. Combined with the detail contribution, the basic reference degree can simultaneously characterize the texture consistency of the feature block and the information contribution ability of the feature point. For example, when the gradient mean of a feature block is high and the detail contribution is large, its basic reference degree increases accordingly, indicating that the feature block has a higher reference value in the fusion process. Therefore, the calculation of the basic reference degree not only takes into account the information gain of the feature point, but also quantifies the texture stability of the local area through the gradient mean, making the subsequent determination of the basic description index more accurate, thereby improving the accuracy of multimodal geological image fusion.

[0082] As an example, the basic reference degree of the target feature block can be determined by the following formula 5: Formula 5 In formula 5, The basic reference degree used to characterize the target feature block a, Used to characterize the first gradient mean of the k-th feature point in the target feature block a, The second gradient mean used to characterize the k-th feature point in the target feature block a, It is used to represent the detail contribution of the kth feature point in the target feature block a. exp is used to represent the inverse normalization operation, and K is used to represent the number of feature points in the target feature block a.

[0083] in, Used to measure the first gradient mean of the kth feature point and its second gradient mean The property of inverse normalization exp is that the smaller the difference, the larger the function value; the larger the difference, the smaller the function value, which means that the feature points close to the mean will have a greater weight. Contribution to details Weighting is performed so that the feature points close to the first gradient mean contribute more to the basic reference degree. Ensure that the value of the basic reference degree Fa is within a reasonable range to facilitate comparison of the basic reference degrees of different blocks.

[0084] Basic reference It reflects the comprehensive importance of the target feature block a after considering the detail contribution of its internal feature points and the difference from the mean. The value indicates that the feature block has more important basic features in the geological image and may play a key role in the overall structure and information expression of the geological image.

[0085] This embodiment effectively utilizes the gradient information and detail contribution of the feature point neighborhood to accurately calculate the basic reference degree of the feature block. This more comprehensively reflects the importance of the feature block in the overall image, improving the accuracy of weight allocation in the subsequent fusion process. Furthermore, this method considers the differences between the original image and the fused image, better preserving the unique information of each modality image, thereby improving the quality and reliability of the final fusion result.

[0086] In some of the above-mentioned schemes of the present invention, when determining the basic description indicators of the geological image based on the basic reference degrees of the feature blocks, if a simple superposition or averaging of the basic reference degrees is directly adopted, the distribution differences of the detail contributions within different feature blocks may be ignored, resulting in the basic description indicators being unable to accurately reflect the overall information characteristics of the geological image, thereby affecting the calculation accuracy of subsequent feature weights.

[0087] In this regard, the present invention further proposes that S204 may specifically include: The detail contribution of the feature points in each feature block is processed by averaging to obtain the detail weight of each feature block; The product of the detail weight of each feature block and the basic reference degree is averaged to obtain the basic description index of the geological image.

[0088] In this embodiment, the basic reference degree of the feature block is determined by the gradient mean difference and detail contribution of the matching feature points. The detail contribution reflects the information contribution ability of the feature point in the fused block. The mean processing eliminates the influence of the outlier value of a single feature point on the weight distribution by calculating the arithmetic mean of the detail contribution of all feature points in the feature block. The product processing combines the detail weight with the basic reference degree, so that the basic description index simultaneously includes the information distribution characteristics within the feature block and the cross-modal gradient consistency. The result after mean processing is used as the basic description index to normalize the information gain of the reference feature point.

[0089] Specifically, after obtaining the base reference degree for each feature block of the target geological image, the method first traverses each feature block and extracts the set of detail contributions of all feature points within that block. The detail weight of each feature block is calculated by averaging the values in this set. Subsequently, the detail weight of each feature block is multiplied by its corresponding base reference degree to obtain a weighted block contribution value. Finally, the weighted contribution values of all feature blocks in the geological image are averaged to obtain the basic description index for the geological image. For example, if a feature block contains three feature points with detail contributions of 0.8, 0.7, and 0.9, respectively, the averaged detail weight is 0.8. If the base reference degree of the block is 1.2, the product is 0.96. When the geological image contains 10 feature blocks and the sum of the product results is 9.6, the basic description index is 0.96. This index effectively integrates the distribution characteristics of the contribution of feature points within the block with cross-modal gradient difference information, providing a more accurate benchmark parameter for subsequent information gain normalization.

[0090] As an example, the detail weight of the feature block can be determined by the following formula 6: Formula 6 In formula 6, Used to characterize the detail weight of the target feature block a, It is used to characterize the detail contribution of the kth feature point in the target feature block a, and K is used to characterize the number of feature points in the target feature block a.

[0091] Among them, by summing , merge the detail contributions of all feature points in the target feature block a to reflect the overall detail information of the target feature block a. Multiply by This is to eliminate the influence of the number of feature points so that the detail weight It can fairly reflect the average detail importance of the feature blocks and facilitate comparison between different feature blocks.

[0092] Detail Weight Reflects the importance of the target feature block a at the detail level. The value indicates that the feature points in the target feature block have a higher detail contribution on average, that is, the target feature block plays a more important role in the detail expression of the geological image.

[0093] The basic description index of geological images can be determined by the following formula 7: Formula 7 In formula 7, The basic descriptive index used to characterize the e-th geological image, Used to characterize the detail weight of the wth feature block in the eth geological image, It is used to characterize the basic reference degree of the wth feature block in the eth geological image, and W is used to characterize the number of feature blocks in the eth geological image.

[0094] in, The importance of details and basic reference of each feature block are combined to reflect the comprehensive importance of the feature block in details and basic features. The mean processing is to eliminate the influence of the number of feature blocks, so that the basic description index It can fairly reflect the average comprehensive importance of the e-th geological image and facilitate comparison between different images.

[0095] Basic descriptive indicators Reflects the overall importance of the e-th geological image in terms of details and basic features. The value indicates that the feature blocks of the geological image have higher detail weight and basic reference degree on average, that is, the geological image plays a more important role in the overall description.

[0096] This embodiment effectively and comprehensively considers the detailed information and overall reference value of feature blocks, thereby obtaining more accurate and comprehensive basic descriptive indicators for geological images. This method not only reflects the detailed features of each local area in a geological image, but also reflects the overall characteristics of the entire image, providing a more reliable foundation for subsequent image fusion and analysis. This improves the accuracy and reliability of digital mapping for geological exploration and facilitates better capture and utilization of key information in geological images.

[0097] In some of the above-mentioned solutions of the present invention, when geological images are divided and fused by feature point matching, if the feature point matching accuracy is insufficient or the feature block division is unreasonable, it may cause distortion in local areas during the fusion process, affecting the accuracy of digital mapping.

[0098] In this regard, the present invention further proposes that S102 may specifically include: Perform feature point detection on each geological image to obtain the feature points in each geological image; Match the feature points in each geological image to obtain the feature point matching relationship; Based on the feature point matching relationship, determining the reference feature points in the target geological image, where the target geological image is any geological image; Region growing is performed using each reference feature point in the target geological image as a seed point until a preset stop-growth condition is reached, thereby obtaining multiple feature blocks corresponding to the target geological image.

[0099] In this embodiment, feature point detection can use a scale-invariant feature transformation algorithm or an accelerated robust feature algorithm to extract feature points with rotation and scale invariance. Feature point matching measures similarity by calculating the Euclidean distance or Hamming distance between feature points, and selects successfully matched feature points as reference feature points. During the region growing process, the seed point serves as the initial growth center, and the region is expanded based on the grayscale difference, gradient change, or texture similarity of adjacent pixels. The preset stop growth conditions may include a region area threshold, a growth number limit, or a feature difference threshold for adjacent regions. The division of feature blocks is based on successfully matched reference feature points to ensure that the block boundaries are consistent with the trend of geological structure changes.

[0100] Specifically, feature point detection first extracts key feature points from geological images and establishes correspondences between different images through a matching algorithm. Reference feature point selection criteria are based on feature points with stable matching relationships across all modal images to avoid errors introduced by isolated feature points. During region growing, the expansion direction of the seed point is constrained based on the gradient direction of adjacent pixels, prioritizing expansion toward regions with continuous textures. When region growth meets preset conditions, expansion stops and a closed feature block boundary is generated. This method guides region growth by matching feature points, allowing block division to simultaneously take into account geometric alignment and texture consistency, improving the reliability of subsequent feature weight calculations.

[0101] As an example, feature point detection is first performed on each geological image to obtain the feature points in each geological image. Feature point detection can use the SIFT algorithm. For example, for a 512×512 pixel geological image, 500 feature points can be detected.

[0102] Then, the feature points in each geological image are matched to obtain the feature point matching relationship. The matching can be performed using the nearest neighbor matching algorithm, with the distance ratio threshold set to 0.7.

[0103] Based on the feature point matching relationship, reference feature points are determined in the target geological image. The target geological image is any geological image. Feature points that have successfully matched other geological images can be selected as reference feature points.

[0104] Region growing is then performed using each reference feature point in the target geological image as a seed point until a preset stop-growth condition is reached, resulting in multiple feature blocks corresponding to the target geological image. Region growing can be performed using a grayscale similarity criterion, with a grayscale difference threshold of 10. The stop-growth condition can be set when the region area reaches 1000 square pixels or when the number of boundary pixels reaches 100.

[0105] This embodiment adaptively partitions feature blocks within each geological image based on the matching relationships between feature points in multimodal geological images. This partitioning method fully considers the correspondence between images of different modalities, resulting in more reasonable partitioning results. Compared to traditional fixed-size partitioning methods, the solution of the present invention better preserves the integrity of geological features, providing a good foundation for subsequent feature weight calculation and image fusion, thereby improving the quality and accuracy of the final fused geological image.

[0106] In some of the above-mentioned schemes of the present invention, when multimodal geological images are fused based on feature weights, the geometric differences between the features of single modal images are not effectively eliminated, resulting in pixel misalignment or deformation in the local area after fusion, affecting the integrity of the geological information.

[0107] In this regard, the present invention further proposes that S104 may specifically include: According to the feature weights of each reference feature point in each geological image, the interpolation method is used to assign corresponding pixel weights to each pixel point in each geological image; Constructing a second geometric transformation model based on the matching relationship of reference feature points between the geological images; Based on the second geometric transformation model, the geological images are resampled to obtain geometrically aligned resampled images; According to the pixel weight of each pixel in the resampled image, the pixel features of each resampled image are fused to obtain the target fused geological image of the target area.

[0108] In this embodiment, the pixel weight allocation process is implemented through interpolation methods, such as bilinear interpolation or cubic spline interpolation, to ensure that the weight changes continuously with the spatial position; the second geometric transformation model is constructed by the coordinate correspondence of the reference feature points, including translation, rotation or scaling transformation parameters; the resampling process adopts the nearest neighbor interpolation or bicubic convolution algorithm to eliminate the geometric misalignment between images; pixel feature fusion adopts weighted average or maximum value retention strategy, and combines the weight of each pixel point to perform multimodal data integration.

[0109] Specifically, in the pixel weight assignment stage, with the high feature weight of the reference feature point as the center, the interpolation method is used to diffuse to the surrounding pixels to form a weight distribution map, so that important areas have a higher fusion priority. In the geometric transformation model construction stage, based on the matching reference feature point pairs, the transformation matrix is calculated by the least squares method to establish the spatial correspondence between images. In the resampling stage, the transformation matrix is used to perform geometric correction on the original image to ensure that images of different modalities are aligned in the same coordinate system. Finally, in the pixel fusion stage, the pixel values of multiple resampled images are weighted according to the weight of each pixel point. For example, the normalized weighted sum of multiple pixel values at the same coordinate position is performed to suppress the distortion caused by geometric misalignment and improve the accuracy of the fusion result.

[0110] As an example, an inverse distance weighted interpolation algorithm is first used to generate a dynamically assigned pixel weight map for each pixel in the geological image based on the feature weight values of the formation reflection feature points. Then, based on the coordinates of the matched reference feature points, an affine transformation model is constructed using the least squares method. This model includes translation, rotation, and non-uniform scaling parameters. During the resampling process, a bilinear interpolation method is used to geometrically correct the visible light image and the infrared thermal imaging image, so that the rock layer boundaries of the two sets of images are aligned at the sub-pixel level in the Cartesian coordinate system. Finally, a weighted averaging strategy is used to fuse the multimodal images. The pixel fusion weights in the fault region are dynamically adjusted by the fracture zone information gain of the corresponding feature point, while the sedimentary region uses an adaptive fusion coefficient based on texture complexity.

[0111] This embodiment effectively addresses the problem of distortion in multimodal geological image fusion in complex terrain. A dynamic pixel weighting mechanism enhances the ability to retain features of rock fracture zones and folds. Precise registration of geometric transformation models eliminates pixel misalignment caused by differences in sensor perspectives in multi-source data. A multi-dimensional feature fusion strategy comprehensively improves the accuracy of the fused image in representing texture details and macroscopic geological structures.

[0112] Based on the geological exploration digital mapping method provided by the present invention, the present invention also provides a specific embodiment of the geological exploration digital mapping system.

[0113] like Figure 3 As shown, the geological exploration digital mapping system 300 provided by the embodiment of the present invention includes an image acquisition module 310 , an image segmentation module 320 , a weight determination module 330 and an image fusion module 340 .

[0114] An image acquisition module 310 is used to acquire a multimodal geological image of a target area; An image segmentation module 320 is configured to segment each geological image into a plurality of feature blocks based on feature point matching relationships between the geological images. The feature blocks are used to represent local image regions obtained by region growing using reference feature points as seed points. Reference feature points are feature points in a geological image that have successfully matched other geological images. A weight determination module 330 is configured to determine a feature weight of each reference feature point in each geological image based on the texture features of each feature block and the information gain of each reference feature point; The image fusion module 340 is used to fuse the multimodal geological images of the target area based on the feature weights of the reference feature points in each geological image to obtain a target fused geological image of the target area.

[0115] In the digital mapping system for geological exploration provided by an embodiment of the present invention, a multimodal geological image of the target area is first acquired. Then, based on the matching relationship between feature points in each image, the image is divided into feature blocks obtained by regional growth using successfully matched reference feature points as seed points. This method can more accurately locate local areas of the image and provide more detailed units for subsequent processing. Then, based on the texture features of each feature block and the information gain of each reference feature point, the feature weights of the reference feature points in each image are determined. This method, which considers the correlation between multimodal images and the feature information gain, can more scientifically measure the importance of each feature in the fusion process and avoid the limitations of relying solely on single-modal image features. Finally, the multimodal geological images are fused based on the feature weights of each reference feature point. Due to the more reasonable weight determination, the features of each modal image can be better coordinated and focused, thereby effectively reducing the distortion of the fused geographic information data in some data dimensions, making the final target fused geological image more accurate, thereby improving the accuracy of digital mapping.

[0116] Based on the geological prospecting digital mapping method provided by the present invention, the present invention also provides a specific embodiment of the geological prospecting digital mapping device.

[0117] Figure 4 The figure shows a hardware structure diagram of a digital surveying and mapping device for geological exploration provided by an embodiment of the present invention.

[0118] The geological exploration digital mapping device may include a processor 401 and a memory 402 storing computer program instructions.

[0119] Specifically, the processor 401 may include a central processing unit, or a specific integrated circuit, or may be configured to implement one or more integrated circuits of the embodiment of the present invention.

[0120] Memory 402 may include a large capacity memory for data or instructions. By way of example and not limitation, memory 402 may include a hard drive, a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus drive, or a combination of two or more of these. Where appropriate, memory 402 may include removable or non-removable (or fixed) media. Where appropriate, memory 402 may be internal or external to the integrated gateway disaster recovery device. In certain embodiments, memory 402 is a non-volatile solid-state memory.

[0121] The processor 401 reads and executes computer program instructions stored in the memory 402 to implement any one of the digital mapping methods for geological exploration in the above embodiments.

[0122] In one example, the geological prospecting digital mapping device may further include a communication interface 403 and a bus 410. Figure 4 As shown, the processor 401 , the memory 402 , and the communication interface 403 are connected via a bus 410 and communicate with each other.

[0123] The communication interface 403 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiment of the present invention.

[0124] Bus 410 comprises hardware, software or both, and couples the components of global positioning device to each other. By way of example and not limitation, bus can comprise accelerated graphics port or other graphics bus, enhanced industry standard architecture bus, front side bus, hypertransport interconnect, industry standard architecture bus, infinite bandwidth interconnect, low pin count bus, memory bus, micro channel architecture bus, peripheral component interconnect bus, serial advanced technology attachment bus, video electronics standard association local bus or other suitable bus or two or more of these combinations. Where appropriate, bus 410 can comprise one or more buses. Although the embodiment of the present invention describes and shows a specific bus, the present invention contemplates any suitable bus or interconnect.

[0125] It should be understood that the present invention is not limited to the specific configurations and processes described above and illustrated in the figures. For the sake of brevity, a detailed description of known methods is omitted. In the above embodiments, several specific steps are described and illustrated as examples. However, the method of the present invention is not limited to the specific steps described and illustrated. Those skilled in the art may make various changes, modifications, and additions, or change the order of the steps after understanding the spirit of the present invention.

[0126] It should also be noted that the exemplary embodiments described herein describe methods or systems based on a series of steps or devices. However, the present invention is not limited to the order of the steps described above. In other words, the steps may be performed in the order described in the embodiments, or in a different order, or several steps may be performed simultaneously.

[0127] The above description is only a specific embodiment of the present invention. Those skilled in the art will clearly understand that for the convenience and brevity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present invention is not limited to this. Any person skilled in the art can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention.

Claims

1. A digital mapping method for geological exploration, characterized in that: The method comprises: Acquire multimodal geological images of the target area; Based on the matching relationship between the feature points of the geological images, each geological image is divided into a plurality of feature blocks, wherein the feature blocks are used to represent a local image region obtained by region growing using a reference feature point as a seed point, wherein the reference feature point is a feature point in the geological image that successfully matches other geological images; Determining the feature weight of each reference feature point in each geological image based on the texture feature of each feature block and the information gain of each reference feature point; Based on the feature weight of each reference feature point in each geological image, the multimodal geological images of the target area are fused to obtain a target fused geological image of the target area.

2. The digital surveying and mapping method for geological exploration according to claim 1, characterized in that: The step of determining the feature weight of each reference feature point in each geological image based on the texture feature of each feature block and the information gain of each reference feature point comprises: Fusing the matching feature blocks between the geological images to obtain a fused block image; Determining the detail contribution of each of the feature points in the fused block image based on the overall information entropy of the fused block image and the conditional entropy of each of the feature points in the fused block image to the fused block image; Determining the basic reference degree of each feature block based on the detail contribution of each feature point in each feature block; Determining a basic description index of the geological image based on a basic reference degree of each of the characteristic blocks in the geological image; The information gain of each reference feature point is divided by the corresponding basic description index of the geological image to obtain the feature weight of each reference feature point in each geological image.

3. The digital surveying and mapping method for geological exploration according to claim 2, characterized in that: The step of fusing the feature blocks that match the geological images to obtain a fused block image includes: Acquiring a matching relationship between feature points of each of the geological images; Based on the feature point matching relationship, a first geometric transformation model is constructed, wherein the geometric transformation model is used to characterize at least one of a translation transformation relationship, a rotation transformation relationship, and a scaling transformation relationship between the geological images; resampling the feature blocks that match each of the geological images based on the first geometric transformation model to obtain resampled blocks that are geometrically aligned with each other; An image fusion operation based on wavelet transform is performed on each of the resampled blocks to obtain the fused block image.

4. The digital surveying and mapping method for geological exploration according to claim 2, characterized in that: The determining, based on the overall information entropy of the fused block image and the conditional entropy of each feature point in the fused block image to the fused block image, the detail contribution of each feature point in the fused block image, includes: Determining the overall information entropy of the fused block image by using the pixel features of each pixel in the fused block image, and determining the block information entropy of each feature block by using the pixel features of each pixel in each feature block, wherein the pixel features include at least one of grayscale, gradient value, and gradient direction; Determine the conditional entropy of the target feature point on the fused block image using the pixel features of the target feature point and the information entropy of each block, where the target feature point is any one of the feature points in the fused block image; The difference between the overall information entropy and the conditional entropy of the target feature point to the fused block image is divided by the overall information entropy to obtain the detail contribution of the target feature point.

5. The digital surveying and mapping method for geological exploration according to claim 2, characterized in that: The determining of the basic reference degree of each feature block based on the detail contribution of each feature point in each feature block includes: Obtaining a first gradient mean of pixels within a preset neighborhood of each feature point in a target feature block, and a second gradient mean of pixels within a preset neighborhood of each corresponding matching feature point in the fused block image, wherein the target feature block is any one of the feature blocks; The basic reference degree of the target feature block is determined by using the first gradient mean values, the second gradient mean values, and the detail contribution of each feature point in the target feature block.

6. The digital surveying and mapping method for geological exploration according to claim 2, characterized in that: Determining the basic description index of the geological image based on the basic reference degree of each feature block in the geological image includes: Performing mean processing on the detail contribution of the feature points in each feature block to obtain the detail weight of each feature block; The product of the detail weight of each feature block and the basic reference degree is averaged to obtain the basic description index of the geological image.

7. The digital surveying and mapping method for geological exploration according to any one of claims 1 to 6, characterized in that: The step of dividing each geological image into a plurality of feature blocks based on the feature point matching relationship between the geological images comprises: Performing feature point detection on each of the geological images to obtain the feature points in each of the geological images; Matching the feature points in each of the geological images to obtain a feature point matching relationship; Determining the reference feature points in a target geological image based on the feature point matching relationship, where the target geological image is any one of the geological images; Region growing is performed using each reference feature point in the target geological image as a seed point until a preset stop-growth condition is reached, thereby obtaining a plurality of feature blocks corresponding to the target geological image.

8. The digital surveying and mapping method for geological exploration according to any one of claims 1 to 6, characterized in that: The step of fusing the multimodal geological images of the target area based on the feature weights of the reference feature points in the geological images to obtain a target fused geological image of the target area includes: According to the feature weight of each reference feature point in each geological image, a corresponding pixel weight is assigned to each pixel point in each geological image using an interpolation method; constructing a second geometric transformation model based on the matching relationship of reference feature points between the geological images; resampling the geological images based on the second geometric transformation model to obtain geometrically aligned resampled images; According to the pixel weight of each pixel point in the resampled image, the pixel features of each resampled image are fused to obtain a target fused geological image of the target area.

9. A digital mapping system for geological exploration, characterized in that: The system comprises: An image acquisition module, used to acquire multimodal geological images of the target area; An image segmentation module is configured to segment each geological image into a plurality of feature blocks based on a matching relationship between feature points of each geological image. The feature blocks are used to represent a local image region obtained by performing region growing using a reference feature point as a seed point. The reference feature point is a feature point in the geological image that successfully matches other geological images. A weight determination module, configured to determine the feature weight of each reference feature point in each geological image based on the texture features of each feature block and the information gain of each reference feature point; The image fusion module is used to fuse the multimodal geological images of the target area based on the feature weight of each reference feature point in each geological image to obtain a target fused geological image of the target area.

10. A digital surveying and mapping device for geological exploration, characterized in that: The apparatus comprises: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the digital mapping method for geological exploration according to any one of claims 1 to 8 is implemented.

Images