Mineral exploration drilling electronic map element sorting method, system and equipment and medium

Abstract

The invention provides a mineral exploration drilling electronic map element sorting method, a mineral exploration drilling electronic map element sorting system, mineral exploration drilling electronic map element sorting equipment and a medium, and belongs to the technical field of geological information processing. Correspondingly splitting the table into a plurality of independent text image blocks and lithologic symbol image blocks according to contents in the cells based on the table lines; recognizing texts in the plurality of independent text image blocks by using an optical character recognition algorithm to obtain text elements in the mineral exploration drilling electronic map; and training the convolutional neural network, inputting the plurality of independent lithologic symbol image blocks into the trained convolutional neural network, and carrying out lithologic symbol identification to obtain different types of lithologic symbol elements. According to the method, the problems of large map size, multiple table rows and columns, complex data structure and dense text tables can be solved, the accuracy of information sorting is effectively improved, and the utilization efficiency and fusion sharing of mineral exploration data are promoted.

Description

Technical Field

[0001] The present invention belongs to the technical field of geological information processing, and in particular relates to a method, system, equipment and medium for sorting elements of electronic maps of mineral exploration boreholes. Background Art

[0002] Data and information at different stages of the mineral resource exploration process are the basis for scientific decision-making in mineral resource exploration. As the difficulty of prospecting increases, rich and high-quality mineral resource exploration data, especially deep detection data revealed by drilling, has become one of the key data for intelligent mineral resource exploration. At present, a large amount of mineral resource exploration drilling data has only reached the level of electronicization and has not yet been fully digitized, becoming a bottleneck of data resources for information utilization, and it is urgent to make up for the shortcomings and release the data value. The application of artificial intelligence theory, methods and technologies can help make up for the shortcomings in the transition from electronic mineral exploration drilling maps to digitalization, give full play to the value of mineral resource exploration big data, and provide important technical accumulation and data support for a new round of mineral resource exploration and evaluation.

[0003] In actual work, electronic drawings of mineral resource exploration drilling holes are mostly hand-drawn or produced by professional software, and there are problems such as inconsistent preset templates and complex drawing structures, and their information extraction work faces many challenges. At present, most automated recognition software only has high recognition accuracy for well-structured text or tables. The information sorting of unstructured elements (pictures, videos, audio and other data) of electronic drawings of mineral resource exploration drilling holes mostly relies on manual work, which is inefficient, low in accuracy and high in cost, limiting the comprehensive utilization of geological data.

[0004] Existing methods for identifying electronic drawings focus on the recognition and digital processing of text, and have high text recognition accuracy. However, for electronic drawings of mineral resource exploration drill holes, the drawings are large in size, have many rows and columns in the table, have a complex data structure, and dense text tables. Existing methods cannot handle this type of drawings well, which limits the information mining and application of unstructured data in geological big data, and is not conducive to promoting the integration and sharing of mineral exploration data. Summary of the invention

[0005] In order to overcome the shortcomings of the above-mentioned prior art, the present invention provides a method for sorting elements of electronic maps of mineral exploration drilling holes, comprising the following steps:

[0006] Identify table lines in electronic drawings of mineral exploration drilling holes, and split the table into multiple independent text image blocks and multiple independent lithology symbol image blocks according to the contents in the cells based on the table lines;

[0007] An optical character recognition algorithm is used to recognize text in multiple independent text image blocks to obtain text elements in mineral exploration drill hole electronic maps. A lithology symbol image database is constructed based on the lithology symbols appearing in the mineral exploration drill hole electronic maps, and the images in the lithology symbol image database are input into a convolutional neural network, which is trained. Multiple independent lithology symbol image blocks are input into the trained convolutional neural network to perform lithology symbol recognition and obtain different types of lithology symbol elements.

[0008] Preferably, the method of identifying table lines in the electronic map of mineral exploration drilling holes and splitting the table into a plurality of independent text image blocks and lithology symbol image blocks according to the contents in the cells based on the table lines comprises the following steps:

[0009] Identify the horizontal and vertical lines, number of rows and columns in the electronic drawings of mineral exploration drill holes, set the window according to the threshold, and filter out the horizontal and vertical lines of the cells by sliding the window;

[0010] Merge the intersection of the horizontal and vertical lines of the cells, calculate the coordinates of the intersection in the mineral exploration drill hole electronic map, restore the table structure of the mineral exploration drill hole electronic map according to the coordinates, and store the table structure of the mineral exploration drill hole electronic map in the form of point coordinates and line coordinates;

[0011] The mineral exploration drilling electronic map is split into multiple independent text image blocks and lithology symbol image blocks according to the coordinates.

[0012] Preferably, before identifying the table lines in the mineral exploration drilling electronic map, the method further includes preprocessing the mineral exploration drilling electronic map, including the following steps:

[0013] Conduct red stamp detection on electronic drawings of mineral exploration drilling holes, extract pixels whose red channel is greater than the threshold, assign them as background values, and eliminate red stamps;

[0014] The electronic maps of mineral exploration drilling holes after removing the red seal are binarized and morphologically operated by the set threshold to eliminate the noise and ambiguity in the electronic maps of mineral exploration drilling holes;

[0015] The page direction of the mineral exploration drilling electronic map after noise and blur are eliminated is detected and rotated, and the image is vertically projected, the histogram peak of the projection image is analyzed, and the mineral exploration drilling electronic map with non-positive direction is adjusted;

[0016] The table in the rectified mineral exploration drill hole electronic map is extracted, the distorted table is corrected, and other background areas are cropped.

[0017] Preferably, the electronic map of mineral exploration drilling holes after eliminating the red stamp is binarized and morphologically operated by a set threshold, specifically: converting the color image into a gray image, binarizing the grayscale image by a set threshold, enhancing the image features by dilation and corrosion operations, and eliminating noise and blur in the electronic map of mineral exploration drilling holes.

[0018] Preferably, after obtaining different types of lithology symbol elements, the method further includes correcting the text elements according to the obtained different types of lithology symbol elements, including the following steps:

[0019] According to the confidence of text element recognition and the confidence of lithology symbol element, the correct value of the stratum is determined;

[0020] Through manual interpretation, check whether the identification content of the electronic map of mineral exploration drilling is correct.

[0021] Preferably, the correct value of the stratum is judged according to the text element recognition confidence and the lithology symbol element confidence, specifically: whether the text element recognition confidence is greater than a set threshold, if so, the stratum is considered to be the stratum described in the text element; if the text element recognition confidence is not greater than the set threshold, whether the lithology symbol element confidence is greater than the set threshold, if so, the stratum is considered to be the stratum corresponding to the lithology image classification result; if the lithology symbol element confidence is still not greater than the set threshold, the text recognition result is used as a temporary output, and an indication that manual inspection is required is given during manual verification.

[0022] The present invention also provides a mineral exploration drilling electronic map element sorting system, comprising:

[0023] A table splitting module is used to identify table lines in electronic drawings of mineral exploration drilling holes, and split the table into multiple independent text image blocks and multiple independent lithology symbol image blocks according to the contents in the cells based on the table lines;

[0024] The element sorting module is used to use an optical character recognition algorithm to recognize text in multiple independent text image blocks to obtain text elements in the electronic map of mineral exploration drilling holes; build a lithology symbol image database based on the lithology symbols appearing in the electronic map of mineral exploration drilling holes, input the images in the lithology symbol image database into a convolutional neural network, train the convolutional neural network, input multiple independent lithology symbol image blocks into the trained convolutional neural network, perform lithology symbol recognition, and obtain different types of lithology symbol elements.

[0025] The present invention also provides a computer device, including a memory and a processor; the memory stores a computer program, and the processor is used to run the computer program in the memory to execute the mineral exploration drilling electronic map element sorting method.

[0026] The present invention also provides a computer-readable storage medium, which stores a computer program. The computer program is suitable for loading by a processor to execute the method for sorting elements of electronic maps of mineral exploration drilling holes.

[0027] The mineral exploration drilling electronic map element sorting method, system, equipment and medium provided by the present invention have the following beneficial effects:

[0028] The present invention identifies table lines in electronic drawings of mineral exploration boreholes, and based on the table lines, splits the table into a plurality of independent text image blocks and lithology symbol image blocks according to the contents in the cells, and obtains independent text image blocks and lithology symbol image blocks. This process can split the elements in the electronic drawings of mineral exploration boreholes, ensure that the text segments in the table are independent, and avoid erroneous recognition problems due to the adhesion of numbers and texts that are too close to each other in the table structure, thereby solving the problems of large map size, many table rows and columns, complex data structure, and dense text tables, thereby facilitating the accurate recognition of different types of elements in complex electronic drawings of mineral exploration boreholes.

[0029] The present invention can obtain the recognition result of text information by recognizing text image blocks; and can obtain the recognition result of lithology graphic symbol information by recognizing lithology symbol image blocks. The present invention can complete the effective utilization of all elements of the map by separately recognizing tables, texts, and lithology symbols, effectively improve the accuracy of information sorting, and promote the utilization efficiency and integration sharing of mineral exploration data. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiment of the present invention and its design scheme, the following briefly introduces the drawings required for this embodiment. 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 creative work.

[0031] Figure 1 A schematic flow chart of an embodiment of a method for sorting element information of electronic drawings of mineral exploration boreholes provided for the implementation of the present invention;

[0032] Figure 2 for Figure 1 Detailed implementation steps of step S1 in FIG.

[0033] Figure 3 for Figure 1 Detailed implementation steps of step S2 in the flowchart;

[0034] Figure 4 for Figure 1Detailed implementation steps of step S4 in FIG.

[0035] Figure 5 for Figure 1 Detailed implementation steps of step S5;

[0036] Figure 6 This is a diagram of common text recognition errors in common OCR algorithms, and errors checked based on corpus, rules, and image recognition;

[0037] Figure 7 It is the result of sorting the element information of electronic drilling maps for mineral exploration, among which: Figure 7 (b) and (c) are Figure 7 Enlarged image of the two red boxes on the left and right in (a). DETAILED DESCRIPTION

[0038] In order to enable those skilled in the art to better understand the technical solution of the present invention and implement it, the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the scope of protection of the present invention.

[0039] In the description of the present invention, it is to be understood that the terms “center”, “longitudinal”, “lateral”, “length”, “width”, “thickness”, “up”, “down”, “front”, “back”, “left”, “right”, “vertical”, “horizontal”, “top”, “bottom”, “inside”, “outside”, “axial”, “radial”, “circumferential”, etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the technical solutions of the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0040] In addition, the terms "first", "second", etc. are used for descriptive purposes only and are not to be understood as indicating or implying relative importance. In the description of the present invention, it should be noted that, unless otherwise clearly specified or limited, the terms "connected" and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to the specific circumstances. In the description of the present invention, unless otherwise specified, "plurality" means two or more, which will not be described in detail here.

[0041] Example

[0042] The present invention provides a method for sorting elements of electronic maps of mineral exploration drilling holes, specifically, Figure 1 As shown, the following steps are included:

[0043] Step S1: pre-processing the electronic map of mineral exploration drilling.

[0044] Among them, the file of the electronic drawing of mineral exploration boreholes can be a graphic file stored in electronic form, which is used to display the stratigraphic and sample attribute information of the exploration boreholes. It is mainly aimed at the electronic drawings obtained by scanning the original paper drawings. It can be an image file (such as JPEG, PNG, TIF, etc.) or a portable file (PDF format). The file contains all the visual elements of the electronic drawing of mineral exploration boreholes, such as text, lines, symbols, etc., and can be in color format or monochrome format. The electronic drawing of mineral exploration boreholes is used to reflect the properties of different strata and rocks passed through by the boreholes during mineral exploration. It comprehensively records the basic information of the boreholes and the geological information of each layer. The geological information includes the name of the stratum, thickness, lithology description, sampling location, sample attributes, etc., and is presented in an intuitive graphical way. In geological exploration and mineral development, the electronic drawing of mineral exploration boreholes is an important basis for geological engineers and researchers to conduct geological interpretation and mineral evaluation.

[0045] It is understandable that the preprocessing of electronic drawings of mineral exploration boreholes directly affects the accuracy and efficiency of subsequent information extraction. In the process of automatically identifying the content of graphic tables and sorting element information, determining the range of the table is a prerequisite for accurately identifying different geological information elements (such as stratigraphic lines, lithological symbols, etc.). Only by clarifying the boundary range can the elements within the boundary be accurately analyzed and processed. Before identifying the content of the graphic table, the original image needs to be subjected to red seal detection, binarization processing and morphological operations to eliminate the interference of non-graphic elements. Only by solving the noise and blur problems in historical maps and avoiding the misjudgment and extraction of irrelevant information can the accuracy and accuracy of recognition be improved. When performing recognition, it is necessary to determine the correct page direction and adjust the appropriate page angle. Only by obtaining the corrected image can the correct table structure and text information be extracted.

[0046] Step S2: identifying table lines in the electronic map of mineral exploration drilling, and splitting the table into a plurality of independent text image blocks and a plurality of independent lithology symbol image blocks according to the contents in the cells based on the table lines.

[0047] Among them, the table refers to the framework structure that constitutes the electronic drawing of mineral exploration drilling, which is mainly composed of horizontal and vertical line segments. The structural rows and columns of the table in the drawing are not necessarily consistent, and there may be more complex merge and split relationships. The length of each row or column of the table may be inconsistent.

[0048] It is understandable that the correct extraction of table structure is of great significance for sorting table element information in drawings. At the same time, the splitting of tables directly affects the accuracy and efficiency of subsequent information extraction. In text recognition, traditional methods are used to simultaneously recognize the entire page of text, which makes it difficult to deal with the problems of large drawing size, many table rows and columns, complex data structure, and dense text tables in electronic drawings of mineral exploration drilling. Through table line recognition, the content of each unit can be distinguished and identified, which helps to improve the accuracy of text recognition. At the same time, by segmenting the lithology symbols according to the table structure, the split symbols can be classified into independent images.

[0049] Step S3: using an optical character recognition algorithm to recognize text in multiple independent text image blocks, and obtaining text elements in the mineral exploration drilling electronic map.

[0050] Among them, text refers to various types of text in the electronic map of mineral exploration drilling, including Chinese characters, English characters, numbers and special symbols. After step S2, all texts are split into independent image blocks, and character detection and recognition are performed separately. However, the age symbols of some strata are composed of combined characters, and the accuracy is poor when they are recognized separately. This type of symbol should be classified and information extracted using the image recognition algorithm in step S4.

[0051] It can be understood that the recognition of basic text mainly depends on the optical character recognition (OCR) algorithm. The existing OCR algorithm is relatively mature. In the embodiment of the present invention, it can be selected according to specific circumstances and needs.

[0052] Step S4: training the convolutional neural network, inputting multiple independent lithology symbol image blocks into the trained convolutional neural network, performing lithology symbol recognition, and obtaining different types of lithology symbol elements.

[0053] Among them, lithology symbols are an important part of the geological information expression in the electronic drawings of mineral exploration boreholes. Lithology symbols usually appear in the form of a variety of geometric primitives to represent different lithology information. For example, lines, dots, text and special symbols are one of the most basic primitives, which can appear in different directions, lengths, sizes and thicknesses to represent key information such as the lithology and composition ratio of the strata. For example, dots represent sand, horizontal lines represent mud, and regular arrangements of multiple dots represent silt, etc. In addition to a single image, lithology symbols can also use a combination of multiple graphics, which can form complex patterns to represent mixed lithology compositions, such as siliceous limestone and sandy dolomite. In an embodiment of the present invention, by classifying lithology symbols in the map based on a method based on a convolutional neural network, different geological strata can be accurately determined. This helps to better restore the text and graphic symbol elements in the map through the distribution of strata.

[0054] It is understandable that in step S3, some combined characters, such as combined geological age symbols, are difficult to obtain high recognition accuracy through conventional OCR algorithms, so this type is recognized by the same method as step S4. By building a lithology image symbol library, the actual semantic information corresponding to the geological age symbol is obtained, so as to sort the element information of this type.

[0055] Specifically, before image symbol recognition, a lithology image symbol library of a certain scale with symbol and name relationship can be established. The same lithology may have multiple lithology symbols, so it is necessary to continuously expand the lithology image symbol library to ensure the accuracy of lithology digitization. In the present invention, after each new map is identified, classified and the result is confirmed, the symbol and its corresponding lithology are entered into the lithology image symbol library, thereby effectively expanding the lithology image symbol library while identifying.

[0056] Step S5: Correct the text recognition results based on the stratigraphic symbol information sorting results and manual review.

[0057] It is understandable that the existing recognition results of all texts, tables, and graphic symbols, whether based on knowledge rules or deep learning, cannot be completely correct. Therefore, it is necessary to correct the recognition results. Correction can include a combination of manual correction and machine-assisted judgment correction, which can improve work efficiency and reduce workload as much as possible while ensuring the accuracy rate.

[0058] Step S6: Combining knowledge rules and entity semantic recognition, the information sorting results are stored in a structured manner.

[0059] Among them, the knowledge rule can be the spatial relationship between each element title and the element. Specifically, in some embodiments, by summarizing the left and right spatial relationship between the texts such as "hole number" and "drill number" and the actual corresponding values, the digital text appearing on the right side of the "hole number" and "drill number" can be stored with the attribute meaning of the hole number. Entity semantic recognition refers to an algorithm based on natural language processing (NLP) that extracts information of interest to users from text and forms structured data. This method is trained based on a large amount of data, is not limited to specific row and column relationships, and has a wider range of applications, but in some embodiments, the matching accuracy is low. The present invention combines knowledge rules and deep learning methods, complements the results of the two, and takes the maximum matching result as the final result.

[0060] It can be understood that, up to step S5, the identified mineral exploration borehole electronic map elements only have content and coordinate information, but no semantic information. Specifically, for example, when storing, "ZK1" cannot be automatically understood as a hole number, nor can the information in each stratum row be automatically associated. Therefore, it is still in a semi-structured state when stored. Such data is not conducive to the sorting and effective use of information. Through step S6, the unstructured and semi-structured mineral exploration borehole electronic map element information is structured and sorted.

[0061] In some embodiments, the attribute meaning of the element information may not be labeled by the preset knowledge rules or the relationship extraction results in the entity semantic recognition algorithm. In this case, information sorting is achieved through human-computer interaction. Exemplarily, by selecting a set of keys and values, the attribute of the value is labeled. For example, by clicking the "ZK1" unit and the hole number attribute in the dictionary, the value of "ZK1" is assigned to the drilling attribute.

[0062] The present invention pre-processes the electronic map of mineral exploration boreholes through step S1, which can ensure that the image that has been corrected, rectified and enhanced enters the subsequent steps for identification, effectively improving the accuracy of the extraction of map element information. The present invention recognizes and extracts the tables, texts and images in the map in steps S2, S3 and S4 respectively, and can extract and sort the elements of the electronic map of mineral exploration boreholes more accurately and in accordance with the process; at the same time, the electronic map element information sorting method of mineral exploration boreholes designed by the present invention based on knowledge rules and deep learning improves the semantic analysis and automatic sorting capabilities of the recognized text, effectively improves the accuracy and efficiency of information sorting, completes the effective use of all elements of the map, and promotes the utilization efficiency and integrated sharing of mineral exploration data.

[0063] For a detailed flowchart of the implementation steps of step S1, see Figure 2 As shown, the following steps are included:

[0064] Step S11: If a red seal appears in the original image, red seal detection and elimination are required to avoid interference with the recognition result. Red seal detection is performed using an image processing-based algorithm, and pixels with red channels greater than a threshold are extracted and assigned as background values, thereby eliminating the red seal.

[0065] In some embodiments, the method for red seal detection may be to identify the range of the red seal through image processing technology. For example, edge detection algorithm is used to identify edge information in the image, so as to determine the range of the red seal that needs to be eliminated. Alternatively, a similarity detection method may be used. For example, the SIFT algorithm and the text similarity detection algorithm are used to obtain the possible red seals. Alternatively, the pixel channel threshold method may be used, for example, the red channel is greater than the set threshold, or the red channel is greater than the green and blue channels, to filter out the red seal range. The first two methods are more effective for all images; the latter has better effects on color images, and the recognition and removal accuracy is greater than the first two. In the embodiment of the present invention, selection can be made according to specific circumstances and needs.

[0066] Step S12: Binarization and morphological operations are performed on the map after the red stamp is removed to enhance the contrast and clarity of the image, including converting the color image into a gray image, binarizing the gray image through a set threshold, and enhancing the image features through expansion and corrosion operations to solve the noise and blur problems in the electronic maps of historical mineral exploration drill holes.

[0067] In some embodiments, noise is removed and images are enhanced by image morphology methods. For example, corrosion operations are used to remove small objects and details and reduce the boundaries of objects in the image. For each pixel, only the smallest value in its neighborhood is retained, or a dilation method can be used to enhance the boundaries of objects, fill small holes, and expand the boundaries of objects in the image. For each pixel, the maximum value in its neighborhood is retained, or an open operation method can be used to first erode and then dilate, which is usually used to remove small objects or noise while retaining the shape of larger objects. Or, for example, a closed operation method can be used to first dilate and then erode to fill small holes and breaks, which can usually make the edges of objects more continuous. In the embodiments of the present invention, selection can be made according to specific circumstances and needs.

[0068] Step S13: Detect and rotate the page direction by vertically projecting the image and observing the horizontal distribution of the text. By analyzing the histogram peak of the projection image, determine the main direction of the text and adjust the images that are not in the positive direction.

[0069] Step S14: Extract the table area, use the deep learning-based table recognition algorithm (PaddlePaddle PP-YOLOv2), correct the distorted table, extract the basic area range of the table, and crop other background areas.

[0070] The present invention performs preprocessing on the map including red stamp removal, binarization and image enhancement, direction detection and correction, table area extraction and the like, and the result is a clear image with offset interference and noise interference removed. This means that through the preprocessing process, the accuracy of element recognition can be improved in an automated and process-based manner.

[0071] For a detailed flowchart of the implementation steps of step S2, see Figure 3 As shown, the following steps are included:

[0072] Step S21: Identify horizontal and vertical lines, obtain the number of rows and columns of the image, set the window according to the threshold, and filter out the horizontal and vertical line parts respectively by sliding the window. When identifying horizontal lines, the height of the window is 1 and the width is the number of columns divided by 20. When identifying vertical lines, the width of the window is 1 and the height is the number of rows divided by 40.

[0073] In some embodiments, affected by the thickness and fragmentation of the lines, the recognition accuracy of the horizontal and vertical lines in the table can be improved by adjusting the window threshold according to specific circumstances and needs.

[0074] Step S22: merge the intersection points of the horizontal and vertical lines, calculate the coordinates of the intersection points in the image, restore the table structure of the mineral exploration drilling electronic map, and store the map table structure in the form of point coordinates and line coordinates.

[0075] Step S23: Split the original image into local areas and save them according to the segmentation coordinates of the line pair image in the table.

[0076] The present invention identifies table lines and extracts table element information. The key to this step is to ensure that the table is correct and the lines of the table are clear and recognizable, which means that the table needs to be corrected, distorted, and enhanced based on image and morphological algorithms. The result will be a data structure with table position coordinate information, indicating the position and morphology of the table. This means that the table can be disassembled by line recognition to solve the problem of difficult recognition of complex tables in electronic drawing materials for mineral exploration drilling. According to the table lines, the cells are split, and the fragmented text is input into the OCR algorithm in blocks for recognition, which means that each text fragment and graphic fragment will be independently identified and matched, solving the problem of dense text being easily glued together and easily misplaced in the general model.

[0077] The present invention recognizes the text in the table by OCR algorithm. The key to this step is to ensure that the text segments in the table are independent, and the structure of the table will not cause the numbers and texts that are too close to each other to stick together, resulting in erroneous recognition problems, which means that table splitting based on table line recognition is necessary.

[0078] For a detailed flowchart of the implementation steps of step S4, see Figure 4 As shown, the following steps are included:

[0079] Step S41: construct a lithology image symbol library of a certain scale for lithology symbols that may appear in electronic drawings of mineral exploration boreholes. The lithology image symbol library is saved via <mineral lithology name, lithology symbol image>.

[0080] Step S42: Divide the lithology image symbol library into a training set, a validation set, and a test set in proportion, extract image features through a convolutional neural network, build a lithology symbol classification model, and use a cross entropy loss function and an optimization algorithm to update the model parameters.

[0081] Step S43: Use the trained model to predict the lithology symbol category in the input mineral exploration drill hole electronic map, and input the category and its prediction probability.

[0082] Step S44: After each mineral exploration drilling electronic map element information is sorted, the sorted lithology images and their corresponding strata are stored in the lithology image symbol library for updating the recognition model.

[0083] The present invention realizes the recognition and extraction of image element information in electronic maps of mineral exploration drilling holes by recognizing lithological symbol images. Image classification based on Convolutional Neural Network (CNN) can select a variety of appropriate CNN architectures, including ResNet, Inception, VGG, etc. The results of each round of recognition are added to the lithological image symbol library. After adding new images every fixed number of times, training can be carried out again to expand the data set and improve the accuracy of the model. Through image recognition, text recognition errors can be assisted in correction and the accuracy of element recognition can be improved.

[0084] For a detailed flowchart of the implementation steps of step S5, see Figure 5 As shown, the following steps are included:

[0085] Step S51: judging the correct value of the stratum according to the confidence of the text element recognition and the confidence of the symbol recognition. Step S51 specifically includes:

[0086] Step S511: determining whether the text element recognition confidence is greater than a set threshold, if so, the stratum is considered to be the stratum described in the text;

[0087] Step S512: if the text element recognition confidence is not greater than the set threshold, then determine whether the lithology symbol element confidence is greater than the set threshold, if so, then consider that the stratum is the stratum corresponding to the lithology image classification result;

[0088] Step S513: If the confidence of the lithology symbol element is still not greater than the set threshold, the text recognition result is used as a temporary output, and during manual verification, it is indicated that manual inspection is required.

[0089] The present invention combines the text element recognition confidence and the lithology symbol element confidence to judge the recognition result of the stratum lithology, which helps to improve the recognition accuracy of strata that are difficult to identify and have critical element information, such as Figure 6 shown.

[0090] Step S52: using manual interaction and manual interpretation to check the identification content of text information in the electronic map of mineral exploration boreholes, including basic information of the boreholes, stratigraphic depth information, stratigraphic description information, and parameter information.

[0091] Step S6 includes the following steps:

[0092] Step S61: A knowledge rule base is formed based on the common key-value spatial relationship in the electronic map of mineral exploration boreholes; wherein, in the electronic map of mineral exploration boreholes, the key-value spatial relationship refers to the relationship between various geological and exploration data and spatial positions in the map. The "key" here usually refers to the identifier of a specific location or feature in the geographic information (such as the borehole number, coordinates, etc.), and the "value" represents the specific data or attributes related to the location (such as geological strata, mineral content, lithological characteristics, etc.). The core of the key-value spatial relationship refers to how the data at different locations in space are related to each other, and how to reveal the regularity between these data through spatial analysis. For example: in the process of mineral exploration, the location of the borehole (such as the borehole number or coordinates) is a "key", and the "value" associated with it is the mineral content at different depths in the borehole. By visualizing these data in the electronic map, the mineral distribution of different boreholes can be displayed, and the relationship between the mineral content and the borehole depth or geographical location can be analyzed.

[0093] Step S62: For each segment of recognized text and image, the classified information is stored in a structured manner according to knowledge rules, semantic entity recognition and relationship extraction algorithms.

[0094] The beneficial effects of the present invention are as follows: preprocessing electronic drawings of mineral exploration boreholes; using a table line recognition algorithm to split the elements in the drawings, obtaining intelligent recognition results of table information, and obtaining independent text blocks and stratigraphic segments; recognizing the text in the drawings through an optical character recognition algorithm to obtain intelligent recognition results of text information; recognizing lithological symbols in the drawings through an image recognition algorithm based on deep learning to obtain intelligent recognition results of graphic symbol information; correcting the text recognition results in combination with stratigraphic symbol information sorting results and manual review; and performing structured storage of information sorting results in combination with knowledge rules and entity semantic recognition. In this way, through the process from preprocessing to digitization and then to structuring, the automatic sorting of the entire process of electronic mapping information of mineral exploration drill holes can be realized; by identifying tables, texts, and pictures in steps, the effective utilization of all elements of the map can be completed, and the recognition result of a certain element can be used to effectively improve the recognition results of other elements, solving the problems of large map size, many rows and columns of tables, complex data structure, and dense text tables; by combining knowledge rules, deep learning and manual verification, semantic analysis and automatic sorting of recognized texts can effectively improve the accuracy of information sorting and promote the utilization efficiency and integrated sharing of mineral exploration data.

[0095] Based on the same inventive concept, an embodiment of the present invention provides a mineral exploration drilling electronic map element information sorting result, such as Figure 7 As shown, Figure 7 (b) and (c) are Figure 7 Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is defined by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and scope of the equivalent elements of the application documents are included in the present invention.

[0096] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined in the present invention can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown in the present invention, but should conform to the widest scope consistent with the principles and novel features of the present invention.

[0097] The present invention also provides a mineral exploration drilling electronic map element sorting system, comprising:

[0098] A table splitting module is used to identify table lines in electronic drawings of mineral exploration drilling holes, and split the table into multiple independent text image blocks and multiple independent lithology symbol image blocks according to the contents in the cells based on the table lines;

[0099] The element sorting module is used to use an optical character recognition algorithm to recognize text in multiple independent text image blocks to obtain text elements in the electronic map of mineral exploration drilling holes; build a lithology symbol image database based on the lithology symbols appearing in the electronic map of mineral exploration drilling holes, input the images in the lithology symbol image database into a convolutional neural network, train the convolutional neural network, input multiple independent lithology symbol image blocks into the trained convolutional neural network, perform lithology symbol recognition, and obtain different types of lithology symbol elements.

[0100] The present invention also provides a computer device, including a memory and a processor; the memory stores a computer program, and the processor is used to run the computer program in the memory to execute a method for sorting elements of electronic maps of mineral exploration drilling holes.

[0101] The present invention also provides a computer-readable storage medium, which stores a computer program. The computer program is suitable for loading by a processor to execute a method for sorting elements of electronic maps of mineral exploration drilling holes.

[0102] The embodiments described above are only preferred specific implementation modes of the present invention, and the protection scope of the present invention is not limited thereto. Any simple changes or equivalent replacements of the technical solutions that can be obviously obtained by any technician familiar with the field within the technical scope disclosed in the present invention belong to the protection scope of the present invention.

Claims

1. A method for sorting elements of electronic drawings of mineral exploration drilling, characterized in that: The steps include: Identify table lines in electronic drawings of mineral exploration drilling holes, and split the table into multiple independent text image blocks and multiple independent lithology symbol image blocks according to the contents in the cells based on the table lines; An optical character recognition algorithm is used to recognize text in multiple independent text image blocks to obtain text elements in mineral exploration drill hole electronic maps. A lithology symbol image database is constructed based on the lithology symbols appearing in the mineral exploration drill hole electronic maps, and the images in the lithology symbol image database are input into a convolutional neural network, which is trained. Multiple independent lithology symbol image blocks are input into the trained convolutional neural network to perform lithology symbol recognition and obtain different types of lithology symbol elements.

2. The method for sorting elements of electronic drawings of mineral exploration drilling according to claim 1 is characterized in that: The method of identifying table lines in the electronic map of mineral exploration drilling holes and splitting the table into a plurality of independent text image blocks and lithology symbol image blocks according to the contents in the cells based on the table lines comprises the following steps: Identify the horizontal and vertical lines, number of rows and columns in the electronic drawings of mineral exploration drill holes, set the window according to the threshold, and filter out the horizontal and vertical lines of the cells by sliding the window; Merge the intersection of the horizontal and vertical lines of the cells, calculate the coordinates of the intersection in the mineral exploration drill hole electronic map, restore the table structure of the mineral exploration drill hole electronic map according to the coordinates, and store the table structure of the mineral exploration drill hole electronic map in the form of point coordinates and line coordinates; The mineral exploration drilling electronic map is split into multiple independent text image blocks and lithology symbol image blocks according to the coordinates.

3. The method for sorting elements of electronic drawings of mineral exploration drilling according to claim 1 is characterized in that: Before identifying the table lines in the mineral exploration drilling electronic map, the method further includes preprocessing the mineral exploration drilling electronic map, including the following steps: Conduct red stamp detection on electronic drawings of mineral exploration drilling holes, extract pixels whose red channel is greater than the threshold, assign them as background values, and eliminate red stamps; The electronic maps of mineral exploration drilling holes after removing the red seal are binarized and morphologically operated by the set threshold to eliminate the noise and ambiguity in the electronic maps of mineral exploration drilling holes; The page direction of the mineral exploration drilling electronic map after noise and blur are eliminated is detected and rotated, and the image is vertically projected, the histogram peak of the projection image is analyzed, and the mineral exploration drilling electronic map with non-positive direction is adjusted; The table in the rectified mineral exploration drill hole electronic map is extracted, the distorted table is corrected, and other background areas are cropped.

4. The method for sorting elements of electronic drawings of mineral exploration drilling according to claim 3 is characterized in that: The method uses a set threshold to perform binarization and morphological operations on the electronic map of mineral exploration drilling holes after eliminating the red stamp, specifically: converting the color image into a gray image, binarizing the gray image through a set threshold, enhancing the image features through expansion and corrosion operations, and eliminating noise and blur in the electronic map of mineral exploration drilling holes.

5. The method for sorting elements of electronic drawings of mineral exploration drilling according to claim 1 is characterized in that: After obtaining different types of lithology symbol elements, the method further includes correcting text elements according to the obtained different types of lithology symbol elements, including the following steps: Obtaining the text element recognition confidence and the lithology symbol element confidence, and judging the correct value of the stratum according to the text element recognition confidence and the lithology symbol element confidence; Through manual interpretation, check whether the identification content of the electronic map of mineral exploration drilling is correct.

6. The method for sorting elements of electronic drawings of mineral exploration drilling according to claim 5 is characterized in that: The correct value of the stratum is judged according to the text element recognition confidence and the lithology symbol element confidence, specifically: whether the text element recognition confidence is greater than the set threshold, if so, the stratum is considered to be the stratum described in the text element; if the text element recognition confidence is not greater than the set threshold, whether the lithology symbol element confidence is greater than the set threshold, if so, the stratum is considered to be the stratum corresponding to the lithology image classification result; if the lithology symbol element confidence is still not greater than the set threshold, the text recognition result is used as a temporary output, and an indication that manual inspection is required is given during manual verification.

7. A mineral exploration drilling electronic map element sorting system, characterized in that: include: A table splitting module is used to identify table lines in electronic drawings of mineral exploration drilling holes, and split the table into multiple independent text image blocks and multiple independent lithology symbol image blocks according to the contents in the cells based on the table lines; The element sorting module is used to use an optical character recognition algorithm to recognize text in multiple independent text image blocks to obtain text elements in the electronic map of mineral exploration drilling holes; build a lithology symbol image database based on the lithology symbols appearing in the electronic map of mineral exploration drilling holes, input the images in the lithology symbol image database into a convolutional neural network, train the convolutional neural network, input multiple independent lithology symbol image blocks into the trained convolutional neural network, perform lithology symbol recognition, and obtain different types of lithology symbol elements.

8. A computer device, characterized in that: It comprises a memory and a processor; the memory stores a computer program, and the processor is used to run the computer program in the memory to execute the method for sorting elements of electronic maps of mineral exploration drilling holes as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is suitable for being loaded by a processor to execute the method for sorting elements of electronic maps of mineral exploration drill holes as described in any one of claims 1-6.

Images