An automatic identification method of rock quality index RQD value based on digital image technology

Abstract

The invention relates to the field of mining and geological engineering technology, and in particular to a method for automatically identifying a rock quality index RQD value based on digital image technology; S1: converting a core photo into a digital image; S2: intercepting and correcting a target core in S1 by using a digital image geometric change technology to realize geometric transformation of the image; S3: covering the inner frame of a core box with respect to the core image in S2 that has been processed by geometric transformation to shield stray information in an edge area; S4: marking fine cracks that cannot be captured in S3 by using crack self-calibration; S5: traversing the core image in sequence, automatically identifying a continuous core section that is not interrupted by a joint surface and identifying its length, calibrating and counting continuous blocks that meet a specified threshold, thereby realizing calculation of the RQD value; the invention utilizes a borehole core photo and digital image processing technology, and combines automatic with manual processing, thereby improving the accuracy of the RQD value calculation result.

Description

Technical Field

[0001] The invention relates to the technical field of mining and geological engineering, and in particular to a method for automatically identifying a rock quality index RQD value based on digital image technology. Background Art

[0002] In rock mechanics and rock engineering, the integrity and structure of the rock mass are important components of engineering design planning and development. The rock quality index RQD value is one of the basic rock quality indicators that must be measured in rock quality evaluation. The traditional RQD value acquisition requires the use of a 75mm diameter diamond drill bit and a double-layer core tube to drill into the rock and continuously coring. The calculation method is to compare the total length of the core section greater than 10cm with the drilling footage length, and the result is expressed as a percentage.

[0003] The RQD value is used in many fields in the mining process:

[0004] In geological exploration, rock quality and rock integrity are evaluated to provide a basis for mineral resource exploration.

[0005] In ore mining, it is used to determine the mining method and mining process to ensure safe and efficient ore mining.

[0006] In the field of rock engineering, it provides guidance for rock support design and engineering projects to ensure the safety of rock structures.

[0007] The RQD value can determine the mechanical properties of rock, select appropriate construction methods and processes, and ensure safe and efficient construction. In addition, the RQD value is used to evaluate rock stability and rock mass engineering properties, and plays a key role in ensuring the reliability of engineering design.

[0008] Traditional RQD value measurement and calculation mainly rely on manual visual inspection and judgment, which is subjective and subject to individual differences. Secondly, traditional methods are time-consuming and labor-intensive, making it difficult to cope with the processing needs of large-scale core samples. At the same time, the lack of automation and digitization limits the efficiency of data processing and analysis.

[0009] In order to overcome these limitations, it is necessary to use modern technologies and methods, such as image processing, machine learning and data mining, to achieve automated, digital and precise RQD value measurement and analysis, so as to improve accuracy, efficiency and comprehensiveness and promote the development of rock engineering.

[0010] Therefore, this application proposes a method for automatically identifying the rock quality index RQD value based on digital image technology. Summary of the invention

[0011] In order to make up for the deficiencies of the prior art and solve the technical problems existing in the background technology, the present invention proposes an automatic identification method of rock quality index RQD value based on digital image technology.

[0012] The present invention is achieved through the following technical solutions:

[0013] A method for automatically identifying rock quality index RQD value based on digital image technology comprises the following steps:

[0014] S1: Convert core photos into digital images;

[0015] S2: intercept and correct the target core in S1 using digital image geometric transformation technology to achieve geometric transformation of the image;

[0016] S3: Cover the inner border of the core box on the core image processed by geometric transformation in S2 to block the stray information in the edge area;

[0017] S4: Use crack autonomous calibration to mark subtle cracks that cannot be captured in S3;

[0018] S5: Traverse the core images in sequence, automatically identify the continuous core segments that are not interrupted by joint surfaces and identify their lengths, calibrate the continuous blocks that meet the specified threshold and perform statistics, so as to realize the calculation of RQD values.

[0019] Preferably, S1 uses digital image processing technology to discretize the coordinate space and the property space of the core photos.

[0020] Preferably, the geometric transformation of the image in S2 includes a first image shearing transformation, an image rotation transformation, an image projection transformation, a second image shearing transformation, and an image scaling transformation;

[0021] The first shearing transformation of the image is to shear out the portion containing the target core, that is, a complete core box;

[0022] The image rotation transformation is to rotate the image according to the actual placement of the core in the core box to facilitate subsequent processing and simplify the calculation process;

[0023] The image projection transformation is to transform the distorted core image according to the rectangular core box, and perform projection correction according to the standard size and geometric shape of the rectangular core box, so that it corresponds to the spatial position and scale of the actual core;

[0024] The second shearing transformation of the image is to further shear the image, retaining only the core box and the core in the image, so as to facilitate the subsequent calibration of the image size;

[0025] The image scaling transformation is to scale the image that has been cut for the second time in proportion so that it corresponds to the actual size.

[0026] Preferably, covering the inner frame of the core box in S3 includes manual covering and automatic covering;

[0027] The manual overlay is to set the number of cycles, click to draw a black straight line to cover the inner frame of the core box, and overcome the error caused by the color of the core and the core box being very similar;

[0028] The automatic coverage is to give a pixel range segment of the x-axis that needs to be identified, only retain the pixel values ​​within the specified range, and set other pixels to 0.

[0029] Preferably, the calibration method in S4 comprises the following steps:

[0030] S41: judging based on the image after morphological operation and the original image;

[0031] S42: Select the starting point and end point of the fine cracks that cannot be fully captured;

[0032] S43: autonomously calibrate the fractures in sequence according to the actual arrangement order of the cores, and draw a line segment with a slope of 0 to deepen and cover the fractures;

[0033] If the slope of the crack to be calibrated is not 0, the line segment drawn between the starting point and the midpoint must pass through the midpoint of the crack.

[0034] Preferably, in S5, the identification of continuous blocks of the image covering the inner frame of the core box and the self-calibrated fractures includes the following steps:

[0035] S51: morphological processing;

[0036] S52: Screening and statistics of continuous blocks;

[0037] Among them, S51 specifically includes the following steps:

[0038] Adjust relevant thresholds as needed to eliminate small particles;

[0039] Adjust the radius of the white connected area to perform a clearing operation;

[0040] To remove small noise or small connected areas in the image, we traverse each white pixel in the image and check whether the square area with the pixel as the upper left corner is within the image boundary;

[0041] If the square area is within the boundary, count the number of white pixels in the area, and set all pixels in the square area to black according to the given threshold, while retaining the original pixel value.

[0042] Preferably, S52 includes the following steps:

[0043] S521: sorting the selected connected regions according to the ordinates, and then traversing the sorted regions;

[0044] For each region, calculate the vertical interval between the region and the current continuous block. If the vertical interval is greater than or equal to the set threshold, add the region to the total length of the current continuous block; otherwise, start a new continuous block.

[0045] S522: Draw the identified continuous blocks on the original image for visual display, and mark the actual continuous block length in the image.

[0046] Preferably, the RQD value is calculated by formula (1):

[0047]

[0048] In the above formula:

[0049] Li is a single core section with a length greater than or equal to 10 cm;

[0050] L is the total length of the borehole in the rock formation;

[0051] RQD is the percentage of the cumulative length of the core to the total length of the drilling footage;

[0052] The calculated RQD value of each image is saved in the specified Excel table.

[0053] The beneficial effects of the present invention are:

[0054] The present invention converts core photos into digital images; and uses digital image geometric change technology to intercept and correct target cores to meet the needs of subsequent image processing and analysis; uses the stray information of the edge of the covering core box frame to reduce the influence of interference factors on the core image analysis; at the same time, for the subtle cracks that cannot be captured, the self-calibration of the cracks is used to ensure that all important cracks are accurately marked through manual selection, thereby improving the accuracy of recognition; and automatically frames the continuous part of the core that is not interrupted by the cracks, traverses the length of the continuous blocks in turn, calibrates the continuous blocks that meet the specified threshold and performs statistics, realizes the RQD value calculation, and thus improves the accuracy of the RQD value calculation result. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 is a flow chart of the present invention;

[0056] Figure 2 is the core image diagram of the present invention;

[0057] Figure 3is the image after the first shearing transformation in the present invention;

[0058] Figure 4 is the image after rotation transformation in the present invention;

[0059] Figure 5 is the image after projection transformation in the present invention;

[0060] Figure 6 is the image after the second shearing transformation in the present invention;

[0061] Figure 7 is the image after scaling transformation in the present invention;

[0062] Figure 8 It is a block diagram of the automatic covering core box in the present invention;

[0063] Fig. 9 It is the identification continuous block diagram in the present invention;

[0064] Fig.10 It is an image diagram after the crack is autonomously selected in the present invention;

[0065] Fig.11 The present invention is to select the cracks and identify the continuous block diagram again;

[0066] Fig.12 It is a schematic diagram of calculating and saving the RQD value in the present invention. DETAILED DESCRIPTION

[0067] The present invention will be further described below in conjunction with specific examples. It should be understood that these examples are only used to illustrate the present invention and are not intended to limit the scope of the present invention. The experimental methods in the following examples without specifying specific conditions are usually carried out under conventional conditions or under conditions recommended by the manufacturer.

[0068] Unless otherwise defined, all professional and scientific terms used herein have the same meanings as those familiar to those skilled in the art. The reagents or raw materials used in the present invention can be purchased through conventional channels. Unless otherwise specified, the reagents or raw materials used in the present invention are used in a conventional manner in the art or in accordance with the product instructions. In addition, any method and material similar to or equivalent to the described content can be applied to the method of the present invention. The present invention is further described according to the accompanying drawings and specific embodiments of the present invention, and the preferred implementation methods and materials described in the present invention are only for demonstration purposes.

[0069] Embodiment 1:

[0070] like Figure 1-12 As shown:

[0071] Automatic identification of rock quality index RQD value:

[0072] It includes digitization of core photos, geometric transformation of images, covering the inner frame of the core box, autonomous calibration of fractures, identification of continuous core blocks and calculation of RQD values.

[0073] The digitization of core photographs is the conversion of core photographs into digital images;

[0074] The geometric transformation of the image uses digital image geometric change technology to intercept and correct the target core to meet the needs of subsequent image processing and analysis;

[0075] Covering the inner frame of the core box is to block the stray information at the edge and reduce the influence of interference factors on the core image analysis;

[0076] Autonomous crack calibration is to manually select the subtle cracks that cannot be captured to ensure that all important cracks are accurately marked, thereby improving the accuracy of recognition;

[0077] The identification of continuous core blocks and the calculation of RQD values ​​are to automatically frame the continuous parts of the core that are not interrupted by cracks, traverse the length of the continuous blocks in sequence, calibrate and count the continuous blocks that meet the specified threshold, so as to realize the calculation of RQD values.

[0078] (1) Digitization of core photographs.

[0079] Based on digital image processing technology, the core photographs are sampled (discretized in coordinate space) and quantized (discretized in property space) to obtain a digital representation of the core photograph, namely, a core digital image.

[0080] (2) Geometric transformation of images.

[0081] The geometric transformation of the image includes the first shearing transformation, image rotation transformation, image projection transformation, image second shearing transformation, and image scaling transformation.

[0082] The first shearing transformation of the image is to cut out the part containing the target core, that is, a complete core box (such as Figure 3 ); Image rotation transformation is to rotate the image according to the actual placement of the core in the core box to facilitate subsequent processing and simplify the calculation process (such as Figure 4 ); Image projection transformation is to transform the distorted core image according to certain criteria (based on the rectangular core box) so that it corresponds to the actual core in a certain way (such as Figure 5 ); The second shearing transformation of the image is to further shear the image, the purpose is to retain only the core box and core in the image, which is convenient for the subsequent calibration of the image size (such as Figure 6); Use image scaling transformation to scale the image after the second shearing in proportion to make it correspond to the actual size. If the size of the image after shearing is M×N pixels, the width of the core box is W (mm) and the length is L (mm), then scale the original image of M×N pixels to L×W pixels, which is convenient for the subsequent statistical analysis of the relevant parameters of the continuous blocks, so that it conforms to the actual situation and meets the requirements of subsequent analysis (such as Figure 7 ).

[0083] (3) Cover the inner frame of the core box.

[0084] Covering the inner border of the core box includes manual covering and automatic covering.

[0085] Taking the starting core in the core image after geometric transformation as the reference, manual coverage is to set the number of cycles, click to draw a black straight line to cover the inner border of the core box, and overcome the error caused by the very similar colors of the core and the core box frame; automatic coverage is to give the pixel range segment of the x-axis to be identified, only keep the pixel values ​​within the specified range, and set other pixels to 0 (such as Figure 8 ). Compared with manual operations, automated coverage can complete tasks quickly, save time, and improve work efficiency.

[0086] (4) Identification of continuous blocks of core.

[0087] The images covering the inner frame of the core box and the self-calibrated fractures are used to identify continuous blocks.

[0088] The first step is to perform morphological processing. Adjust the relevant thresholds as needed to eliminate small particles; adjust the radius of the white connected area to perform a clearing operation; to remove small-sized noise or small connected areas in the image, traverse each white pixel in the image and check whether the square area with the pixel as the upper left corner is within the image boundary. If the square area is within the boundary, calculate the number of white pixels in the area, and decide whether to set all pixels in the square area to black (value 0) or retain the original pixel value based on the given threshold. In the final processed image, the local area where the number of white pixels is lower than the threshold is turned black, while other areas remain unchanged.

[0089] The second step is to filter and count the continuous blocks. First, sort the selected connected areas according to the vertical coordinates, and then traverse the sorted areas. For each area, calculate the vertical interval between it and the current continuous block. If the vertical interval is greater than or equal to the set threshold, add the area to the total length of the current continuous block; otherwise, start a new continuous block. Finally, draw the identified continuous blocks on the original image for visualization (such as Fig. 9 ), and its actual continuous block length is marked in the image.

[0090] (5) Autonomous calibration of cracks.

[0091] The calibration method is to first judge the image after morphological operation and the original image, select the starting point and end point of the fine cracks that cannot be fully captured under complex rock structures or unsatisfactory light conditions, and then perform self-calibration of the cracks according to the actual placement order of the cores, and draw a line segment with a slope of 0 to deepen and cover the cracks (such as Fig.10 ).

[0092] When the slope of the crack to be calibrated is not 0, the line segment drawn between the starting point and the midpoint must pass through the midpoint of the crack.

[0093] (6) Calculation of RQD value.

[0094] The rock mass RQD value index refers to the percentage of the cumulative length of the core equal to or greater than 10 cm in the core to the total length of the drilling footage. The formula is as follows:

[0095]

[0096] Where Li is the length of a single core section ≥ 10 cm

[0097] L——Total length of the borehole in the rock formation

[0098] The RQD value of each image can be calculated by the formula and saved in a specified Excel table (such as Fig.12 ).

[0099] Any matters not described in the present invention are applicable to the prior art.

Claims

1. A method for automatically identifying the rock quality index RQD value based on digital image technology, characterized in that: The following steps are involved: S1: Convert core photos into digital images; S2: intercept and correct the target core in S1 using digital image geometric transformation technology to achieve geometric transformation of the image; The geometric transformations of the image in S2 include the first image shearing transformation, image rotation transformation, image projection transformation, second image shearing transformation, and image scaling transformation; The first shearing transformation of the image is to shear out the portion containing the target core, that is, a complete core box; The image rotation transformation is to rotate the image according to the actual placement of the core in the core box to facilitate subsequent processing and simplify the calculation process; The image projection transformation is to transform the distorted core image according to the rectangular core box, and perform projection correction according to the standard size and geometric shape of the rectangular core box, so that it corresponds to the spatial position and scale of the actual core; The second shearing transformation of the image is to further shear the image, retaining only the core box and the core in the image, so as to facilitate the subsequent calibration of the image size; The image scaling transformation is to scale the image after the second shearing in proportion to make it correspond to the actual size; S3: Cover the inner border of the core box on the core image processed by geometric transformation in S2 to block the stray information in the edge area; S4: traverse the core images in sequence, automatically identify the continuous core segments that are not interrupted by joint surfaces and identify their lengths, and mark and count the continuous blocks that meet the specified threshold; S5: Use crack autonomous calibration to mark the subtle cracks that cannot be captured in S4, so as to calculate the RQD value; The calibration method includes the following steps: S41: judging based on the image after morphological operation and the original image; S42: Select the starting point and end point of the fine cracks that cannot be fully captured; S43: autonomously calibrate the fractures in sequence according to the actual arrangement order of the cores, and draw a line segment with a slope of 0 to deepen and cover the fractures; If the slope of the crack to be calibrated is not 0, the line segment drawn between the starting point and the midpoint must pass through the midpoint of the crack; The identification of continuous blocks of images includes the following steps: S51: morphological processing; S52: Screening and statistics of continuous blocks; Among them, S51 specifically includes the following steps: Adjust relevant thresholds as needed to eliminate small particles; Adjust the radius of the white connected area to make it clearer; To remove small noise or small connected areas in the image, we traverse each white pixel in the image and check whether the square area with the pixel as the upper left corner is within the image boundary; If the square area is within the boundary, calculate the number of white pixels in the area, and set all pixels in the square area to black according to the given threshold, while retaining the original pixel value; S52 includes the following steps: S521: sorting the selected connected regions according to the vertical coordinates, and then traversing the sorted regions; For each region, calculate the vertical interval between the region and the current continuous block. If the vertical interval is greater than or equal to the set threshold, add the region to the total length of the current continuous block; otherwise, start a new continuous block. S522: Draw the identified continuous blocks on the original image for visual display, and mark the actual continuous block length in the image.

2. The method for automatically identifying the rock quality index RQD value based on digital image technology according to claim 1 is characterized in that: S1 uses digital image processing technology to discretize the coordinate space and property space of core photos.

3. The method for automatically identifying the rock quality index RQD value based on digital image technology according to claim 1 is characterized in that: Covering the inner border of the core box in S3 includes manual covering and automatic covering; The manual overlay is to set the number of cycles, click to draw a black straight line to cover the inner frame of the core box, and overcome the error caused by the color of the core and the core box being very similar; The automatic coverage is to give a pixel range segment of the x-axis that needs to be identified, only retain the pixel values ​​within the specified range, and set other pixels to 0.

4. The method for automatically identifying the rock quality index RQD value based on digital image technology according to claim 1 is characterized in that: The RQD value is calculated using formula (1): In the above formula: Li is a single core section with a length greater than or equal to 10 cm; L is the total length of the borehole in the rock formation; RQD is the percentage of the cumulative length of the core to the total length of the drilling footage; The calculated RQD value of each image is saved in the specified Excel table.

Images