Method for determining particle size distribution in a rock pile

By constructing a volumetric relief and using neural networks to calculate rock mass piece sizes, the method enhances the accuracy of granulometric composition analysis, addressing safety and precision issues in rock mass collapse determination, leading to optimized blasting operations and reduced costs.

WO2024242584A1PCT designated stage expired Publication Date: 2024-11-28OBSCHESTVO S OGRANICHENNOI OTVETSTVENNOSTYU DAVTEKH
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
PCT/RU2023/050275
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-05-25
Filing Date
2023-11-27
Publication Date
2024-11-28

AI Technical Summary

Technical Problem

Existing methods for determining the granulometric composition of rock mass collapse, such as those using stereoscopic cameras and neural networks, face challenges in accuracy and safety due to limitations in calculating distances and geometric sizes, leading to reduced precision in blasting operations.

Method used

The method involves obtaining a stereo image of the rock mass collapse using a stereoscopic camera, constructing a volumetric relief by determining pixel spacing and distances, and using a neural network to calculate the size and number of rock mass pieces, with calibration coefficients derived from approximations to convert pixel sizes to linear dimensions, enhancing accuracy and safety.

Benefits of technology

This approach significantly increases the accuracy of determining the size and number of rock mass pieces, thereby improving the precision of granulometric composition analysis, optimizing blasting operations and reducing costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure IMGF000008_0001
    Figure IMGF000008_0001
  • Figure IMGF000009_0001
    Figure IMGF000009_0001
  • Figure IMGF000009_0002
    Figure IMGF000009_0002
Patent Text Reader

Abstract

The invention relates to methods for determining the particle size distribution in rock piles using images obtained with a stereoscopic camera. The essence of the invention consists in that in the method for determining the particle size distribution in a rock pile, the geometric parameters of fragments in the rock pile are determined on the basis of a three-dimensional model of the surface of the pile. The technical result toward which the invention is directed is that of increasing the accuracy with which the dimensions and number of pieces of rock are determined by a computing unit of an electronic device.
Need to check novelty before this filing date? Find Prior Art

Description

[0001] Method for determining the granulometric composition of rock mass breakdown

[0002] The invention relates to methods for determining the granulometric composition of rock mass breakdown based on images obtained by a stereoscopic camera, and can be used for automatic measurement of the fractional composition of rock mass destroyed or crushed as a result of blasting and other operations in quarries, faces and other mining industry facilities.

[0003] Analysis of the granulometric composition of blasted rock mass is carried out to optimize time and material costs during mining operations, as well as other works. The results of the analysis allow mining industry specialists to determine the parameters of blasting operations, such as the pitch and depth of well drilling, and to calculate with high accuracy the amount of explosive that must be placed in the wells. Accurate determination of the parameters of blasting operations will reduce the costs of excessive well drilling and the excessive amount of explosives used.

[0004] A method for determining the granulometric composition of a rock mass collapse is known from the prior art [RU2388998, publication date: 20.10.2009]. When performing the method, an arbitrarily oriented scaling rectangle with arbitrarily selected side lengths is placed in any accessible place, the rock mass collapse is photographed at any angle, the photoplanogram is entered into a computer, a quadrangular calculation contour of an arbitrary size and position is formed on it, the areas of the rock mass pieces are outlined, size classes are specified within the calculation contour, and using a computer program, the granulometric composition of the crushed rock mass is calculated by assigning the areas of the outlined rock mass pieces to the area of ​​the calculation contour.

[0005] The disadvantage of the known technical solution is the low safety and accuracy of determining the granulometric composition of the rock mass collapse due to the need to use a scaling rectangle during photography, which is ineffective at small angles of inclination of the collapse due to the length of the photography scene in perspective, as well as the need for personnel to be in dangerous areas of the collapse when placing the scaling rectangle, which significantly reduces the technological effectiveness of the method.

[0006] The prototype chosen is a method for determining the granulometric composition of rock mass breakdown [JP7117273, publication date: 12.08.2022]. In the proposed method, a stereoscopic image of rock mass breakdown is obtained using a stereoscopic camera, sections are selected on the obtained image using a personal computer, after which the sizes and number of pieces of rock mass breakdown are determined pixel by pixel using a neural network, and then the granulometric composition of the rock mass breakdown is determined.

[0007] The advantage of the prototype over the known technical solution is its higher technological efficiency due to obtaining images of the collapse using a stereoscopic camera and determining the size and number of pieces by using a neural network without the need to place a scaling rectangle on the collapse, as a result of which the accuracy and safety of this method increases. However, a disadvantage of the prototype may be the insufficiently high accuracy of determining the size and number of pieces of rock mass by the neural network, since the distance to the fragment may not be calculated accurately, and therefore its geometric size may not be calculated accurately, as a result of which the accuracy of determining the fraction and number of pieces of the collapse and its granulometric composition may decrease.

[0008] The technical problem that the invention is aimed at solving is the need to increase the accuracy of determining the granulometric composition of rock mass breakdown.

[0009] The technical result that the invention is aimed at achieving consists in increasing the accuracy of determining the size and quantity of pieces of rock mass by the computing unit of the electronic device.

[0010] The essence of the invention is as follows.

[0011] The method for determining the granulometric composition of the rock mass breakdown is as follows:

[0012] - using a stereoscopic camera, a stereo image of the rock mass collapse is obtained;

[0013] - based on the obtained stereo image of the rock mass collapse, a volumetric relief of the rock mass collapse surface is constructed, which is a set of distances from the stereoscopic camera matrix to the selected sections of the rock mass collapse surface and the corresponding row and column numbers in the image matrix, for which:

[0014] - determine the value of pixel spacing of a pair of stereo images for selected sections of rock mass breakdown;

[0015] - determine the distance from the matrix of the stereoscopic camera to the surface of the selected sections of the rock mass collapse based on the value of the pixel spacing of the pair of stereo images, the value of the angular size of the matrix pixel and the value of the camera stereo base;

[0016] - based on the constructed volumetric relief of the surface of the rock mass collapse, the sizes of the pieces of rock mass collapse and their quantity are determined using a neural network;

[0017] - determine the granulometric composition of the rock mass breakdown based on the size of the pieces of rock mass breakdown and their quantity.

[0018] A stereoscopic camera provides the construction of an image with spatial depth using two identical lenses placed at a distance of the stereo base from each other. The computing unit can be a data processing module of the stereoscopic camera itself, or the computing unit of any other device, including a tablet computer, smartphone, personal computer, server, etc.

[0019] A stereo image is a stereo pair of flat images of the same scene obtained at different offset angles. A stereo image of a rock mass collapse can be obtained by installing a stereoscopic camera opposite the collapse using camera mounting devices or by fixing it to stationary objects. Since all objects, including fragments of the rock mass, are represented in the image obtained by the camera as a group of pixels, it is necessary to ensure the transformation of pixel sizes into linear ones, for which a volumetric relief of the surface of the rock mass collapse is initially constructed. The volumetric relief is a set of distances from the matrix of the stereoscopic camera to the surface of the selected sections of the rock mass collapse and the corresponding numbers of rows and columns in the image matrix.To construct a volumetric relief of the surface on a stereo image, sections of the rock mass collapse are selected and the value of the pixel spacing of the stereo image pair is determined for the selected sections. Since any image is a matrix consisting of a set of rows and columns, the value of the pixel spacing provides the ability to obtain data on the value of the shift of two flat images in some rows or columns relative to each other. In this case, to bind the selected sections of the rock mass collapse surface to the numbers of rows and columns in the image matrix, a square grid superimposed on the stereo image or other methods can be used. In this case, sections of the rock mass collapse surface can be selected near the nodes of the square grid.In this case, the value of pixel spacing can be calculated based on the shift of the maximum of the row-by-row or column-by-column cross-correlation function (CCF) relative to zero, into which the contents of the rows or columns of the selected section on the stereo image with the same row and column number in the image matrix are substituted.

[0020] After this, the distance from the matrix of the stereoscopic camera to the surface of each selected section of the rock mass collapse is determined. For this, the previously obtained value of the pixel spacing of the pair of stereo images, the value of the angular size of the matrix pixel and the value of the stereo base of the camera are used. In this case, for each selected section and the corresponding distance, their own value of pixel spacing is calculated. The value of the angular size (angular resolution) of the matrix pixel means the ratio of the full angle of view of the camera to its resolution along the corresponding axis. The stereo base means the distance between the optical centers of the lenses of the stereoscopic camera, information on which is contained in the technical documentation for the stereoscopic camera used.

[0021] The geometric parameters are determined based on the constructed volumetric relief of the rock mass surface. To do this, an approximation of the slope relief is first constructed for the entire surface of the collapse in the image, which allows determining the slope angle to the optical axis of the cameras and calculating calibration coefficients for converting the pixel sizes of the pieces into linear ones.

[0022] The approximation can be constructed using the least squares method (LSM) known in mathematics by solving a differential equation. It is possible to construct a two-dimensional approximation in the central plane of a vertical section, in which the parameters of a straight line with the smallest standard deviation from the surface relief are calculated. Additionally, to improve the accuracy of determining the size and number of rock mass collapse pieces, the computing unit of the electronic device can construct a three-dimensional approximation by a plane over the entire volume in the frame, in which the calculated plane parameters will have a minimum standard deviation from the uneven collapse surface and will allow obtaining the most accurate values ​​of the accuracy of measuring the size of pieces with a relatively low computational resource intensity.

[0023] Based on the approximation parameters, the distance from the stereoscopic camera matrix to each point of the collapse surface in the stereo image is then determined and, based on this, the equivalent pixel spacing for this point is calculated, thereby obtaining calibration coefficients for each point. The resulting calibration coefficients are used to determine the sizes of all detected pieces.

[0024] The constructed approximation is then used to determine calibration coefficients for all sections of the rock mass collapse surface, which allows the construction of a volumetric relief based on the VCF only once, which reduces the resource intensity of the calculations.

[0025] The sizes and quantity of rock mass collapse pieces are determined based on the constructed volumetric relief of the rock mass collapse surface using a neural network, which allows determining the contours of the rock mass collapse pieces' sizes and counting the number of pixels inside each fragment. The number of pixels can be reduced to the linear dimensions of the contour based on the cubic model of the fragment. To improve the accuracy of determining the sizes and quantity of rock mass collapse pieces by the computing unit of the electronic device, fuzzy neural networks can be used as a neural network.

[0026] Based on the data obtained, the granulometric composition of the rock mass breakdown is determined. For this purpose, statistical analysis can be applied, as a result of which the data are displayed in the form of graphs, histograms and diagrams based on the calculation of the number of pieces of mass of different sizes. The invention can be made from known materials using known means, which indicates its compliance with the patentability criterion "industrial applicability".

[0027] The invention is characterized by a previously unknown set of essential features from the state of the art, distinguished by the fact that:

[0028] - based on the obtained stereo image of the rock mass collapse, a volumetric relief of the rock mass collapse surface is constructed, for which:

[0029] - determine the value of pixel spacing of a pair of stereo images for selected sections of rock mass breakdown;

[0030] - determine the distance from the matrix of the stereoscopic camera to the surface of the selected sections of the rock mass collapse based on the value of the pixel spacing of the pair of stereo images, the value of the angular size of the matrix pixel and the value of the camera stereo base;

[0031] - the size and number of pieces of rock mass breakdown are determined on the basis of the constructed volumetric relief of the surface of the rock mass breakdown using a neural network.

[0032] The set of essential features of the invention provides the possibility of constructing a volumetric relief of the surface of the rock mass collapse in a discrete form, on the basis of which calibration coefficients are calculated with high accuracy, allowing the pixel dimensions of the sizes of the pieces of rock mass to be converted into linear ones, which in turn allows the contours to be determined with high accuracy, and on their basis the dimensions, as well as the number of pieces of rock collapse, by means of a neural network.

[0033] This ensures the achievement of a technical result consisting in increasing the accuracy of determining the size and quantity of pieces of rock mass breakdown by the computing unit of the electronic device, thereby increasing the accuracy of determining the granulometric composition of the rock mass breakdown.

[0034] The invention has a set of essential features previously unknown in the state of the art, which indicates its compliance with the patentability criterion of “novelty”.

[0035] The principle of constructing a volumetric relief of the surface of a rock mass collapse proposed by the authors of the present invention is not known from the prior art, which represents a set of distances from the matrix of a stereoscopic camera to the surface of selected sections of the rock mass collapse and the corresponding numbers of rows and columns in the image matrix, on the basis of which the sizes and number of pieces of the rock mass collapse are subsequently determined, making it possible to determine with high accuracy the granulometric composition of the rock mass collapse based on a stereo image.

[0036] In view of this, the invention meets the patentability criterion of “inventive step”.

[0037] The invention is illustrated by the following figures.

[0038] Fig. 1 — Schematic representation of the relief of the rock mass collapse, which shows the distances from the matrix of the stereoscopic camera to the selected sections of the collapse, the plane of the three-dimensional approximation of the slope relief, as well as the optical axis of the stereoscopic camera.

[0039] Fig. 2 - Algorithm for determining the granulometric composition of rock mass breakdown using an electronic computing device.

[0040] To illustrate the possibility of implementation and a more complete understanding of the essence of the invention, an embodiment of it is presented below, which can be changed or supplemented in any way, while the present invention is in no way limited to the presented embodiment.

[0041] The system for determining the granulometric composition of the rock mass breakdown includes a stereoscopic camera 100 mounted on a tripod opposite the rock mass breakdown, and a computing unit of the electronic device.

[0042] The method for determining the granulometric composition of rock mass breakdown is carried out as follows.

[0043] After drilling and blasting operations are carried out, a stereoscopic camera 100 is installed opposite the resulting collapse and at stage 200 a stereo image of the collapse surface of the rock mass is obtained, which is then processed using a computing unit, performing a stereoscopic analysis of the obtained image. At the same time, a computing unit based on a neural network performs a contour analysis of the images and determines the number of rock pieces in the frame and the pixel dimensions (the number of pixels in the contour) P s for each contour of a piece of rock. Based on the cubic model of the fragment, the linear dimensions of the pieces of rock can be determined as In the process of processing the stereo image, initially at step 300, the value of the pixel spacing of the stereo image pair for the selected rock mass collapse sections is determined. To bind the selected rock mass collapse surface sections to specific row and column numbers in the image matrix, the nodes of the square lattice superimposed on the image are used.

[0044] Since a digital stereo image is a rectangular matrix consisting of a certain number of rows and columns, then to determine the magnitude of DR, a row-by-row or column-by-column (in the vertical dimension) mutual correlation function (MCF) is calculated within the areas of the surface selected at the nodes of a square lattice, in the form: or in discrete form:

[0045] 5(i) = .j SljS2 i+j , Where:

[0046] S1 and S2 are the contents of the same lines of two images of surface areas selected near the nodes of a square lattice.

[0047] Since when shooting the same scene with a stereoscopic camera 100 from different angles, the contents of the same lines of two images will have a high level of mutual correlation, a clearly visible maximum can be discerned in the VCF. In this case, the shift of the VCF maximum relative to zero is equal to the shift of the lines relative to each other, i.e., the pixel spacing. Thus, for each section of the camber surface, its own pixel spacing of the LR is calculated.

[0048] After this, at stage 400, a volumetric relief of the surface of the rock mass collapse is constructed. For this, the values ​​of the distances H are determined п (pos. 410) from the matrix of the stereoscopic camera to the surface of each selected section of the rock mass collapse.

[0049] For this purpose, the obtained value of the pixel spacing DR of a pair of stereo images, the value of the angular size of the matrix pixel and the value of the camera stereo base (the distance between the optical centers of the camera lenses) are used. The value of the last parameter is taken from the technical documentation for the stereoscopic camera used.

[0050] To determine the range of H п For each selected section, use the following expression:

[0051] ЛЬ - the value of the camera's stereo base; a p - the value of the angular size of a pixel of the camera matrix, which is calculated as the ratio of the full angle of view of the camera to its resolution along the corresponding axis.

[0052] At stage 500, the sizes and quantity of pieces of rock mass collapse are determined. For this, the obtained values ​​of distances H are used п, and on their basis, a three-dimensional approximation of the relief 505 of the slope formed by the plane 510 is constructed to the entire volume in the frame, and the angle of inclination of the plane 510 to the optical axis 515 of the camera lens 100 is determined. In the case of constructing a three-dimensional approximation, the calculated parameters of the plane have a minimum standard deviation from the uneven surface of the collapse and allow obtaining the most accurate values ​​of the accuracy of measuring the sizes of pieces with a relatively low resource intensity of calculations.

[0053] Based on the parameters of plane 510, the distance H is calculated п from the matrix of the stereoscopic camera 100 to each point of the camber surface and on its basis the equivalent pixel spacing DR' is calculated for a given point using the following expression: arctan ( — ), where: And

[0054] ДЬ - the value of the camera's stereo base; a р - the value of the angular size of a pixel of the camera matrix.

[0055] Based on the value of DR', the calibration coefficient I is determined for a given point using the following expression:

[0056] ДЬ - the size of the camera's stereo base;

[0057] DR' is the equivalent pixel spacing for a given point.

[0058] Based on the obtained calibration coefficient I, the sizes Lk of all detected piece sizes are determined using the expression:

[0059] b / c = I - Pk, where:

[0060] Рк - pixel dimensions of the fragment.

[0061] The constructed approximation allows to determine calibration coefficients for all sections with high accuracy, without the need to conduct a costly calculation of the VCF for each individual section. In this case, based on the value of DR' for each section of the surface, the calibration coefficient for this section is determined and the sizes of all detected pieces are determined.

[0062] At step 600, the granulometric composition of the rock mass breakdown is determined, for which a statistical method is used. To present the results, the sizes of the pieces and their number are displayed on graphs, tables or histograms, thereby obtaining information on the granulometric composition of the breakdown.

[0063] This ensures the achievement of a technical result consisting in increasing the accuracy of determining the size and quantity of pieces of rock mass breakdown by the computing unit of the electronic device, thereby increasing the accuracy of determining the granulometric composition of the rock mass breakdown.

Claims

Invention formula 1. A method for determining the granulometric composition of rock mass breakdown, which consists in the following: - using a stereoscopic camera, a stereo image of the rock mass collapse is obtained; - based on the obtained stereo image of the rock mass collapse, a volumetric relief of the rock mass collapse surface is constructed, which is a set of distances from the stereoscopic camera matrix to the selected sections of the rock mass collapse surface and the corresponding row and column numbers in the image matrix, for which: - determine the value of pixel separation of a pair of stereo images for selected sections of rock mass breakdown; - determine the distance from the matrix of the stereoscopic camera to the surface of each selected section of the rock mass collapse based on the value of the pixel spacing of the pair of stereo images, the value of the angular size of the matrix pixel and the value of the camera stereo base; - based on the constructed volumetric surface relief, the size and number of pieces of rock mass breakdown are determined using a neural network; - determine the granulometric composition of the rock mass breakdown based on the size of the pieces of rock mass breakdown and their quantity.

2. The method according to paragraph 1, in the course of which the value of the pixel spacing of a pair of stereo images is calculated on the basis of the shift of the maximum of the row-by-row or column-by-column cross-correlation function relative to zero, into which the contents of the rows or columns of the selected section on the stereo image with the same number are substituted.

3. The method according to paragraph 1, during which, in the process of determining the size and quantity of pieces of rock mass breakdown, a three-dimensional approximation of the slope relief is constructed with a plane to the entire surface of the breakdown on a stereo image.

4. The method according to paragraph 1, in the process of determining the size and quantity of pieces of rock mass breakdown, fuzzy neural networks are used. AND

Citation Information

Patent Citations

  • System for crushing large mass of rock

    JP2009235781A

  • Method for detecting broken fragments, method for dividing broken fragments, device for detecting broken fragments, and device for dividing broken fragments

    JP7117273B2

  • Definition method of content of effective component in imploded mountain mass at its excavation at motions

    RU2386032C1

  • Method for detection of grain size composition in crushed rock of mines

    RU2388998C2

  • Method for determining particle size distribution of blown-up rock mass in quarry

    SU926282A1