Blasting lumpiness identification and blasting effect evaluation method and system based on image processing

Abstract

The invention discloses a blasting lumpiness identification and blasting effect evaluation method and system based on image processing, and relates to the field of blasting engineering.The method comprises the steps that S1, a high-definition camera and a laser radar are carried on an unmanned aerial vehicle and connected with a computer, and the unmanned aerial vehicle enters a working state after debugging is completed; s2, after the unmanned aerial vehicle arrives at a designated position, acquiring image information of the rock mass through a high-definition camera, acquiring distance information between the unmanned aerial vehicle and the rock mass through a laser radar, and wirelessly transmitting the acquired image data and distance data to a computer; s3, the computer processes the image data through an image processing module, and rock mass lumpiness information is extracted; and S4, the computer processes the rock mass lumpiness information through a blasting effect evaluation module to obtain an evaluation result. The influence of factors such as the size and distribution of rock blocks after mine blasting can be solved, the best technical and economic effects are achieved through blasting effect evaluation, and a certain reference value is provided for development of the blasting technology.

Description

Technical Field

[0001] The present invention relates to the technical field of blasting engineering, and more particularly to a method and system for blasting fragmentation recognition and blasting effect evaluation based on image processing. Background Art

[0002] With the rapid development of society and economy, my country has built a large number of important blasting projects, involving transportation tunnels, mining, and other projects. As the number of projects increases, theories and construction methods, equipment, and instruments are also developing rapidly. However, in complex geological conditions, the blasting process needs to be adjusted according to actual conditions. Currently, the construction process still has a strong degree of uncertainty, making it a complex and high-risk system project.

[0003] Open-pit mines offer advantages such as large production scale, high mining efficiency, and low costs. Blasting is the primary method of mining. As a common excavation method in engineering projects, blasting technology is widely used in mining and other projects. While blasting brings high engineering economic benefits, it also creates a series of problems. When the blasting intensity is high, the resulting rock masses are too small, increasing the amount of dust, causing serious pollution, affecting the mining environment, and limiting the ability to achieve sustainable development. When the blasting intensity is low, the resulting rock masses are too large, increasing the cost of secondary crushing and the difficulty of transportation. Therefore, selecting the appropriate blasting method, adjusting blasting parameters, and evaluating blasting effects are the primary solutions to the major challenges of high cost and difficulty in mining. Currently, many methods for identifying rock mass size generally lack accuracy and persuasiveness. Using scientific methods and technologies to guide construction is a major trend in future engineering.

[0004] Digital and intelligent equipment can quickly analyze the size and distribution of rock fragmentation, allowing for timely adjustments to blasting plans, which directly impacts project quality. Traditional rock fragmentation analysis relies on manual field measurements, which are often susceptible to natural influences. The shift from manual to digital methods is crucial for intelligently extracting rock information, evaluating blasting effectiveness, and optimizing blasting plans. Unmanned aerial vehicle (UAV) technology, coupled with digital photography and image processing, provides a promising method for efficiently acquiring rock surface data during engineering. Compared to manual measurement methods, it offers improved safety and speed, and can also capture data and rock information difficult to measure manually. After identifying rock fragmentation, analysis of rock fragmentation allows for a scientific and rational evaluation of blasting effectiveness in mines, minimizing the impact of blasting on rock mass damage and other issues, which has significant engineering significance for underground engineering construction.

[0005] Therefore, how to solve the influence of factors such as rock block size and distribution after mine blasting, achieve the best technical and economic effects through blasting effect evaluation, and provide certain reference value for the development of blasting technology is a technical problem that technical personnel in this field urgently need to solve. Summary of the Invention

[0006] In view of this, the present invention provides a method and system for blasting fragmentation recognition and blasting effect evaluation based on image processing, which solves the problems existing in the background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solutions:

[0008] A method for identifying blasting fragmentation and evaluating blasting effects based on image processing comprises the following steps:

[0009] S1: Install the high-definition camera and lidar on the drone and connect them to the computer. After debugging, they will enter the working state;

[0010] S2: After the UAV arrives at the designated location, it collects image information of the rock mass through a high-definition camera, obtains the distance information between the UAV and the rock mass through a laser radar, and wirelessly transmits the collected image data and distance data to a computer;

[0011] S3: The computer processes the image data through the image processing module to extract the rock mass block information;

[0012] S4: The computer processes the rock mass block information through the blasting effect evaluation module to obtain the evaluation results.

[0013] Optionally, in S3, extracting rock mass fragmentation information specifically includes the following steps:

[0014] S31: converting the color image captured by the high-definition camera into a gray image by grayscale operation;

[0015] S32: Use the improved fuzzy C-means clustering algorithm to segment the gray image;

[0016] S33: Use the connected domain denoising method based on the least squares method to remove tiny noise points in the segmented image;

[0017] S34: The basic equation of the inter-class dispersion matrix is derived through formula derivation:

[0018] S B (s,t)=ω a (μ a -μ) T (μ a -μ)+ω b (μ b -μ)T (μ b -μ)

[0019] Tr(S B (s,t))=ω a [(μ ai -μ i ) 2 +(μ a j-μ j )]+ω b [(μ bi -μ i ) 2 +(μ b j-μ j )]

[0020] Where: s represents a feature dimension in the feature space, t represents another feature dimension, S B (s, t) represents the element located in row s and column t in the inter-class scatter matrix; ω a represents the prior probability of category a, μ a represents the mean vector of category a, μ represents the overall mean vector; ω b represents the prior probability of category b, μ b represents the mean vector of category b; T represents the transpose operation of the matrix, and Tr(·) represents the trace of the matrix;

[0021] When Tr(S B (s 1 ,t 1 ))=max{Tr(S B (s,t))},0≤s,t≥L-1, (s 1 ,t 1 ) is the optimal threshold;

[0022] S35: Combined with the improved Canny algorithm, the pixel area is calculated using the image processing algorithm. The actual area of the crack is calculated based on the proportional relationship between the pixel area and the actual size.

[0023] Optionally, in S32, the improved fuzzy C-means clustering algorithm comprehensively utilizes local and global spatial information, constructs a joint membership function Y through combination, and calculates a new cluster center X:

[0024]

[0025] Where: u mn represents the local membership, μ mn represents the global membership; p represents the fuzzy index, b and q represent the relative importance parameters of controlling the global membership and local membership, N represents the number of pixels, and x c Represents pixels.

[0026] Optionally, in S34, during the confirmation of the optimal threshold, the pixels of the entire image are retrieved once, and an improvement is made based on the two-dimensional OTSU algorithm to automatically obtain two thresholds T1 and T2. The specific steps are as follows:

[0027] Reduce the size of the denoised image by pooling to retain the image features;

[0028] Segment the pooled image into four parts of the same size. The four parts of the image include the four corners of the original image. Starting from the corners, select three-quarters of the image in the diagonal direction. The four segmented parts all have the common part of the original image.

[0029] The two-dimensional OTSU algorithm is used to process the four parts of the image segmentation respectively to obtain four sets of corresponding thresholds, and the average value of the four sets of thresholds is taken as the new threshold.

[0030] Optionally, in S4, obtaining the evaluation result specifically includes the following steps:

[0031] S41: Calculate the number of large blocks of the rock mass based on the rock mass blockiness information obtained in S3, set a threshold range based on the connected domain principle, screen the rock mass that meets the conditions, and use the obtained information as a new indicator;

[0032] S42: Use the hierarchical analysis method to determine the hierarchy of the new indicators, compare the indicators of each layer pairwise and analyze them layer by layer, and calculate the consistency test index CI; obtain the average random consistency index RI by taking the average of the eigenvalues of the random judgment matrix after multiple calculations; calculate the consistency index When CR<0.1, it means that the judgment matrix passes the consistency, otherwise it will be revised;

[0033] S43: After determining the indicator weights through the hierarchical analysis method, the fuzzy comprehensive evaluation method is carried out to synthesize the fuzzy comprehensive evaluation result vector. The formula is as follows:

[0034]

[0035] Where: Q represents the result of vector multiplication and matrix multiplication, w m Represents the mth element in the left row vector, r mn represents the element in the mth row and nth column of the intermediate matrix, b n Represents the nth element in the result vector on the right.

[0036] Optionally, S5 is further included to establish a GUI system interface, which specifically includes the following steps:

[0037] S51: Establish the system interface, which is developed using the GUI development toolbox built into the Matlab software. It is mainly divided into the following four areas:

[0038] (1) Image display area: four display frames are used to display the effect images of different functions;

[0039] (2) Image processing area: consists of loading image files, binarization processing, image segmentation and image denoising;

[0040] (3) Blasting effect evaluation area: It consists of the maximum value of block size area, average value of block size area, large block rate and effect evaluation;

[0041] (4) Save and exit area: Save the processed image and information, and click Exit to close everything;

[0042] S52: In Matlab, enter deploytool in the command window, select Application Compiler, add the application source code, add associated files, select the packaging method, and complete the packaging operation.

[0043] A blasting fragmentation recognition and blasting effect evaluation system based on image processing, which executes any of the above-mentioned blasting fragmentation recognition and blasting effect evaluation methods based on image processing, comprises: an unmanned aerial vehicle, a high-definition camera, a laser radar, and a computer, wherein the computer comprises an image processing module, a blasting effect evaluation module, and a GUI system interface;

[0044] A high-definition camera is mounted on the drone to collect image information of the rock mass and transmit it wirelessly to a computer;

[0045] The laser radar is mounted on the UAV to obtain the distance information between the UAV and the rock mass and transmit it wirelessly to the computer;

[0046] Image processing module, used to process the image information of the rock mass and extract the rock mass block information;

[0047] The blasting effect evaluation module is used to process the rock mass block information and obtain the evaluation results;

[0048] GUI system interface, used to display rock mass image processing information and evaluation results.

[0049] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a method and system for blasting fragmentation recognition and blasting effect evaluation based on image processing, which has the following beneficial effects:

[0050] (1) The present invention obtains rock mass block information in an automated manner, which can replace manual methods to a great extent, reduce manpower and time spent, and improve the speed and efficiency of information acquisition;

[0051] (2) The image processing method used in the present invention can accurately identify rock mass blockiness and extract geometric information, which has important guiding significance for subsequent engineering construction (such as blasting engineering);

[0052] (3) The present invention can provide guidance and help when performing blasting design, improve design deficiencies, and analyze blasting effects in real time to provide reference;

[0053] (4) By combining drone technology, the present invention can visually observe the mine and use image processing technology to identify the rock mass blockiness, which to a certain extent improves efficiency and reduces the uncertainty brought by manual methods, and has certain reference significance for engineering construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0055] Figure 1 A flow chart of the method for blasting fragmentation recognition and blasting effect evaluation based on image processing provided by the present invention;

[0056] Figure 2 The image processing flow chart provided by the present invention;

[0057] Figure 3 A flow chart for evaluating blasting effects provided by the present invention;

[0058] Figure 4 A diagram of the comprehensive evaluation system for blasting effects provided by the present invention;

[0059] Figure 5 A schematic diagram of the system interface provided by the present invention;

[0060] Figure 6 This is a structural diagram of the blasting fragmentation recognition and blasting effect evaluation system based on image processing provided by the present invention. DETAILED DESCRIPTION

[0061] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0062] The embodiment of the present invention discloses a method for blasting fragmentation recognition and blasting effect evaluation based on image processing, such as Figure 1 As shown, the following steps are included:

[0063] S1: Install the high-definition camera and lidar on the drone and connect them to the computer. After debugging, they will enter the working state;

[0064] S2: After the UAV arrives at the designated location, it collects image information of the rock mass through a high-definition camera, obtains the distance information between the UAV and the rock mass through a lidar (the image size is calculated by the distance), and wirelessly transmits the collected image data and distance data to a computer;

[0065] S3: The computer processes the image data through the image processing module and extracts the rock mass block information, such as Figure 2 As shown, the specific steps include:

[0066] S31: Grayscaling is used to convert the color image captured by the high-definition camera, which contains too much information, into a gray image with relatively less information. This is done to increase the computational speed, require less space than a color image, and highlight the target area to be processed to a certain extent. The weighted average method is to perform a weighted average of the R, G, and B values according to certain weights. That is, the weights of R, G, and B are averaged based on their importance or other requirements. By taking different weights, a grayscale image different from the maximum and average methods is obtained. The expression is as follows:

[0067] R=G=W=W R R+W G G+W B B

[0068] Where: R, G and B are the three component values of the image, W R 、W G and W B are the weight values of R, G and B respectively. Generally speaking, when W R =0.3, W G =0.59, W B When =0.11, the best grayscale image can be obtained.

[0069] S32: Use the improved fuzzy C-means (FCM) clustering algorithm to segment the gray image; the fuzzy C-means (FCM) clustering algorithm classifies and divides data based on the size of the membership. The similarity between objects classified into the same category is large, while the similarity between different categories is small. It is a soft clustering method.

[0070] The Fuzzy C-means (FCM) clustering algorithm takes into account the fuzzy membership relationship between sample points and cluster centers, so that each sample point can belong to multiple cluster centers instead of just one. At the same time, the Fuzzy C-means (FCM) clustering algorithm is also an algorithm based on the objective function. Suppose there is a rock image with cracks with a total of n pixels, that is, given a data set containing n pixels: x = {x1, x2, ..., x n}, the objective function of the fuzzy C-means (FCM) clustering algorithm is defined as follows:

[0071]

[0072] Where: u mn is pixel X m The fuzzy membership relative to the mth cluster, p is the fuzzy index, d 2 mn is the nth pixel X n and the mth cluster center v m Euclidean clustering between .

[0073] In order to find the minimum value of the objective function J, the Lagrange multiplier method is used for the objective function under the condition of satisfying the constraints, and J is set to m and u mn The partial derivative is 0, and the membership matrix u can be obtained mn , cluster center v m , the specific formula is as follows:

[0074]

[0075] The specific steps of the fuzzy C-means (FCM) clustering algorithm are as follows: first, input the data of the rock mass blockiness image into the algorithm, set the number of cluster categories, fuzzy index p and initialize the cluster center, etc.; set the threshold for stopping the iteration and the maximum number of iterations; update the membership and calculate the cluster center; determine whether the conditions of the standardized measurement function are met. If so, stop the algorithm and output the membership and cluster center; divide the image pixels according to the maximum membership principle to achieve image segmentation.

[0076] The Fuzzy C-means (FCM) clustering algorithm only considers the grayscale information of the image, not the spatial information of the image. Therefore, the local spatial information of the image is weighted with the grayscale information and then incorporated into the objective function, making the algorithm have good segmentation performance even for noisy images. In order to utilize local spatial information, the probability function f of pixel x is introduced to represent the probability of adjacent pixels in the i-th cluster:

[0077]

[0078] Where: N k is the square neighborhood window centered on pixel x in the spatial domain, and M is the square neighborhood N k area.

[0079] The probability function f is merged into the global membership function to generate a new membership value set for each pixel. The modified objective function of the FCM algorithm based on local spatial information is:

[0080]

[0081] Where: α is the preference for the neighboring effect, is the Euclidean distance between the average intensity of the nth pixel neighborhood and the cluster center; if α = 0, the formula reverts to the FCM algorithm; if α > 0, when the neighboring membership values of other clusters are small, the dependence on the adjacent cluster will cause μ to be larger.

[0082] Derived membership μ mn for:

[0083]

[0084] In order to comprehensively utilize local and global spatial information, a joint membership function Y is constructed by combination, and the new cluster center X is calculated:

[0085]

[0086] Where: u mn represents the local membership, μ mn represents the global membership, u cn 、μ cn with u mn 、μ mn has the same meaning; p represents the fuzzy index, b and q represent the relative importance parameters of controlling global membership and local membership, N represents the number of pixels, and x c Represents pixels.

[0087] S33: Use the connected domain denoising method based on the least squares method to remove tiny noise points in the segmented image;

[0088] After segmentation using the improved Fuzzy C-means (FCM) clustering algorithm, small noise points are often present in the image. Therefore, image denoising is used to remove these small noise points. The image denoising algorithm used in this embodiment is a connected domain denoising method based on the least squares method, a commonly used image processing method. This method assumes that noise points are usually isolated and not connected to other pixels, while the target in a real image is usually a continuous area. The filtered connected domains are then resynthesized into an image to obtain an image without the noise points. The specific steps are as follows:

[0089] (1) Find all connected domains in the rock mass fracture image and obtain the number of pixels contained in each of their connected domains;

[0090] (2) Determine the threshold range of the algorithm by using the least squares method;

[0091] (3) Determine the threshold range of the algorithm based on the median value of the threshold;

[0092] (4) The connected domain of the rock mass fracture image is processed according to the limitation of the threshold range.

[0093] When there is not much noise in the image and the size of the connected domain where the noise is located is significantly different from the size of the connected domain where the target is located, the least squares method can achieve more accurate results. The expression of the least squares method is:

[0094]

[0095] According to the connected domain data in the rock mass fracture image, the least squares method is used to determine the threshold range. This method is suitable for situations with less noise and has better effects when the size of the connected domains varies greatly.

[0096] S34: In order to accurately calculate the area size, the Canny algorithm is used to identify contour information. The output of the Canny algorithm is a binary image, in which white pixels represent the detected edges and black pixels represent the background. The Canny algorithm can effectively suppress noise and details, has high accuracy and stability in edge detection, can effectively process various types of images, and has high computational efficiency. In the Canny algorithm, a Gaussian filter is used. It is a linear filter obtained by sampling and normalizing a two-dimensional Gaussian function. The mathematical expression of the two-dimensional Gaussian function used by the Gaussian filter in the Canny algorithm is as follows:

[0097]

[0098] S=G×I

[0099] Where: G represents the Gaussian filter obtained by operating the two-dimensional Gaussian function, I represents the image to be detected, and S represents the smoothed image.

[0100] The magnitude and direction of the gradient of each pixel in the smoothed image are calculated. The vertical and horizontal components of the gradient of the pixel on the edge are:

[0101]

[0102] Pixel gradient size g θ and direction θ are:

[0103]

[0104] Based on the given high and low thresholds, the strong and weak edges are divided and isolated weak edges are eliminated to obtain the final detection result. In the Canny algorithm, two high and low thresholds T1 and T2 are set. Points with gradient values greater than T1 are considered strong edge points, and vice versa.

[0105] The two thresholds T1 and T2 of the Canny algorithm need to be manually input and determined, so the OTSU algorithm is used to automatically determine the threshold. The principle of the OTSU algorithm is to select a value from the grayscale value range of the image as the threshold, and then use the threshold to classify the image types on the image into two types, and calculate the maximum variance coefficient between the two types. The inter-class variance between the two classes is used as the standard. When the variation value between the two classes of the image reaches the maximum, the threshold is the optimal threshold.

[0106] Let f(x,y) and g(x,y) be the grayscale value of the image at (x,y) and the grayscale average of the k×k neighborhood, respectively. Then:

[0107]

[0108] Let f(x,y)=i,g(x,y)=j,set M i,j is the total number of times vector (i, j) occurs, and the probability density calculation formula is:

[0109]

[0110] Where: M i,j is the total number of image pixels, and N is the image grayscale level.

[0111] Find the optimal two-dimensional threshold (s, t). After formula derivation, the basic equation of the inter-class scatter matrix is:

[0112] S B (s,t)=ω a (μ a -μ) T (μ a -μ)+ω b (μb -μ) T (μ b -μ)

[0113] Tr(S B (s,t))=ω a [μ ai -μ i ) 2 +(μ aj -μ j )]+ω b [μ bi -μ i ) 2 +(μ bj -μ j )]

[0114] Where: s represents a feature dimension in the feature space, t represents another feature dimension, S B (s, t) represents the element located in row s and column t in the inter-class scatter matrix; ω a represents the prior probability of category a, μ a represents the mean vector of category a, μ represents the overall mean vector; ω b represents the prior probability of category b, μ b represents the mean vector of category b; T represents the transpose operation of the matrix, and Tr(·) represents the trace of the matrix;

[0115] When Tr(S B (s 1 ,t 1 ))=max{Tr(S B (s,t))},0≤s,t≥L-1, (s 1 ,t 1 ) is the optimal threshold; although the two-dimensional OTSU algorithm has accurate image segmentation results, it has complex steps, complex calculations, and takes a long time. In order to meet the threshold, the pixels of the entire image must be retrieved once. Therefore, based on the two-dimensional OTSU algorithm, improvements are made to speed up the processing. The specific steps are as follows:

[0116] (1) Reduce the size of the denoised image by pooling, retaining the image features and only reducing the pixel area, which speeds up the retrieval process to a certain extent;

[0117] (2) Segment the pooled image into four parts of the same size. The four parts of the image include the four corners of the original image. Starting from the corners and moving diagonally, select three-quarters of the image. The four segmented images all have the common part of the original image, and basically retain the image features. However, some image features are still not included in the segmented images, and the pixel area of the image is also reduced.

[0118] (3) Use the two-dimensional OTSU algorithm to process the four parts of the image segmentation respectively, obtain four sets of corresponding thresholds, and take the average value of the four sets of thresholds as the new threshold.

[0119] The above method can greatly speed up the processing and maintain accuracy.

[0120] S35: Combined with the improved Canny algorithm, the pixel area is calculated using the image processing algorithm. The actual area of the crack is calculated based on the proportional relationship between the pixel area and the actual size.

[0121] Area calculation is performed using a modified Canny algorithm after image denoising. First, the boundaries are identified, and pixel areas are calculated using an image processing algorithm. Finally, the true area of the crack is calculated based on the proportional relationship between the pixel area and the actual size. This pixel-based crack area calculation method is applicable to a variety of irregular crack shapes and can be implemented using digital image processing techniques. Accurate area calculation involves the following steps: The image is scaled using scale markers or other known size information; an edge detection algorithm is used to extract the crack outline and mark the area; and the pixel count function in the image processing software or programming language is used to count the number of pixels within the area. This gives the pixel area of the region. The actual area is then calculated by multiplying the pixel area by the actual area of each pixel (determined by the image resolution and scale calibration).

[0122] S4: The computer processes the rock mass fragmentation information through the blasting effect evaluation module to obtain the evaluation results, such as Figure 3 As shown, the specific steps include:

[0123] S41: Calculate the number of large blocks of rock mass based on the rock mass block information obtained in S3, set the threshold range based on the connected domain principle, screen the rock mass that meets the conditions, and use the obtained information as a new indicator; specifically, use the obtained area information as the large block rate indicator C1. At the same time, during the construction process, the data such as the drilling length C2 and the unit consumption of explosives C3 can be obtained as reference indicators for blasting effect evaluation, and then use the average value C4 and median C5 of the rock mass block area and the integrity C6 of the surrounding rock mass (including road surface flatness) as corrections to prevent the problem of inaccurate evaluation results caused by the excessively large rock mass block area.

[0124] The evaluation of blasting effect usually considers three aspects: first, safety. The impact of blasting operations on the surrounding environment cannot be ignored; second, quality. Blasting quality can be considered based on the blasting purpose, blasting method, blasting object and surrounding environment. Generally speaking, it is mainly considered from the perspective of rock blockiness; finally, economic benefits, which are mainly considered from the perspective of explosives consumption. Therefore, the above indicators are selected to analyze and evaluate the blasting effect.

[0125] Analyzing the actual situation of open-pit blasting, the comprehensive evaluation index A1 of blasting effect is usually divided into three first-level indicators: quality factor B1, economic factor B2 and safety factor B3. The first-level indicators are further divided into multiple second-level indicators, corresponding to the evaluation indicators of the blasting effect, such as Figure 4 The figure shows the comprehensive evaluation system of blasting effect.

[0126] S42: Use the hierarchical analysis method to determine the hierarchy of the new indicators, compare the indicators of each layer pairwise and analyze them layer by layer to obtain the weight of each indicator. Generally, the 1 to 9 scaling method is used to achieve the quantification of the matrix;

[0127] If an indicator A1 in the previous level contains B1, N2, ..., B m ,The constructed judgment matrix is shown in Table 1.

[0128] Table 1 Judgment Matrix

[0129]

[0130] Among them, a ij Indicates the importance between two indicators. If the former is important, then a ij >1, if they are equally important, then a ij =1, as shown in Table 2.

[0131] Table 2 a of indicators of different importance levels ij Assignment

[0132]

[0133] Calculate the consistency test index CI:

[0134]

[0135] Where: max =n, CI=0; the larger the CI value, the worse the consistency of the judgment matrix, and vice versa;

[0136] The average random consistency index RI is obtained by calculating the average of the eigenvalues of the random judgment matrix multiple times. When the order of the constructed judgment matrix (the number of evaluation indicators) is different, the RI value is also different, as shown in Table 3;

[0137] Table 3 RI values of different matrix orders

[0138]

[0139] Calculating consistency index When CR<0.1, it means that the judgment matrix passes the consistency, otherwise it will be revised;

[0140] The weights of the AB layers are determined according to the established evaluation model, as shown in Table 4.

[0141] Table 4A-B layer weights

[0142]

[0143] Based on the blasting effect classification of related industries, the blasting effect is initially divided into five levels: very good, good, fair, poor, and bad. For ease of expression, the blasting effect is quantitatively analyzed and the stability is expressed in percentages, as shown in Table 5.

[0144] Table 5 Blasting effect grade table

[0145]

[0146]

[0147] S43: After determining the index weights through the hierarchical analysis method, carry out the fuzzy comprehensive evaluation method. First, establish the fuzzy comprehensive evaluation method: determine the factor domain of the evaluation object, set U = {u1,u2,…,u m} is a collection of various indicators that can characterize the characteristics of the evaluation target, and the comment level domain is determined; use V = [v1, v2, ..., v n} represents the evaluation level of various indicators that can characterize the characteristics of the evaluation target. This parameter can be adjusted and is determined by the specific indicator system. It is generally divided into 3 to 5 levels. In this embodiment, the evaluation level is divided into 5 levels, and all levels will correspond one-to-one with a fuzzy subset. Establish an overall fuzzy relationship matrix R, and all evaluation indicators and the evaluated target construct a fuzzy relationship matrix R as follows:

[0148]

[0149] Determine the weight vector of the evaluation factor, assuming that the weight vector set of the indicator is W = (w1, w2, ..., w m ), in this set, w represents the weight of index u in the fuzzy subset. In the process of determining the weight vector of the evaluation factor, this embodiment applies the hierarchical analysis method to obtain the weight and performs normalization processing. The formula is as follows:

[0150]

[0151] Synthesize the fuzzy comprehensive evaluation result vector and perform matrix synthesis. The formula is as follows:

[0152]

[0153] Where: Q represents the result of vector multiplication and matrix multiplication, w m Represents the mth element in the left row vector, r mn represents the element in the mth row and nth column of the intermediate matrix, b n Represents the nth element in the result vector on the right.

[0154] S5: Establishing the GUI system interface, specifically including the following steps:

[0155] S51: Establish system interface, such as Figure 5 As shown, development is performed using the built-in GUI development toolbox in Matlab software, which is mainly divided into the following four areas:

[0156] (1) Image display area: four display frames are used to display the effect images of different functions;

[0157] (2) Image processing area: consists of loading image files, binarization processing, image segmentation and image denoising;

[0158] (3) Blasting effect evaluation area: It consists of the maximum value of block size area, average value of block size area, large block rate and effect evaluation;

[0159] (4) Save and exit area: Save the processed image and information, and click Exit to close everything;

[0160] The operation process and functions of this system are as follows:

[0161] ①This system can process a variety of image formats and supports the processing and recognition of most common images. Obtain images through the device, click "Load image file" to import the image into this system, and then implement subsequent functions;

[0162] ② After the image is loaded successfully, you can perform image processing module operations, including grayscale, binarization, image segmentation and image denoising functions. You can select the corresponding function according to your needs;

[0163] ③ After image processing is completed, the blasting effect evaluation module is used to calculate the maximum block size area, average block size area and large block rate. At the same time, other parameters are input, and finally the blasting effect evaluation function is used to complete the blasting effect evaluation;

[0164] ④ After the processing task is completed, save the image and data. After use, click "Exit Program" to end the operation.

[0165] S52: In Matlab, packaging an executable program is accomplished using built-in functionality. First, open Matlab, enter deploytool in the command window, select Application Compiler, add the application source code, add associated files, and select the packaging method.

[0166] and Figure 1 Corresponding to the above method, the embodiment of the present invention also provides a blasting fragmentation recognition and blasting effect evaluation system based on image processing, which is used to Figure 1 The specific implementation of the method in the embodiment of the present invention is a blasting fragmentation recognition and blasting effect evaluation system based on image processing, which can be applied to computer terminals or various mobile devices, such as Figure 6 As shown, it specifically includes: a drone 1, a high-definition camera 4, a laser radar 7, and a computer 5. The computer 5 includes an image processing module 2, a blasting effect evaluation module 3, and a GUI system interface 6;

[0167] The high-definition camera 4 is mounted on the drone 1 and is used to collect image information of the rock mass and transmit it wirelessly to the computer 5. Specifically, the high-definition camera 4 installed on the bottom of the drone 1 can change its direction and angle through manual operation.

[0168] The laser radar 7 is mounted on the UAV 1 and is used to obtain the distance information between the UAV and the rock mass and transmit it wirelessly to the computer 5;

[0169] Image processing module 2 is used to process the image information of the rock mass and extract the rock mass blockiness information. Image processing refers to the use of algorithms on computers to process images in different scenarios, mainly including pre-processing methods such as grayscale, binarization, and histogram equalization. In the process of collecting rock mass images, there are many factors that interfere with the subsequent image analysis. Therefore, when extracting the main information in the image, image pre-processing operations are used to improve the image quality and highlight the characteristic areas. Image pre-processing plays an important role in image analysis.

[0170] Blasting effect evaluation module 3 is used to process rock fragmentation information and obtain evaluation results. Specifically, the blasting effect and guidance plan are judged based on established indicators. The blasting effect and guidance plan are determined based on parameters such as the maximum value of the rock fragmentation area, the minimum value of the rock fragmentation area, the average value and variance of the rock fragmentation area, and the drilling length, explosive consumption per unit, and shovel loading efficiency indicators are added for evaluation. The rock blasting evaluation effect is initially divided into five categories. The category is determined by the calculated values based on the established indicators, and the blasting effect is then evaluated.

[0171] GUI system interface 6 is used to display rock image processing information and evaluation results. Specifically, the system interface is programmed using Matlab software. Matlab is a powerful mathematical computing software and a commonly used tool in the field of image processing. It provides many image processing-related functions and toolboxes, which can easily implement various image processing algorithms. Matlab uses a popular graphical user interface, which integrates command input, execution, modification and debugging. It is called an integrated development environment. The operation is very intuitive and convenient, and it has excellent visualization capabilities. It can easily generate various images and charts to display and analyze image processing results. Matlab provides a powerful GUI development toolbox, such as GUIDE (GUI Development Environment), which can quickly create a customized graphical interface. Users do not need to write code tediously, but can build a GUI through simple drag and drop and configuration operations. This system interface is mainly divided into an image processing module, a blasting effect evaluation module and other functions.

[0172] During use, the drone 1 first arrives at the designated location, measures the distance between it and the rock mass through the laser radar 7, and then uses the high-definition camera 4 to obtain image information of the rock mass. The image information and distance information are transmitted to the computer 5 through a wireless connection. The computer 5 uses the image processing module 2 to extract the rock mass block information, and then processes the rock mass block information extracted by the image processing module 2 through the blasting effect evaluation module 3 to evaluate the blasting effect.

[0173] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0174] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for identifying blasting fragmentation and evaluating blasting effects based on image processing, characterized in that: The following steps are involved: S1: Install the high-definition camera and lidar on the drone and establish a connection with the computer. After debugging, they will enter the working state; S2: After the UAV arrives at the designated location, it collects image information of the rock mass through a high-definition camera, obtains the distance information between the UAV and the rock mass through a laser radar, and wirelessly transmits the collected image data and distance data to a computer; S3: The computer processes the image data through the image processing module to extract the rock mass block information; S4: The computer processes the rock mass block information through the blasting effect evaluation module to obtain the evaluation results.

2. The method for blasting fragmentation recognition and blasting effect evaluation based on image processing according to claim 1, characterized in that: In S3, the rock mass fragmentation information is extracted, which specifically includes the following steps: S31: converting the color image captured by the high-definition camera into a gray image by grayscale operation; S32: Use the improved fuzzy C-means clustering algorithm to segment the gray image; S33: Use the connected domain denoising method based on the least squares method to remove tiny noise points in the segmented image; S34: The basic equation of the inter-class dispersion matrix is derived through formula derivation: S B (s,t)=ω a (m a -m) T (m a -m)+w b (m b -m) T (m b -μ)Tr(S B (s,t))=ω a [(m ai -m i ) 2 +(m aj -μj)]+ω b [(m bi -m i ) 2 +(m bj -m j )] Where: s represents a feature dimension in the feature space, t represents another feature dimension, S B (s, t) represents the element located in row s and column t in the inter-class scatter matrix; ω a represents the prior probability of category a, μ a represents the mean vector of category a, μ represents the overall mean vector; ω b represents the prior probability of category b, μ b represents the mean vector of category b; T represents the transpose operation of the matrix, and Tr(·) represents the trace of the matrix; When Tr(S B (s 1 ,t 1 ))=max{Tr(S B (s,t))},0≤s,t≥L-1, (s 1 ,t 1 ) is the optimal threshold; S35: Combined with the improved Canny algorithm, the pixel area is calculated using the image processing algorithm. The actual area of the crack is calculated based on the proportional relationship between the pixel area and the actual size.

3. The method for blasting fragmentation recognition and blasting effect evaluation based on image processing according to claim 2, characterized in that: In S32, the improved fuzzy C-means clustering algorithm comprehensively utilizes local and global spatial information, constructs a joint membership function Y through combination, and calculates the new cluster center X: Where: u mn represents the local membership, μ mn represents the global membership; p represents the fuzzy index, b and q represent the relative importance parameters of controlling the global membership and local membership, N represents the number of pixels, and x c Represents pixels.

4. The method for blasting fragmentation recognition and blasting effect evaluation based on image processing according to claim 2, characterized in that: In S34, during the confirmation of the optimal threshold, the pixels of the entire image are retrieved once, and an improvement is made based on the two-dimensional OTSU algorithm to automatically obtain two thresholds T1 and T2. The specific steps are as follows: Reduce the size of the denoised image by pooling to retain the image features; Segment the pooled image into four parts of the same size. The four parts of the image include the four corners of the original image. Starting from the corners, select three-quarters of the image in the diagonal direction. The four segmented parts all have the common part of the original image. The two-dimensional OTSU algorithm is used to process the four parts of the image segmentation respectively to obtain four sets of corresponding thresholds, and the average value of the four sets of thresholds is taken as the new threshold.

5. The method for blasting fragmentation recognition and blasting effect evaluation based on image processing according to claim 1, characterized in that: In S4, the evaluation results are obtained, which specifically includes the following steps: S41: Calculate the number of large blocks of the rock mass based on the rock mass blockiness information obtained in S3, set a threshold range based on the connected domain principle, screen the rock mass that meets the conditions, and use the obtained information as a new indicator; S42: Use the hierarchical analysis method to determine the hierarchy of the new indicators, compare the indicators of each layer pairwise and analyze them layer by layer, and calculate the consistency test index CI; obtain the average random consistency index RI by taking the average of the eigenvalues of the random judgment matrix after multiple calculations; calculate the consistency index When CR<0.1, it means that the judgment matrix passes the consistency, otherwise it will be revised; S43: After determining the indicator weights through the hierarchical analysis method, the fuzzy comprehensive evaluation method is carried out to synthesize the fuzzy comprehensive evaluation result vector. The formula is as follows: Where: Q represents the result of vector and matrix multiplication, w m Represents the mth element in the left row vector, r mn represents the element in the mth row and nth column of the intermediate matrix, b n Represents the nth element in the result vector on the right.

6. The method for blasting fragmentation recognition and blasting effect evaluation based on image processing according to claim 1, characterized in that: It also includes S5, establishing a GUI system interface, which specifically includes the following steps: S51: Establish the system interface, which is developed using the GUI development toolbox built into the Matlab software. It is mainly divided into the following four areas: (1) Image display area: four display frames are used to display the effect images of different functions; (2) Image processing area: consists of loading image files, binarization processing, image segmentation and image denoising; (3) Blasting effect evaluation area: It consists of the maximum value of block size area, average value of block size area, large block rate and effect evaluation; (4) Save and exit area: Save the processed image and information, and click Exit to close everything; S52: In Matlab, enter deploytool in the command window, select Application Compiler, add the application source code, add associated files, select the packaging method, and complete the packaging operation.

7. A blasting fragmentation recognition and blasting effect evaluation system based on image processing, characterized in that: The method for identifying blasting fragmentation and evaluating blasting effect based on image processing according to any one of claims 1 to 6 is implemented, comprising: a drone, a high-definition camera, a laser radar, and a computer, wherein the computer comprises an image processing module, a blasting effect evaluation module, and a GUI system interface; A high-definition camera is mounted on the drone to collect image information of the rock mass and transmit it wirelessly to a computer; The laser radar is mounted on the UAV to obtain the distance information between the UAV and the rock mass and transmit it wirelessly to the computer; Image processing module, used to process the image information of the rock mass and extract the rock mass block information; The blasting effect evaluation module is used to process the rock mass block information and obtain the evaluation results; GUI system interface, used to display rock mass image processing information and evaluation results.

Images