CN112950659A - Image processing method for fracturing propping agent microscopic amplification - Google Patents
Image processing method for fracturing propping agent microscopic amplification Download PDFInfo
- Publication number
- CN112950659A CN112950659A CN202110035853.9A CN202110035853A CN112950659A CN 112950659 A CN112950659 A CN 112950659A CN 202110035853 A CN202110035853 A CN 202110035853A CN 112950659 A CN112950659 A CN 112950659A
- Authority
- CN
- China
- Prior art keywords
- image
- particle
- edge
- point
- fracturing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 18
- 238000003199 nucleic acid amplification method Methods 0.000 title claims abstract description 8
- 230000003321 amplification Effects 0.000 title claims abstract description 7
- 239000002245 particle Substances 0.000 claims abstract description 205
- 238000000034 method Methods 0.000 claims abstract description 72
- 238000012545 processing Methods 0.000 claims abstract description 40
- 238000000605 extraction Methods 0.000 claims abstract description 33
- 238000011049 filling Methods 0.000 claims abstract description 31
- 238000001514 detection method Methods 0.000 claims abstract description 26
- 238000009499 grossing Methods 0.000 claims abstract description 25
- 238000003709 image segmentation Methods 0.000 claims abstract description 25
- 239000011159 matrix material Substances 0.000 claims description 48
- 238000001914 filtration Methods 0.000 claims description 22
- 230000011218 segmentation Effects 0.000 claims description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 5
- 238000010606 normalization Methods 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 230000008030 elimination Effects 0.000 claims description 3
- 238000003379 elimination reaction Methods 0.000 claims description 3
- 239000013589 supplement Substances 0.000 claims description 3
- 230000001629 suppression Effects 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 238000000386 microscopy Methods 0.000 claims 8
- 238000005259 measurement Methods 0.000 abstract description 9
- 239000003795 chemical substances by application Substances 0.000 description 64
- 238000010586 diagram Methods 0.000 description 11
- 238000005070 sampling Methods 0.000 description 10
- 238000003708 edge detection Methods 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 4
- 238000005429 filling process Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000003247 decreasing effect Effects 0.000 description 2
- 239000012530 fluid Substances 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 238000011179 visual inspection Methods 0.000 description 2
- 241000252254 Catostomidae Species 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 230000010339 dilation Effects 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 238000005315 distribution function Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 239000008187 granular material Substances 0.000 description 1
- 125000001475 halogen functional group Chemical group 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 239000007769 metal material Substances 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 239000011148 porous material Substances 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 239000011435 rock Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
- G06T5/30—Erosion or dilatation, e.g. thinning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration using histogram techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10056—Microscopic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20048—Transform domain processing
- G06T2207/20056—Discrete and fast Fourier transform, [DFT, FFT]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Geometry (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an image processing method for fracturing propping agent microscopic amplification, which comprises the following steps: acquiring image information of the amplified fracturing proppant particles; carrying out smoothing processing, image segmentation, hole filling and area marking on the image information to obtain a processed particle image; adopting a Canny operator detection method to carry out edge extraction on the processed particle image; and obtaining the geometric characteristics of the particle image after edge extraction, wherein the geometric characteristics comprise area, perimeter, maximum inscribed circle radius and minimum circumscribed circle radius. According to the method, the particle images of the collected fracturing propping agent are subjected to smoothing treatment, image segmentation, hole filling and area marking treatment, so that the precision of the particle images of the fracturing propping agent is improved, the labor cost is greatly saved, and the measurement efficiency and the measurement precision are improved.
Description
Technical Field
The invention belongs to the technical field of detection of fracturing propping agents of oil and gas fields, and particularly relates to an image processing method for micro-amplification of the fracturing propping agents.
Background
Fracturing technology is well known and widely used as an emerging effective means of modifying oil and gas fields. The fracturing technology is that the rock stratum generates fractures of different degrees by utilizing artificially applied ultrahigh pressure, and then fluid is injected into the fractures, so that a channel which is easy for oil and gas to flow through can be formed. In order to keep the fractures created after fracturing open at all times, a granular fracturing proppant is typically injected into the fluid.
The fracturing support adopted according to the requirements of the fracturing process must be solid particles with certain roundness and sphericity, and the higher the sphericity value of the particles, the better. The quality and the performance of the fracturing propping agent play a crucial role in the diversion degree of oil and gas and directly determine the oil and gas yield. Currently, the evaluation of roundness and sphericity is usually determined by manual calibration through comparison of sphericity and roundness templates through visual inspection, that is, generally, under a microscope, the magnification of the microscope is adjusted to be 30-40 times or proppant particles are photographed through a photomicrograph technology, then the contrast is performed on the proppant particles and a sphericity and roundness template (Krumbein-Sloss template) of an industry specified standard, and the sphericity and roundness of each proppant particle are determined according to collected images of all particles.
However, the proppant particle images collected by a microscope are directly used for manual visual calibration, the particle images are easy to be adhered, and the images are easy to be influenced by external factors, so that the measured results of the sphericity value and the circularity value are inaccurate; the probability that the manual measurement is interfered by human is also high, so that the basis is insufficient, and the traceability is poor.
Disclosure of Invention
The invention provides an image processing method for fracturing propping agent microscopic amplification, which aims to solve the problems and solve the problems that the manual visual inspection calibration is carried out on propping agent particle images directly acquired by a microscope, the particle images are easy to adhere, and the images are easy to be influenced by external factors, so that the measured results of sphericity and circularity are inaccurate; the probability that the manual measurement is interfered by human is also high, so that the basis is insufficient, and the traceability is poor.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
an image processing method for fracturing proppant microscopical magnification, comprising the steps of:
acquiring image information of the amplified fracturing proppant particles;
carrying out smoothing processing, image segmentation, hole filling and area marking on the image information to obtain a processed particle image;
adopting a Canny operator detection method to carry out edge extraction on the processed particle image;
and obtaining the geometric characteristics of the particle image after edge extraction, wherein the geometric characteristics comprise area, perimeter, maximum inscribed circle radius and minimum circumscribed circle radius.
The method further comprises the following steps before the acquiring the image information of the amplified fracturing proppant particles:
the fracturing propping agent particles are sucked and conveyed to a vacuum negative pressure sucker on a vacuum negative pressure device through a vacuum negative pressure device, and a plurality of sucker holes are uniformly distributed on the vacuum negative pressure sucker;
and (4) carrying out image amplification on the fracturing propping agent particles on the air negative pressure device through a microscope.
The smoothing process specifically comprises:
smoothing the acquired particle image of the fracturing propping agent by adopting Gaussian filtering, wherein the smoothing is used for optimizing the edge of the image and reducing the influence of noise elimination on the edge, and the Gaussian filtering comprises the following steps:
1) firstly, carrying out gray level transformation on an original image of the collected fracturing propping agent, and converting the original image into a gray level image;
2) according to the formula
Generating a Gaussian sequence;
where P (z) represents the gray scale value, μ represents the mean or expected value of z, σ represents the standard deviation of z, and σ represents the square of the standard deviation2Is the variance of z;
and filtering the gray level image by using a Gaussian sequence.
The image segmentation specifically comprises the following steps:
performing image segmentation on the smoothed particle image by adopting an Otsu optimal threshold segmentation method, wherein the method comprises the following steps of:
1) carrying out histogram statistics on the particle image;
2) obtaining the average value of pixel points in the particle image according to the histogram;
3) according to the histogram statistics, carrying out normalization processing on the histogram;
4) obtaining an inter-class variance matrix according to the histogram to obtain a maximum inter-class variance;
5) obtaining a threshold value of the maximum value of the inter-class variance through the maximum inter-class variance, and determining an optimal threshold value;
6) and segmenting the particle image of the fracturing propping agent according to the optimal threshold value.
The hole filling specifically comprises:
carrying out hole filling processing on the particle image subjected to image segmentation processing through an expansion algorithm, wherein the hole filling processing is used for eliminating unnecessary holes in a fracturing propping agent particle image area, and the hole filling processing comprises the following steps:
recording the positions of neighborhood points in the particle image; detecting neighborhood point pixels; and filling holes in the particle image according to the neighborhood point pixels.
The area marks are specifically:
carrying out region marking processing on the particle image by adopting an 8-neighborhood scanning method, wherein the region marking is used for distinguishing and classifying a plurality of particles in the particle image; the area marking process includes the steps of:
1) initializing the region serial numbers of all pixel points of the image to be 0, scanning line by line from the leftmost point in the granular image, taking the scanned point with the first pixel value of 0 as a starting point, and marking the region serial number of the point to be 1;
(2) scanning the next point; if the pixel value of the point is 0, judging the condition of each point in the eight-neighborhood of the point; if the area serial number of one point is not 0, the area serial number of the point is the same as the area serial number of the point; if the sequence numbers of the regions of all the points in the eight neighborhoods of the point are 0, the point belongs to a new region, and the sequence number of the region is the maximum value of the sequence number of the current region plus 1;
(3) after all points in the particle image are scanned, marking different fracturing proppant particles in the fracturing proppant particle image according to the region positions.
The method for extracting the edge of the particle image by Canny operator detection comprises the following steps:
1) after the gray level of the particle image is transformed, Gaussian filtering is carried out to optimize the edge of the image and reduce the influence of noise on the edge; the mathematical expression satisfying the first criterion is:
the expression satisfying the second criterion is:
when the scale of f is changed, let fw(x) F (x/w), yielding:
2) and (3) solving the gradient value of each pixel point by using a first derivative operator, wherein the formula is as follows:
setting the smoothed image matrix as I [ I, j ], the matrix expression of the partial derivatives in x and y directions is:
3) the method for searching the edge point by adopting the non-maximum suppression method comprises the following operations: the maximum gradient value of the gradient of the adjacent pixel in the gradient direction is reserved and marked as 1, 1 represents a possible edge point, other marks are 0, 0 is a non-edge point, and the expression is as follows:
N[i,j]＝NMS(M[i,j],ξ[i,j]) (6)
wherein, the zero value point of the formula (6) is the non-edge point; searching for real edge points by using the contrast of the edge points and the non-edge points;
4) the method for detecting the edge of the particle image by adopting double-threshold detection processing comprises the following steps: first, two thresholds T are sethAnd TlThen detecting the image processed by the steps 1)2)3) and adding all the image larger than the threshold value ThIs classified as one, will be greater than a threshold value TlRemain as another class; after double-threshold detection, two threshold edge image matrixes T are obtainedh[i,j]And Tl[i,j](ii) a Matrix T with high threshold detectionh[i,j]Based on a matrix T obtained by low threshold detectionl[i,j]To supplement this, the two matrices are combined to extract the edges of the more complete image.
The geometric feature extraction comprises the steps of extracting the area, the perimeter, the maximum inscribed circle radius and the minimum circumscribed circle radius of a particle image of the fracturing propping agent;
extracting the area of the particle image by adopting a contrast matrix method to extract the number of pixel points occupied by a target area, namely the boundary of the area and all the pixel points contained in the boundary; the method comprises the following steps:
1) let K be 1, 2.., K, and establish a matrix M K × ones (M, n); k represents the number of particles, m × n is the size of the image;
2) comparing the matrix I corresponding to the image with the established matrix M to generate a brand new matrix NewM obtained by comparing elements at corresponding positions on the matrix I and the matrix M, and if the corresponding positions are consistent, marking the position as 1, and if the corresponding positions are inconsistent, marking the position as 0;
3) forming a matrix NewM processed in the step 2) into a matrix only containing elements 0 and 1, summing the elements of the matrix NewM, wherein the sum is the number of pixels marked as k, and multiplying the sum by a conversion coefficient to obtain the area of the particle image.
The extraction of the maximum inscribed circle radius and the minimum circumscribed circle radius of the particle image comprises the following steps:
assuming that the minimum condition circle center of the fracture proppant edge image is (a, b), the objective function is F (a, b), and the formula is as follows:
wherein: rmaxDenotes the radius, R, furthest from (a, b) to the edgeminRepresents the nearest radius from (a, b) to the edge; x is the number ofmax，ymaxIs and RmaxCorresponding coordinates; x is the number ofmin，yminIs and RminCorresponding coordinates;
the objective function of the minimum circumscribed circle is:
the objective function of the maximum inscribed circle is:
when F (a, b) in formula (9) takes the minimum value, then (a, b) is the center coordinates of the minimum circumscribed circle of the minimum condition; when F (a, b) in equation (9) takes the maximum value, then (a, b) is the center coordinates of the maximum inscribed circle of the minimum condition.
The perimeter extraction of the particle image is to extract the boundary length of each particle in the particle image; the method for calculating the perimeter of the particle image by adopting the eight-direction chain code method comprises the following steps:
1) recording the number of continuous perimeter pixel points in the vertical direction: recording the number of the continuous pixel points in the vertical direction as Ny；
2) Recording the number of continuous perimeter pixel points in the horizontal direction: recording the number of continuous pixels in the horizontal direction and recording as Nx；
3) Calculating the total SN of the connected boundary pixels by adopting a formula No＝SN-Nx-NyCalculating the chain code number N of odd codeso；
4) Using a formula
And calculating the eight-connected chain code perimeter of the fracturing proppant image.
The method has the advantages that the method acquires the image information of the amplified fracturing propping agent particles, performs smoothing processing, image segmentation, hole filling and area marking on the image information to obtain processed particle images, and performs edge extraction on the processed particle images by adopting a Canny operator detection method to finally obtain the geometric characteristics of the edge-extracted particle images. According to the method, the particle images of the collected fracturing propping agent are subjected to smoothing treatment, image segmentation, hole filling and area marking treatment, so that the precision of the particle images of the fracturing propping agent is improved, the labor cost is greatly saved, and the measurement efficiency and the measurement precision are improved.
Drawings
FIG. 1 is a flow chart of an image processing method of the present invention for fracturing proppant microsampling;
FIG. 2 is a schematic structural diagram of the vacuum chuck device of the present invention;
FIG. 3 is a flow chart of the Gaussian filtering process of the present invention;
FIG. 4 is a flow chart of the thresholding process of the present invention;
FIG. 5 is an image of a fractured proppant particle image of the present invention after a thresholding process;
FIG. 6 is a schematic view of the expansion process of the hole filling process of the present invention;
FIG. 7 is a flow chart of the hole filling process of the present invention;
FIG. 8 is an image of fractured proppant particles after hole filling of the present invention;
FIG. 9 is a schematic diagram of a region 8 of a pixel in a particle image according to the present invention;
FIG. 10 is a flow chart of zone tagging of the present invention;
FIG. 11 is an image of fractured proppant particles after connected zone marking of the present invention;
FIG. 12 is a flow chart of edge detection according to the present invention;
FIG. 13 is a schematic view of an edge type of the present invention;
FIG. 14 is a diagram illustrating the edge detection results of the Canny edge operator of the present invention for particles;
FIG. 15 is a schematic diagram of an 8-way chain code according to the present invention;
FIG. 16 is a diagram illustrating the result of image processing according to the present invention;
wherein, 1 is a sucker base, and 2 is a vacuum negative sucker.
Detailed Description
An image processing scheme for fracturing proppant micro-magnification provided by the embodiments of the present invention will be described in detail below by way of several specific examples.
Referring to fig. 1, the invention discloses an image processing method for fracturing proppant microscopic magnification, which comprises the following steps:
acquiring image information of the amplified fracturing proppant particles;
carrying out smoothing processing, image segmentation, hole filling and area marking on the image information to obtain a processed particle image;
adopting a Canny operator detection method to carry out edge extraction on the processed particle image;
and obtaining the geometric characteristics of the particle image after edge extraction, wherein the geometric characteristics comprise area, perimeter, maximum inscribed circle radius and minimum circumscribed circle radius.
In the above embodiment, when image information of fracturing proppant particles is collected, the fracturing proppant particles can be sucked and conveyed to a vacuum negative pressure sucker on a vacuum negative pressure device through the vacuum negative pressure device, and a plurality of sucker holes are uniformly distributed on the vacuum negative pressure sucker; and carrying out image amplification on the fracturing propping agent particles on the air negative pressure device through a microscope. The cross section of the vacuum negative pressure sucker 2 is circular or rectangular, and a plurality of sucker holes are uniformly distributed on the vacuum negative pressure sucker 2. According to the invention, through testing the vacuum negative pressure suckers 2 with different shapes and diameters, the number of pore diameter distribution points of the circular vacuum negative pressure sucker 2 is small, and according to the structure of the microscope objective table, the circular vacuum negative pressure sucker 2 can fully observe fracturing propping agent particles distributed on the sucker only by swinging in four directions, so that the observation is not facilitated, and the operation is complex; the aperture distribution of the rectangular vacuum negative pressure sucker 2 is more, the requirement of sampling points is met, the structure of the object stage of the microscope is matched, observation of fracturing propping agent particles distributed on the vacuum negative pressure sucker 2 can be completed only in the left direction or the right direction or the upper direction and the lower direction, and the operation is simple. Therefore, the rectangular vacuum chuck 2 is preferably used. The aperture of sucking disc hole be 0.4mm, the interval between the adjacent sucking disc hole is 1.5mm, can guarantee like this to gather the quantity of fracturing propping agent sample, also can ensure simultaneously that the propping agent granule can not make image processing become more complicated owing to apart from too near or direct adhesion phenomenon. Collecting sample particles of fracturing propping agent by a sampling mode of a vacuum negative pressure device, amplifying the fracturing propping agent on a vacuum negative pressure sucker by using a microscope, collecting particle images of each fracturing propping agent, and processing the particle images of the fracturing propping agents; smoothing, image segmentation, hole filling and area marking are carried out on the particle image to obtain a processed particle image; then, performing edge extraction on the processed particle image by adopting a Canny operator detection method; and extracting geometric features of the processed particle image, including the area, the perimeter, the maximum inscribed circle radius and the minimum circumscribed circle radius. The invention adopts the vacuum negative pressure device for sampling, the particles are positioned uniformly, the particles are prevented from being adhered, the stability is high, and the working efficiency of the microscope for collecting the particle images is improved. The invention also carries out smoothing treatment, image segmentation, hole filling and area marking treatment on the collected particle image of the fracturing propping agent, improves the precision of the fracturing propping agent particle image, greatly saves the labor cost, and improves the measurement efficiency and the measurement precision.
The method adopts a vacuum negative pressure device for sampling, the fracturing propping agent particles are sucked and sent to a vacuum negative pressure sucker 2 of a vacuum negative pressure sucker device of a microscope, vacuum negative pressure generated by a fan is used as conveying power, the fracturing propping agent bulk particles are sucked and sent to a specified positioning point to be detected by using negative pressure airflow, and the negative pressure pneumatic conveying can be controlled according to the size of the fracturing propping agent; the invention selects the oil-free vacuum pump head as the vacuum negative pressure device, the device can start air extraction from atmospheric pressure and can directly discharge the extracted body to the atmosphere, no oil or other working media exist in the pump cavity, and the limit pressure of the pump is the same magnitude as or close to that of the oil-sealed vacuum device. Specifically, the microscope of the invention adopts a high-end research-grade SteREO discovery.V8 model under the Karl Chuiss brand, and adopts a full-circle annular reflection light LED cold light source, so that the collected image is uniform in illumination, free of shadow and clear in image, the error is reduced for processing the image, and the accuracy is improved. The microscope provided by the invention collects the particle images of the fracturing propping agent under the magnification of 30 times, and the fracturing propping agent particle images collected under the magnification of 30 times are clear and have moderate visual field, so that the follow-up image processing work is facilitated to observe.
Referring to fig. 2, a schematic view of the vacuum suction cup device of the present invention is shown. The vacuum negative pressure sucker device comprises a sucker base 1 and a vacuum negative pressure sucker 2 arranged on the sucker base. The vacuum negative pressure sucker device is suitable for particle objects with uneven sizes and weights, particles are uniformly positioned and distributed, a particle sieve is not needed, the system stability is high, the operation is simple, and the working efficiency is high. The vacuum negative pressure sucker device is arranged at a particle to-be-detected positioning point of the microscope. The purpose of sucking and conveying the particle sample through the vacuum negative pressure device is to avoid mutual adhesion between fracturing propping agent particles, so that the particle images of the fracturing propping agent collected by a microscope are caused to have an adhesion phenomenon, and then the calculation results of the roundness and sphericity of the fracturing propping agent are influenced.
Specifically, the vacuum negative pressure device sucks and sends a fracturing propping agent particle sample to a vacuum negative pressure sucker 2 of a vacuum negative pressure sucker device of a microscope, the microscope collects an obtained fracturing propping agent particle image, and the particle image is subjected to smoothing treatment, image segmentation, hole filling and area marking to obtain a treated particle image; then, performing edge extraction on the processed particle image by adopting a Canny operator detection method; and extracting geometric features of the processed particle image, including the area, the perimeter, the maximum inscribed circle radius and the minimum circumscribed circle radius.
The invention adopts the vacuum negative pressure sucker device arranged below the microscope for sampling, and has great difficulty in meeting the requirements of rapidness, uniformity, stability and the like in the sampling and placing process because the fracturing propping agent has small diameter, light weight and various shapes and is easy to be polymerized into one piece. If the fracturing propping agent can not effectively separate the particles of the pressure propping agent in the sampling process, the phenomena of undersize intervals between the particles or mutual adhesion and the like occur, so that the adhesion phenomenon of the collected propping agent images can be caused, the difficulty and complexity of image processing are increased, the misjudgment of the propping agent image edges is easily caused in the image segmentation process, the precision of the propping agent edges is further influenced, and the measurement of the roundness and sphericity of the pressure propping agent is obviously influenced. The tray in the traditional sampling mode adopts a mode that a particle sieve and a flat plate are tightly combined up and down, wherein the particle sieve is arranged on the upper layer, and the flat plate is arranged on the lower layer of the particle sieve. The flat plate is engraved with sunken positioning points which correspond to holes of the particle sieve one by one, when in use, the propping agent is put into the particle sieve, the redundant particles are removed by shaking, then the particle sieve is taken away, the propping agent to be measured is left at the position of the flat plate positioning points, and the reading is taken for photographing. However, different particle sieves are required to be selected for particles with different sizes, and experiments show that the particles are overlapped after the particle sieves are removed, and the particles roll, so that the operation difficulty is high. In addition, the material selection difficulty of the tray is also higher, the groove is difficult to be formed after the metal material is tested, and the halo phenomenon can occur under a microscope after the grooves are formed in other materials (such as organic glass) so as to influence the judgment of the sphericity. Therefore, the invention adopts a vacuum negative pressure device for sampling, and uses the vacuum negative pressure generated by a fan as conveying power to suck bulk particles such as fracturing propping agent and the like onto a vacuum negative pressure sucker 2 at a to-be-detected positioning point of a microscope by using negative pressure airflow, and the microscope collects particle images of the fracturing propping agent.
It should be noted that the invention collects sample particles of fracturing propping agent by a sampling mode of a vacuum negative pressure device, amplifies the fracturing propping agent on a vacuum negative pressure sucker 2 by a microscope, collects particle images of each fracturing propping agent, then carries out smoothing treatment, image segmentation, hole filling and area marking on the particle images of the fracturing propping agent, and then carries out edge extraction and geometric feature extraction on the treated particle images of the fracturing propping agent. According to the invention, the acquired particle image of the fracturing propping agent is processed, so that the precision of the particle image of the fracturing propping agent acquired by a microscope is improved, and a foundation is laid for the subsequent calculation of the sphericity and roundness of the fracturing propping agent.
Further, smoothing the acquired particle image of the fracturing propping agent by adopting Gaussian filtering, wherein the smoothing is used for optimizing the edge of the image and reducing the influence of noise elimination on the edge, and the Gaussian filtering comprises the following steps:
1) firstly, carrying out gray level transformation on an original image of the collected fracturing propping agent, and converting the original image into a gray level image;
2) according to the formula
Generating a Gaussian sequence;
where P (z) represents the gray scale value, μ represents the mean or expected value of z, σ represents the standard deviation of z, and σ represents the square of the standard deviation2Is the variance of z;
and filtering the gray level image by using a Gaussian sequence.
Specifically, gaussian filtering is a linear smooth filtering, that is, gaussian filtering is a process of performing weighted average on the whole image, and values of all pixel points included in the image are obtained by performing weighted average to a certain degree on the pixel points themselves and values of other pixel points in the neighborhood of the pixel points.
Common gaussian noise includes thermal noise (white gaussian noise), shot noise, and the like. The statistical properties of which follow a normal distribution. The formula expression is as follows:
where P (z) represents the gray value, μ represents the mean or expected value of z, σ represents the standard deviation of z, and the square of the standard deviation σ2Is the variance of z.
The one-dimensional zero mean gaussian function is:
where σ determines the width of the gaussian function.
By theoretical analysis, the gaussian function has the following five properties: (1) the two-dimensional Gaussian distribution function has rotational symmetry; according to practical experience analysis, the effective information of each direction of a certain image cannot be predicted in general, so that before filtering, it cannot be determined which direction needs more smoothing operation. This property of the gaussian function means that the gaussian smoothing filter processes the image uniformly, i.e. the filter smoothes the noise in all directions to the same extent. (2) The gaussian function is a single valued function; this property means that each pixel value can be obtained by weighting the gaussian filter to some degree. (3) The fourier transform spectrum of the gaussian function is single-lobed; this property means that the smoothed image is not contaminated by unwanted high frequency signals, while the active part is preserved. (4) The width of the Gaussian filter is represented by a parameter sigma; the larger σ, the wider the band of the gaussian filter, and the better the smoothing degree. By adjusting the smoothness parameter σ, information to be smoothed and effective information can be effectively distinguished. (5) The Gaussian function has separability; the two-dimensional gaussian function may be derived by mathematical derivation of a one-dimensional gaussian function.
Referring to fig. 3, a flow chart of the gaussian filtering process of the present invention is shown. The method adopts the Gaussian smoothing filter to perform Gaussian filtering on the particle image of the fracturing propping agent, and the Gaussian smoothing filter not only has good filtering processing performance on the image in a spatial domain, but also is very effective in a frequency domain. In the actual image processing, an engineer can set effective parameters according to requirements, so that the accuracy of Gaussian filtering is improved.
Further, the image segmentation specifically comprises:
performing image segmentation on the smoothed particle image by adopting an Otsu optimal threshold segmentation method, wherein the method comprises the following steps of:
1) carrying out histogram statistics on the particle image;
2) obtaining the average value of pixel points in the particle image according to the histogram;
3) according to the histogram statistics, carrying out normalization processing on the histogram;
4) obtaining an inter-class variance matrix according to the histogram to obtain a maximum inter-class variance;
5) obtaining a threshold value of the maximum value of the inter-class variance through the maximum inter-class variance, and determining an optimal threshold value;
6) and segmenting the particle image of the fracturing propping agent according to the optimal threshold value.
In particular, image segmentation is a process of optimizing valuable information in particle images of fracturing proppants, by which the images are more easily identified and analyzed. The invention adopts threshold segmentation to carry out image segmentation on the particle image, and the maximum characteristic of the threshold segmentation is to segment the target image and the background image which occupy different gray levels. The project mainly adopts an Otsu optimal threshold segmentation method, and the method specifically comprises the following operations:
the method assumes that the gray level of a certain image is S, and the average gray level of the neighborhood is also S level. And calculating the average gray value of the neighborhood one by one to form a binary group of the gray value and the average gray value of the neighborhood. Assume that the frequency of occurrence of the doublet (i, j) is fijThen the corresponding joint probability density pijComprises the following steps:
and: p is a radical ofijSatisfy the requirement of
Suppose there are two classes C in the image0And C1If the threshold is (s, t), the probabilities of two types of occurrences are:
and two classes C0And C1The corresponding gray level mean vector is:
the total gray level mean vector on the two-dimensional histogram of the image is:
the between-class variance matrix is:
will S(s,t)Trace t ofr(S(s,t)) As measure of dispersion between classes:
similar to the one-dimensional Otsu method, the corresponding segmentation threshold value is the optimal threshold value(s) when the inter-metric degree criterion takes the maximum value*,t*) Then the optimal threshold satisfies:
referring to fig. 4, a flow chart of the threshold segmentation process of the present invention is shown. Firstly, histogram statistics is carried out on an input image, then the average value of pixel points in the image is calculated, then normalization processing is carried out on the histogram according to the histogram statistics, an inter-class variance matrix is obtained according to the histogram, the maximum inter-class variance is obtained, the threshold value capable of obtaining the maximum value of the inter-class variance is the corresponding optimal threshold value, and then the image is segmented according to the optimal threshold value. An image of proppant particles segmented by the optimal thresholding of the invention is shown in figure 5.
Further, the hole filling specifically comprises:
carrying out hole filling processing on the particle image subjected to image segmentation processing through an expansion algorithm, wherein the hole filling processing is used for eliminating unnecessary holes in a fracturing propping agent particle image area, and the hole filling processing comprises the following steps:
recording the positions of neighborhood points in the particle image; detecting neighborhood point pixels; and filling holes in the particle image according to the neighborhood point pixels.
According to the characteristic analysis after image segmentation, the method adopts an expansion algorithm to fill the holes of the proppant particle image after the image segmentation, so that preparation is made for subsequent edge extraction, and the extraction of wrong edges is avoided. The dilation may make the binary image more voluminous. The fullness of the binary image is generally controlled by a preset structural element. The mathematical expression for the expansion operation is:
and (x, y) is the coordinate of a pixel point in the image, A represents a set A, B represents a set B, and i and j are the displacement of the structural element in the image. Fig. 6 shows a schematic diagram of the expansion process of the hole filling process of the present invention, wherein (a) is an original image, (b) is a structural element, (c) is a translation diagram, and (d) is an output image. When the origin of the structural element is translated to the position E in the figure (E represents a gray area), the original image value is kept unchanged; when the origin of the structural element is translated to the F position (in the drawing "+") in (c), the original image is superimposed by 1, and the output image is finally obtained as shown in (d). As can be seen from FIG. 6, the source image is fuller in the expansion process, and it can be clearly seen that the image subjected to the expansion operation is obviously diffused, so that the effects of expanding outwards and filling the cavity can be achieved. When the image is segmented, the selected threshold value cannot be completely suitable for all pixel points in the image, so that holes are generated in the segmented image, and if the holes are not processed, subsequent edge detection and geometric feature extraction are influenced definitely. Referring to fig. 7, a flow chart of the hole filling process of the present invention is shown. The hole filling processing process firstly records the positions of neighborhood points in the image, then detects the pixels of the neighborhood points, and finally fills the holes in the image according to the pixels of the neighborhood points. An image of the fracture proppant particles after hole filling is shown in fig. 8.
Further, the area markers are specifically:
carrying out region marking processing on the particle image by adopting an 8-neighborhood scanning method, wherein the region marking is used for distinguishing and classifying a plurality of particles in the particle image; the area marking process includes the steps of:
1) initializing the region serial numbers of all pixel points of the image to be 0, scanning line by line from the leftmost point in the granular image, taking the scanned point with the first pixel value of 0 as a starting point, and marking the region serial number of the point to be 1;
(2) scanning the next point; if the pixel value of the point is 0, judging the condition of each point in the eight-neighborhood of the point; if the area serial number of one point is not 0, the area serial number of the point is the same as the area serial number of the point; if the sequence numbers of the regions of all the points in the eight neighborhoods of the point are 0, the point belongs to a new region, and the sequence number of the region is the maximum value of the sequence number of the current region plus 1;
(3) after all points in the particle image are scanned, marking different fracturing proppant particles in the fracturing proppant particle image according to the region positions.
Specifically, the area marking plays a very important role in digital image processing, and the area marking is a processing procedure of assigning the same mark to all pixels in the same connected area in an image, and assigning different marks to different connected areas. The invention mainly adopts an 8-neighborhood scanning method, and the specific process is as follows:
if a certain pixel point in the image is P and the coordinate thereof is (m, N), the eight adjacent points of the pixel point P form eight neighborhoods, and N is used8(p) is as shown in FIG. 9.
As shown in fig. 10, a flow chart of region labeling of the neighborhood algorithm is that, firstly, an initialization pixel point is labeled, a scanned first pixel value of 0 is used as a starting point through 8-neighborhood search, and a region serial number of the point is labeled as 1; continuing to scan, and accumulating the same pixel points; and by analogy, after all the points in the image are scanned, the marking is finished. An image of proppant particles marked by connected regions is shown in fig. 11.
Further, edge extraction is carried out on the particle image by Canny operator detection, and the method comprises the following steps:
1) after the gray level of the particle image is transformed, Gaussian filtering is carried out to optimize the edge of the image and reduce the influence of noise on the edge; the mathematical expression satisfying the first criterion is:
the expression satisfying the second criterion is:
when the scale of f is changed, let fw(x) F (x/w), yielding:
2) and (3) solving the gradient value of each pixel point by using a first derivative operator, wherein the formula is as follows:
setting the smoothed image matrix as I [ I, j ], the matrix expression of the partial derivatives in x and y directions is:
3) the method for searching the edge point by adopting the non-maximum suppression method comprises the following operations: the maximum gradient value of the gradient of the adjacent pixel in the gradient direction is reserved and marked as 1, 1 represents a possible edge point, other marks are 0, 0 is a non-edge point, and the expression is as follows:
N[i,j]＝NMS(M[i,j],ξ[i,j]) (6)
wherein, the zero value point of the formula (6) is the non-edge point; searching for real edge points by using the contrast of the edge points and the non-edge points;
4) the method for detecting the edge of the particle image by adopting double-threshold detection processing comprises the following steps: first, two thresholds T are sethAnd TlThen detecting the image processed by the steps 1)2)3) and adding all the image larger than the threshold value ThIs classified as one, will be greater than a threshold value TlRemain as another class; after double-threshold detection, two threshold edge image matrixes T are obtainedh[i,j]And Tl[i,j](ii) a Matrix T with high threshold detectionh[i,j]Based on a matrix T obtained by low threshold detectionl[i,j]To supplement this, the two matrices are combined to extract the edges of the more complete image.
Referring to fig. 12, a flowchart of the edge detection of the present invention is shown. An edge refers to the set of pixels in an image whose surrounding pixels have step-wise and roof-like (inflection points of change in gray value increasing to decreasing) changes in gray value. Edges exist between objects and background, objects and objects, regions and regions. The gray level of the pixel is selected as a reference, and the gradient value change condition of the gray level of the pixel reflects the distribution condition of the edge points. That is, the edge represents an apparent contour feature of the object, while the edge is also an important reference feature for image analysis.
Referring to fig. 13, an edge type diagram of the present invention is shown. According to the change of the image gray scale, the edge of the image can be divided into two types: the first is a step-jump-like edge, which is mainly located where the difference in gray values is large, i.e. jumps directly from one gray level to another gray level, which is much higher than it. The slope at which the step edge is inclined at an angle close to 90 deg., as shown in fig. 13 (a). The second is a roof-like edge which is located at a turn where the gray value increases from decreasing, i.e., in the edge area, the gray value slowly increases to a certain extent and then slowly decreases, as shown in fig. 13 (b).
Referring to fig. 12, edge detection is mainly performed by performing edge connection by mutually complementing matrixes obtained after gaussian filtering, gradient calculation, threshold setting and edge detection with a corresponding set threshold, taking an actual binary image of fracturing proppant particles as an example, the edge of the particles is detected by using a Canny edge operator in Matlab, and a schematic diagram of the detection result is shown in fig. 14.
Further, the geometric feature extraction comprises extracting the area, the perimeter, the maximum inscribed circle radius and the minimum circumscribed circle radius of the particle image of the fracturing proppant;
extracting the area of the particle image by adopting a contrast matrix method to extract the number of pixel points occupied by a target area, namely the boundary of the area and all the pixel points contained in the boundary; the method comprises the following steps:
1) let K be 1, 2.., K, and establish a matrix M K × ones (M, n); k represents the number of particles, m × n is the size of the image;
2) comparing the matrix I corresponding to the image with the established matrix M to generate a brand new matrix NewM obtained by comparing elements at corresponding positions on the matrix I and the matrix M, and if the corresponding positions are consistent, marking the position as 1, and if the corresponding positions are inconsistent, marking the position as 0;
3) forming a matrix NewM processed in the step 2) into a matrix only containing elements 0 and 1, summing the elements of the matrix NewM, wherein the sum is the number of pixels marked as k, and multiplying the sum by a conversion coefficient to obtain the area of the particle image.
In particular, the basic geometric characteristics of a fracturing proppant include a range of parameters related to the morphological characteristics and size, shape, dimensions, etc. of the fracturing proppant. The basic geometric characteristics of the fracturing propping agent are obtained, and a foundation can be laid for measuring the roundness and sphericity of the subsequent fracturing propping agent. The method extracts information such as the area, the perimeter, the maximum inscribed circle radius and the minimum circumscribed circle radius of the particle image of the treated fracturing propping agent. The area of the particle image is extracted by adopting a contrast matrix method, and the area of the target area in the binary image is the number of pixel points occupied by the target area, namely the boundary of the area and the number of all pixel points contained in the boundary.
Further, the extraction of the maximum inscribed circle radius and the minimum circumscribed circle radius of the particle image comprises the following steps:
assuming that the minimum condition circle center of the fracture proppant edge image is (a, b), the objective function is F (a, b), and the formula is as follows:
wherein: rmaxDenotes the radius, R, furthest from (a, b) to the edgeminRepresents the nearest radius from (a, b) to the edge; x is the number ofmax，ymaxIs and RmaxCorresponding coordinates; x is the number ofmin，yminIs and RminCorresponding coordinates;
the objective function of the minimum circumscribed circle is:
the objective function of the maximum inscribed circle is:
when F (a, b) in formula (9) takes the minimum value, then (a, b) is the center coordinates of the minimum circumscribed circle of the minimum condition; when F (a, b) in equation (9) takes the maximum value, then (a, b) is the center coordinates of the maximum inscribed circle of the minimum condition.
Specifically, according to the minimum condition principle in the national standard 'shape and position tolerance-detection regulation', the minimum condition center of the proppant edge image is set to be (a, b), and the objective function is F (a, b) and the expression of F (a, b).
Fig. 16 is a schematic diagram of the image processing result of the present invention. Wherein (a) is a microscopically acquired particle image of the fracturing proppant; (b) is a gaussian filtered particle image of the fracturing proppant; smoothing the particle image to eliminate noise interference; (c) the particle image after threshold segmentation; segmenting the image by a threshold value to extract proppant particles in the particle image; (d) the particle image is processed by filling the hole; removing unnecessary holes in the area of the proppant particles through an expansion operation; (e) the result after area marking, edge extraction, maximum inscribed circle and minimum circumscribed circle extraction; zone markers are used primarily to distinguish between different frac proppant particles; the edge extraction and the extraction of the maximum inscribed circle and the minimum circumscribed circle provide a basis for the calculation of the roundness and sphericity of the subsequent fracturing proppant particles.
Further, the perimeter extraction of the particle image is to extract the boundary length of each particle in the particle image; the method for calculating the perimeter of the particle image by adopting the eight-direction chain code method comprises the following steps:
1) recording the number of continuous perimeter pixel points in the vertical direction: recording the number of the continuous pixel points in the vertical direction as Ny；
2) Recording the number of continuous perimeter pixel points in the horizontal direction: recording the number of continuous pixels in the horizontal direction and recording as Nx；
3) Calculating the total SN of the connected boundary pixels by adopting a formula No＝SN-Nx-NyCalculating the chain code number N of odd codeso；
4) Using a formula
And calculating the eight-connected chain code perimeter of the fracturing proppant image.
The perimeter is an important parameter in image detection, and the perimeter in the particle image refers to the boundary length of each particle. The commonly used four-direction and eight-direction chain code method is used for calculating the perimeter, and is characterized in that a series of connected straight line segments with specific length and direction are used for representing the boundary of an object. The invention adopts the eight-direction chain code method to calculate the perimeter, and the schematic diagram of the eight-direction chain code is shown in fig. 15.
Claims (10)
1. An image processing method for fracturing proppant microscopic magnification, characterized by comprising the following steps:
acquiring image information of the amplified fracturing proppant particles;
carrying out smoothing processing, image segmentation, hole filling and area marking on the image information to obtain a processed particle image;
adopting a Canny operator detection method to carry out edge extraction on the processed particle image;
and obtaining the geometric characteristics of the particle image after edge extraction, wherein the geometric characteristics comprise area, perimeter, maximum inscribed circle radius and minimum circumscribed circle radius.
2. The image processing method for fracturing proppant microscopy magnification as set forth in claim 1, further comprising, prior to said acquiring image information of the magnified fracturing proppant particles:
the fracturing propping agent particles are sucked and conveyed to a vacuum negative pressure sucker on a vacuum negative pressure device through a vacuum negative pressure device, and a plurality of sucker holes are uniformly distributed on the vacuum negative pressure sucker;
and (4) carrying out image amplification on the fracturing propping agent particles on the air negative pressure device through a microscope.
3. The image processing method for fracturing proppant microscopy magnification as set forth in claim 1, characterized in that the smoothing process is specifically:
smoothing the acquired particle image of the fracturing propping agent by adopting Gaussian filtering, wherein the smoothing is used for optimizing the edge of the image and reducing the influence of noise elimination on the edge, and the Gaussian filtering comprises the following steps:
1) firstly, carrying out gray level transformation on an original image of the collected fracturing propping agent, and converting the original image into a gray level image;
2) root of herbaceous plantAccording to the formula
Generating a Gaussian sequence;
where P (z) represents the gray scale value, μ represents the mean or expected value of z, σ represents the standard deviation of z, and σ represents the square of the standard deviation2Is the variance of z;
and filtering the gray level image by using a Gaussian sequence.
4. The image processing method for fracturing proppant microscopy magnification as set forth in claim 1, wherein the image segmentation is specifically:
performing image segmentation on the smoothed particle image by adopting an Otsu optimal threshold segmentation method, wherein the method comprises the following steps of:
1) carrying out histogram statistics on the particle image;
2) obtaining the average value of pixel points in the particle image according to the histogram;
3) according to the histogram statistics, carrying out normalization processing on the histogram;
4) obtaining an inter-class variance matrix according to the histogram to obtain a maximum inter-class variance;
5) obtaining a threshold value of the maximum value of the inter-class variance through the maximum inter-class variance, and determining an optimal threshold value;
6) and segmenting the particle image of the fracturing propping agent according to the optimal threshold value.
5. The image processing method for fracturing proppant microscopy magnification as set forth in claim 1, wherein the hole filling is specifically:
carrying out hole filling processing on the particle image subjected to image segmentation processing through an expansion algorithm, wherein the hole filling processing is used for eliminating unnecessary holes in a fracturing propping agent particle image area, and the hole filling processing comprises the following steps:
recording the positions of neighborhood points in the particle image; detecting neighborhood point pixels; and filling holes in the particle image according to the neighborhood point pixels.
6. The image processing method for fracturing proppant microscopy magnification as set forth in claim 1, characterized in that the zone markers are in particular:
carrying out region marking processing on the particle image by adopting an 8-neighborhood scanning method, wherein the region marking is used for distinguishing and classifying a plurality of particles in the particle image; the area marking process includes the steps of:
1) initializing the region serial numbers of all pixel points of the image to be 0, scanning line by line from the leftmost point in the granular image, taking the scanned point with the first pixel value of 0 as a starting point, and marking the region serial number of the point to be 1;
(2) scanning the next point; if the pixel value of the point is 0, judging the condition of each point in the eight-neighborhood of the point; if the area serial number of one point is not 0, the area serial number of the point is the same as the area serial number of the point; if the sequence numbers of the regions of all the points in the eight neighborhoods of the point are 0, the point belongs to a new region, and the sequence number of the region is the maximum value of the sequence number of the current region plus 1;
(3) after all points in the particle image are scanned, marking different fracturing proppant particles in the fracturing proppant particle image according to the region positions.
7. The image processing method for fracturing proppant microscopy magnification according to claim 1, characterized by using Canny operator detection for edge extraction of particle images, comprising the steps of:
1) after the gray level of the particle image is transformed, Gaussian filtering is carried out to optimize the edge of the image and reduce the influence of noise on the edge; the mathematical expression satisfying the first criterion is:
the expression satisfying the second criterion is:
when the scale of f is changed, let fw(x) F (x/w), yielding:
2) and (3) solving the gradient value of each pixel point by using a first derivative operator, wherein the formula is as follows:
setting the smoothed image matrix as I [ I, j ], the matrix expression of the partial derivatives in x and y directions is:
3) the method for searching the edge point by adopting the non-maximum suppression method comprises the following operations: the maximum gradient value of the gradient of the adjacent pixel in the gradient direction is reserved and marked as 1, 1 represents a possible edge point, other marks are 0, 0 is a non-edge point, and the expression is as follows:
N[i,j]＝NMS(M[i,j],ξ[i,j]) (6)
wherein, the zero value point of the formula (6) is the non-edge point; searching for real edge points by using the contrast of the edge points and the non-edge points;
4) the method for detecting the edge of the particle image by adopting double-threshold detection processing comprises the following steps: first, two thresholds T are sethAnd TlThen detecting the image processed by the steps 1)2)3) and adding all the image larger than the threshold value ThIs classified as one, will be greater than a threshold value TlRemain as another class; after double-threshold detection, two threshold edge image matrixes T are obtainedh[i,j]And Tl[i,j](ii) a Matrix T with high threshold detectionh[i,j]Based on a matrix T obtained by low threshold detectionl[i,j]To supplement this, the two matrices are combined to extract the edges of the more complete image.
8. The image processing method for fracturing proppant microscopically expanding as set forth in claim 1, wherein the geometric feature extraction includes extracting an area, a perimeter, a maximum inscribed circle radius and a minimum circumscribed circle radius of a particle image of the fracturing proppant;
extracting the area of the particle image by adopting a contrast matrix method to extract the number of pixel points occupied by a target area, namely the boundary of the area and all the pixel points contained in the boundary; the method comprises the following steps:
1) let K be 1, 2.., K, and establish a matrix M K × ones (M, n); k represents the number of particles, m × n is the size of the image;
2) comparing the matrix I corresponding to the image with the established matrix M to generate a brand new matrix NewM obtained by comparing elements at corresponding positions on the matrix I and the matrix M, and if the corresponding positions are consistent, marking the position as 1, and if the corresponding positions are inconsistent, marking the position as 0;
3) forming a matrix NewM processed in the step 2) into a matrix only containing elements 0 and 1, summing the elements of the matrix NewM, wherein the sum is the number of pixels marked as k, and multiplying the sum by a conversion coefficient to obtain the area of the particle image.
9. The image processing method for fracturing proppant microscopy magnification as set forth in claim 7, characterized in that the maximum inscribed circle radius and minimum circumscribed circle radius extraction of the particle image comprises the steps of:
assuming that the minimum condition circle center of the fracture proppant edge image is (a, b), the objective function is F (a, b), and the formula is as follows:
wherein: rmaxDenotes the radius, R, furthest from (a, b) to the edgeminRepresents the nearest radius from (a, b) to the edge; x is the number ofmax，ymaxIs and RmaxCorresponding coordinates; x is the number ofmin，yminIs and RminCorresponding coordinates;
the objective function of the minimum circumscribed circle is:
the objective function of the maximum inscribed circle is:
when F (a, b) in formula (9) takes the minimum value, then (a, b) is the center coordinates of the minimum circumscribed circle of the minimum condition; when F (a, b) in equation (9) takes the maximum value, then (a, b) is the center coordinates of the maximum inscribed circle of the minimum condition.
10. The image processing method for fracturing proppant microscopy magnification as set forth in claim 1, characterized in that the perimeter extraction of the particle image is an extraction of the boundary length of each particle in the particle image; the method for calculating the perimeter of the particle image by adopting the eight-direction chain code method comprises the following steps:
1) recording the number of continuous perimeter pixel points in the vertical direction: recording the number of the continuous pixel points in the vertical direction as Ny；
2) Recording the number of continuous perimeter pixel points in the horizontal direction: recording the number of continuous pixels in the horizontal direction and recording as Nx；
3) Calculating the total SN of the connected boundary pixels by adopting a formula No＝SN-Nx-NyCalculating the chain code number N of odd codeso；
4) Using a formula
And calculating the eight-connected chain code perimeter of the fracturing proppant image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110035853.9A CN112950659B (en) | 2021-01-12 | 2021-01-12 | Image processing method for microscopic amplification of fracturing propping agent |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110035853.9A CN112950659B (en) | 2021-01-12 | 2021-01-12 | Image processing method for microscopic amplification of fracturing propping agent |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112950659A true CN112950659A (en) | 2021-06-11 |
| CN112950659B CN112950659B (en) | 2024-06-28 |
Family
ID=76235263
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110035853.9A Active CN112950659B (en) | 2021-01-12 | 2021-01-12 | Image processing method for microscopic amplification of fracturing propping agent |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112950659B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113450292A (en) * | 2021-06-17 | 2021-09-28 | 重庆理工大学 | High-precision visual positioning method for PCBA parts |
| CN113706566A (en) * | 2021-09-01 | 2021-11-26 | 四川中烟工业有限责任公司 | Perfuming spray performance detection method based on edge detection |
| CN118379301A (en) * | 2024-06-27 | 2024-07-23 | 大石桥市鹏飞高科耐材有限公司 | Hydration degree appearance detection method for regenerated granule |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102749437A (en) * | 2012-07-24 | 2012-10-24 | 东北石油大学 | Method for researching fractal particle size of multistage fracturing propping agent and fracture characteristic parameter of rock mass |
| CN104819915A (en) * | 2015-01-14 | 2015-08-05 | 中国石油天然气股份有限公司 | Method for testing roundness and sphericity of fracturing propping agent |
| CN106442230A (en) * | 2016-10-11 | 2017-02-22 | 中国石油大学(华东) | Fracturing propping agent roundness and sphericity detecting method based on image processing technology |
-
2021
- 2021-01-12 CN CN202110035853.9A patent/CN112950659B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102749437A (en) * | 2012-07-24 | 2012-10-24 | 东北石油大学 | Method for researching fractal particle size of multistage fracturing propping agent and fracture characteristic parameter of rock mass |
| CN104819915A (en) * | 2015-01-14 | 2015-08-05 | 中国石油天然气股份有限公司 | Method for testing roundness and sphericity of fracturing propping agent |
| CN106442230A (en) * | 2016-10-11 | 2017-02-22 | 中国石油大学(华东) | Fracturing propping agent roundness and sphericity detecting method based on image processing technology |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113450292A (en) * | 2021-06-17 | 2021-09-28 | 重庆理工大学 | High-precision visual positioning method for PCBA parts |
| CN113706566A (en) * | 2021-09-01 | 2021-11-26 | 四川中烟工业有限责任公司 | Perfuming spray performance detection method based on edge detection |
| CN113706566B (en) * | 2021-09-01 | 2024-03-12 | 四川中烟工业有限责任公司 | Edge detection-based perfuming and spraying performance detection method |
| CN118379301A (en) * | 2024-06-27 | 2024-07-23 | 大石桥市鹏飞高科耐材有限公司 | Hydration degree appearance detection method for regenerated granule |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112950659B (en) | 2024-06-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112950659A (en) | Image processing method for fracturing propping agent microscopic amplification | |
| US8077958B2 (en) | Computer-aided pathological diagnosis system | |
| CN109255757B (en) | Method for segmenting fruit stem region of grape bunch naturally placed by machine vision | |
| CN107292310B (en) | Visual positioning and automatic reading method for circular pointer type dial plate | |
| CN109409355B (en) | Novel transformer nameplate identification method and device | |
| CN110120042B (en) | Crop image pest and disease damage area extraction method based on SLIC super-pixel and automatic threshold segmentation | |
| CN108376403B (en) | Grid colony image segmentation method based on Hough circle transformation | |
| CN109345554A (en) | A kind of adhesion mushroom vision in-situ measuring method based on RGB-D camera | |
| CN114067114B (en) | Area nuclear segmentation counting method based on algae cell morphology | |
| CN112132153B (en) | Tomato fruit identification method and system based on clustering and morphological processing | |
| CN110287898A (en) | A kind of optical satellite remote sensing image cloud detection method of optic | |
| CN108052886A (en) | A kind of puccinia striiformis uredospore programming count method of counting | |
| CN113256580A (en) | Automatic identification method for target colony characteristics | |
| CN107274413A (en) | A kind of steel microstructure's crystal boundary extraction method | |
| CN109035227A (en) | The system that lung tumors detection and diagnosis is carried out to CT image | |
| CN112861654B (en) | Machine vision-based famous tea picking point position information acquisition method | |
| CN115294527B (en) | Subway tunnel damage detection method based on computer vision | |
| CN117853722A (en) | Steel metallographic structure segmentation method integrating superpixel information | |
| CN107516315B (en) | Tunneling machine slag tapping monitoring method based on machine vision | |
| CN116523898A (en) | Tobacco phenotype character extraction method based on three-dimensional point cloud | |
| CN112884746A (en) | Character defect intelligent detection algorithm based on edge shape matching | |
| CN107507192B (en) | Portable device for in-vivo measurement of rice sword leaf included angle and in-vivo measurement method | |
| CN112734743B (en) | Automatic measuring method and device for fetal head circumference | |
| CN108269264B (en) | Denoising and fractal method of bean kernel image | |
| CN114152211B (en) | Microscopic image processing-based roundness measurement method for fracturing propping agent |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |