Method for measuring dissociation degree of mineral under mirror based on improved Gaussian pyramid
By improving the objective-subjective dissociation measurement method of Gaussian pyramid, the particle size of the mineral objective is automatically measured, which solves the problem of time-consuming and inaccurate traditional measurements and improves the accuracy and reliability of measurements.
Patent Information
- Application Number
- CN202510044775.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-10
- Publication Date
- 2025-05-09
AI Technical Summary
Traditional mineral dissociation measurements rely on manual observation, which is time-consuming and prone to subjective errors, resulting in reduced accuracy and reliability.
Using the improved mineral objective-based dissociation measurement method of Gaussian pyramid, automated measurement is achieved through Gaussian pyramid segmentation, segmentation and edge extraction of target minerals and associated minerals, application of edge coordinates and calculation of dissociation.
Improve the accuracy and reliability of understanding the dissociation measurement, reduce artificial errors, shorten measurement time, and enhance the advancement of mineral processing technology and the utilization rate of mineral resources.
Smart Images
Figure CN119959113A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of mineral processing, and in particular to a method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid. Background Art
[0002] In mineral processing, mineral dissociation degree is a key parameter that directly affects the separation efficiency and the quality of the final product. Dissociation degree refers to whether the mineral particles are completely separated from their associated gangue minerals. A high dissociation degree means that the mineral particles are well separated from the gangue, which plays an important role in the selection of related processing technology in the subsequent mineral processing process.
[0003] Traditionally, mineral dissociation degree is measured by manual observation, which is time-consuming and prone to subjective errors, thus reducing accuracy and reliability.
[0004] Therefore, it is necessary to provide a method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid to solve the above technical problems. Summary of the invention
[0005] The invention provides a method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid, which solves the problem that the degree of dissociation of minerals is traditionally measured by manual observation, which is time-consuming and prone to subjective errors, thereby reducing the accuracy and reliability.
[0006] To solve the above technical problems, the present invention provides a method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid, comprising the following steps: S1: acquiring an image; S2: segmenting the improved Gaussian pyramid; S3: segmenting the target mineral and the associated minerals respectively, and selecting the target mineral particles in the target mineral image at the same time; S4: binarizing the target mineral image; S5: extracting the edge coordinates of the target mineral; S6: applying the edge coordinates of the target mineral to the associated mineral image to determine whether there are associated minerals around the edge coordinates of the target mineral; S7: counting the number of monomer dissociated and associated mineral particles; S8: calculating the degree of dissociation.
[0007] The S2 comprises the following steps: S21: Multi-scale input construction: The input image is downsampled by Gaussian pyramid to obtain multi-scale images, and the resolution of each layer is gradually reduced. The image under the mineral microscope in this paper is downsampled into three layers, and the resolution of each layer is reduced to represent different image details; I: 1 , I 2 .....I l} S22: VEA encoder network is applied to each layer: Use the variational autoencoder encoder network for each layer of the Gaussian pyramid to get Z l The mean and variance of z l =u l +σ l ·∈,∈~N(0,I) The encoder learns the mean and variance of the latent variables. This step establishes the connection between each layer of the pyramid and the latent space. S23: Fusion in latent space: Put different levels into the latent space and fuse them through upsampling and concatenation to form a richer latent space; z fused =z 1 +Upsample(z l -1)+…+Upsample l (z 0 ) S24: Decoder reconstructs the image: The decoding network of the variational autoencoder is used to reconstruct the image from the fused latent variables. The decoder generates images of different scales based on the latent space variables, and gradually applies them to the low-resolution images to reconstruct the full-resolution image.
[0008] Preferably, the S3 includes the following steps: S31: analyzing the grayscale image obtained after segmentation using the improved Gaussian pyramid, there are differences in the images obtained from different minerals or under different shooting conditions, selecting the grayscale threshold of the corresponding mineral, and segmenting the target mineral and the gangue associated mineral images respectively; S32: assigning different color masks to the target mineral and the associated mineral so that the target mineral and the associated mineral are both a single color, and at the same time, box-selecting and calculating all particles in the target mineral, and counting the number of boxes used.
[0009] Preferably, in S4, the obtained target mineral image and associated mineral image are binarized to obtain image I; the formula is:
[0010] Preferably, in said S5, the edges of the target mineral particles in the target mineral image are extracted by using the Sobel edge detection algorithm on the target mineral image, and the specific steps are as follows: S51: Input image: Let the input image be I(x,y), whose size is M×N (the number of rows is M and the number of columns is N); S52: Calculate dark channel image: D(x,y)=min(Ir(x,y),Ig(x,y),Ib(x,y)) Where Ir(x, y), Ig(x, y), and Ib(x, y) represent the pixel values of the image in the red, green, and blue channels respectively; for grayscale images, D(x, y) = I(x, y) S53: Sobel operator: Use the Sobel operator to calculate the gradient of the image. The Sobel operator has two directions: horizontal and vertical. The horizontal Sobel operator is: Sobel operator in the vertical direction: Perform a convolution operation on the dark channel image to calculate the gradient: Horizontal gradient Gx(x, y); Horizontal gradient Gy(x, y); S54: Calculate edge strength: The edge strength of each pixel is calculated using the horizontal and vertical gradients: S55: Output the image.
[0011] Preferably, in the mask image of the target mineral under the microscope, the edge points of all particles are extracted respectively, Cci is the edge point set of all mineral particle contours, and ci is the edge point set of one particle, as shown in the figure below; Cci={ci,j}.
[0012] Preferably, in S7, the edge point of each particle in the target mineral image is used in the associated mineral image, and all points are traversed within a range of 6 pixels with a radius from the center point (x, y) with the point as the center of the circle. If the pixel value in the area is not zero, it means that the mineral particle has not dissociated. At the same time, the point is located to the particle of the mineral framed in step 2, and all points are traversed in a loop until the dissociation and associated number of the target mineral particles and other associated mineral particles are counted, which specifically includes the following steps: S71: traverse each edge point: For each edge point (xi, yi): S72: Define the search area: traverse all points in a circular area with (xi, yi) as the center and a radius of 6 pixels. Let the search area be Ri: Ri={(x,y)|(x-xi)2+(y-yi)2<6 2} S73: Check the pixel value in the area: For each point (x, y) in the search area Ri, check the pixel value I (x, y): S74: Determine the dissociation state: If there exists (x, y)∈Ri such that I(x, y)≠0, then the particle is not dissociated: S75: Positioning to the mineral particle: Positioning the point (xi, yi) to the mineral particle framed in step 2.
[0013] Preferably, when calculating the dissociation degree, the monomer dissociation degree J of the target mineral is calculated by the following formula using the data obtained above: Where: L is the number of monomer-dissociated mineral particles; M is the total number of mineral particles.
[0014] Preferably, in the process of extracting the edge coordinates of the target mineral in step S5 of the method for measuring the degree of dissociation of minerals under a microscope based on the improved Gaussian pyramid, a computer needs to be used to display the image, and the computer includes: a computer display screen; A movable groove, the movable groove is opened on one side of the surface of the computer display screen, and a limit groove is opened on the other side of the surface of the computer display screen, a reciprocating screw rod is rotatably installed inside the movable groove, an adjusting block is fixedly connected to the top of the reciprocating screw rod, and a movable device is threadedly engaged on the outer surface of the reciprocating screw rod, and the movable device includes a threaded block, a connecting plate, a sliding groove, a sliding rod and a limit slider; An observation device, the observation device is slidably mounted inside the moving device, and the observation device includes a moving block, a connecting plate and a magnifying glass; A limiting groove is provided on the other side of the surface of the computer display screen, and a limiting rod is fixedly installed inside the limiting groove.
[0015] Preferably, the threaded block is threadably engaged with the outer surface of the reciprocating screw, the connecting plate is fixedly installed on the front side of the threaded block, the limit slider is fixedly installed on one side of the surface of the connecting plate, the slide groove is opened in the middle of the front side of the connecting plate, the slide rod is fixedly installed in the middle of the inside of the slide groove, the moving block is slidably installed on the outer surface of the slide rod, the connecting plate is fixedly connected to one side of the surface of the moving block, and the magnifying glass is fixedly installed on the surface of the connecting plate.
[0016] Compared with the related art, the method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid provided by the present invention has the following beneficial effects: The invention provides a method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid. The method comprises the following steps: segmenting a mineral image under a microscope by using the improved Gaussian pyramid; segmenting a target mineral and an associated mineral image separately, setting the pixel values of the remaining points to zero and giving different color masks; converting the target mineral image into a color image, and extracting the edge of a target mineral particle; applying the edge point of the target mineral to the associated mineral image, judging whether there is a point with a non-zero pixel value around a position corresponding to the edge point of the target mineral, and judging that the mineral particle corresponding to the edge point has been dissociated if the pixel value is zero, otherwise it is associated; traversing all points, finding all dissociated mineral particles and associated mineral particles, and calculating the degree of dissociation of the target mineral; the invention adopts the improved Gaussian pyramid to measure the mineral image under a microscope, and the measuring method is simple. Compared with a clustering algorithm and a manual measurement method, the method can more quickly and accurately measure the particle size of the mineral under a microscope. The method improves the accuracy and reliability of the degree of dissociation measurement, reduces human errors, shortens the measurement time, and is helpful to promote the progress of mineral processing technology and improve the utilization rate of mineral resources. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] Figure 1 A schematic flow chart of a first embodiment of a method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid provided by the present invention; Figure 2 This is a schematic diagram of pyrite and magnetite images; Figure 3 This is a schematic diagram of the segmentation results of the improved Gaussian pyramid algorithm; Figure 4 Image diagram after adding mask to pyrite; Figure 5 Schematic diagram of the image after adding a mask for magnetite; Figure 6 A schematic diagram of marking and counting pyrite particles; Figure 7 Schematic diagram of the edge of the target mineral; Figure 8 A schematic structural diagram of a second embodiment of a method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid provided by the present invention; Fig. 9 for Figure 8 An enlarged schematic diagram of point A is shown; Fig.10 for Figure 8 The enlarged schematic diagram of point B is shown; Fig.11 for Fig. 9 Schematic diagram of the observation device structure shown.
[0018] Numbers in the figure: 1. computer display screen, 2. support base, 3. moving device, 31. threaded block, 32. connecting plate, 33. slide groove, 34. slide rod, 35. limit slider, 4. moving groove, 5. reciprocating screw rod, 6. adjustment block, 7. observation device, 71. moving block, 72. connecting plate, 73. magnifying glass, 8. limit groove, 9. limit rod. DETAILED DESCRIPTION The present invention will be further described below in conjunction with the accompanying drawings and implementation modes.
[0019] First embodiment Please refer to Figure 1 , Figure 2 , Figure 3 , Figure 4 , Figure 5 , Figure 6 and Figure 7 ,in, Figure 1 A schematic flow chart of a first embodiment of a method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid provided by the present invention; Figure 2 This is a schematic diagram of pyrite and magnetite images; Figure 3 This is a schematic diagram of the segmentation results of the improved Gaussian pyramid algorithm; Figure 4 Image diagram after adding mask to pyrite; Figure 5 Schematic diagram of the image after adding a mask for magnetite; Figure 6 A schematic diagram of marking and counting pyrite particles; Figure 7 The diagram is a schematic diagram of the edge of the target mineral. A method for measuring the dissociation degree of minerals under a microscope based on an improved Gaussian pyramid includes the following steps: S1: acquiring an image; S2: segmenting with an improved Gaussian pyramid; S3: segmenting the target mineral and the associated minerals respectively, and selecting the target mineral particles in the target mineral image; S4: binarizing the target mineral image; S5: extracting the edge coordinates of the target mineral; S6: applying the edge coordinates of the target mineral to the associated mineral image to determine whether there are associated minerals around the edge coordinates of the target mineral; S7: counting the number of monomer dissociated and associated mineral particles; S8: calculating the dissociation degree.
[0020] The S2 comprises the following steps: S21: Multi-scale input construction: The input image is downsampled by Gaussian pyramid to obtain multi-scale images, and the resolution of each layer is gradually reduced. The image under the mineral microscope in this paper is downsampled into three layers, and the resolution of each layer is reduced to represent different image details; I: 1 , I 2 .....I l} S22: VEA encoder network is applied to each layer: Use the variational autoencoder encoder network for each layer of the Gaussian pyramid to get Z l The mean and variance of z l =u l +σ l ∈, ∈~N(0, I) are learned by the encoder and represent the mean and variance of the latent variable. This step establishes the connection between each layer of the pyramid and the latent space; S23: Fusion in latent space: Put different levels into the latent space and fuse them through upsampling and concatenation to form a richer latent space; z fused =z l +Upsample(z l -1)+…+Upsample l (z 0 ) S24: Decoder reconstructs the image: The decoding network of the variational autoencoder is used to reconstruct the image from the fused latent variables. The decoder generates images of different scales based on the latent space variables, and gradually applies them to the low-resolution images to reconstruct the full-resolution image.
[0021] The S3 includes the following steps: S31: analyzing the grayscale image obtained after segmentation using the improved Gaussian pyramid. There are differences in the images obtained from different minerals or under different shooting conditions. The grayscale threshold of the corresponding mineral is selected to segment the target mineral and the gangue associated mineral images respectively; S32: assigning different color masks to the target mineral and the associated mineral so that the target mineral and the associated mineral are both a single color. At the same time, all particles in the target mineral are framed and calculated, and the number of frames used is counted.
[0022] Preferably, in S4, the obtained target mineral image and associated mineral image are binarized to obtain image I; the formula is:
[0023] In S5, the target mineral particle edges in the target mineral image are extracted: by using the Sobel edge detection algorithm on the target mineral image, the specific steps are as follows: S51: Input image: Assume that the input image is I(x, y), and its size is M×N (the number of rows is M and the number of columns is N); S52: Calculate dark channel image: D(x,y)=min(Ir(x,y),Ig(x,y),Ib(x,y)) Where Ir(x, y), Ig(x, y), and Ib(x, y) represent the pixel values of the image in the red, green, and blue channels respectively; for grayscale images, D(x, y) = I(x, y). S53: Sobel operator: Use the Sobel operator to calculate the gradient of the image. The Sobel operator has two directions: horizontal and vertical. The horizontal Sobel operator is: Sobel operator in the vertical direction: Perform a convolution operation on the dark channel image to calculate the gradient: Horizontal gradient Gx(x, y); Horizontal gradient Gy(x, y); S54: Calculate edge strength: The edge strength of each pixel is calculated using the horizontal and vertical gradients: S55: Output the image.
[0024] In the target mineral microscopic mask image, the edge points of all particles are extracted respectively, Cci is the edge point set of all mineral particle contours, and ci is the edge point set of one particle, as shown in the figure below; Cci={ci,j}.
[0025] In the step S7, the edge point of each particle in the target mineral image is used in the associated mineral image, and the point is used as the center of the circle, and all points are traversed within a range of 6 pixels with a radius from the center point (x, y). If the pixel value in the area is not zero, it means that the mineral particle has not dissociated. At the same time, the point is located to the particle of the mineral framed in step 2, and all points are traversed in a loop until the dissociation and associated number of the target mineral particle and other associated mineral particles are counted, which specifically includes the following steps: S71: traverse each edge point: For each edge point (xi, yi): S72: Define the search area: traverse all points in a circular area with (xi, yi) as the center and a radius of 6 pixels. Let the search area be Ri: Ri={(x,y)|(x-xi)2+(y-yi)2≤62} S73: Check the pixel value in the area: For each point (x, y) in the search area Ri, check the pixel value I (x, y): S74: Determine the dissociation state: If there exists (x, y)∈Ri such that I(x, y)≠0, then the particle is not dissociated: S75: Positioning to the mineral particle: Positioning the point (xi, yi) to the mineral particle framed in step 2.
[0026] Preferably, when calculating the dissociation degree, the monomer dissociation degree J of the target mineral is calculated by the following formula using the data obtained above: Where: L is the number of monomer-dissociated mineral particles; M is the total number of mineral particles.
[0027] With the help of pictures and text, we first select an image of pyrite under a microscope as shown in the figure. The yellow part is pyrite and the gray part is magnetite. Figure 2 shown.
[0028] The image is segmented using the improved Gaussian pyramid. The segmentation results are as follows: Figure 3 shown.
[0029] Assign different pseudo colors to magnetite and pyrite, such as Figure 4 shown.
[0030] Select and mark the target particles in the target mineral image, such as Figure 6 As shown, a total of 516 frames are used.
[0031] Extract mineral edges such as Figure 7 shown.
[0032] All points in the extracted target mineral edge image are extracted, and these points are used to compare with the pixel points in the associated mineral image. When the pixel value of the point in the corresponding associated mineral image is not zero within the area with a radius of 6 with a certain point as the center of the circle in the target mineral edge image, it means that the point has not dissociated from the associated mineral. At the same time, return to the image selected in step 3, locate the particle represented by the position of the center of the circle, and determine that the particle has not dissociated. Loop through all points until all extracted edge points are calculated.
[0033] By calculation, there are 516 pyrite particles, of which 32 are dissociated from the mineral and 484 are associated. According to the dissociation degree calculation formula, the dissociation degree of pyrite in this image is 6%.
[0034] The working principle of the method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid provided by the present invention is as follows: When working, the improved Gaussian pyramid is first used to segment the mineral microscopic image; the target mineral and associated mineral images are separately segmented, and the pixel values of the remaining points are set to zero and given different color masks; the target mineral image is converted into a color image, and the edges of the target mineral particles are extracted; the edge points of the target mineral are applied to the associated mineral image to determine whether there are points with non-zero pixel values around the positions corresponding to the target mineral edge points. If they are zero, it is determined that the mineral particles corresponding to the edge points have been dissociated, otherwise they are associated; all points are traversed to find all dissociated mineral particles and associated mineral particles, and the dissociation degree of the target mineral is calculated.
[0035] Compared with the related art, the method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid provided by the present invention has the following beneficial effects: The invention provides a method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid. The method comprises the following steps: segmenting a mineral image under a microscope by using the improved Gaussian pyramid; segmenting a target mineral and an associated mineral image separately, setting the pixel values of the remaining points to zero and giving different color masks; converting the target mineral image into a color image, and extracting the edge of a target mineral particle; applying the edge point of the target mineral to the associated mineral image, judging whether there is a point with a non-zero pixel value around a position corresponding to the edge point of the target mineral, and judging that the mineral particle corresponding to the edge point has been dissociated if the pixel value is zero, otherwise it is associated; traversing all points, finding all dissociated mineral particles and associated mineral particles, and calculating the degree of dissociation of the target mineral; the invention adopts the improved Gaussian pyramid to measure the mineral image under a microscope, and the measuring method is simple. Compared with a clustering algorithm and a manual measurement method, the method can more quickly and accurately measure the particle size of the mineral under a microscope. The method improves the accuracy and reliability of the degree of dissociation measurement, reduces human errors, shortens the measurement time, and is helpful to promote the progress of mineral processing technology and improve the utilization rate of mineral resources.
[0036] Second embodiment Please refer to Figure 8 , Fig. 9 , Fig.10 and Fig.11 Based on the method for measuring the degree of dissociation of a mineral under a microscope based on an improved Gaussian pyramid provided in the first embodiment of the present application, the second embodiment of the present application proposes another method for measuring the degree of dissociation of a mineral under a microscope based on an improved Gaussian pyramid. The second embodiment is only a preferred embodiment of the first embodiment, and the implementation of the second embodiment will not affect the independent implementation of the first embodiment.
[0037] Specifically, the difference of the method for measuring the degree of dissociation of a mineral under a microscope based on an improved Gaussian pyramid provided in the second embodiment of the present application is that a computer needs to be used to display an image in the process of extracting the edge coordinates of the target mineral in step S5 of the method for measuring the degree of dissociation of a mineral under a microscope based on an improved Gaussian pyramid, and the computer includes: a computer display screen 1; a moving groove 4, the moving groove 4 is opened on one side of the surface of the computer display screen 1, and a limiting groove 8 is opened on the other side of the surface of the computer display screen 1, a reciprocating screw 5 is rotatably installed inside the moving groove 4, and an adjusting block 6 is fixedly connected to the top of the reciprocating screw 5, and a moving device 3 is threadedly engaged on the outer surface of the reciprocating screw 5, and the moving device 3 includes a threaded block 31, a connecting plate 32, a sliding groove 33, a sliding rod 34 and a limiting slider 35; An observation device 7, which is slidably mounted inside the moving device 3, and includes a moving block 71, a connecting plate 72 and a magnifying glass 73; A limiting groove 8 is provided on the other side of the surface of the computer display screen 1 , and a limiting rod 9 is fixedly installed inside the limiting groove 8 .
[0038] The threaded block 31 is threadably engaged with the outer surface of the reciprocating screw rod 5, the connecting plate 32 is fixedly installed on the front side of the threaded block 31, the limiting slider 35 is fixedly installed on one side of the surface of the connecting plate 32, the slide groove 33 is opened in the middle of the front side of the connecting plate 32, the slide rod 34 is fixedly installed in the middle of the inside of the slide groove 33, the moving block 71 is slidably installed on the outer surface of the slide rod 34, the connecting plate 72 is fixedly connected to one side of the surface of the moving block 71, and the magnifying glass 73 is fixedly installed on the surface of the connecting plate 72.
[0039] The working principle of the method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid provided by the present invention is as follows: During operation, the reciprocating screw 5 is first driven to rotate by rotating the adjustment block 6, and the reciprocating screw 5 rotates the threaded engagement moving device 3 to adjust the height of the magnifying glass 73, and then the lateral position of the magnifying glass 73 is adjusted by the moving block 71 and the sliding rod 34, and the magnifying glass 73 is adjusted to move to any position on the surface of the computer display screen 1.
[0040] Compared with the related art, the method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid provided by the present invention has the following beneficial effects: The present invention provides a method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid. The method comprises the following steps: rotating an adjusting block 6 to drive a reciprocating screw 5 to rotate, and rotating the reciprocating screw 5 to engage a moving device 3 to adjust the height of a magnifying glass 73. Then, a moving block 71 and a sliding rod 34 are used to adjust the lateral position of the magnifying glass 73, so that an operator can conveniently adjust the movement of the magnifying glass 73 to any position on the surface of a computer display screen 1. The operator extracts all points in an extracted target mineral edge image, and when using these points for a comparison step with pixel points in an associated mineral image, the operator can observe subtle parts of the image by moving the magnifying glass without enlarging the image, thereby avoiding blurring of the enlarged image, improving the accuracy of observation, and improving the practicality of the device.
[0041] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.
Claims
1. A method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid, characterized in that: The following steps are involved: S1: Acquire an image; S2: Segment with an improved Gaussian pyramid; S3: Segment the target mineral and associated minerals separately, and select the target mineral particles in the target mineral image; S4: Binarize the target mineral image; S5: Extract the target mineral edge coordinates; S6: Apply the target mineral edge coordinates to the associated mineral image to determine whether there are associated minerals around the target mineral edge coordinates; S7: Count the number of monomer dissociated and associated mineral particles; S8: Calculate the degree of dissociation.
2. A method for measuring the degree of dissociation under a mineral microscope based on an improved Gaussian pyramid according to claim 1, characterized in that: The S2 comprises the following steps: S21: Multi-scale input construction: The input image is downsampled by Gaussian pyramid to obtain multi-scale images, and the resolution of each layer is gradually reduced. The image under the mineral microscope in this paper is downsampled into three layers, and the resolution of each layer is reduced to represent different image details; I:{I1,I2......I l } S22: VEA encoder network is applied to each layer: Use the variational autoencoder encoder network for each layer of the Gaussian pyramid to get z l The mean and variance of WITH l =u l +σ l ·∈,∈~N(0,I) The encoder learns the mean and variance of the latent variables. This step establishes the connection between each layer of the pyramid and the latent space. S23: Fusion in latent space: Put different levels into the latent space and fuse them through upsampling and concatenation to form a richer latent space; from fused =from l +Upsample(from l -1)+…+Upsample l (from 0) S24: Decoder reconstructs the image: The decoding network of the variational autoencoder is used to reconstruct the image from the fused latent variables. The decoder generates images of different scales based on the latent space variables, and gradually applies them to the low-resolution images to reconstruct the full-resolution image.
3. A method for measuring the degree of dissociation under a mineral microscope based on an improved Gaussian pyramid according to claim 1, characterized in that: The S3 The method comprises the following steps: S31: analyzing the grayscale image obtained after segmentation using the improved Gaussian pyramid, and there are differences in the images obtained from different minerals or under different shooting conditions, and selecting the grayscale threshold of the corresponding mineral to segment the target mineral and the gangue associated mineral images respectively; S32: assigning different color masks to the target mineral and the associated mineral so that the target mineral and the associated mineral are both in a single color, and at the same time, selecting and calculating all the particles in the target mineral, and counting the number of frames used.
4. A method for measuring the degree of dissociation under a mineral microscope based on an improved Gaussian pyramid according to claim 1, characterized in that: In S4, the obtained target mineral image and associated mineral image are binarized to obtain image I; the formula is:
5. A method for measuring the degree of dissociation under a mineral microscope based on an improved Gaussian pyramid according to claim 1, characterized in that: In S5, the target mineral particle edges in the target mineral image are extracted: by using the Sobel edge detection algorithm on the target mineral image, the specific steps are as follows: S51: Input image: Assume that the input image is I(x, y), and its size is M×N (the number of rows is M and the number of columns is N); S52: Calculate dark channel image: D(x,y)=min(Ir(x,y),Ig(x,y),Ib(x,y)) Where Ir(x, y), Ig(x, y), and Ib(x, y) represent the pixel values of the image in the red, green, and blue channels, respectively; for a grayscale image, D(x, y) = I(x, y). S53: Sobel operator: Use the Sobel operator to calculate the gradient of the image. The Sobel operator has two directions: horizontal and vertical. The horizontal Sobel operator is: Sobel operator in the vertical direction: Perform a convolution operation on the dark channel image to calculate the gradient: Horizontal gradient Gx(x, y); Horizontal gradient Gy(x, y) S54: Calculate edge strength: The edge strength of each pixel is calculated using the horizontal and vertical gradients: S55: Output the image.
6. A method for measuring the degree of dissociation under a mineral microscope based on an improved Gaussian pyramid according to claim 1, characterized in that: In the target mineral microscopic mask image, the edge points of all particles are extracted respectively, Cci is the edge point set of all mineral particle contours, and ci is the edge point set of one particle, as shown in the figure below; Cci={ci,j}.
7. A method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid according to claim 1, characterized in that: In the step S7, the edge point of each particle in the target mineral image is used in the associated mineral image, and the point is used as the center of the circle, and all points are traversed within a range of 6 pixels with a radius from the center point (x, y). If the pixel value in the area is not zero, it means that the mineral particle has not dissociated. At the same time, the point is located to the particle of the mineral framed in step 2, and all points are traversed in a loop until the dissociation and associated number of the target mineral particle and other associated mineral particles are counted, which specifically includes the following steps: S71: traverse each edge point: For each edge point (xi, yi): S72: Define the search area: traverse all points in a circular area with (xi, yi) as the center and a radius of 6 pixels. Let the search area be Ri: Ri={(x,y)|(x-xi)2+(y-yi)2≤6 2 } S73: Check the pixel value in the area: For each point (x, y) in the search area Ri, check the pixel value I (x, y): S74: Determine the dissociation state: If there exists (x, y)∈Ri such that I(x, y)≠0, then the particle is not dissociated: S75: Positioning to the mineral particle: Positioning the point (xi, yi) to the mineral particle framed in step 2.
8. A method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid according to claim 1, characterized in that: When calculating the dissociation degree, the monomer dissociation degree J of the target mineral is calculated by the following formula using the above-obtained data: Where: L is the number of monomer-dissociated mineral particles; M is the total number of mineral particles.
9. A method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid according to claim 1, characterized in that: In the process of extracting the target mineral edge coordinates in step S5 of the method for measuring the degree of dissociation of minerals under a microscope based on the improved Gaussian pyramid, a computer needs to be used to display an image, and the computer includes: a computer display screen; A movable groove, the movable groove is opened on one side of the surface of the computer display screen, and a limit groove is opened on the other side of the surface of the computer display screen, a reciprocating screw rod is rotatably installed inside the movable groove, an adjusting block is fixedly connected to the top of the reciprocating screw rod, and a movable device is threadedly engaged on the outer surface of the reciprocating screw rod, and the movable device includes a threaded block, a connecting plate, a sliding groove, a sliding rod and a limit slider; An observation device, which is slidably mounted inside the moving device, and includes a moving block, a connecting plate and a magnifying glass; A limiting groove is provided on the other side of the surface of the computer display screen, and a limiting rod is fixedly installed inside the limiting groove.
10. A method for measuring the degree of dissociation of minerals under a microscope based on an improved Gaussian pyramid according to claim 9, characterized in that: The threaded block is threadably engaged with the outer surface of the reciprocating screw rod, the connecting plate is fixedly installed on the front side of the threaded block, the limit slider is fixedly installed on one side of the surface of the connecting plate, the slide groove is opened in the middle of the front side of the connecting plate, the slide rod is fixedly installed in the middle of the inside of the slide groove, the moving block is slidably installed on the outer surface of the slide rod, the connecting plate is fixedly connected to one side of the surface of the moving block, and the magnifying glass is fixedly installed on the surface of the connecting plate.