Rock mineral component automatic segmentation method based on image processing and multi-channel correlation

Abstract

The present application relates to geology, rock engineering field, specifically to a kind of rock mineral component automatic segmentation method based on image processing and multi-channel correlation.In the aspect of geological research and lithology identification, quantitative analysis of rock mineral component is crucial, the present application is based on rock digital image, first, the original rock sample image is cropped, size is determined, noise reduction and other pretreatment;Then, using pixel row curve and digital image interactive method, the initial segmentation threshold of mineral component in rock is determined, and the initial segmentation of rock mineral component is carried out;Finally, the correlation between multi-channel is used again, the mineral components with similar color and texture are secondary segmented, and the accuracy of rock mineral component segmentation is improved.The present application improves the accuracy of rock mineral component analysis, not only can make geology practitioner understand rock material composition and category faster and better, but also can provide high-precision fine numerical model for rock material numerical simulation analysis.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of geology and rock engineering, and particularly relates to a rock mineral component automatic segmentation method based on image processing and multi-channel correlation. BACKGROUND

[0002] In the field of geological research and lithology identification, quantitative analysis of rock mineral components is crucial. Traditional rock mineral analysis mainly relies on manual experience identification and microscopic observation, and the results are limited by the subjective experience of the observer. The color of minerals in rocks is the result of the interaction of their chemical composition, crystal structure, and light. This color is usually caused by the presence of specific elements in the mineral or the crystal structure.

[0003] With the rapid development of computer vision technology, automatic identification and segmentation of rock mineral components based on image analysis has attracted much attention. This method can automatically identify and quantify mineral components in rock optical images, improving the efficiency and accuracy of rock mineral component analysis. However, due to the color diversity and fuzzy boundaries of minerals in rock optical images, current single-channel image processing methods cannot accurately segment various mineral components.

[0004] Through analysis and learning of rock images, using multi-channel rock and texture features, and through image processing algorithms, high-precision automatic segmentation of mineral components in rock optical images is achieved. Compared with traditional methods, this method can effectively handle image noise and more accurately identify mineral boundaries, thereby improving the accuracy of rock mineral component analysis. The development of this technology not only enables geologists to better understand the composition and classification of rock materials, but also provides high-precision numerical models for numerical simulation analysis of rock materials. SUMMARY

[0005] The purpose of the present application is to solve the technical problems and challenges in the automatic mineral component segmentation process of true-color rock images in the prior art, which directly affect the accuracy of image segmentation and the feasibility in practical scenarios. A rock mineral component automatic segmentation method based on image processing and multi-channel correlation is proposed.

[0006] To achieve the above purpose, the present application adopts the following technical solutions:

[0007] A rock mineral component automatic segmentation method based on image processing and multi-channel correlation, comprising the following steps:

[0008] Step 1, prepare a rock sample;

[0009] The rock sample is mainly divided into fine-grained and coarse-grained rock according to the size of the rock mineral grain size;

[0010] For fine-grained rock samples, clean the rock surface with water, use a rock cutting machine to cut the fine-grained rock sample into a thin slice, and polish it to be transparent; the thickness is 0.03mm, and the area is 22mm×22mm;

[0011] For coarse-grained rock samples, select a rock mass with clear mineral crystals, clean the rock surface with water, and use a cutting machine to cut the sample into a flat surface;

[0012] Step 2, optical image acquisition of rock sample;

[0013] For fine-grained rock, place the fine-grained rock slice on the microscope stage, select appropriate eyepiece and objective lens to magnify the sample, adjust the light source to ensure uniform illumination on the sample, and use a high-resolution camera connected to the microscope for image capture;

[0014] For coarse-grained rock, use an optical camera or scanner to take a photo or scan the flat surface of the rock to obtain the image information of the rock surface;

[0015] Step 3, image preprocessing of rock sample;

[0016] Crop the coarse-grained and fine-grained rock images, only keep the area including the rock mineral components, and remove the environmental influence area at the edge of the image;

[0017] When the image is too large, adjust the image resolution to reduce the image size to improve the calculation efficiency of subsequent segmented images;

[0018] Use digital filtering algorithms such as Gaussian filtering and median filtering to remove random noise in the image;

[0019] Step 4, RGB channel separation of rock sample image;

[0020] The rock sample image in step 3 is composed of two-dimensional arrays of R, G, and B color channels. According to the data structure of R, G, and B channels, separate each channel data representing mineral component information, and each array represents a color channel;

[0021] Step 5, threshold determination for mineral component segmentation;

[0022] In the rock sample image, take any row of all pixel points in the rock sample image as a pixel row, randomly select 1 pixel row, which contains the R, G, and B three-channel pixel value curve drawn by the pixel row, and the R, G, and B colors in the pixel row represent different mineral component information;

[0023] According to the rock sample image and the mineral component information corresponding to the pixel row, a boundary between adjacent mineral components is distinguished, and a pixel interval and a segmentation threshold corresponding to different mineral components in each channel are determined according to the boundary;

[0024] Step 6, determination of the multi-channel correlation of the mineral component segmentation;

[0025] Through step 5, effective segmentation of the mineral components of most rock images can be achieved. For rock images that are difficult to be effectively segmented, further segmentation is performed by using the correlation between R, G and B multi-channels. The most commonly used method is to perform four arithmetic operations on R, G and B three channels;

[0026] Step 7, multi-channel segmentation and multi-channel correlation result fusion;

[0027] Through the threshold segmentation of step 5 and the multi-channel correlation of step 6, the segmented mineral component image and the segmentation image of each mineral component are obtained, so that the composition and category of the rock material are obtained;

[0028] Step 8, quantitative analysis of the mineral components;

[0029] For each mineral component in the rock image, the proportion of the mineral component in the image is calculated (formula 1) to obtain the quantitative composition analysis of the rock:

[0030]

[0031] a- the proportion of a single mineral component, P m - the total number of pixels of a single mineral component, P r - the total number of pixels of the rock image;

[0032] To provide a high-precision fine numerical model for numerical simulation analysis of rock materials.

[0033] As a further preferred scheme, in step 4, the color sample image is separated into RGB three channels, instead of using a single gray value channel.

[0034] As a further preferred scheme, in step 5, the threshold is determined by using the pixel row and the color image interaction method;

[0035] The specific steps are as follows: first, select 1 pixel row, read the R, G and B three channel pixel value information in the pixel row, and draw the pixel value information into a continuous curve; then, use the image information and the curve interaction to preliminarily determine the R, G and B value interval of each mineral component.

[0036] As a further preferred scheme, in step 6, the mineral component segmentation is determined by using the multi-channel correlation;

[0037] The specific steps are as follows: for mineral components with similar textures and colors and unable to be well segmented, the interrelation between different channels is analyzed; the difference between R, G and B of a certain mineral component is different, and the interrelation between R, G and B of the certain mineral component can be used to further accurately segment the mineral components in the rock.

[0038] Advantages

[0039] Compared with the prior art, the rock mineral component automatic segmentation method based on image processing and multi-channel correlation has the following advantages:

[0040] 1. The RGB three-channel segmentation method provided by the present application can better distinguish different mineral components in the rock by using multi-channel information, and compared with the missegmentation problem often caused by a single gray threshold value, the RGB three-channel segmentation method improves the accuracy of mineral component segmentation.

[0041] 2. The present application considers the correlation between RGB multi-channels, and only through the threshold segmentation of each channel, missegmentation caused by similar color or texture is easy to cause, and the correlation between each channel is established, which can make up for the above shortcomings of missegmentation.

[0042] 3. The present application can be applied to the component segmentation of different rock types, and the multi-channel correlation analysis can adapt to the characteristics of different types of rocks and minerals, and improve the generalization ability of the algorithm when processing diversified samples. When facing complex rock mineral combinations, multi-channel analysis provides more information and dimensions, which can flexibly cope with and accurately segment. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 is a flowchart of the rock mineral component automatic segmentation method based on image processing and multi-channel correlation of the present application;

[0044] Figure 2 is a rock image (a) original sample image (b) image after mean filtering and noise reduction;

[0045] Figure 3 is a rock mineral component pixel row gray value and RGB value image;

[0046] Figure 4 is a rock mineral component segmentation result and each mineral component shape image. DETAILED DESCRIPTION

[0047] The rock mineral component automatic segmentation method based on image processing and multi-channel correlation of the present application is specifically described from the following key technical problems:

[0048] 1. Rock optical image quality. The image is the only source of information for segmenting rock mineral components. If the image has obvious noise and quality problems, it will have a negative impact on the image processing algorithm and increase the difficulty of segmenting rock mineral components.

[0049] 2. Variety of mineral components and ambiguity of mineral boundaries. Mineral component types are diverse, and determining mineral boundaries by artificial methods has strong artificial differences.

[0050] 3. Missegmentation of mineral components with similar gray values. Some mineral components have significant differences in color images, but are very similar in gray space. For example, a certain mineral component color (R-220, G-200, B-180), another mineral component color (R-180, G-200, B-220), and another mineral component color (R-200, G-200, B-200). The above three colors are the same gray value in the gray image, but they are three different minerals.

[0051] 4. Low accuracy of non-associated multi-channel threshold segmentation. After determining the segmentation threshold of each multi-channel, most mineral components can be effectively segmented, but the segmentation effect of mineral components with similar texture and color is not good. The simple threshold range of non-associated multi-channel cannot directly distinguish these mineral components with similar texture and color.

[0052] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all.

[0053] Embodiment 1

[0054] An automatic rock mineral component segmentation method based on image processing and multi-channel association, the specific steps are as follows:

[0055] Step 1, prepare a coarse-grained granite sample to make its surface flat.

[0056] Step 2, obtain the surface image, the image is an RGB color image, as shown in Figure 2 (a).

[0057] Step 3, adjust the image to 256x256 pixel size. In order to reduce the influence of non-natural factors on the rock surface (such as knife marks, miscellaneous colors, etc.) and optimize the coherence and integrity of the particles, a 2x2 template is used to perform median filtering on the sample image, as shown in Figure 2 (b).

[0058] Step 4, RGB three-channel separation is performed on the image, the color rock sample image is converted into a multi-dimensional array of 256x256x3, and then R, G and B three channels are individually separated into 256x256 arrays.

[0059] Step 5, as shown in the single-channel gray image (such as Figure 3 , it is found that the gray values of component 2 and component 3 are very close, and it is difficult to form effective segmentation. The 128th line is selected as a reference pixel line, and the 256 pixel values of each channel R, G and B in the pixel value information of the 128th line are plotted as a curve, as shown in Figure 3 . Through the interactive analysis of the segmentation line of adjacent mineral components in the image and the pixel curve, the R, G and B three-channel segmentation threshold values of component 1 and component 3 are 120, and the R, G and B three-channel segmentation threshold values of component 3 and component 2 are 160, that is, 0≤R, G, B<=120 is component 1, 120<R, G, B<160 is component 3, and 160≤R, G, B≤255 is component 2. The preliminary segmentation image is shown in Figure 2 (c).

[0060] Step 6, through the initial segmentation Figure 2 (c), it is found that R, G and B>=160 contain part of the R, G and B values of component 3, so it is necessary to segment the R, G and B values of R-B<10 into component 3 particles according to the difference between the R, G and B values of component 3 and component 2, to obtain the final accurate mineral component segmentation image, as shown in Figure 2 (d).

[0061] Step 7, through steps 5 and 6, the rock mineral component segmentation result and the shape image of each mineral component can be obtained, as shown in Figure 4 .

[0062] Step 8, the proportion information of the segmented image is counted, and the proportions of component 1, component 2 and component 3 are 15%, 32% and 53%, respectively.

[0063] The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can make equivalent replacement or change according to the technical solution and the inventive concept of the present application within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. An automatic segmentation method for rock mineral components based on image processing and multi-channel correlation, comprising the following steps, characterized in that: Step 1: Prepare rock samples; Rock samples are classified into fine-grained and coarse-grained rocks based on the size of the rock mineral grains. For fine-grained rock samples, clean the rock surface with water, cut the fine-grained rock samples into thin slices using a rock cutter, and polish them until they are translucent. For coarse-grained rock samples, select rock blocks with clearly visible mineral crystals, clean the rock surface with clean water, and use a cutting machine to cut the sample into a flat surface; Step 2: Acquisition of optical images of the rock sample; For fine-grained rocks, place the thin section of fine-grained rock on the microscope stage, select the eyepiece and objective lens to magnify the sample, adjust the light source to ensure that the light source illuminates the sample evenly, and use a high-resolution camera connected to the microscope to capture images. For coarse-grained rocks, use an optical camera or scanner to photograph or scan the flat surface of the rock to obtain image information of the rock surface; Step 3: Rock sample image preprocessing; Cropping coarse-grained and fine-grained rock images to retain only the areas containing rock mineral components, and removing environmentally affected areas at the image edges; When the image is too large, the image size can be reduced by adjusting the image resolution to improve the computational efficiency of subsequent image segmentation. Digital filtering algorithms are used to remove random noise from images; Step 4: Separate the RGB channels of the rock sample image; The rock sample image in step 3 is composed of two-dimensional arrays of three color channels: R, G, and B. Based on the data structure of the R, G, and B channels, the data of each channel representing mineral composition information is separated, and each array represents a color channel. Step 5: Determine the threshold for mineral component segmentation; In the rock sample image, all the pixels in any row of the rock sample image are taken as the pixel row. One pixel row is randomly selected. The pixel row contains the R, G, and B three-channel pixel value curves of the pixel row. The R, G, and B colors in the pixel row represent different mineral composition information. Based on the rock sample image and the mineral composition information corresponding to the pixel row, the boundary line between adjacent mineral components is distinguished, and the pixel interval and segmentation threshold corresponding to different mineral components in each channel are determined according to the boundary line. Step 6: Determine the multi-channel correlation of mineral component segmentation; Step 5 enables effective segmentation of mineral components in most rock images. For rock images that are difficult to segment effectively, the correlation between the R, G, and B channels will be used for further segmentation by performing four arithmetic operations on the R, G, and B channels. Step 7: Multi-channel segmentation and multi-channel correlation result fusion; By further segmenting through threshold segmentation in step 5 and multi-channel correlation in step 6, the segmented mineral composition map and the segmented image of each mineral composition are obtained, thereby obtaining the composition and category of the rock material. Step 8: Quantitative analysis of mineral components; For each mineral component in the rock image, its proportion in the image is calculated to obtain a quantitative compositional analysis of the rock: a-Proportion of single mineral components, P m - Total number of pixels for a single mineral component, P r - The total number of pixels in the rock image; It provides a high-precision microscopic numerical model for numerical simulation analysis of rock materials.

2. The automatic segmentation method for rock mineral components based on image processing and multi-channel correlation according to claim 1, characterized in that: In step 4, the color sample image was separated into RGB three channels instead of using a single grayscale channel.

3. The automatic segmentation method for rock mineral components based on image processing and multi-channel correlation according to claim 1, characterized in that: Step 5 uses a method of interacting with pixel rows and color images to determine the threshold; The steps are as follows: First, select any row of pixels, read the pixel value information of the R, G, and B channels in the row, and plot the pixel value information as a continuous curve; then, use the interaction between image information and curve to preliminarily determine the R, G, and B value ranges for each mineral component.

4. The automatic segmentation method for rock mineral components based on image processing and multi-channel correlation according to claim 1, characterized in that: In step 6, the mineral component segmentation is determined using multi-channel correlation. The steps are as follows: For mineral components with similar textures and colors that cannot be well segmented, analyze the interrelationships between different channels; if the differences between R, G, and B values ​​of a certain mineral component are different, further refine the segmentation of mineral components in the rock by analyzing the correlation between R, G, and B values ​​of a certain mineral component.

Images