Loess soil unit digital reconstruction and soil pore searching method

Abstract

The invention belongs to the technical field of three-dimensional graph generation, and particularly relates to a loess soil unit digital reconstruction and soil pore searching method. The method mainly comprises six parts of soil CT scanning, three-dimensional reconstruction image construction, three-dimensional stereogram segmentation and correction, binarization processing, soil pore three-dimensional stereogram reconstruction and soil pore feature analysis. According to the method, high-precision soil basic parameters can be provided for the unsaturated soil moisture movement, solute transport simulation and other processes, and a scientific basis is provided for improving the simulation and prediction precision of water circulation, carbon, nitrogen and the like. Three-dimensional reconstruction of loess soil units can be achieved, a digital and visual scheme is provided for soil structure research, agricultural soil management and agricultural water saving implementation, agricultural development is facilitated, and the system is suitable for the environment.

Description

Technical Field

[0001] The invention belongs to the technical field of three-dimensional graphics generation, and in particular relates to a loess soil unit digital reconstruction and soil pore search method. Background Art

[0002] Loess is a special Quaternary continental loose deposit formed by sedimentation and consolidation under special natural environment. It is mainly distributed in the arid and semi-arid zones in the mid-latitudes of the northern hemisphere of the world's continents, covering about 10% of the earth's land surface. In my country, the distribution area of ​​loess reaches 640,000 square kilometers, accounting for 6.3% of the country's land area. Loess is easy to cultivate and form fertile fields and fertile soil due to its fine particles, soft soil and rich nutrients, which is very conducive to agricultural development. However, due to its very loose structure, it is easy to be eroded and cause soil and water loss. As an important component of water resources, soil water has become one of the most important factors in terrestrial ecosystems and the basis for the survival of all terrestrial plants. Many physical, chemical and biological processes in the soil often require a certain amount of water to proceed. Soil water can not only provide plants with the water they need for growth, but also provide the nutrients they need. At the same time, agricultural pollution caused by excessive application of chemical fertilizers and irrigation with reclaimed water in areas with water shortages is very prominent. These pollutants enter the soil with soil water as the medium, become soil solutes, and migrate with the movement of soil water. Therefore, it is of great significance to study the retention and movement of soil water.

[0003] At present, 3D reconstruction technology is an important field of computer application and has been widely used in industrial production and social life. In addition, CT technology is one of the latest scientific and technological achievements of the contemporary era. CT scanning technology has gradually been applied from the medical field to soil science, providing a new technical means for studying soil properties. In civil engineering, the microstructure of clay soil is also an important factor in determining the mechanical properties of soil. How to realize the digital reconstruction of loess soil units will help to better alleviate soil loss and degradation and protect the ecological environment.

[0004] When applying mathematical and physical methods to quantitatively analyze the three main soil water consumption and absorption processes of soil water infiltration, evaporation, and water flow in the soil-plant-atmosphere continuum, hydraulic motion parameters are an indispensable part, whether using analytical or numerical solutions. Hydraulic motion parameters include specific water capacity C (the inverse of the slope of the water characteristic curve), hydraulic conductivity K, and diffusivity D, and the quantitative relationship between the three is D = K / C. There are two main methods for obtaining hydraulic motion parameters: direct method and indirect method. The direct method refers to directly obtaining hydraulic motion parameters through experiments. Among them, the direct methods for obtaining soil water characteristic curves mainly include tensiometer method, pressure membrane instrument method, sand funnel method, etc.; the direct methods for obtaining hydraulic conductivity mainly include instantaneous profile method, vertical infiltration flux method and vertical soil column stable evaporation method; the direct method for obtaining diffusivity is horizontal soil column infiltration method. Since the direct acquisition of soil hydraulic movement parameters is expensive and time-consuming, and is affected by a variety of soil physical and chemical parameters (soil texture, structure, bulk density, organic matter content, etc.), it is difficult to obtain a function that accurately expresses soil hydraulic movement parameters. In addition, most of the research on soil hydraulic movement parameters by scholars focuses on disturbed soils, and there is less research on undisturbed soils in the field. Since there are obvious differences in soil structure between disturbed soils and undisturbed soils in the field, the results obtained from disturbed soil tests will cause large errors if used for undisturbed soils in the field. The indirect method mainly refers to the process of constructing a function between soil-related physical and chemical parameters and soil hydraulic movement parameters (soil transfer function, Pedo-transferFunctions), and inputting relevant physical and chemical parameters to obtain hydraulic movement parameters.

[0005] By understanding the current status of research, on the basis of the three-dimensional reconstruction method, relevant scholars such as Lv Fei, Liu Jianli, Cheng Yanan, Xiang Fangyuan, etc. have established network models based on CT imaging technology, and have made progress in more accurate prediction of soil properties. We believe that the methods in this field have the following two shortcomings: ① CT scanning is suitable for large-scale research. Since the surface of the soil sample itself is not smooth enough and the surface material is not uniform enough, although CT scanning imaging and analysis speed are fast and the operation process is relatively simple, these all affect the correctness of CT scanning, resulting in unstable results of digital reconstruction of loess soil units. ② After constructing the three-dimensional pore space structure based on the continuous slice images obtained by CT scanning the original soil, it is easy to cause the three-dimensional reconstructed model to simulate hysteresis and have limitations in water absorption and drainage.

[0006] In summary, we believe that most of the soil hydraulic parameters derived from soil transfer functions do not consider the influence of soil structure, which leads to large deviations in the simulation and prediction of hydrological processes (precipitation infiltration, land evaporation, surface runoff, groundwater recharge) and biogeochemical processes. Studies at the scale of soil columns and soil profiles have shown that soil structure has an important influence on soil hydraulic parameters. Summary of the invention

[0007] In order to overcome the problems existing in the above-mentioned technologies, the present invention provides a method for digital reconstruction of loess soil units and soil pore search.

[0008] The present invention determines the quantitative relationship between soil structure and soil hydraulic movement parameters, constructs a loess soil hydraulic parameter dataset containing soil structural information, clarifies the influence of soil structure on soil hydraulic movement parameters, and realizes digital reconstruction of the three-dimensional structure of loess soil taking into account soil structural changes.

[0009] The technical solution adopted by the present invention is: a method for reconstructing three-dimensional loess soil CT images and searching for soil pores, comprising the following steps:

[0010] A method for digitally reconstructing loess soil units and searching for soil pores comprises the following steps:

[0011] (1) Soil CT scanning: Based on the existing database, soil samples of different shapes, sizes, and structures are collected at fixed points in the field; the samples are placed on the scanning platform of the CT scanner, and the parameters of the CT scanner are set according to the particle size characteristics of the sample loess soil and the required resolution. The CT scanning equipment is started, and different soil structures are scanned using the CT scanner and focused ion beam-scanning electron microscope to obtain multiple continuous multi-dimensional CT scan images;

[0012] (2) constructing a three-dimensional reconstruction image: selecting continuous CT scan images from the CT images obtained in step (1), inputting the captured projection data into an image processing platform, using the CT images as sampling data, selecting the same position of each image, and interpolating the sampling data according to equal time intervals to construct a three-dimensional reconstruction image;

[0013] (3) Segmenting, denoising, correcting, binary processing, and preprocessing the three-dimensional reconstruction image constructed by step (2) to ensure the accuracy of the image and facilitate subsequent images to more realistically reflect the physical properties of the soil;

[0014] (4) Binarization: using two basic mathematical morphology operators, erosion and dilation, to erode the binary image with spherical structures of nine diameter levels and then dilate them with structures of the same size to remove pores smaller than the corresponding level, thus obtaining the pore size distribution corresponding to different levels;

[0015] (5) Reconstructing a three-dimensional soil pore image: extracting pore units from the three-dimensional reconstruction image after binarization processing in step (4) and redrawing them to obtain a reconstructed three-dimensional soil pore image;

[0016] (6) Analyzing soil pore characteristics: The three-dimensional image is divided into pore bodies and pore throats through the processing of step (5); after this division, the three-dimensional macroporous structure characterization of the soil can be achieved by controlling the volume rendering in the image;

[0017] (7) The time step is set to 1000, and the output is performed every 1000 steps. The CT image selected in step (2) is binarized, and the pore area after binarization is compared with the pore area before binarization. If the difference in area value is greater than 0.1, the threshold is reset to conduct the test;

[0018] (8) When the difference between the calculated pore area and the unbinarized pore area is within 0.1, it can be considered that the binarized pore area is equal to the true pore area, so this threshold is selected as the binarization threshold.

[0019] Furthermore, the image processing platform can also be one of C++ combined with opencv image processing technology and python platform BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 is a flow chart of the method of the present invention;

[0021] Figure 2 This is a flow chart of digital reconstruction of loess soil units in the present invention;

[0022] Figure 3 For the image processing and three-dimensional reconstruction of the present invention

[0023] Among them: a. Image acquisition; b. Image clipping; c. Preprocessing: noise removal and intensity correction; d. Binary processing; e. Three-dimensional macropore reconstruction. DETAILED DESCRIPTION

[0024] like Figures 1 to 3 As shown, the method for digitally reconstructing loess soil units and searching for soil pores described in this embodiment includes the following steps:

[0025] (1) Based on the existing database, soil samples of different shapes, sizes, and structures are collected at fixed points in the field; the samples are placed on the scanning platform of the CT scanner, and the parameters of the CT scanner are set according to the particle size characteristics of the sample loess soil and the required resolution. The CT scanning equipment is started, and different soil structures are scanned using the CT scanner and focused ion beam-scanning electron microscope to obtain multiple continuous multi-dimensional CT scan images;

[0026] (2) constructing a three-dimensional reconstruction image: selecting continuous CT scan images from the CT images obtained in step (1), inputting the captured projection data into an image processing platform, using the CT images as sampling data, selecting the same position of each image, and interpolating the sampling data according to equal time intervals to construct a three-dimensional reconstruction image;

[0027] (3) Segmenting, denoising, correcting, binary processing, and preprocessing the three-dimensional reconstruction image constructed by step (2) to ensure the accuracy of the image and facilitate subsequent images to more realistically reflect the physical properties of the soil;

[0028] (4) Binarization: that is, using two basic mathematical morphology operators, corrosion and expansion, to corrode the binary image with spherical structures of 9 diameter levels and then use structures of the same size to perform expansion operations, eliminating pores smaller than the corresponding level, and obtaining the pore size distribution corresponding to different levels; the formula is expressed as:

[0029]

[0030] Where: B(c,y,z) is the pixel value of the binarized image at position (x,y,z); I(x,y,z) is the pixel grayscale value of the original CT image at position (x,y,z); T is the set threshold;

[0031] Where: B(c,y,z) is the pixel value of the binarized image at position (x,y,z); I(x,y,z) is the pixel grayscale value of the original CT image at position (x,y,z). T is the set threshold.

[0032] (5) Reconstructing a three-dimensional soil pore image: extracting pore units from the three-dimensional reconstruction image after binarization processing in step (4) and redrawing them to obtain a reconstructed three-dimensional soil pore image;

[0033] (6) Analyzing soil pore characteristics: The three-dimensional image is divided into pore bodies and pore throats through the processing of step (5); after this division, the three-dimensional macroporous structure characterization of the soil can be achieved by controlling the volume rendering in the image; the formula is expressed as:

[0034] The calculation formula of the equivalent diameter (ED) of each macropore is expressed as:

[0035]

[0036] The soil macroporosity (MAP) calculation formula is expressed as:

[0037] MAP=V p / V t

[0038] Where Vp represents the volume of macropores in the model, and Vt represents the total volume of the sample;

[0039] P(ε) 1 / D =α 0 ε (1-D) / D [A(ε)] 1 / 2 =α 0 ε[A(ε)] 1 / 2

[0040] Where P is the perimeter of the macropore, α 0 is the constant of fractal characteristics, A is the surface area, and ε is the fractal coefficient;

[0041] By further simplifying the formula, we can get:

[0042] P(ε) 1 / D ∝A(ε) 1 / 2

[0043] Here, the radius R of the macropore can replace the perimeter P, and R is proportional to ε; therefore, the formula can be further expressed as:

[0044]

[0045] In the formula, β / 2 is the slope of the fitting line of lnR and lnA, and C is a constant; therefore, the fractal dimension of soil pore morphology can be expressed as D=3-β / 2.

[0046] (7) The time step is set to 1000, and the output is performed every 1000 steps. The CT image selected in step (2) is binarized, and the pore area after binarization is compared with the pore area before binarization. If the difference in area value is greater than 0.1, the threshold is reset to conduct the test;

[0047] (8) When the difference between the calculated pore area and the unbinarized pore area is within 0.1, it can be considered that the binarized pore area is equal to the true pore area, so this threshold is selected as the binarization threshold.

[0048] The image processing platform can also be one of C++ combined with opencv image processing technology and python platform.

Claims

1. A method for digital reconstruction of loess soil units and soil pore search, characterized in that: The steps include: (1) Soil CT scanning: Based on the existing database, soil samples of different shapes, sizes, and structures are collected at fixed points in the field; the samples are placed on the scanning platform of the CT scanner, and the parameters of the CT scanner are set according to the particle size characteristics of the sample loess soil and the required resolution. The CT scanning equipment is started, and different soil structures are scanned using the CT scanner and focused ion beam-scanning electron microscope to obtain multiple continuous multi-dimensional CT scan images; (2) constructing a three-dimensional reconstruction image: selecting continuous CT scan images from the CT images obtained in step (1), inputting the captured projection data into an image processing platform, using the CT images as sampling data, selecting the same position of each image, and interpolating the sampling data according to equal time intervals to construct a three-dimensional reconstruction image; (3) Segmenting, denoising, correcting, binary processing, and preprocessing the three-dimensional reconstruction image constructed by step (2) to ensure the accuracy of the image and facilitate subsequent images to more realistically reflect the physical properties of the soil; (4) Binarization: using two basic mathematical morphology operators, erosion and dilation, to erode the binary image with spherical structures of nine diameter levels and then dilate them with structures of the same size to remove pores smaller than the corresponding level, thus obtaining the pore size distribution corresponding to different levels; (5) Reconstructing a three-dimensional soil pore image: extracting pore units from the three-dimensional reconstruction image after binarization processing in step (4) and redrawing them to obtain a reconstructed three-dimensional soil pore image; (6) Analyzing soil pore characteristics: The three-dimensional image is divided into pore bodies and pore throats through the processing of step (5); after this division, the three-dimensional macroporous structure characterization of the soil can be achieved by controlling the volume rendering in the image; (7) The time step is set to 1000, and the output is performed every 1000 steps. The CT image selected in step (2) is binarized, and the pore area after binarization is compared with the pore area before binarization. If the difference in area value is greater than 0.1, the threshold is reset to conduct the test; (8) When the difference between the calculated pore area and the unbinarized pore area is within 0.1, it can be considered that the binarized pore area is equal to the true pore area, so this threshold is selected as the binarization threshold.

2. According to claim 1, a method for three-dimensional reconstruction of soil CT images and soil pore search, characterized in that: The image processing platform can also be one of C++ combined with opencv image processing technology and python platform.

Images