CN116894851A - Soil CT image pore extraction method and system - Google Patents
Soil CT image pore extraction method and system Download PDFInfo
- Publication number
- CN116894851A CN116894851A CN202311159455.3A CN202311159455A CN116894851A CN 116894851 A CN116894851 A CN 116894851A CN 202311159455 A CN202311159455 A CN 202311159455A CN 116894851 A CN116894851 A CN 116894851A
- Authority
- CN
- China
- Prior art keywords
- soil
- threshold
- image
- value
- segmentation
- 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
- 239000002689 soil Substances 0.000 title claims abstract description 250
- 239000011148 porous material Substances 0.000 title claims abstract description 63
- 238000000605 extraction Methods 0.000 title claims abstract description 30
- 238000000034 method Methods 0.000 claims abstract description 66
- 230000011218 segmentation Effects 0.000 claims abstract description 60
- 238000003709 image segmentation Methods 0.000 claims abstract description 44
- 238000013178 mathematical model Methods 0.000 claims abstract description 18
- 238000004422 calculation algorithm Methods 0.000 claims description 32
- 230000003044 adaptive effect Effects 0.000 claims description 16
- 238000004590 computer program Methods 0.000 claims description 15
- 238000005070 sampling Methods 0.000 claims description 13
- 238000009826 distribution Methods 0.000 claims description 11
- 238000005315 distribution function Methods 0.000 claims description 11
- 230000010355 oscillation Effects 0.000 claims description 10
- 238000002591 computed tomography Methods 0.000 claims description 8
- 230000001186 cumulative effect Effects 0.000 claims description 7
- 238000003860 storage Methods 0.000 claims description 7
- 239000011800 void material Substances 0.000 claims description 2
- 238000012545 processing Methods 0.000 abstract description 7
- 230000033228 biological regulation Effects 0.000 abstract description 2
- 239000000543 intermediate Substances 0.000 description 30
- 230000008569 process Effects 0.000 description 20
- 230000000694 effects Effects 0.000 description 9
- 230000006870 function Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- 238000005457 optimization Methods 0.000 description 6
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 238000011002 quantification Methods 0.000 description 5
- 238000003708 edge detection Methods 0.000 description 4
- 230000004075 alteration Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000002595 magnetic resonance imaging Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000001105 regulatory effect Effects 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 238000012800 visualization Methods 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000031018 biological processes and functions Effects 0.000 description 1
- 230000002860 competitive effect Effects 0.000 description 1
- 230000009514 concussion Effects 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000011549 displacement method Methods 0.000 description 1
- 238000011066 ex-situ storage Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000004260 weight control Methods 0.000 description 1
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
- G06T7/136—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
- G06T7/11—Region-based segmentation
-
- 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
- G06T2207/10072—Tomographic images
- G06T2207/10081—Computed x-ray tomography [CT]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/30—Assessment of water resources
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
A soil CT image pore extraction method and system relate to the field of soil image processing. Solves the problem that the existing soil pore extraction is easily affected by noise, resulting in inaccurate soil CT image pore extraction. The method comprises the following steps: establishing a mathematical model based on multi-threshold soil CT image segmentation, and carrying out image segmentation to obtain segmented images; constructing an initial threshold set for soil CT image segmentation; updating the initial threshold set of the soil CT image segmentation according to the self-adaptive weight searching strategy to obtain a local optimal threshold set; searching the global according to the differential evolution strategy, and updating a local optimal threshold set; calculating the information entropy of each threshold value; iteratively updating until the iteration times are greater than or equal to a preset value to obtain a global optimal segmentation threshold set; and dividing the original soil CT image according to the optimal threshold value in the global optimal dividing threshold value set to obtain soil pores. The invention is applied to the field of soil regulation.
Description
Technical Field
The invention relates to the field of soil image processing, in particular to a soil CT image pore extraction method.
Background
Soil pore quantification is one of the basic problems of analyzing the structural stability and connectivity of soil, and provides important theoretical basis and technical support for regulating the physical, chemical and biological processes of soil. The main aim of quantification of soil porosity is to obtain the change of soil pore structure by analyzing the number, the size, the morphology and the distribution of the soil pores, so as to regulate and control soil management measures. Quantitative analysis of soil porosity is a very complex analysis problem, and in recent years, many methods for calculating the porosity of soil have been studied, such as soil specific gravity, water level method, dry density method, gas displacement method, and the like. Although to some extent the global optimal solution can be found, these methods also have significant drawbacks: the studied problems can only be solved by adopting a linear objective function and a constraint condition, but the linear approximation operation on the quantitative problem of the soil porosity can lead to larger errors of the final calculation result. Therefore, the above method is not suitable for solving the problem of quantification of soil porosity. The development of CT technology and magnetic resonance imaging (magnetic resonance imaging, MRI) technology directly realizes the 3D visualization and quantification of undisturbed pores, so that the result has higher robustness. In pore studies, CT techniques are currently the most widely used method, mainly by ex situ studies to obtain as-situ pore structure features. The CT scanning technology is adopted to scan the soil, so that the true soil undisturbed structure of the soil can be obtained, and therefore, the related parameters of the soil pore structure are calculated, and theorem analysis of the soil pores is realized. In addition to the resolution, contrast effects of the image itself, the choice of segmentation method is also directly related to the accuracy and effectiveness of the quantized data. Because the background of the soil image is complex, the noise is more, and the extraction of soil pores, namely image segmentation, is needed before the pore structure is quantized.
The conventional image segmentation generally adopts an edge detection method to segment the image, and mainly detects boundary information in the image to segment the image. Common edge detection algorithms include Canny edge detection, sobel operator, laplacian operator, and the like. These methods can determine the edge position by detecting the gradient change of the image gray value. However, in the soil image, the edge is not necessarily clear, and there are factors such as noise and texture variation, so that the edge detection method is easily interfered, and problems such as breakage and false detection are generated.
Disclosure of Invention
Aiming at the problem that the existing soil pore extraction is susceptible to noise and the soil CT image pore extraction is inaccurate, the invention provides a soil CT image pore extraction method, which comprises the following steps:
a method for extracting soil CT image voids, the method comprising:
s1: establishing a mathematical model based on multi-threshold soil CT image segmentation, and carrying out image segmentation according to the mathematical model to obtain a segmented image;
s2: constructing a soil CT image segmentation initial threshold set according to the segmented image;
s3: updating the initial threshold set of the soil CT image segmentation according to the self-adaptive weight searching strategy to obtain a local optimal threshold set;
S4: searching the global according to a differential evolution strategy, and updating the local optimal threshold set;
s5: calculating the information entropy of each threshold in the step S3 and the step S4, and reserving the selection schemes of n thresholds with the top information entropy ranks as a threshold set of t+1 iterations;
s6: iterating the steps S3 to S5 until the iteration times are greater than or equal to a preset value, and obtaining a global optimal segmentation threshold set;
s7: and dividing the original soil CT image according to the optimal threshold value in the global optimal division threshold value set to obtain soil pores.
Further, there is also provided a preferred mode, wherein the initial threshold set for segmenting the CT image of the soil in the step S2 includes: and the multiple groups of threshold selection schemes comprise different soil image threshold parameters, and the soil image threshold parameters consist of gray value parameters of soil CT scanning images.
Further, there is also provided a preferred mode, wherein the step S3 includes:
calculating an adaptive weight search strategy control factor:
calculating oscillation factor:
Threshold selection scheme in updated threshold set：
，
wherein ,an ith threshold selection scheme representing t threshold searches, >Current local optimum representing t threshold searches, < >>Representing a random scheme in a threshold set in T times of threshold search, T representing an index of the current gray value search times, and T representing the total iteration times. />Is [0,1 ]]Random parameters between->Is a random factor->Representing an adaptive weight search policy control factor, +.>Representing a random scaling factor in an adaptive weight search strategy,/->Information entropy indicating the ith threshold selection scheme, < ->For the current optimal information entropy->For worst information entropy->Representing the median value of the entropy of the information in the threshold set, +.>And b is the search radius of the algorithm for carrying out random search in the global.
Further, there is also provided a preferred mode, wherein the step S4 includes:
rearranging the threshold value set randomly;
vacating the position where the current threshold selection scheme is arranged;
randomly generating a scaling factor;
randomly sampling in a gray space of a soil CT image to obtain a gray value intermediate;
setting a boundary for the gray value intermediate;
randomly selecting the dimension of the gray value to be exchanged;
traversing each dimension of the threshold selection scheme, and updating the threshold selection scheme of the new individual according to the conditions of the dimension to be exchanged and the crossover probability.
Further, there is provided a preferred mode, wherein the step of randomly sampling in a gray scale space of a soil CT image to obtain a gray scale value intermediate includes:
wherein ,threshold selection schemes that centralize rearranged thresholds at 1,2,3 positions.
Further, there is also provided a preferred manner, wherein the calculating the information entropy of each threshold in the step S4 includes:
…
wherein ,、/>、/>is a segmentation threshold parameter; />Gray level of CT image of soil>Probability distribution of (2); />Entropy of soil image information corresponding to 1 st threshold interval,>the entropy of the soil image information corresponding to the 2 nd threshold interval,entropy of soil image information corresponding to the Kth threshold interval,>cumulative distribution function of soil gray value for 1 st threshold interval,/for the soil gray value>Cumulative distribution function of soil gray value for the 2 nd threshold interval,>accumulating distribution function for soil gray value of Kth threshold value interval, < >>Is the maximum gray level of the soil image.
Further, there is also provided a preferred mode, wherein the step S6 includes:
，
wherein ,for the optimal threshold set, ++>For maximum information entropy in threshold set, +.>When the threshold value set TH is applied to the soil image, the soil image information entropy corresponding to the z-TH threshold value section is obtained.
Based on the same inventive concept, the invention also provides a soil CT image pore extraction system, which comprises:
the image segmentation unit is used for establishing a mathematical model based on multi-threshold soil CT image segmentation, and carrying out image segmentation according to the mathematical model to obtain a segmented image;
an initial threshold set constructing unit, configured to construct a soil CT image segmentation initial threshold set according to the segmented image;
the self-adaptive weight searching strategy updating unit is used for updating the initial threshold set of the soil CT image segmentation according to the self-adaptive weight searching strategy to obtain a local optimal threshold set;
the differential evolution strategy searching unit is used for searching the global according to the differential evolution strategy and updating the local optimal threshold set;
the computing unit is used for computing the information entropy of each threshold value in the self-adaptive weight search strategy updating unit and the differential evolution strategy searching unit, and reserving the selection schemes of n threshold values with the top ranking of the information entropy as a threshold value set of t+1 iterations;
the iteration unit is used for iterating the self-adaptive weight searching strategy updating unit to the calculating unit until the iteration times are greater than or equal to a preset value, so as to obtain a global optimal segmentation threshold set;
And the soil pore acquisition unit is used for dividing the original soil CT image according to the global optimal dividing threshold set optimal threshold to acquire soil pores.
Based on the same inventive concept, the present invention also provides a computer readable storage medium for storing a computer program for executing a soil CT image aperture extraction method as described in any one of the above.
Based on the same inventive concept, the invention also provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes a soil CT image pore extraction method according to any one of the above.
The invention has the advantages that:
the invention solves the problem that the existing soil pore extraction is affected by noise easily, so that the soil CT image pore extraction is inaccurate.
According to the soil CT image pore extraction method, a mathematical model based on multiple thresholds is firstly established, and the soil CT image is segmented according to the model. The multi-threshold segmentation can segment the image into a plurality of different regions, including soil pores and other soil tissue. By using multiple thresholds in combination with the characteristics of the pixel gray values, the porous and non-porous regions can be more accurately partitioned. The embodiment also uses the self-adaptive weight searching strategy to update the initial threshold value set, and the strategy can dynamically adjust the weight of each threshold value according to the effect of the current threshold value set so as to better adapt to the characteristics of different soil images. Through the adjustment of the self-adaptive weight, the effective segmentation of pores and other structures can be maintained, and the loss of other important information in the CT image of the soil can be reduced as much as possible. Further, a differential evolution strategy is used for searching the global, and a local optimal threshold set is updated. Differential evolution is an optimization algorithm that performs a global search through a simulated evolution process to find better threshold combinations. Therefore, the segmentation result can be further optimized, the accurate extraction of the pores is improved, and important information in the original image is reserved. The embodiment also calculates the information entropy of each threshold value in each iteration process, and selects n threshold values with top information entropy ranks as a threshold value set of the next iteration. Information entropy is a measure of the gray value distribution of an image pixel, and can reflect the complexity and information content of the image. The threshold combination with the top information entropy ranking is reserved, so that the segmentation result can be ensured to have higher accuracy, and more original image information can be reserved.
The invention is applied to the field of soil regulation.
Drawings
Fig. 1 is a flowchart of a soil CT image pore extraction method according to an embodiment;
FIG. 2 is a flow chart of an optimal threshold selection scheme for extracting soil CT image void problems according to an eleventh embodiment;
fig. 3 is an input CT image of soil to be segmented according to the eleventh embodiment;
fig. 4 is an image obtained by multi-threshold segmentation of a CT image of soil according to the eleventh embodiment;
fig. 5 is a soil pore map obtained by applying the minimum value of the gray threshold parameter to the image after multi-threshold segmentation according to the eleventh embodiment.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments.
Embodiment one, this embodiment will be described with reference to fig. 1. The embodiment of the method for extracting the soil CT image pores comprises the following steps:
s1: establishing a mathematical model based on multi-threshold soil CT image segmentation, and carrying out image segmentation according to the mathematical model to obtain a segmented image;
S2: constructing a soil CT image segmentation initial threshold set according to the segmented image;
s3: updating the initial threshold set of the soil CT image segmentation according to the self-adaptive weight searching strategy to obtain a local optimal threshold set;
s4: searching the global according to a differential evolution strategy, and updating the local optimal threshold set;
s5: calculating the information entropy of each threshold in the step S3 and the step S4, and reserving the selection schemes of n thresholds with the top information entropy ranks as a threshold set of t+1 iterations;
s6: iterating the steps S3 to S5 until the iteration times are greater than or equal to a preset value, and obtaining a global optimal segmentation threshold set;
s7: and dividing the original soil CT image according to the optimal threshold value in the global optimal division threshold value set to obtain soil pores.
In the embodiment, a mathematical model based on multiple thresholds is first established, and a soil CT image is segmented according to the model. The multi-threshold segmentation can segment the image into a plurality of different regions, including soil pores and other soil tissue. By using multiple thresholds in combination with the characteristics of the pixel gray values, the porous and non-porous regions can be more accurately partitioned. The embodiment also uses the self-adaptive weight searching strategy to update the initial threshold value set, and the strategy can dynamically adjust the weight of each threshold value according to the effect of the current threshold value set so as to better adapt to the characteristics of different soil images. Through the adjustment of the self-adaptive weight, the effective segmentation of pores and other structures can be maintained, and the loss of other important information in the CT image of the soil can be reduced as much as possible. Further, a differential evolution strategy is used for searching the global, and a local optimal threshold set is updated. Differential evolution is an optimization algorithm that performs a global search through a simulated evolution process to find better threshold combinations. Therefore, the segmentation result can be further optimized, the accurate extraction of the pores is improved, and important information in the original image is reserved. The embodiment also calculates the information entropy of each threshold value in each iteration process, and selects n threshold values with top information entropy ranks as a threshold value set of the next iteration. Information entropy is a measure of the gray value distribution of an image pixel, and can reflect the complexity and information content of the image. The threshold combination with the top information entropy ranking is reserved, so that the segmentation result can be ensured to have higher accuracy, and more original image information can be reserved.
According to the method, through the combination of strategies such as multi-threshold segmentation, self-adaptive weight search, differential evolution, information entropy evaluation and the like, the soil pores can be extracted, and meanwhile, the information content contained in the original soil CT image can be well reserved. This provides a more comprehensive and detailed description and analysis of the CT image of the soil while obtaining accurate pore structure. Meanwhile, the multi-threshold segmentation method is adopted, the characteristics of different pixel gray values in the image are comprehensively considered, and the image can be more accurately segmented into different areas, so that the soil pore structure is extracted.
The embodiment introduces an adaptive weight search strategy, and can be dynamically updated according to the performance of the current threshold set. Therefore, the threshold value set can be automatically adjusted in the segmentation process, the segmentation effect is improved, and the method is better suitable for the characteristics and the complexity of different soil images. And searching the global by using a differential evolution strategy, and updating the local optimal threshold set. The global optimization strategy can better explore and utilize information in the image, improves the overall accuracy of segmentation, and is helpful for avoiding the situation of sinking into local optimization. Updating and optimizing the threshold value set in an iterative mode, and selecting the threshold value with the top rank for the next iteration by calculating the information entropy of each threshold value. The automated iteration process can gradually optimize the threshold set until the preset iteration times are reached, and the globally optimal segmentation threshold set is obtained. In the iterative process, the scheme adopts information entropy as an evaluation index, and the selection scheme of n thresholds which are ranked at the top is reserved. The preferred strategy can screen out the threshold combination with better effect, and improve the quality and reliability of the segmentation result.
The soil CT image segmentation method of the embodiment provides an accurate, self-adaptive and global optimization method to obtain a soil pore structure with high visualization and quantification capability.
In a second embodiment, the method for extracting a soil CT image aperture according to the first embodiment is further defined, and the initial threshold set for segmenting the soil CT image in step S2 includes: and the multiple groups of threshold selection schemes comprise different soil image threshold parameters, and the soil image threshold parameters consist of gray value parameters of soil CT scanning images.
The embodiment adopts a plurality of groups of threshold selection schemes, which is helpful for considering the characteristics of CT images of different soils and the diversity of gray value distribution. Because the nature and composition of the soil sample varies, the use of a single threshold value cannot be applied to all images. Therefore, by constructing a plurality of groups of threshold selection schemes, the method can be better suitable for the characteristics of various soil CT images, and the robustness and adaptability of the algorithm are enhanced. The gray value parameter of the soil CT scan image is an essential element constituting the soil image, and is also an important index for image segmentation. These gray value parameters reflect the gray scale characteristics of pixels in different regions and can be used to distinguish between porous and non-porous regions. By using the soil image threshold parameters, the threshold can be set more accurately, and a more accurate and effective segmentation result can be achieved.
According to the embodiment, by using a plurality of groups of threshold selection schemes, personalized threshold selection can be performed according to the characteristics of different images, and the accuracy of the segmentation result is further improved. CT images of different soil samples are different, so that a threshold selection scheme with high adaptability is needed for segmentation, and a pore area and other structural areas are better extracted.
In an actual scene, the gray distribution in the soil CT image may vary greatly due to factors such as soil properties, water content, and the like. According to the method, multiple groups of threshold selection schemes are constructed and gray value parameters are considered, so that the diversity of gray distribution of the image can be fully considered, and the segmentation algorithm has better adaptability and expandability.
In an actual scene, characteristics and quality of a soil CT image may be affected by various factors, such as noise, uneven brightness, and the like. By using a plurality of groups of threshold selection schemes, the robustness of the algorithm can be improved to a certain extent, so that the algorithm has a certain fault tolerance to image quality change, and the stability and reliability of a segmentation result are ensured.
In the embodiment, a plurality of groups of threshold selection schemes are used, and gray value parameters of the soil CT scanning images are used as soil image threshold parameters for segmentation, so that the segmentation accuracy is improved, the diversity of image gray distribution is considered, and the robustness of an algorithm is enhanced. Therefore, the method can be better suitable for the characteristics of CT images of different soils, extracts soil pores and retains the information of original images.
In the third embodiment, the present embodiment is further defined by the soil CT image pore extraction method described in the second embodiment, and the step S3 includes:
calculating an adaptive weight search strategy control factor:
calculating oscillation factor:
Threshold selection scheme in updated threshold set：
，
wherein ,an ith threshold selection scheme representing t threshold searches,>current local optimum representing t threshold searches, < >>Representing a random scheme in a threshold set in T times of threshold search, T representing an index of the current gray value search times, and T representing the total iteration times. />Is [0,1 ]]Random parameters between->Is a random factor->Representing an adaptive weight search policy control factor, +.>Representing a random scaling factor in an adaptive weight search strategy,/->Information entropy indicating the ith threshold selection scheme, < ->For the current optimal information entropy->For worst information entropy->Representing the median value of the entropy of the information in the threshold set, +.>And b is the search radius of the algorithm for carrying out random search in the global. Further, a, b gradually decrease to 0 along with the increase of the iteration times.
In the embodiment, a self-adaptive weight searching strategy is adopted to conduct global searching on the gray value parameters of the soil CT image so as to update the position of the threshold value set in the gray level whole space of the soil image, and the updated current local optimal threshold value set is obtained.
The present embodiment adopts an adaptive weight search strategy to search in the full gray level space of the soil image, and is not limited to a specific range. Therefore, the missing of potential effective threshold parameters can be avoided, the global searching capability of the algorithm is improved, and the separation result is more accurate and comprehensive. The self-adaptive weight search strategy carries out self-adaptive adjustment on the search weight according to the current search result, so that the information in the search space can be more effectively explored and utilized. By calculating the control factors, the search weight can be dynamically adjusted according to the current search condition and the current search result, and more accurate and efficient parameter search is realized. The oscillation factor is calculated to avoid excessive oscillation or sinking into a locally optimal solution during the search process. The concussion factor is calculated, so that the searching process can be regulated and controlled, the explorability of searching and the utilization of a local optimal solution are balanced, and the robustness and the stability of an algorithm are improved. And obtaining the updated current local optimal threshold set through global searching and self-adaptive weight adjustment. The threshold selection scheme through the optimization process can be more suitable for the characteristics of the soil CT image, and the accuracy and stability of segmentation are improved. The updated threshold value set can better distinguish porous areas from non-porous areas, and the quality and effect of soil image segmentation are improved.
According to the method, the self-adaptive weight searching strategy is adopted to search in the whole soil image gray level space, so that potential effective threshold parameters can be explored more comprehensively, the global searching capability of an algorithm is improved, and important information is prevented from being missed due to the limitation of the searching range. The calculation of the adaptive weight control factor may adaptively adjust the search weight based on the current search results. Therefore, the key point and the range of the search can be dynamically adjusted according to the progress of the search, and the search efficiency and the search accuracy are improved. By calculating the oscillation factors, the search process can be controlled, and the situation that excessive oscillation or local optimal solutions are trapped is avoided. Thus, the stability and the robustness of the algorithm can be improved, and the separation result is more reliable and accurate. And obtaining the updated current local optimal threshold set through global searching and self-adaptive adjustment. Therefore, a threshold selection scheme which is more suitable for the characteristics of the CT image of the soil can be obtained, and the segmentation accuracy and stability are improved, so that the soil pore region is segmented better.
According to the embodiment, the self-adaptive weight searching strategy is adopted to conduct global searching on gray value parameters of the soil CT image, and control factors and oscillation factors are calculated to adjust the gray value parameters, so that the purposes of improving global searching capacity, self-adaptively adjusting searching weights, avoiding oscillation and sinking into a local optimal solution and obtaining an updated threshold selection scheme are achieved. Thus, the soil image can be segmented more accurately, and the segmentation quality and effect are improved.
In a fourth embodiment, the present embodiment is further defined by the soil CT image pore extraction method described in the second embodiment, and the step S4 includes:
rearranging the threshold value set randomly;
vacating the position where the current threshold selection scheme is arranged;
randomly generating a scaling factor;
randomly sampling in a gray space of a soil CT image to obtain a gray value intermediate;
setting a boundary for the gray value intermediate;
randomly selecting the dimension of the gray value to be exchanged;
traversing each dimension of the threshold selection scheme, and updating the threshold selection scheme of the new individual according to the conditions of the dimension to be exchanged and the crossover probability.
In order to improve the performance of the self-adaptive weight search strategy and avoid the algorithm from being sunk into a local optimal solution too early, the embodiment introduces a differential evolution search strategy. The present embodiment effectively applies differential evolution search to the problem of CT segmented images of soil.
Specific: the threshold selection schemes are rearranged randomly in order in the threshold set.
The position where the current threshold selection scheme is displaced is emptied (not currently involved in generating the differential intermediate).
Randomly generated scaling factorsThe method is used for adjusting the value range of the gray value midbody. By introducing the scaling factor, the gray value intermediate can be fluctuated within a certain range, and the diversity of searching is increased.
Production of grey value intermediates in the grey value space of soil imagesThis intermediate is a new threshold selection scheme.
In order to prevent the intermediate from exceeding the boundary of the soil gradation value, the intermediate needs to be subjected to boundary restriction. If the intermediate value exceeds the boundary of the gray value, it is adjusted to the boundary value.
A random number is generated, i.e. the number of the dimension of gray values to be exchanged is selected for determining which dimension of gray values is to be exchanged.
Traversing each dimension of the threshold selection scheme if the current dimension is the dimension to be swapped or the random probability is less than the crossover probabilityThe current dimension value of the new individual is equal to the segmentation threshold value of the corresponding dimension of the segmentation threshold intermediate, otherwise, the dimension value of the new threshold selection scheme is equal to the corresponding segmentation threshold value of the current individual。
The method combines the global searching capability of the differential evolution strategy, and updates the local optimal value in the threshold selection scheme to expect to achieve a better threshold selection scheme, thereby improving the effect and performance of the algorithm. The differential evolution strategy enables the generation of a new threshold selection scheme to have certain randomness and diversity through the crossing and mutation operation of a plurality of individuals, thereby being beneficial to global search and avoiding sinking into a local optimal solution.
This embodiment rearranges the threshold set randomly and randomly selects the dimension of the gray values to be exchanged, and samples randomly in the gray space, which increases the randomness and diversity of the algorithm. Therefore, the search space can be explored more comprehensively, the situation that a local optimal solution is trapped is avoided, and the global search capability is improved. And setting a boundary for the gray value intermediate to ensure that the threshold selection scheme is changed within a reasonable range. This can limit excessive variation in the search process, avoid invalid searches and result instability, and improve the robustness of the algorithm. The scaling factor is randomly generated for controlling the search step size. Therefore, the exploration breadth and depth in the searching process can be adjusted, global searching and local searching are balanced, and the searching efficiency of the algorithm and the accuracy of results are improved.
According to the implementation mode, the position where the current threshold selection scheme is arranged is emptied, each dimension of the threshold selection scheme is traversed according to the dimension to be exchanged and the condition of the crossover probability, and the threshold selection scheme of a new individual is updated. The aim of the operation is to obtain new individuals by exchanging and updating each dimension of the threshold selection scheme, further optimize the threshold selection scheme, enable the threshold selection scheme to be more suitable for gray features of the soil CT image, and improve the accuracy of segmentation. The goal is to increase the diversity and randomness of the algorithm by rearranging the set of thresholds randomly, selecting randomly the dimension of the gray values to be exchanged, and sampling randomly in the gray space. Therefore, the search space can be explored more comprehensively, the situation that a local optimal solution is trapped is avoided, and the global search capability of an algorithm is improved. By randomly generating the scaling factor, the size of the search step may be controlled. Therefore, the exploration breadth and depth in the searching process can be adjusted, global searching and local searching are balanced, and the searching efficiency of the algorithm and the accuracy of results are improved.
The implementation increases diversity and randomness, boundary setting and controls search step length; the method aims at updating a threshold selection scheme, improving segmentation accuracy and increasing global searching capability through randomness and diversity. The operations can optimize the threshold selection scheme, so that the threshold selection scheme is more in line with the characteristics of the soil CT image, and the quality and stability of the segmentation result are improved.
In a fifth embodiment, the present embodiment is further defined by the method for extracting a soil CT image aperture according to the fourth embodiment, wherein the step of randomly sampling in a gray space of the soil CT image to obtain a gray value intermediate includes:
wherein ,threshold selection schemes that centralize rearranged thresholds at 1,2,3 positions.
The present embodiment can generate different intermediates by randomly sampling in the gray value space, thereby generating diversified threshold selection schemes. The mathematical model of the process is:
wherein ,threshold selection schemes that centralize rearranged thresholds at 1,2,3 positions.
In the embodiment, different gray value intermediates can be obtained by randomly sampling in a gray value space. The diversity intermediates generated in this way can introduce more diversity in the searching process, and the diversity and the randomness of the algorithm are increased. This helps to avoid trapping in the locally optimal solution and improves the global search capability of the algorithm. The generation of the gray value intermediate is realized by random sampling, and the sampling is uniformly performed on the gray space, so that the whole gray range is covered. Such an operation may enable the threshold selection scheme to explore a wider search space, discover more possibilities, and improve the search results of the algorithm.
The present embodiment generates diversified gray value intermediates in order to generate diversified threshold selection schemes. The multiple threshold selection scheme facilitates segmentation of the image from different angles, capturing different features and information. By introducing diversity, the robustness of segmentation can be improved, and the deviation of the result is reduced, so that a more accurate soil CT image segmentation result is obtained. By randomly sampling in the grey value space and generating different intermediates, the aim is to introduce more randomness in the search of the threshold selection scheme. The method can help the algorithm to jump out of the local optimal solution, explore a wider search space and improve the global search capability of the algorithm. Through global searching, a better threshold selection scheme is found, and the quality of a segmentation result is improved.
In a sixth embodiment, the present embodiment is further defined by the method for extracting a soil CT image aperture according to the second embodiment, and the calculating the information entropy of each threshold in step S5 includes:
…
wherein ,、/>、/>is a segmentation threshold parameter; />Gray level of CT image of soil>Probability distribution of (2); />Entropy of soil image information corresponding to 1 st threshold interval, >The entropy of the soil image information corresponding to the 2 nd threshold interval,entropy of soil image information corresponding to the Kth threshold interval,>cumulative distribution function of soil gray value for 1 st threshold interval,/for the soil gray value>Cumulative distribution function of soil gray value for the 2 nd threshold interval,>accumulating distribution function for soil gray value of Kth threshold value interval, < >>Is the maximum gray level of the soil image.
The soil CT image gray value cumulative distribution function of each category is defined as
…
In a seventh embodiment, the present embodiment is further defined by the soil CT image pore extraction method described in the second embodiment, and the step S6 includes:
，
wherein ,for the optimal threshold set, ++>For maximum information entropy in threshold set, +.>When the threshold value set TH is applied to the soil image, the soil image information entropy corresponding to the z-TH threshold value section is obtained.
An eighth embodiment is a soil CT image aperture extraction system according to the present embodiment, the system including:
the image segmentation unit is used for establishing a mathematical model based on multi-threshold soil CT image segmentation, and carrying out image segmentation according to the mathematical model to obtain a segmented image;
an initial threshold set constructing unit, configured to construct a soil CT image segmentation initial threshold set according to the segmented image;
The self-adaptive weight searching strategy updating unit is used for updating the initial threshold set of the soil CT image segmentation according to the self-adaptive weight searching strategy to obtain a local optimal threshold set;
the differential evolution strategy searching unit is used for searching the global according to the differential evolution strategy and updating the local optimal threshold set;
the computing unit is used for computing the information entropy of each threshold value in the self-adaptive weight search strategy updating unit and the differential evolution strategy searching unit, and reserving the selection schemes of n threshold values with the top ranking of the information entropy as a threshold value set of t+1 iterations;
the iteration unit is used for iterating the self-adaptive weight searching strategy updating unit to the calculating unit until the iteration times are greater than or equal to a preset value, so as to obtain a global optimal segmentation threshold set;
and the soil pore acquisition unit is used for dividing the original soil CT image according to the global optimal dividing threshold set optimal threshold to acquire soil pores.
The computer-readable storage medium according to the ninth embodiment is a computer-readable storage medium storing a computer program for executing the soil CT image aperture extraction method according to any one of the first to seventh embodiments.
The computer device according to the tenth embodiment includes a memory and a processor, the memory storing a computer program, and the processor executes the soil CT image aperture extraction method according to any one of the seventh embodiments when the processor executes the computer program stored in the memory.
Embodiment eleven, this embodiment will be described with reference to fig. 2 to 5. The present embodiment provides a specific example of the soil CT image pore extraction method according to the first embodiment, and is also used for explaining the second embodiment to the seventh embodiment, specifically:
the invention provides a method for extracting soil CT image pores, which can better retain the information content contained in an original soil CT image, and is specific when extracting soil pores
Establishing a mathematical model for multi-threshold soil CT image segmentation, and establishing an optimal segmentation threshold selection objective function for a control target by using the maximum information entropy of the segmented image;
in the experimental process, the soil CT image I is provided with L gray level parameters, and threshold parameters are segmented=The soil CT image I can be divided into +.>Categories, denoted- >，/>. For 8-bit soil CT image, soil gray level +.>Therefore, the value range is [0,255]. Thus, if->Then indicate gray level +.>The pixels in (a) belong to the category->。
Establishing an objective function of optimal image segmentation by taking the maximum information entropy of the segmented image as a control target;
the multi-threshold image segmentation target with the maximized soil CT image information entropy is to find the optimal threshold setMaximizing the entropy of the threshold set information, expressed as
Order theFor the gray level of the gray pixels of a given soil CT image, then +.>。/>For the total number of pixels of the soil CT image I, < >>Gray level histogram of CT image of soil>The gray level of the CT image of the soil is +.>Pixel number of>Gray level of CT image of soil>Probability distribution of (1)
The entropy value of the image information corresponding to each soil CT image threshold value is
…
The soil CT image gray value cumulative distribution function of each category is defined as
…
An initial threshold set is constructed, wherein the initial threshold set comprises a plurality of groups of threshold selection schemes, each group of threshold selection schemes consists of a plurality of different soil image threshold parameters, each soil image threshold parameter consists of gray value parameters of a soil CT scanning image, the soil CT image segmentation threshold set is initialized based on constraint conditions, and parameters used for updating the threshold set subsequently are initialized specifically as follows:
S2.1: parameter initialization, determining the maximum iteration times in the methodNumber of threshold selection schemes in threshold set +.>Wherein the number of threshold schemes in the threshold set should be chosen appropriately so that the random initialization can cover the entire search space and is therefore set to 40 in the study. Determining the upper bound of threshold parameters of each multidimensional soil image based on constraint conditions>Lower bound->Each threshold selection scheme consists of one +.>Vector constitution of dimensions. Initializing adaptive weight search strategiesGlobal parameter->Scaling factor in differential search strategy>Upper and lower bounds and crossover probability of->. And calculates the information entropy of each threshold selection scheme.
S2.2: an initial threshold set is constructed, wherein the initial threshold set comprises a plurality of groups of threshold selection schemes, each group of threshold selection schemes consists of a plurality of different soil image threshold parameters, each soil image threshold parameter consists of gray value parameters of a soil CT scanning image, the soil CT image segmentation threshold set is initialized based on constraint conditions, and parameters to be used in the process of updating the threshold set subsequently are initialized. The threshold selection scheme in the threshold set X is randomly initialized in the upper and lower bounds, and the initialization formula is as follows:
wherein ,is->The j-th dimension threshold of the threshold selection scheme, < >>Is->Person->Lower bound of soil image threshold of dimension, +.>Is->Person->Upper bound of soil image threshold of dimension, +.>The gray level threshold random scaling factor of the CT image of the uniformly distributed soil is greater than or equal to 0 and less than 1, N is the number of threshold selection schemes, and D is the total category number of the segmented image.
(3) Updating the soil CT image segmentation threshold set by utilizing a self-adaptive weight searching strategy, searching threshold parameters near the maximum threshold of the information entropy in the current threshold set, and performing global searching on the threshold parameters to update the position of the segmentation threshold set in the whole threshold parameter solution space to obtain an updated threshold set, and executing a boundary replacement strategy on the updated threshold set to obtain the current local optimal threshold set, wherein the method is specifically shown in fig. 2:
s3.1: and performing global search on the gray value parameters of the soil CT image by adopting a self-adaptive weight search strategy so as to update the position of the threshold value set in the gray level full space of the soil image and obtain an updated current local optimal threshold value set.
Calculating an adaptive weight search strategy control factor according to formula (1):
(1)
Calculating the oscillation factor according to formulas (2) to (5) :
(2)
(3)
(4)
(5)
Threshold selection scheme in updated threshold set according to equation (6)：/>
(6)
wherein ,an ith threshold selection scheme representing t threshold searches,>current local optimum representing t threshold searches, < >>Representing a random scheme in a threshold set in T times of threshold search, T representing an index of the current gray value search times, and T representing the total iteration times. />Is [0,1 ]]Random parameters between->Is a random factor->Representing an adaptive weight search policy control factor, +.>Representing a random scaling factor in an adaptive weight search strategy,/->Information entropy indicating the ith threshold selection scheme, < ->Representing the current optimal information entropy and worst information entropy, < ->Representing the median value of the entropy of the information in the threshold set.
And S3.2, calculating the information entropy value of each threshold selection scheme in the updated threshold set, and setting the threshold selection scheme position with the minimum current information entropy value as the optimal threshold selection scheme. And judging the updated gray level thresholds one by adopting a boundary replacement strategy, marking the gray level exceeding the boundary in the threshold selection scheme, and replacing the marked gray level with the threshold selection scheme with the maximum current information entropy.
Searching the global by utilizing a differential evolution strategy, updating a soil CT image segmentation threshold set to obtain a current local optimal threshold set, respectively calculating the information entropy value of each threshold selection scheme obtained by the self-adaptive weight search strategy and the differential evolution search strategy, and iteratively reserving the threshold selection scheme with the information entropy ranking at the front, and iterating the steps (3) and (4) until the iteration times are larger than or equal to a preset value, and stopping preferred selection to obtain the global optimal segmentation threshold set;
Executing a differential evolution search strategy on the threshold set to obtain a current local optimal threshold set, and executing the following steps for each scheme in the threshold set:
the threshold selection schemes are rearranged randomly in order in the threshold set.
The position where the current threshold selection scheme is displaced is emptied (not currently involved in generating the differential intermediate).
Randomly generated scaling factorsThe method is used for adjusting the value range of the gray value midbody. By introducing a scaling factor, the gray value can be madeThe intermediate fluctuates in a certain range, and the diversity of searching is increased.
Production of grey value intermediates in the grey value space of soil imagesThis intermediate is a new threshold selection scheme. By randomly sampling in the grey value space, different intermediates can be generated, resulting in a diversified threshold selection scheme. The mathematical model of the process is:
(7)
wherein ,threshold selection schemes that centralize rearranged thresholds at 1,2,3 positions.
In order to prevent the intermediate from exceeding the boundary of the soil gradation value, the intermediate needs to be subjected to boundary restriction. If the intermediate value exceeds the boundary of the gray value, it is adjusted to the boundary value.
A random number is generated, i.e. the number of the dimension of gray values to be exchanged is selected for determining which dimension of gray values is to be exchanged.
Traversing each dimension of the threshold selection scheme if the current dimension is the dimension to be swapped or the random probability is less than the crossover probabilityAnd if the new individual current dimension value is equal to the segmentation threshold value of the corresponding dimension of the segmentation threshold intermediate, otherwise, the new threshold selection scheme dimension value is equal to the corresponding segmentation threshold value of the current individual.
Judging the information entropy of the updated threshold selection schemeWhether or not to be better than the threshold selection scheme before updating +.>For selecting information entropyThe better valued threshold selection scheme serves as the threshold selection scheme to be used at the next iteration. If the information entropy of the updated threshold selection scheme is better, a new threshold selection scheme is +.>Set to gray value intermediate->The method comprises the steps of carrying out a first treatment on the surface of the Otherwise, a new threshold selection scheme +.>Set to the current threshold selection scheme->。
The information entropy of the soil CT segmented image is used as an objective function, the information entropy value of each threshold selection scheme obtained through the self-adaptive weight search strategy and the differential evolution search strategy is calculated respectively, and N threshold selection schemes with the top information entropy rank are reserved in a competitive mode and are used as a threshold set of t+1 iterations.
And (3) continuing to iterate the steps (3) and (4) until the preset value is reached, and stopping preferred selection to obtain the global optimal segmentation threshold set.
The minimum value of the soil image threshold parameter in the global optimal segmentation threshold set is the optimal threshold in the pore segmentation process, and the threshold is applied to the original soil CT image, so that the pores of the soil can be extracted while the information quantity of the original soil CT image is reserved to the maximum extent.
Fig. 3, 4 and 5 show effect diagrams of processing images by applying the soil CT image pore extraction method provided by the present invention. Wherein, fig. 3 shows an input soil CT image to be segmented, and fig. 4 shows an image obtained by multi-threshold segmentation of the soil CT image; fig. 5 is a graph of soil porosity obtained by applying a minimum value of gray threshold parameters to an image after multi-threshold segmentation. As can be seen from fig. 3 to fig. 5, after the soil CT image is segmented by the multi-threshold image, the image can still better retain the information of the original soil CT image, and on this basis, the pore structure in the original soil CT image can be better restored by using the soil pore obtained by the minimum value of the gray threshold parameter for the soil CT image.
The embodiment provides a soil CT image threshold segmentation algorithm based on a self-adaptive weight search strategy, a differential evolution search strategy and a boundary replacement strategy, wherein the differential evolution search strategy encourages information exchange between threshold selection schemes, enriches diversity of threshold selection schemes in a threshold set, and improves the searching degree of a threshold parameter solution space; the self-adaptive weight searching strategy and the boundary replacing strategy enable the algorithm to converge more quickly, a better threshold selection scheme is provided for the whole algorithm, stability of the algorithm is improved, and the capability of solving the problem of extracting the soil CT image pores is improved.
While the preferred embodiments of the present disclosure have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the disclosure.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present disclosure without departing from the spirit or scope of the disclosure. Thus, the present disclosure is intended to include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
It will be appreciated by those skilled in the art that embodiments of the present disclosure may be provided as a method, system, or computer program product. Accordingly, the present disclosure may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present disclosure may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
The present disclosure is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the disclosure. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks. These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Finally, it should be noted that: the above embodiments are only for illustrating the technical aspects of the present disclosure and not for limiting the scope thereof, and although the present disclosure has been described in detail with reference to the above embodiments, it should be understood by those of ordinary skill in the art that: various alterations, modifications, and equivalents may be suggested to the specific embodiments of the invention, which would occur to persons skilled in the art upon reading the disclosure, are intended to be within the scope of the appended claims.
Claims (10)
1. A method for extracting soil CT image pores, the method comprising:
s1: establishing a mathematical model based on multi-threshold soil CT image segmentation, and carrying out image segmentation according to the mathematical model to obtain a segmented image;
S2: constructing a soil CT image segmentation initial threshold set according to the segmented image;
s3: updating the initial threshold set of the soil CT image segmentation according to the self-adaptive weight searching strategy to obtain a local optimal threshold set;
s4: searching the global according to a differential evolution strategy, and updating the local optimal threshold set;
s5: calculating the information entropy of each threshold in the step S3 and the step S4, and reserving the selection schemes of n thresholds with the top information entropy ranks as a threshold set of t+1 iterations;
s6: iterating the steps S3 to S5 until the iteration times are greater than or equal to a preset value, and obtaining a global optimal segmentation threshold set;
s7: and dividing the original soil CT image according to the optimal threshold value in the global optimal division threshold value set to obtain soil pores.
2. The method according to claim 1, wherein the initial set of threshold values for the segmentation of the CT image in step S2 comprises: and the multiple groups of threshold selection schemes comprise different soil image threshold parameters, and the soil image threshold parameters consist of gray value parameters of soil CT scanning images.
3. The method for extracting voids from a CT image of soil according to claim 2, wherein said step S3 comprises:
calculating an adaptive weight search strategy control factor:
calculating oscillation factor:
Threshold selection scheme in updated threshold set：
，
wherein ,an ith threshold selection scheme representing t threshold searches,>current local optimum representing t threshold searches, < >>Representing a random scheme in a threshold set in a T-time threshold search, T representing an index of the number of current gray value searches, T representing the total number of iterations,/>Is [0,1 ]]Random parameters between->Is a random factor->Representing an adaptive weight search policy control factor, +.>Representing a random scaling factor in an adaptive weight search strategy,/->Information entropy indicating the ith threshold selection scheme, < ->For the current optimal information entropy->For worst information entropy->Representing the median value of the entropy of the information in the threshold set, +.>And b is the search radius of the algorithm for carrying out random search in the global.
4. The method for extracting voids from a CT image of soil according to claim 2, wherein said step S4 comprises:
Rearranging the threshold value set randomly;
vacating the position where the current threshold selection scheme is arranged;
randomly generating a scaling factor;
randomly sampling in a gray space of a soil CT image to obtain a gray value intermediate;
setting a boundary for the gray value intermediate;
randomly selecting the dimension of the gray value to be exchanged;
traversing each dimension of the threshold selection scheme, and updating the threshold selection scheme of the new individual according to the conditions of the dimension to be exchanged and the crossover probability.
5. The method for extracting voids from a soil CT image according to claim 4, wherein said randomly sampling in a gray scale space of the soil CT image to obtain a gray scale value intermediate comprises:
wherein ,threshold selection schemes that centralize rearranged thresholds at 1,2,3 positions.
6. The method according to claim 2, wherein calculating the entropy of each threshold in step S5 comprises:
…
wherein ,、/>、/>is a segmentation threshold parameter; />Gray level of CT image of soil>Probability distribution of (2);entropy of soil image information corresponding to 1 st threshold interval,>entropy of soil image information corresponding to the 2 nd threshold interval,>entropy of soil image information corresponding to the Kth threshold interval, >Cumulative distribution function of soil gray value for 1 st threshold interval,/for the soil gray value>A distribution function is accumulated for the soil gradation value of the 2 nd threshold interval,accumulating distribution function for soil gray value of Kth threshold value interval, < >>Is the maximum gray level of the soil image.
7. The method for extracting voids from a CT image of soil according to claim 2, wherein said step S6 comprises:
，
wherein ,for the optimal threshold set, ++>For maximum information entropy in threshold set, +.>When the threshold value set TH is applied to the soil image, the soil image information entropy corresponding to the z-TH threshold value section is obtained.
8. A soil CT image aperture extraction system, the system comprising:
the image segmentation unit is used for establishing a mathematical model based on multi-threshold soil CT image segmentation, and carrying out image segmentation according to the mathematical model to obtain a segmented image;
an initial threshold set constructing unit, configured to construct a soil CT image segmentation initial threshold set according to the segmented image;
the self-adaptive weight searching strategy updating unit is used for updating the initial threshold set of the soil CT image segmentation according to the self-adaptive weight searching strategy to obtain a local optimal threshold set;
the differential evolution strategy searching unit is used for searching the global according to the differential evolution strategy and updating the local optimal threshold set;
The computing unit is used for computing the information entropy of each threshold value in the self-adaptive weight search strategy updating unit and the differential evolution strategy searching unit, and reserving the selection schemes of n threshold values with the top ranking of the information entropy as a threshold value set of t+1 iterations;
the iteration unit is used for iterating the self-adaptive weight searching strategy updating unit to the calculating unit until the iteration times are greater than or equal to a preset value, so as to obtain a global optimal segmentation threshold set;
and the soil pore acquisition unit is used for dividing the original soil CT image according to the global optimal dividing threshold set optimal threshold to acquire soil pores.
9. A computer readable storage medium for storing a computer program for executing a soil CT image void extraction method according to any one of claims 1 to 7.
10. A computer device, characterized by: comprising a memory and a processor, said memory having stored therein a computer program, which when executed by said processor performs a soil CT image aperture extraction method according to any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311159455.3A CN116894851B (en) | 2023-09-11 | 2023-09-11 | Soil CT image pore extraction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311159455.3A CN116894851B (en) | 2023-09-11 | 2023-09-11 | Soil CT image pore extraction method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116894851A true CN116894851A (en) | 2023-10-17 |
| CN116894851B CN116894851B (en) | 2023-11-14 |
Family
ID=88312360
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311159455.3A Active CN116894851B (en) | 2023-09-11 | 2023-09-11 | Soil CT image pore extraction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116894851B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106546521A (en) * | 2016-10-12 | 2017-03-29 | 北京师范大学 | A kind of method that soil macropore spacial framework is quantified based on CT scan technology |
| CN110827306A (en) * | 2019-10-18 | 2020-02-21 | 重庆师范大学 | Purple soil image segmentation and extraction method based on self-adaptive density peak value clustering |
| CN110910347A (en) * | 2019-10-18 | 2020-03-24 | 宁波大学 | Image segmentation-based tone mapping image no-reference quality evaluation method |
| CN111932556A (en) * | 2020-08-11 | 2020-11-13 | 中国科学院微小卫星创新研究院 | Multilevel threshold image segmentation method |
| CN112686914A (en) * | 2021-02-04 | 2021-04-20 | 长春光华学院 | Plant canopy image segmentation algorithm based on improved symbiont search algorithm |
| CN112884716A (en) * | 2021-01-28 | 2021-06-01 | 中国空气动力研究与发展中心超高速空气动力研究所 | Method for strengthening characteristics of ultra-high-speed impact damage area |
-
2023
- 2023-09-11 CN CN202311159455.3A patent/CN116894851B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106546521A (en) * | 2016-10-12 | 2017-03-29 | 北京师范大学 | A kind of method that soil macropore spacial framework is quantified based on CT scan technology |
| CN110827306A (en) * | 2019-10-18 | 2020-02-21 | 重庆师范大学 | Purple soil image segmentation and extraction method based on self-adaptive density peak value clustering |
| CN110910347A (en) * | 2019-10-18 | 2020-03-24 | 宁波大学 | Image segmentation-based tone mapping image no-reference quality evaluation method |
| CN111932556A (en) * | 2020-08-11 | 2020-11-13 | 中国科学院微小卫星创新研究院 | Multilevel threshold image segmentation method |
| CN112884716A (en) * | 2021-01-28 | 2021-06-01 | 中国空气动力研究与发展中心超高速空气动力研究所 | Method for strengthening characteristics of ultra-high-speed impact damage area |
| CN112686914A (en) * | 2021-02-04 | 2021-04-20 | 长春光华学院 | Plant canopy image segmentation algorithm based on improved symbiont search algorithm |
Non-Patent Citations (4)
| Title |
|---|
| ERIK CUEVAS ET AL.: "A novel multi-threshold segmentation approach based on differential evolution optimization", EXPERT SYSTEMS WITH APPLICATIONS, pages 5265 * |
| 孙路 等: "基于信息熵的图像分割阈值迭代改进算法", 计算机应用与软件, vol. 25, no. 10, pages 225 - 226 * |
| 李鑫 等: "基于X-ray CT的古土壤孔裂隙识别与表征", 水土保持通报, vol. 38, no. 6, pages 224 - 230 * |
| 马冬梅 等: "一种差分进化二维熵的KFCM图像分割算法", 西北师范大学学报（自然科学版）, vol. 54, no. 4, pages 47 - 51 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116894851B (en) | 2023-11-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109509197B (en) | Method, device and equipment for dividing region of interest and storage medium | |
| CN105118090B (en) | A kind of point cloud filtering method of adaptive complicated landform structure | |
| CN105335965B (en) | Multi-scale self-adaptive decision fusion segmentation method for high-resolution remote sensing image | |
| CN115641327B (en) | Building engineering quality supervision and early warning system based on big data | |
| CN114724022B (en) | Method, system and medium for detecting farmed fish shoal by fusing SKNet and YOLOv5 | |
| CN109472801A (en) | It is a kind of for multiple dimensioned neuromorphic detection and dividing method | |
| CN113536697A (en) | Bearing residual life prediction method based on improved residual error network and WGAN | |
| CN114022368A (en) | Pavement disease data enhancement method based on generation of countermeasure network | |
| CN113837086A (en) | Reservoir phishing person detection method based on deep convolutional neural network | |
| CN116894851B (en) | Soil CT image pore extraction method and system | |
| CN114299490B (en) | Tumor microenvironment heterogeneity evaluation method | |
| CN117592662B (en) | Ecological fishery resource evaluation system based on data analysis | |
| CN110517766A (en) | Identify the method and device of encephalatrophy | |
| CN110517271A (en) | A kind of image level collection dividing method of prior shape constraint | |
| CN115222139A (en) | Method and device for predicting concrete consumption of transmission line tower engineering | |
| CN115408929A (en) | Data processing method and device for tunnel driving construction prediction | |
| CN111626256A (en) | High-precision diatom detection and identification method and system based on scanning electron microscope image | |
| CN110742687A (en) | Method for non-invasive prediction of growth rate of pituitary adenoma tumor | |
| CN115600749B (en) | Groundwater level prediction method and device and electronic equipment | |
| CN111862003B (en) | Medical image target information acquisition method, device, equipment and storage medium | |
| CN116226623B (en) | Mark layer division method and device based on SegNet segmentation model and computer equipment | |
| Ding et al. | An object-based method for Chinese landform types classification | |
| CN109102517A (en) | A kind of method of Image Edge-Detection, system and associated component | |
| CN114863168B (en) | Remote sensing image classification method, device and equipment of self-adaptive recursive incremental model | |
| CN117253238B (en) | Training sample determination, neural network training and layout optimization method and device |
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 |