A visual rapid detection method for hydraulic fractures in underground boreholes

Abstract

The present invention discloses a method for rapid visualization detection of underground borehole hydraulic fractures. According to a new design for rapid visualization detection of coal and rock hydraulic fractures, borehole television testing, video recording and automatic fracture measurement and evaluation are combined, so that not only the shape and size of the hydraulic fractures on the borehole wall can be intuitively seen, but also the position of the hydraulic fractures in the borehole wall and the fracture expansion morphology can be automatically located, forming a borehole hydraulic fracture detection and evaluation system, forming an innovative ultra-long hole hydraulic fracture rapid detection method, breaking through the technical defects of the current borehole hydraulic fracture ink method and simple peeping, and solving the difficult problem of evaluating and measuring the expansion of roof borehole hydraulic fractures.

Description

Technical Field

[0001] The invention relates to the field of crack detection, and in particular to a method for rapid visualization of hydraulic crack detection in underground boreholes. Background Art

[0002] Borehole observation is a fast, intuitive and effective way to test the effect of hydraulic fractures, and is the development direction of roof fracturing management and evaluation technology. With the large-scale application of coal rock hydraulic fracturing technology in coal mine safety management and mining technology, new and higher requirements have been put forward for the visualization, informatization, and theoretical progress of image processing technology of fracturing effects. There are insurmountable difficulties in the detection of hydraulic fractures in boreholes by the ink method, such as environmental constraints, inability to see hydraulic fractures, inaccurate measurement of borehole inclination and depth, and continuous distribution of visual crack extension. In particular, there is a lack of integrated theory and method for existing borehole peek and hydraulic fracture detection, low level of visualization and automation detection technology, and backward technology, which seriously affect the detection and rapid evaluation of the safety effect of roof fracturing. Based on this, a visualization detection method for underground borehole hydraulic fractures is proposed. It is aimed at realizing the dynamic observation of borehole fractures and the dynamic analysis of the expansion of underground in-situ borehole hydraulic fractures in complex environments. On the basis of borehole fracture photography, an innovative three-dimensional space transformation method is adopted. The three-dimensional space hydraulic fracture parameters can be transformed into two-dimensional graphic parameters through discretization, and the spatial position of dynamic hydraulic fractures, including length, orientation and angle, can be automatically detected to form a hydraulic fracture expansion map, which replaces the traditional ink pad method and breaks through the technical bottleneck of borehole peeping observation in hydraulic fracture analysis.

[0003] Compared with existing technologies:

[0004] (1) Compared with the ink pad method

[0005] The traditional ink method cannot intuitively detect the true position of the borehole hydraulic fractures on the hole wall and the crack trajectory, expansion, dynamic continuity and inclination, and is greatly affected by environmental factors, which brings difficulties to the rapid detection and continuity evaluation of coal rock hydraulic fractures.

[0006] (2) Compared with drilling TV

[0007] The current borehole television can visually see the shape and size of the hydraulic cracks in the borehole wall, but it is difficult to determine the specific location distribution of the hydraulic cracks on the borehole wall, and it is difficult to dynamically locate and describe the spatial position of the hydraulic cracks. Summary of the invention

[0008] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method for rapid visualization detection of hydraulic fractures in downhole boreholes.

[0009] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is:

[0010] A method for rapid visualization detection of hydraulic fractures in underground boreholes, comprising the following steps:

[0011] S1. Use a drilling probe to conduct a panoramic observation of the inside of the borehole to obtain a drilling image;

[0012] S2, preprocessing the borehole image obtained in step S1, increasing the number of horizontal and vertical sampling points of the borehole image, thereby improving the overall image resolution;

[0013] S3, using a local statistical variable threshold segmentation method to segment the borehole image preprocessed in step S2 to obtain a binary image, and performing morphological extraction on the binary image to obtain edge features of the crack;

[0014] S4, identifying the average depth position of the stratum where the fracture is located according to the edge characteristics of the fracture, and using Hough transform to convert the clean binary image after edge detection into a discrete two-dimensional parameter space;

[0015] S5. Different types of cracks are classified according to parameter attributes in a discrete two-dimensional parameter space.

[0016] Furthermore, the local statistical variable threshold segmentation method in S3 is specifically:

[0017] S301, define a variable local threshold using the local standard deviation and local mean of any point (x, y) in the preprocessed drilling image, expressed as:

[0018] T x，y =aσ xy +bμ xy

[0019] Among them, σ xy and μ xy are the neighborhood N centered at (x,y) in the image. xy The standard deviation and mean of the pixel set included, a and b are non-negative constants; if the global mean μ is used T ,but

[0020] T x，y =aσ xy +bμ T

[0021] S302, through the neighborhood N centered at point (x, y) xy The local standard deviation and variance of all pixels in the preprocessed drilling image are calculated to obtain the local feature combination;

[0022] S303, combining local features by using AND logic, and calculating local thresholds by weights to obtain a binary image, which is expressed as:

[0023]

[0024] Among them, I b (x, y) is a binary image, I(x, y) is the preprocessed drilling image, and μ is the local mean μ xy Or the global mean μ T .

[0025] Furthermore, the specific method of performing morphological extraction on the binary image in S3 to obtain edge features of the crack is:

[0026] S311, the binary image obtained in step S303 is processed by morphological operators to obtain a clean binary image, which is expressed as:

[0027]

[0028] Among them, I cb is a clean binary image, I b is a binary image, B is a structural element;

[0029] S312, using edge detection to eliminate all non-edge sets, retaining the crack edge set that crosses the entire image, expressed as:

[0030]

[0031] Among them, E is the final edge image.

[0032] Furthermore, the specific method of identifying the average depth position of the formation where the crack is located according to the edge characteristics of the crack in S4 is:

[0033] S401, construct a complete cycle crack curve in the plane image I (i, j), expressed as:

[0034]

[0035] Where A is the amplitude, is the initial phase, d is the average depth position of the crack;

[0036] S402: Determine two pixel points I that are half a period apart based on the symmetry relationship of the sine curve. 0 (i, j) and I 1 The relationship between the (h, k) ordinates is expressed as:

[0037] |jk|=T / 2

[0038] S403, traverse the plane image I (i, j), match all the vertical coordinate edge pixels in the image that meet the vertical coordinate relationship determined in step S402 into point pairs, and calculate the pixel point I 0 (i, j) and I 1 The horizontal coordinate of the key point M of the (h, k) connecting line is used to obtain the average depth position of the crack, which is expressed as:

[0039] d=(i+h) / 2.

[0040] Furthermore, the specific method of using Hough transform in S4 to convert the clean binary image after edge detection into a discrete two-dimensional parameter space is:

[0041]

[0042] Among them, A∈(0,A max ), A max is the diagonal length of the drilling image, and d is the average depth of the crack.

[0043] Furthermore, the S5 specifically includes the following steps:

[0044] S51, for any point (i, j) on the drilling image, the amplitude of the crack is calculated, and as long as A satisfies the parameter range, the corresponding accumulator unit contribution is increased by one;

[0045] S52, in Apply maximum value detection in space to find local peaks and select possible candidate crack curves;

[0046] S53. Apply a significance verification process to suppress erroneous curves and select the most significant curve among the candidate crack curves.

[0047] The present invention has the following beneficial effects:

[0048] This method is designed according to the new visualization rapid detection of coal and rock hydraulic fractures, and combines borehole television testing, video recording and automatic crack measurement and evaluation. It can not only intuitively see the shape and size of the hydraulic fractures on the borehole wall, but also automatically locate the position of the hydraulic fractures in the borehole wall and the crack expansion morphology, forming a borehole hydraulic fracture detection and evaluation system, forming an innovative ultra-long hole hydraulic fracture rapid detection method, breaking through the current technical defects of the borehole hydraulic fracture ink method and simple peeping, and solving the evaluation and measurement problems of the roof borehole hydraulic fracture expansion. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the process of the method for rapid visualization of hydraulic fractures in downhole boreholes of the present invention.

[0050] Figure 2This is a schematic diagram of the intersection of a hydraulic fracture and a cylindrical borehole according to an embodiment of the present invention.

[0051] Figure 3 Schematic diagram of complete and partial intersection of a borehole and a crack according to an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0053] A method for rapid visualization of hydraulic fracture detection in underground boreholes, such as Figure 1 As shown, the following steps are included:

[0054] S1. Use a drilling probe to conduct a panoramic observation of the inside of the borehole to obtain a drilling image;

[0055] In this embodiment, the borehole observation head and the host are first connected together, and then the host display is turned on. Then the borehole observation head and the pneumatic push rod are connected together, and the borehole observation head is sent into the borehole, and the recording function is turned on. The air pump control system is controlled to allow the air pump to inflate the pneumatic lifting and pushing rod, so that the detection head can penetrate into the borehole for panoramic observation of the borehole. The automatic measurement module and the automatic positioning memory detection module on the observation head are installed with a large number of sensors in the equipment, which can realize automatic positioning of the crack position in the hole wall. The length, size, shape, expansion angle and direction of the crack are positioned and memorized, and the position distribution of the water pressure crack is determined based on the center line of the water flow on the hole wall until the bottom of the hole is reached. The pneumatic lifting and pushing rod is deflated by controlling the air pump control system, so that the detection head slowly withdraws from the borehole to complete the observation, and then the recording is turned off.

[0056] S2, preprocessing the borehole image obtained in step S1, increasing the number of horizontal and vertical sampling points of the borehole image, thereby improving the overall image resolution;

[0057] In this embodiment, a bilinear interpolation method is applied to preprocess the borehole image to increase the number of horizontal and vertical sampling points of the borehole image and thus improve the overall image resolution.

[0058] S3, using a local statistical variable threshold segmentation method to segment the borehole image preprocessed in step S2 to obtain a binary image, and performing morphological extraction on the binary image to obtain edge features of the crack;

[0059] Mathematical morphology is used to identify the crack patterns in borehole images. First, the borehole amplitude image is modeled. The natural cracks intersecting the wellbore are shown as single-cycle sinusoidal pattern features in the two-dimensional image after being unfolded along the plane. For borehole images with less noise in the background, the traditional global threshold segmentation and edge detection technology are used for edge extraction; however, for borehole images with more complex textures, the variable threshold segmentation method based on local statistics is first used to segment the borehole image. Secondly, morphological processing is applied to the segmented binary image, which effectively suppresses the noise in the background and identifies the crack features. At the same time, morphology is used to extract the main edge features of the cracks.

[0060] like Figure 2 , Figure 3 To track these intersection curves in the borehole image, we must find the unit normal vector of the given plane event. The shape of the intersection curve is obtained under the condition of . Normal vectors are usually parameterized using two angles, the inclination angle θ d and azimuth

[0061]

[0062] Point in a plane Passing through the origin, the following equation must be satisfied:

[0063]

[0064] To represent a plane at any depth, the point Must be given an offset: where α is the distance from the origin to the plane in the same direction. In this plane, the following equations are satisfied:

[0065]

[0066] Assumptions On the well wall, it is expressed by the following formula:

[0067]

[0068] in and z are independent coordinates on the wellbore wall, and R is the wellbore radius. When the plane intersects the wellbore wall, it satisfies, Combining equations (3) and (4), we can see that the curve generated at this time is as follows:

[0069]

[0070] This curve can also be written in parametric form as the following sine curve:

[0071]

[0072] in,

[0073] d=α / n z (7)

[0074]

[0075]

[0076] We call A the amplitude, is the initial phase, and d is the average depth of the crack. They are related to the inclination angle θ d The relationship is given by the following formula:

[0077] θ d =arctan(A / R) (10)

[0078] The amplitude A and phase of the sine wave are They respectively determine the inclination and strike (azimuth) of the crack plane.

[0079] Therefore, we need to detect such sinusoids in the drilling image and evaluate their parameters A, and d.

[0080] For the threshold segmentation process, first convert the original borehole image into a grayscale image, and then use the threshold value represented by the feature of interest in the image to convert the grayscale image into a binary image (black and white). Sometimes the noise and non-uniform illumination in the image will have a certain impact on the image segmentation. In this case, using the global threshold to segment the image often fails and does not achieve the ideal segmentation effect. At this time, we apply the variable threshold method based on local statistics to segment the borehole image. Therefore, the variable threshold segmentation method based on local statistics is specifically:

[0081] S301, define a variable local threshold using the local standard deviation and local mean of any point (x, y) in the preprocessed drilling image, expressed as:

[0082] T x，y =aσ xy +bμ xy (11)

[0083] Among them, σ xy and μ xy are the neighborhood N centered at (x, y) in the image. xy The standard deviation and mean of the pixel set included, a and b are non-negative constants; if the global mean μ is used T ,but

[0084] Tx，y =aσ xy +bμ T (12)

[0085] The local threshold segmentation method is to calculate the local standard deviation and local mean of pixels in a certain size neighborhood around any pixel in an image, and use the linear combination of the two to represent the local threshold of the image. Let σ xy and μ xy Respectively represent the neighborhood N centered at (x, y) in an image xy Contains the standard deviation and mean of a set of pixels, which are used to describe the local contrast and average grayscale of the image, respectively.

[0086] S302, through the neighborhood N centered at point (x, y) xy The local standard deviation and variance of all pixels in the preprocessed drilling image are calculated to obtain the local feature combination;

[0087] The image after segmentation using a variable threshold is:

[0088]

[0089] Where I(x, y) is the input image. xy The local standard deviation and mean are calculated for all pixels (x, y) in the input image.

[0090] S303, combining local features by using AND logic, and calculating local thresholds by weights to obtain a binary image, which is expressed as:

[0091]

[0092] Among them, I b (x, y) is a binary image, I(x, y) is the preprocessed drilling image, and μ is the local mean μ xy Or the global mean μ T .

[0093] Binary image I b The morphological operator shown in the following formula (15) can be used for processing. The operator is constructed by a structure element B of size 3×3. This morphological operation can maintain a continuous set of active pixels in a large range, and similar to a cleaning filter, it can also eliminate a continuous set of active pixels in a small range. Finally, a clean binary image can be obtained after processing using the morphological operator.

[0094]

[0095] Among them, I cbis a clean binary image.

[0096] The crack edge is understood as a set of continuous active pixels that cross the entire image horizontally. And if the crack has a certain width, it can also be identified from the position of its top and bottom edges. Here we propose an edge detection tool that can be represented by mathematical morphological operations:

[0097]

[0098] Among them, E is the final edge image, which only contains the clean binary image I cb In this way, we eliminate all discontinuous or small sets (edges) and only keep the set of crack edges that cross the entire image horizontally.

[0099] S4, identifying the average depth position of the stratum where the fracture is located according to the edge characteristics of the fracture, and using Hough transform to convert the clean binary image after edge detection into a discrete two-dimensional parameter space;

[0100] First, the statistical distribution method is used to estimate the location of the crack depth d, including the following aspects:

[0101] S401, construct a complete cycle crack curve in the plane image I (i, j), and the depth position estimation of the crack is realized by means of a statistical distribution histogram. Assuming that there is exactly one complete cycle crack curve in the image plane I (i, j), the mathematical equation can be described as:

[0102]

[0103] Where A is the amplitude, is the initial phase, in radians, and d is the average depth of the crack. (i, j) is any pixel position in the image. Note that the units of A and d are pixels. For a single-cycle sine curve, its period T is consistent with the width of the drilling image, that is, the following formula:

[0104]

[0105] Where W is the width of the drill hole image.

[0106] S402: Determine two pixel points I that are half a period apart based on the symmetry relationship of the sine curve. 0 (i, j) and I 1 The relationship between the ordinates (h, k) is based on the symmetry of the sine curve. 0 (i, j) and I 1 The (h, k) ordinates satisfy the following relationship:

[0107] |jk|=T / 2 (19)

[0108] S403, traverse the plane image I (i, j), match all the vertical coordinate edge pixels in the image that meet the vertical coordinate relationship determined in step S402 into point pairs, and calculate the pixel point I 0 (i, j) and I 1 The horizontal coordinate of the key point M of the (h, k) connecting line is used to obtain the average depth position of the crack, which is expressed as:

[0109] d=(i+h) / 2 (20)

[0110] The horizontal coordinate of the midpoint M is exactly the average depth position d where the crack is located. By counting the local peak position of the accumulator at the depth position, the depth value parameter of the crack can be obtained.

[0111] When the depth d of the sinusoidal mode is determined, only two free parameters remain: amplitude A and initial phase. According to formula (17), we have the following form:

[0112]

[0113] If A∈(0, A max ), A max is the diagonal length of the image, but is usually no greater than the height of the image. The unit of is radians, and For any point (i, j) on the borehole image, the amplitude of the crack is calculated. As long as A satisfies the parameter range, the corresponding accumulator unit contribution is increased by one. A) Apply maximum detection in space to find local peaks and select possible candidate crack curves.

[0114] S5. Classify different types of cracks according to parameter attributes in the discrete two-dimensional parameter space. Specifically, the steps include:

[0115] S51, for any point (i, j) on the drilling image, the amplitude of the crack is calculated, and as long as A satisfies the parameter range, the corresponding accumulator unit contribution is increased by one;

[0116] S52, in Apply maximum value detection in space to find local peaks and select possible candidate crack curves;

[0117] S53. Apply a significance verification process to suppress erroneous curves and select the most significant curve among the candidate crack curves.

[0118] The peaks in parameter space are very widely distributed. False detections can occur due to the presence of noise on the extended sides of the peaks. Methods should be chosen to suppress this noise and select only the most significant curves. A key step is to remove the candidate curves that support the accepted detection from the original image. This method prevents the same curve segment from being used multiple times as support for the detection candidate curve in the image. Usually the curves with the largest amplitude in parameter space are picked first. Therefore, the amplitude of the correct event is larger than the incorrect event.

[0119] The present invention uses specific embodiments to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

[0120] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A method for rapid visualization detection of hydraulic fractures in underground boreholes, characterized in that: The steps include: S1. Use a drilling probe to conduct a panoramic observation of the inside of the borehole to obtain a drilling image; S2, preprocessing the borehole image obtained in step S1, increasing the number of horizontal and vertical sampling points of the borehole image, thereby improving the overall image resolution; S3, using a local statistical variable threshold segmentation method to segment the borehole image preprocessed in step S2 to obtain a binary image, and performing morphological extraction on the binary image to obtain edge features of the crack, wherein the local statistical variable threshold segmentation method is specifically: S301, using any point in the pre-processed drilling image ( x,y ) defines a variable local threshold, which is expressed as: in, and are the neighborhoods centered at (x,y) in the image. The standard deviation and mean of the set of pixels contained, is a non-negative constant; if the global mean is used ,but S302, by point ( x,y ) is the center of the neighborhood The local standard deviation and variance of all pixels in the preprocessed drilling image are calculated to obtain the local feature combination; S303, combining local features by using AND logic, and calculating local thresholds by weights to obtain a binary image, which is expressed as: in, is a binary image, is the preprocessed drilling image, is the local mean or global mean ; S4. Identify the average depth of the formation where the fracture is located based on the edge characteristics of the fracture, specifically in the following way: S401, in the plane image A complete cycle of crack curve is constructed in: Where A is the amplitude, is the initial phase, d is the average depth position of the crack; S402: Determine two pixel points half a period apart based on the symmetry relationship of the sine curve and The relationship of the ordinate is expressed as: S403, traversing the plane image , all the vertical coordinate edge pixels in the image that satisfy the vertical coordinate relationship determined in step S402 are matched into point pairs, and the pixel points are calculated and The horizontal coordinate of the point M of the connecting line gives the average depth position of the crack, which is expressed as: And use Hough transform to transform the clean binary image after edge detection into a discrete two-dimensional parameter space; S5. Different types of cracks are classified according to parameter attributes in a discrete two-dimensional parameter space.

2. A method for rapid visualization detection of downhole hydraulic fractures according to claim 1, characterized in that: The specific method of performing morphological extraction on the binary image in S3 to obtain the edge features of the crack is: S311, the binary image obtained in step S303 is processed by morphological operators to obtain a clean binary image, which is expressed as: in, is a clean binary image, is a binary image, is a structural element; S312, using edge detection to eliminate all non-edge sets, retaining the crack edge set that crosses the entire image, expressed as: Among them, E is the final edge image.

3. A method for rapid visualization detection of downhole hydraulic fractures according to claim 1, characterized in that: The specific method of using Hough transform in S4 to convert the clean binary image after edge detection into a discrete two-dimensional parameter space is: in, , is the diagonal length of the drilling image, is the average depth of the cracks.

4. A method for rapid visualization detection of downhole hydraulic fractures according to claim 1, characterized in that: Described S5 specifically comprises the following steps: S51. For any point on the drilling image Calculate the amplitude of the crack. As long as A satisfies the parameter range, the corresponding accumulator unit contribution is increased by one. S52, in Apply maximum value detection in space to find local peaks and select possible candidate crack curves; S53. Apply a significance verification process to suppress erroneous curves and select the most significant curve among the candidate crack curves.

Images