CN104866856A - Imaging log image solution cave information picking method based on connected domain equivalence pair processing - Google Patents

Abstract

The invention discloses a method for picking up karst cave information in imaging logging images based on equivalent pairs of connected domains, which belongs to the field of picking up karst cave information in imaging logging images, and includes the following methods: A. Imaging logging based on equivalent pair processing Quantitative marking of connected domains of dissolved pores in well images, picking up the parameter information of each connected domain, including the length and width of the connected domain, the radius of the circumscribed circle of the connected domain, the radius of the inscribed circle, roundness and sorting coefficient; B. Based on the surface porosity curve The reflected dissolution vug development level is used to stratify the imaging logging images. The beneficial effects of the present invention are as follows: the image connected domain marking algorithm based on equivalence pair processing has the advantages of fast and non-repetitive marking. Using this algorithm, the dissolved hole connected domain can be accurately marked from the binary image, and then each connected domain Pick up target information, including hole size, sorting coefficient, connected domain area (face ratio), roundness, etc.

Description

A Method of Extracting Cave Information from Imaging Logging Image Based on Equivalent Pair Processing of Connected Domains

technical field

The invention belongs to the field of picking up karst cave information in imaging well logging images, and in particular relates to a method for picking up karst cave information in imaging well logging images based on equivalent pair processing of connected domains.

Background technique

1. Definition of abbreviations and key terms

1.1 Connected domain: It refers to a set composed of several pixels. The pixels in this set have the following characteristics: the gray level of all pixels is less than or equal to the connected domain level; the pixels in the same connected domain are connected in pairs, that is, in Between any two pixels there exists a path consisting entirely of elements of this set.

1.2 Equivalent pair: Due to the influence of the shape of the cave and the scanning sequence, it is initially considered to be two disconnected areas. With the deepening of the marking scanning process, it is found that they actually belong to the same connected area. This phenomenon can be called is an equivalence pair phenomenon.

It can be seen that when the graph is scanned and marked, the box area will generate equivalent pairs, as shown in Figure 1. The white area in the figure is actually the same connected area, but due to the influence of the scanning order and the marking algorithm, this area is marked as two connected areas '8' and '14', as shown in Figure 2, which is the equivalent pair Phenomenon.

1.3 Face ratio: the percentage of the cave area in the image to the entire image area.

2. Relevant knowledge introduction:

2.1 Formation micro-resistivity scanning imaging logging (FMI) uses multiple rows of button-shaped small electrodes on the multi-electrode plate to emit currents to the wellbore formation. Due to the differences in the rock composition, structure and fluid contained in the electrodes, the current fluctuations are caused. The change of the current reflects the change of the rock resistivity around the borehole wall, and the borehole wall imaging of the resistivity can be displayed accordingly. The image changes from white (high resistance) to yellow, and then to black (low resistance). The darker the color, the lower the resistivity (as shown in Figure 4). The four white strips in the figure are blank areas not covered by the instrument, and the black curve The holes are cracks on the borehole wall, and the black spots are dissolution holes.

2.2 The connected domain of the image is actually the connectivity problem between pixels. In a two-dimensional image, assuming that the target pixel has the same pixel value as an adjacent pixel, the two pixels are said to be connected. When studying connectivity, it is first necessary to determine whether the neighborhood connectivity rules are 4-neighborhood connectivity or 8-neighborhood connectivity. The 4-neighborhood connection focuses on the upper, lower, left, and right four positions of the target pixel (Figure 5). 8 Neighborhood connectivity selects all adjacent pixels of the target pixel in the 3*3 matrix in two-dimensional space, that is, in addition to the upper, lower, left, and right points, it also includes 4 positions of upper left, upper right, lower left, and lower right point (Fig. 6), the algorithm of the present invention adopts the 8-neighborhood connectivity rule.

3. Introduction of existing technology:

3.1 Intelligent fracture information picking method in imaging logging images (Yan Jianping 2009)

After the imaging logging data is processed, high-resolution images of the whole borehole wall can be obtained, and the morphological feature of the intersection line between the fracture surface and the wellbore on the image appears as a single-period sinusoidal curve, and the point-line duality of the Hough transform is applied The crack angle and other information in the image can be picked up efficiently, and better processing results can be obtained. The crack angle, initial phase and other information can be picked up well, as shown in Figure 8.

Although this method can pick up fracture information very well, it ignores cave information. Moreover, it is not very suitable for formations with well-developed caves and few fractures.

3.2 Automatic cave detection method based on borehole imaging logging images (Tian Jinwen 1999)

The method includes using the Roberts operator to detect the edge of the cave, and using the method based on the direction curve to track and refine the edge, so as to realize the automatic extraction and quantitative calculation of the cave. This method obtains the direction sequence of feature points by extracting feature points, tracking them, and uses their corresponding direction curves to identify caves. Its algorithm has less calculation amount, fast judgment speed, good real-time processing and adaptability, as shown in Figure 10.

Although this method can identify karst caves well, it is limited to the detection and identification of karst caves, and the relevant information of karst caves cannot be extracted, so it cannot further provide data support for reservoir evaluation of formations with karst caves.

Contents of the invention

Aiming at the deficiencies of the prior art, the present invention provides a method for picking up karst cave information in imaging logging images based on connected domain equivalent pair processing, which can effectively solve the problems caused by not identifying karst cave information or not extracting karst cave information even though it can be identified. However, it is impossible to provide data support for reservoir evaluation of formations with karst caves.

In order to solve the above problems, the technical solution adopted by the present invention is as follows: a method for picking up karst cave information in imaging logging images based on connected domain equivalent pair processing, comprising the following steps:

A. Quantitative labeling of connected domains of dissolved vugs in imaging logging images based on equivalent pair processing, picking up the parameter information of each connected domain, including the length and width of the connected domain, the radius of the circumscribed circle of the connected domain, the radius of the inscribed circle, roundness and Sorting coefficient;

B. Based on the development degree of dissolution pores reflected by the surface porosity curve, the imaging logging image is layered, and then the number of connected domains, distribution of connected domain sorting coefficients, roundness distribution, inscribed circles and circumscribed circles are counted for each layer Heterogeneity information of radius distribution.

As preferably, A specifically includes the following steps:

1.1 Image preprocessing: first, the original image is grayscaled to obtain a grayscale image; after grayscale processing, the image is subjected to median filtering; then the filtered image is thresholded to obtain a binarized image;

1.2 Connected domain labeling and information picking: Firstly, use the image connected domain labeling algorithm based on equivalent pairs to perform connected domain labeling processing on the binary image to obtain the marked image, and at the same time pick up the length and width of each connected domain in the marked image , the parameter information of inscribed circle radius, circumscribed circle radius and roundness; then pick up the face ratio and face ratio histogram of the connected domain marked image according to the pixels in the depth domain.

As preferably, B specifically includes the following steps:

2.1 Define the length and width of the connected domain, the radius of the inscribed circle, and the radius of the circumscribed circle: I. The length of the connected domain refers to the number of pixels between the leftmost pixel and the rightmost pixel of the connected domain, including the left and right pixels; II. Connectivity Domain width refers to the number of pixels between the top pixel and the bottom pixel of the connected domain, including the top and bottom pixels; It is drawn as a rectangle, and the radius of the inscribed circle has two values, one refers to 1/2 pixel of the short side of the rectangle, that is _{, S inside} R, and the other refers to 1/2 pixel of the long side of the rectangle, That is _{, L in} R; IV. Radius of the circumscribed circle of the connected domain, the pixels on the left and right sides and the top and bottom pixels of the connected domain can be drawn into a rectangle, and the radius of the circumscribed circle refers to 1/2 pixel of the diagonal of the rectangle, namely From this definition _, output these parameters of each connected domain in the image through program design, and at the same time count the length and width distribution of the connected domain, the radius distribution of the inscribed circle, the radius distribution of the circumscribed circle, and the circularity distribution information of the connected domain in an image ;

2.2 Define connected domain circularity:

The closeness of the rectangle formed by the pixels on the left and right sides of the connected domain and the pixels on the top and bottom to the circle is defined as the "circularity of the connected domain", which is calculated by the following formula:

connected domain circularity

In the formula, R _{inner S} - the radius of the short inscribed circle, R _outer - the radius of the circumscribed circle; the meaning of the denominator: the ratio of the radius of the inscribed circle to the radius of the circumscribed circle in a square is It can be considered as a standard;

2.3 Define connected domain sorting coefficient:

Connected domain sorting coefficient S _CO ＝PC ₂₅ /PC ₇₅

In the formula, PC ₂₅ /CP ₇₅ represents the ratio of the radius value corresponding to the cumulative frequency of 25% to the radius value corresponding to the cumulative frequency of 75% on the cumulative curve of inscribed circle radius and circumscribed circle radius in the connected domain; combined with the concept of sediment sorting coefficient , according to the size of S _C0 , the sorting level of the connected domain of holes can also be divided: S _C0 =1~2.5, good sorting; S _C0 =2.5~4.0, medium sorting; S _C0 >4.0, poor sorting.

As a preference, 1.1 The specific method of performing threshold segmentation on the filtered image is as follows: Calibrate the imaging logging image through the core, so that the image after threshold segmentation reflects the information of core dissolution pores and fractures.

Preferably, the face ratio histogram of 1.2 is a face ratio histogram with an interval of 10 to 30 pixels.

Preferably, the face ratio histogram of 1.2 is a face ratio histogram at intervals of 20 pixels.

The beneficial effects of the present invention are as follows:

1. The image connected domain marking method based on equivalent pair processing has the advantages of fast and non-repetitive marking. With this method, the dissolved hole connected domain can be accurately marked from the binary image, and then the target information of each connected domain can be picked up. Including pore size, sorting coefficient (variance), connected domain area (face porosity), roundness, etc.

2. Use the surface porosity curve reflecting the development degree of dissolution vugs to stratify the image. On this basis, the surface porosity value, pore size, roundness, sorting coefficient (variance) distribution, etc. of each layer of dissolution vugs can be picked up. Homogeneity information can better quantitatively reflect the development degree and heterogeneity of dissolution pores and vugs in the depth domain of FMI images, which is helpful to more accurately evaluate the geological characteristics of wellbore, and is also a useful exploration for using FMI images to quantitatively pick up wellbore rock-cavity information.

Description of drawings

Figure 1 binary image;

Figure 2 Equivalent pair image before processing;

Figure 3 Equivalent pair processed image;

Figure 4 FMI image;

Fig. 5 4 connection schematic diagram;

Fig. 6 8 connectivity schematic diagram;

Figure 7 Human-computer interaction to pick up cracks (Yan Jianping 2009);

Figure 8 Improved Hough transform to pick up cracks (Yan Jianping 2009);

Figure 9 Original image (Tian Jinwen 1999);

Figure 10 The image of the cave picking result (Tian Jinwen 1999);

Figure 11 FMI Binary Image Connected Domain Labeling Algorithm Flowchart;

Figure 12 Different position identification of binary image array;

Figure 13 "2" location points focus on the neighborhood points when detecting the connected domain;

Figure 14 "3" location points focus on the neighborhood points when detecting the connected domain;

Figure 15 "4" location points focus on the neighborhood points when detecting the connected domain;

Figure 16 Focus on the neighborhood points when detecting the connected domain of the "5" position point;

Figure 17 The original image of electrical imaging;

Figure 18 grayscale image;

Figure 19 median filter image;

Figure 20 binary image;

Figure 21 Connected domain marking processing;

Figure 22 Connected domain marker zoom;

Figure 23 Face ratio curve;

Figure 24 20pixel surface area distribution;

Figure 25 Radius distribution of inscribed circles in connected domains;

Figure 26 Radius distribution of the circumscribed circle of the connected domain;

Figure 27 Connected domain circularity distribution;

Figure 28 The original image of the electrical imaging example;

Figure 29 Example grayscale image;

Figure 30 Example median filtering;

Figure 31 Example binarization map;

Figure 32 Example face ratio curve (layered);

Figure 33 Example 20pixel surface area distribution;

Figure 34 Connected domains are labeled and layered.

Detailed ways

In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be described in further detail below with reference to the accompanying drawings and examples.

Principle and idea of connected domain labeling method based on equivalence pair processing

As shown in Figure 11, an image logging image cave information extraction method based on equivalent pair processing of connected domains is a method of processing imaging logging images and extracting relevant cave information using the connected domain labeling method based on equivalent pairs .

The input of the method of the present invention is the image (Er) array Count[][] of FMI after grayscale, filtering, binarization and inversion, wherein the pixel value is not 0 or 1, and the array represents black respectively when the image is displayed. (background color) and white (target color), that is, the white area represents the cave. The output is the new structure array NCount[][] after the connected domain is marked according to Count[][] and the image restored by NCount[][]. Different values in the result array indicate that they belong to different connected domains.

After the 8-neighborhood connectivity rules are determined, the binary values of the image are scanned in the order from left to right and from top to bottom. During processing, each target pixel can determine its own connectivity only by judging the relationship between the pixel and surrounding pixels whose connectivity has been determined. However, different location points need to pay attention to different domain points. Figures 12-16 show all possible location points and the neighborhood points that need to be paid attention to when determining connectivity.

However, due to the influence of the shape of the cave and the scanning sequence, the two areas that were initially considered to be disconnected, as the marking and scanning process deepened, it was found that they actually belonged to the same connected area, as shown in Figure 14-16. All three types are possible. yields an equivalence pair. Equivalence pairs must be handled in the program, otherwise this will seriously affect the marking effect.

It can be seen from Figure 1 that when scanning and marking Figure 1, the box area will generate equivalent pairs. As shown in Figure 2, the white area in the figure is actually the same connected area, but due to the influence of scanning order and marking algorithm , this area is marked as two connected domains '8' and '14'. This result not only affects the accurate statistics of the number of dissolved caves and dissolved pores, but also affects the accurate extraction of the characteristic information of dissolved pores. Therefore, in the process of image marking, when an equivalent pair appears, it needs to be processed in time to avoid problems in the subsequent information picking work.

Here, take FIG. 15 as an example to illustrate the processing of equivalence pairs. Put the marked values of the four marked pixels in the new array B[], assuming that there are two non-zero and unequal values in B[], then they are an equivalent pair. At this time, shill NCount[x][y] (the pixel mark value at position 4) is equal to the first non-zero value in B[], and then find that all mark values in NCount[][] are equal to B[] The second non-zero value points and record their positions, and finally assign the value of NCount[x][y] to them one by one. So far, this equivalence pair has been processed. As for the case where the non-zero value in B[] is greater than two, it can be handled in this way.

Figure 3 shows the results of the equivalent pair processing in the red box area, which is uniformly marked as '7', excluding the influence of the equivalent pair.

Dissolution hole mark, information pickup:

1. Steps of marking and picking up information

1.1 Image preprocessing: first, grayscale processing is performed on the original image (Fig. 17) to obtain a grayscale image (Fig. 18); after grayscale processing, median filtering is performed on the image (Fig. 19); Threshold segmentation is performed on the image of the core (the core is used to calibrate the imaging logging image, so that the image after the threshold segmentation can accurately reflect the information of the core dissolution pores and fractures), and a binary image (Fig. Information parameter picking lays the foundation.

1.2 Connected domain labeling and information picking: First, use the above-mentioned equivalent pair-based image connected domain labeling algorithm to perform connected domain labeling processing on the binary image (Fig. 20) to obtain the marked image (Fig. 21). Each connected domain picks up its length, width, inscribed circle radius, circumscribed circle radius, roundness and other parameter information (see below for its detailed definition and description); the connected domain marked image is actually a binary image, but the difference is The connected domains are continuously marked, and the information of each connected domain can be quantitatively picked up through the program. The enlarged image is shown in Figure 22, and the marking of the connected domains can be clearly seen; then the connected domains are marked with images (or binary values (Fig. 20) to pick up the surface porosity according to the pixels in the depth domain (Fig. 23), and the histogram of the surface porosity at intervals of 20 pixels (20pixel) (Fig. 24), in order to understand the development degree of dissolution pores in the depth domain and the actual processing process. Image processing lays the foundation for layering.

2. Mark and pick up information parameters

2.1 Connected domain length, width, inscribed circle radius, and circumscribed circle radius: I. Connected domain length refers to the number of pixels between the leftmost pixel and the rightmost pixel of the connected domain (including left and right pixels); II. Connectivity Domain width refers to the number of pixels between the top pixel and the bottom pixel of the connected domain (including the top and bottom pixels); III. The radius of the inscribed circle of the connected domain refers to the pixels on the left and right sides and the top and bottom pixels of the connected domain It can be drawn as a rectangle, and the radius of the inscribed circle has two values, one refers to 1/2 pixel of the short side of the rectangle ( _{S inside} R), and the other refers to 1/2 pixel of the long side of the rectangle ( _{L in} R); IV. The radius of the circumscribed circle of the connected domain, the pixels on the left and right sides of the connected domain and the top and bottom pixels can be drawn into a rectangle, and the circumscribed circle radius refers to 1/2 pixel of the diagonal of the rectangle ( _outside R). From this definition, these parameters of each connected domain in the image (Fig. 21) can be output through program design, and at the same time, the length and width distribution of the connected domain, the radius distribution of the inscribed circle (Fig. 25), and the radius distribution of the circumscribed circle can be counted in an image. (Fig. 26), the roundness distribution of connected domains (Fig. 27), and other information, the distribution of these parameters can reflect the development degree of pores in the wellbore formation.

2.2 Connected domain roundness: roundness is originally described in sedimentary petrology as the degree to which the original edges and corners of rock clastic grains are rounded, and it is an important structural feature of clastic grains. It has nothing to do with the shape of the particle, just a function of the sharpness of the edges and corners. Roundness geometrically reflects the corner curvature in the image of the particle's largest projected surface. Weld (1932) proposed the following roundness calculation formula:

Roundness R _O ＝(∑r/n)/R (1)

In the formula, r-the radius of the inscribed circle of the corner, n-the number of corners, R-the maximum radius of the inscribed circle of the particle.

In the FMI connected domain identification image of imaging logging, we want to describe the closeness of the dissolved vug connected domain to a circle, so we want to learn from the definition and formula of the roundness of clastic particles, but in fact this definition and the roundness of the connected domain we want to describe It is different in concept, and the parameters involved in the formula, such as the corner angle, are not convenient for program design in the process of image processing. Therefore, we make the rectangle composed of the leftmost and leftmost pixels and the top and bottom pixels of the connected domain close to the circle Degree, defined as "connected domain circularity", is calculated by the following formula:

connected domain circularity

In the formula, _RinnerS -short inscribed circle radius, Router- _{circumscribed} circle radius. The meaning of the denominator: the ratio of the radius of the inscribed circle to the radius of the circumscribed circle in a square is Think of it as a standard. After obtaining the ratio of the inscribed circle and circumscribed circle radius of all connected domain rectangles, comparing it with the standard value can more intuitively see how close the connected domain shape is to a circle.

2.3 Sorting coefficient of connected domain: The sorting coefficient is originally a parameter indicating the degree of sediment sorting, which reflects the uniformity of particle size, or the dispersion of sediment around the concentration trend. The given sorting coefficient formula is :

S ₀ =P ₂₅ /P ₇₅ (3)

In the formula, P ₂₅ and P ₇₅ represent the particle diameters corresponding to 25% and 75% of the cumulative particle content on the cumulative curve, respectively. We want to learn from this formula to characterize the sorting of the connected domain scale of dissolved pores in the image, so the above formula is modified as:

Connected domain sorting coefficient S _CO ＝PC ₂₅ /PC ₇₅ (4)

In the formula, PC ₂₅ /CP ₇₅ represents the ratio of the radius value corresponding to the cumulative frequency of 25% to the radius value corresponding to the cumulative frequency of 75% on the cumulative curve of inscribed circle radius and circumscribed circle radius in the connected domain. Through calculation, the sorting coefficient of the radius of the inscribed circle (R _{inner S} ) in the image in Figure 21 is 2.25, and the sorting coefficient of the radius of the circumscribed circle is 2.6. Of course, PC ₂₅ or PC ₇₅ may not appear in the actual image processing process, and it is also feasible to use the probability statistics parameter "variance" to characterize the sorting property to a certain extent. Combined with the concept of sediment sorting coefficient, the sorting level of the connected domain of vugs can also be divided according to the size of S _C0 : S _C0 = 1-2.5, good sorting; S _C0 = 2.5-4.0, moderate sorting; S _C0 > 4.0, poor sorting.

Specific examples:

Figures 28-34 are examples of processing according to the above method. After the processing, image layering and connected domain information picking are performed: firstly, the image is processed by using the surface porosity curve in the depth domain of the image (Figure 32) combined with the development degree of dissolution pores reflected in Figure 33 Layering (Figure 34); at the same time, pick up parameter information such as its length, width, inscribed circle radius, circumscribed circle radius, and roundness for each connected domain in the marked image; then mark the connected domain image (Figure 34) by Depth domain stratification results, the inscribed circle radius distribution, circumscribed circle radius distribution, roundness distribution and related parameter information of the connected domain in each layer are statistically obtained respectively (Table 1). The quantitative parameter information of the FMI image can better quantitatively reflect the development degree and heterogeneity of dissolution vugs in the depth domain of the FMI image, and help to more accurately evaluate the geological characteristics of the wellbore.

Table 1 Picking data of connected domain parameter information in image layering

Those skilled in the art will appreciate that the embodiments described here are to help readers understand the implementation method of the present invention, and it 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 modifications and combinations based on the technical revelations disclosed in the present invention without departing from the essence of the present invention, and these modifications and combinations are still within the protection scope of the present invention.

Claims (6)

1. A method for picking up karst cave information in imaging logging images based on connected domain equivalence pairs, it is characterized in that, comprising the following steps: A. Quantitative labeling of connected domains of dissolved vugs in imaging logging images based on equivalent pair processing, picking up the parameter information of each connected domain, including the length and width of the connected domain, the radius of the circumscribed circle of the connected domain, the radius of the inscribed circle, roundness and Sorting coefficient; B. Based on the development degree of dissolution pores reflected by the surface porosity curve, the imaging logging image is layered, and then the number of connected domains, distribution of connected domain sorting coefficients, roundness distribution, inscribed circles and circumscribed circles are counted for each layer Heterogeneity information of radius distribution. 2. The method according to claim 1, wherein A specifically comprises the following steps: 1.1 Image preprocessing: first, the original image is grayscaled to obtain a grayscale image; after grayscale processing, the image is subjected to median filtering; then the filtered image is thresholded to obtain a binarized image; 1.2 Connected domain labeling and information picking: Firstly, use the image connected domain labeling algorithm based on equivalent pairs to perform connected domain labeling processing on the binary image to obtain the marked image, and at the same time pick up the length and width of each connected domain in the marked image , the parameter information of inscribed circle radius, circumscribed circle radius and roundness; then pick up the face ratio and face ratio histogram of the connected domain marked image according to the pixels in the depth domain. 3. according to the described method of claim 1 and 2, it is characterized in that, B specifically comprises the following steps: 2.1 Define the length and width of the connected domain, the radius of the inscribed circle, and the radius of the circumscribed circle: I. The length of the connected domain refers to the number of pixels between the leftmost pixel and the rightmost pixel of the connected domain, including the left and right pixels; II. Connectivity Domain width refers to the number of pixels between the top pixel and the bottom pixel of the connected domain, including the top and bottom pixels; It is drawn as a rectangle, and the radius of the inscribed circle has two values, one refers to 1/2 pixel of the short side of the rectangle, that is _{, S inside} R, and the other refers to 1/2 pixel of the long side of the rectangle, That is _{, L in} R; IV. Radius of the circumscribed circle of the connected domain, the pixels on the left and right sides and the top and bottom pixels of the connected domain can be drawn into a rectangle, and the radius of the circumscribed circle refers to 1/2 pixel of the diagonal of the rectangle, namely From this definition _, output these parameters of each connected domain in the image through program design, and at the same time count the length and width distribution of the connected domain, the radius distribution of the inscribed circle, the radius distribution of the circumscribed circle, and the roundness distribution information of the connected domain in an image ; 2.2 Define connected domain circularity: The closeness of the rectangle formed by the pixels on the left and right sides of the connected domain and the pixels on the top and bottom to the circle is defined as the "circularity of the connected domain", which is calculated by the following formula: connected domain circularity In the formula, R _{inner S} - the radius of the short inscribed circle, R _outer - the radius of the circumscribed circle; the meaning of the denominator: the ratio of the radius of the inscribed circle to the radius of the circumscribed circle in a square is It can be considered as a standard; 2.3 Define connected domain sorting coefficient: Connected domain sorting coefficient S _C0 ＝PC ₂₅ /PC ₇₅ In the formula, PC ₂₅ /CP ₇₅ represents the ratio of the radius value corresponding to the cumulative frequency of 25% to the radius value corresponding to the cumulative frequency of 75% on the cumulative curve of inscribed circle radius and circumscribed circle radius in the connected domain; combined with the concept of sediment sorting coefficient , according to the size of S _C0 , the sorting level of the connected domain of holes can also be divided: S _C0 =1~2.5, good sorting; S _C0 =2.5~4.0, medium sorting; S _C0 >4.0, poor sorting. 4. method according to claim 2, it is characterized in that, 1.1 the specific method that threshold value segmentation is carried out to the image after filtering is as follows: calibrate imaging logging image by rock core, make the figure after choosing threshold value segmentation reflect core dissolution hole, fracture Information. 5. The method according to claim 2, characterized in that the face ratio histogram of 1.2 is a face ratio histogram with an interval of 10 to 30 pixels. 6. The method according to claim 5, wherein the face count histogram of 1.2 is a face count histogram at intervals of 20 pixels.