CN112085720A - Detection and characterization method for connected domain of slot and hole - Google Patents

Abstract

The invention discloses a method for detecting and characterizing a connected domain of a slot and a hole, which comprises the following steps: s1, extracting the coherence of the three-dimensional seismic data, calculating a first characteristic value of the coherence, and constructing a first characteristic value three-dimensional data volume; s2 selecting five neighborhoods as connected domain analysis point neighborhoods; s3 connected domain marking, wherein the marking algorithm adopts a one-time scanning method of double-queue recursive search, and comprises the following steps: s31, searching all non-zero value positions to form a large queue; s32, establishing and marking a small queue of five neighborhoods and connected domains in the large queue; s33, judging whether the large queue data point is communicated with the five neighborhoods from the first point, and marking if the large queue data point is communicated with the five neighborhoods; s34, removing abnormal value output marker slices and recording the serial number of the connected domain; s4, marking the connected domain boundary through longitudinal scanning and transverse scanning, calculating the centroid, and calculating the included angle and the distance between the centroid and the boundary point by taking the centroid as the central point to realize the anticlockwise sequencing of the boundary point; s5 calculates the connected component area. The method can improve the interpretation precision and efficiency of the fracture-cavity and provide basic support for reservoir evaluation.

Description

Detection and characterization method for connected domain of slot and hole

Technical Field

The invention relates to the field of seismic data processing and reservoir description, in particular to a method for detecting and characterizing a slot-hole connected domain based on a first eigenvalue of a coherent body and five-neighborhood double-queue recursive search.

Background

The fracture-cavity reservoir is an important oil and gas reservoir and has an important position in oil and gas exploration and development at home and abroad. It features high porosity and permeability of reservoir, good connectivity of seam and hole, and high oil and gas reserves and output. In the western oil area of China, the reservoirs are distributed a lot, and a plurality of high-yield oil and gas fields of Tarim, Changqing and Xinjiang all belong to the reservoirs. The fracture-cavity reservoir profile has typical characteristics, namely beaded seismic reflection characteristics, but the reflection characteristics are strong so as to be convenient for manual interpretation and pickup; on the seismic horizontal slice, the interpretation of the fracture-holes is also limited by the size, number, complex boundaries, background interference, etc. of the fracture-holes.

Disclosure of Invention

Based on the technical problems, the invention provides a method for detecting and characterizing a fracture-cavity connected domain based on a first eigenvalue of a coherent body and five-neighborhood double-queue recursive search, which can improve the precision and efficiency of reservoir prediction and provide a basis for the exploration and development of fracture-cavity oil and gas reservoirs.

The technical solution adopted by the invention is as follows:

a method for detecting and characterizing a fracture-cavity connected domain comprises the following steps:

s1 extracting the coherence of the three-dimensional seismic data and calculating a first characteristic value of the coherence to construct a first characteristic value data body;

s2 selecting five neighborhoods as connected domain analysis point neighborhoods;

s3 connected domain marking, wherein the marking algorithm adopts a one-time scanning method of double-queue recursive search, and comprises the following steps:

s31, searching all non-zero value positions to form a large queue;

s32, establishing and marking a small queue of five neighborhoods and connected domains in the large queue;

s33, starting from the first point, judging whether the large queue data point is communicated with the five neighborhoods, if so, marking, and if not, reducing the length of the large queue by one, and recursing until the length of the large queue is zero;

s34 removing abnormal values, outputting marked slices and recording the serial numbers of connected domains;

s4, marking the connected domain boundary through longitudinal scanning and transverse scanning, calculating the centroid, and calculating the included angle and the distance between the centroid and the boundary point by taking the centroid as the central point to realize the anticlockwise sequencing of the boundary point;

s5 calculates the connected component area.

The specific content of step S1 is as follows:

for a certain earthquake three-dimensional data volume, N sampling points of adjacent J channels are taken to form an NxJ earthquake sub-volume, and a matrix D is used_N×JTo express, the matrix C is composed of D_J×J：

Calculating a first eigenvalue λ of C₁And constructing a first characteristic value data body.

The specific content of step S31 is as follows:

extracting first characteristic value data according to horizontal slices or stratum slices, setting a threshold coefficient, carrying out threshold processing, and setting zeros smaller than a threshold value;

firstly, according to row i and then according to column j, finding out all non-zero value positions to form a large queue; line number defined as line (k), track number as cdp (k), k ═ 1,2, …, NP; k is the sequence number of the searched non-zero value data, and NP is the total number of the non-zero value data.

The specific content of step S32 is as follows:

starting from the first point of the large queue, i is LINE (1), and j is CDP (1), judging whether the five neighborhood points are communicated, and establishing a corresponding small queue with the longest point of 5 points; setting the first connected domain as F as 1, setting the initial value of the point number contained in the F-th connected domain as NL (F) as 0, setting the head position L of the neighborhood small queue as 1, and setting the mark Y (i, j) as F;

(1) judging and identifying a left side point; if the value X (i, j-1) of the point (i, j-1) is greater than zero and is unmarked, i.e., X (i, j-1) >0& Y (i, j-1) ═ 0, then set Y (i, j-1) ═ F, and find out the point from the queue, set the neighborhood small queue location point L ═ L +1, and set the neighborhood point connectivity flag G1 ═ 1, nl (F) ═ nl (F) + 1; otherwise, setting the neighborhood point non-communication mark G1 as 0, and executing the next step (2);

(2) judging and identifying a right side point; if X (i, j +1) >0& Y (i, j +1) ═ 0, then Y (i, j +1) ═ F, and then the neighborhood small queue location point L ═ L +1, and then the neighborhood point connectivity flag G2 ═ 1, nl (F) ═ nl (F) +1, because this point immediately follows point (i, j), it is not necessary to search for and replace it; otherwise, setting the neighborhood point non-communication mark G2 as 0, and executing the next step (3);

(3) judging and identifying a lower point on the left side; if X (i +1, j-1) >0& Y (i +1, j-1) ═ 0, then set Y (i +1, j-1) ═ F, find out the point from the queue, replace it to neighborhood small queue location point L ═ L +1, set the neighborhood point connectivity flag G3 ═ 1, nl (F) ═ nl (F) + 1; otherwise, setting the neighborhood point non-communication mark G3 as 0, and executing the next step (4);

(4) judging and identifying a lower point; if X (i +1, j) >0& Y (i +1, j) ═ 0, then set Y (i +1, j) ═ F, find out the point from the queue, set neighborhood small queue location point L ═ L +1, set the neighborhood point connectivity flag G4 ═ 1, nl (F) ═ nl (F) + 1; otherwise, setting the neighborhood point non-communication mark G4 as 0, and executing the next step (5);

(5) judging and identifying a lower right point; if X (i +1, j +1) >0& Y (i +1, j +1) ═ 0, then set Y (i +1, j +1) ═ F, find out the point from the queue, set neighborhood small queue location point L ═ L +1, set the neighborhood point connectivity flag G5 ═ 1, nl (F) ═ nl (F) + 1; otherwise, if the neighborhood point non-communication flag G5 is set to 0, step S33 is executed.

The specific content of step S33 is as follows:

large queue length minus 1: NP-1;

(1) if NP is 0, go to step S34;

(2) otherwise, finishing and dequeuing the 1 st point of the large queue, and reordering the queue; the queue is shifted left 1 bit from 2 to NP;

1) if G1+ G2+ G3+ G4+ G5 is 0, then F is F +1, indicating the end of the first connected domain;

2) otherwise, the first connected domain mark is not finished;

cases 1) and 2) both return to step S32, continuing the determination and identification.

In the above step S3: the connected body identification can record the number and the number of connected points of each connected domain, and unreasonable values are removed during output; the user can set the maximum and minimum connection points by himself, so that small connecting bodies without mining value and unreasonable connection areas can be removed.

The contents of the longitudinal and transverse scanning in step S4 are as follows:

longitudinal scanning is firstly carried out to obtain longitudinal boundary points, and boundary marking is carried out; then transversely scanning, and if the boundary points are not marked, supplementing the boundary points; finally, the boundary points are output according to a reverse time needle sequence, the initial value of the boundary mark array is set to be zero (Y (i, j) ═ 0, i ═ 1, M, j ═ 1, N), and the initial value l ═ 0 of the boundary point queue;

longitudinal scanning method

1) Taking out the point X (1, j) of the jth column and row 1, if the point is not zero, it is directly the boundary point, and the mark Y (1, j) is 1, l +1, line (l) is 1, cdp (l) is j;

2) rows i-2 to i-M-1, if X (i, j) >0 and X (i +1, j) >0 or X (i-1, j) ═ 0, then (i, j) is a boundary point, let Y (i, j) ═ 1, l ═ l +1, line (l) ═ i, cdp (l) ═ j, allowing multiple boundary points to appear in a column;

3) taking out a point X (M, j) in the jth column and mth row, if the point is not zero, the point is directly a boundary point, and the mark Y (M, j) is 1, l +1, line (l) is M, cdp (l) is j;

② transverse scanning method

1) Taking out the point X (i,1) on the ith row and the 1 st column, if not marked and the point value is not zero, directly the boundary point, marking Y (i,1) ═ 1, l ═ l +1, line (l) ═ i, cdp (l) ═ 1;

2) columns from j 2 to j N-1, if unlabeled and if X (i, j) >0 and X (i, j +1) ═ 0 or X (i, j-1) ═ 0, (i, j) is a boundary point, let Y (i, j) ═ 1, l ═ l +1, line (l) ═ i, cdp (l) ═ j, allowing multiple boundary points to appear on a row;

3) taking out the point X (i, N) on the ith row and nth column, if not marked and if the point value is not zero, directly being a boundary point, marking Y (i, N) 1, l +1, line (l) i, cdp (l) N;

finally, a boundary point queue (line (l), cdp (l)) is obtained, wherein l is 1,2, …, LP; LP is the result of the previous boundary point count.

The centroid calculation and boundary point reverse-time ranking algorithm in the step S4 includes the following steps:

1) calculating the centroid o of the connected domain polygon, and taking the centroid as the central point of the anticlockwise rotation; centroid calculation formula:

2) calculating the included angle between each boundary point p and the centroid o;

wherein, LINE (p) represents the line number of the boundary point p, CDP (p) represents the track number of the boundary point p;

3) sorting the included angles from small to large by a bubbling method, if the included angles are the same, calculating the distance between the included angles and the centroid, and arranging the small in front of the centroid;

the connected component area in step S5 is calculated as follows:

for n points (x)₁,y₁),(x₂,y₂),...,(x_n,y_n) The points of the formed polygon are sequentially ordered in a counterclockwise way, and the area calculation formula is as follows:

the beneficial technical effects of the invention are as follows:

in the long-term research and exploration process, the inventor finds that the first characteristic value data body of the coherent body can well reserve and concentrate the fracture-cave information, eliminate useless surrounding rocks and background information, and can be used as target processing data for identifying the fracture-cave. The method comprises the steps of extracting the layer-following or horizontal slices, automatically detecting the connected space of a seam-cave reservoir by using a connected domain identification method, marking the boundary of the connected domain by longitudinal and transverse scanning, performing reverse-time needle sequencing on connected domain polygons by centroid positioning and angle calculation of each boundary point, and finally calculating the area of the connected domain.

The invention can complete the automatic detection of the connected domain aiming at the fracture-cavity reservoir, improve the interpretation precision and efficiency of the fracture-cavity, realize the quantitative characterization of the reservoir and provide basic support for the reservoir evaluation.

Drawings

FIG. 1 is a flowchart of a specific implementation of a method for detecting and characterizing a connected domain of a slot and a hole based on a first eigenvalue of a coherent body and a five-neighborhood double-queue recursive search according to the present invention;

FIG. 2 is a seismic horizontal slice with a fracture hole;

fig. 3 is a slice diagram of the extracted first feature value (corresponding to fig. 2);

FIG. 4 is a selection and comparison graph of neighborhoods;

FIG. 5 is a diagram of a connector model for process inspection;

FIG. 6 is a graph of the model 1 st neighborhood labeling results;

FIG. 7 is a diagram of theoretical models for boundary labeling, centroid calculation, and area calculation;

FIG. 8 is a graph of model boundary identification results;

FIG. 9 is a boundary diagram of models sorted directly by angle;

fig. 10 is a diagram showing the favorable reservoir identification results of eagle mountain group fracture holes in a certain work area, wherein (a) shows downward extension for 10ms, and (b) shows downward extension for 25 ms;

FIG. 11 is a diagram of actual data centroid and reservoir area calculations.

Detailed Description

In order to make the aforementioned and other objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with figures are described in detail below.

As shown in fig. 1, a method for detecting and characterizing a slot-hole connected domain based on a first eigenvalue of a coherent body and five-neighborhood double-queue recursive search includes the following steps:

s1 for a seismic three-dimensional data volume, N samples of adjacent J channels are taken to form an NxJ seismic sub-volume, and a matrix D is used_N×JTo express, the matrix C is composed of D_J×J：

Calculating a first eigenvalue λ of C₁And constructing a first characteristic value data body which can be used for detecting the fracture-cavity body. Fig. 3 is a first eigenvalue slice corresponding to fig. 2, and it can be seen that there is less surrounding rock and background information. However, the number of connected bodies in the research area cannot be known, and the area of the connected bodies cannot be known, so that the connected domain marking method is required to be used for identification.

And S2 selecting a neighborhood of connected domain analysis points. The neighborhood of the connected domain analysis commonly used comprises four neighborhoods and eight neighborhoods, the four neighborhoods have lower precision and are generally used less, and the eight neighborhoods are mostly used. The measuring line of the fracture-cavity type connected domain exists from the boundary, so that the five-neighborhood analysis domain is adopted, mirror image edge expansion processing is not needed to be carried out on the boundary, and the calculation amount can be reduced compared with that of the eight neighborhoods, as shown in figure 4.

S3 connected domain labels. The marking algorithm adopts a one-time scanning method of double-queue recursive search, which can save storage space and improve calculation efficiency. The method comprises the following five steps:

s31: and extracting first characteristic value data according to the horizontal slice or the stratum slice, setting a threshold coefficient, carrying out threshold processing, and setting zero values smaller than the threshold value. The threshold value is usually set to 5% of the maximum value, so that a part of the non-slot background information can be removed.

S32: a large non-zero value queue is established. And (4) finding out all non-zero value positions according to the row i and then the column j to form a large queue. Line number is defined as LINE (k) and lane number is defined as CDP (k). k is 1,2, …, NP, k is the sequence number of the searched non-zero value data, and NP is the total number of the non-zero value data.

S33: and establishing and marking a small queue of five neighborhoods and connected domains. Starting from the first point (i is LINE (1) and j is CDP (1)), whether the first point is communicated with the five adjacent points is judged, and a corresponding small queue with the maximum length of 5 points is established. Let the first connected domain be denoted as F ═ 1, the initial value of the point number included in the F-th connected domain be nl (F) ═ 0, the head position L of the neighborhood small queue be 1, and the mark Y (i, j) ═ F.

(1) And judging and identifying the left side point. If the value X (i, j-1) of the point (i, j-1) is greater than zero and is unmarked, i.e., X (i, j-1) >0& Y (i, j-1) ═ 0, then set Y (i, j-1) ═ F, and find out the point from the queue, set the neighborhood small queue location point L ═ L +1, and set the neighborhood point connectivity flag G1 ═ 1, nl (F) ═ nl (F) + 1; otherwise, the neighborhood point non-communication flag G1 is set to 0, and the next step (2) is executed.

(2) And judging and identifying the right side point. If X (i, j +1) >0& Y (i, j +1) ═ 0, then Y (i, j +1) ═ F, and then the neighborhood small queue location point L ═ L +1, and then the neighborhood point connectivity flag G2 ═ 1, nl (F) ═ nl (F) +1, because this point immediately follows point (i, j), it is not necessary to search for and replace it; otherwise, the neighborhood point non-communication flag G2 is set to 0, and the next step (3) is executed.

(3) And judging and identifying the lower point on the left side. If X (i +1, j-1) >0& Y (i +1, j-1) ═ 0, then set Y (i +1, j-1) ═ F, find out the point from the queue, replace it to neighborhood small queue location point L ═ L +1, set the neighborhood point connectivity flag G3 ═ 1, nl (F) ═ nl (F) + 1; otherwise, the neighborhood point non-communication flag G3 is set to 0, and the next step (4) is executed.

(4) And judging and identifying a lower point. If X (i +1, j) >0& Y (i +1, j) ═ 0, then set Y (i +1, j) ═ F, find out the point from the queue, set neighborhood small queue location point L ═ L +1, set the neighborhood point connectivity flag G4 ═ 1, nl (F) ═ nl (F) + 1; otherwise, the neighborhood point non-communication flag G4 is set to 0, and the next step (5) is executed.

(5) And judging and identifying the lower right point. If X (i +1, j +1) >0& Y (i +1, j +1) ═ 0, then set Y (i +1, j +1) ═ F, find out the point from the queue, set neighborhood small queue location point L ═ L +1, set the neighborhood point connectivity flag G5 ═ 1, nl (F) ═ nl (F) + 1; otherwise, if the neighborhood point non-communication flag G5 is set to 0, step 4 is executed.

S34: large queue length minus 1: NP-1

(1) If NP is 0, go to S4;

(2) otherwise, the large queue finishes at point 1, dequeues, and the queue reorders. The queue is shifted left 1 bit from 2 to NP.

1) If G1+ G2+ G3+ G4+ G5 is 0, then F is F +1, indicating the end of the first connected domain;

2) otherwise, it indicates that the first connected domain flag has not ended.

Cases 1) and 2) both return to step 3, continue to judge and identify.

And (5) finishing the whole marking and outputting the marked slice.

In S33, a location finding and permutation algorithm is used:

1) search algorithm

The method mainly comprises the following two steps: firstly, finding out the line number of the corresponding position from the long queue, and secondly, saving the corresponding position.

2) Permutation algorithm

The method mainly comprises the following two steps: one is that the queue ends from L +1 to L2, moving 1 bit to the right; and secondly, inserting the neighborhood points into a small queue.

The connected body mark can record the number and the number of connected points of each connected domain, and unreasonable values are removed during output. The user can set the maximum and minimum connecting points by himself, so that small connecting bodies without mining value and unreasonable connecting areas (such as the upper left part of fig. 3) can be removed.

A No. 1 connected domain and a No. 3 connected domain are designed, one north-west direction and one north-east direction respectively correspond to two karst caves in different directions, the sizes of the regions are not the same, and 8 nonzero values and 9 nonzero values are respectively arranged; no. 2 connected domain is northeast, 20 nonzero values are arranged in the region, the two connected domains correspond to two main connected structures, and a simulated cave is communicated with a river, and the figure 5 shows that the simulated cave is communicated with the other cave. The algorithm is illustrated with the five neighborhoods of the 1 st connected domain, point 1 (2, 2). Starting from the first point (2,2), it is determined whether the five neighborhood points are connected. Let the label F equal to 1, the neighbor small queue generation head position L equal to 1, and Y (i, j) equal to F. First (2,3) is found, since it follows immediately thereafter, the location is not swapped. Then find (3,3) in the neighborhood, find it in the big queue, save temporarily, then shift the element between the element and the original small queue to the right by one position, and put the saved (3,3) behind the small queue, get the big queue and the small queue after replacing. And (2) finding the neighborhoods of (2 and 2), finding other two elements to form a queue of 3 elements, and finishing the finding of the small queue. FIG. 6 shows the results after the first neighborhood search, where the triangle points are identified small queue points.

S4 connected domain boundary markers. Longitudinal scanning is firstly carried out to obtain longitudinal boundary points, and boundary marking is carried out; then transversely scanning, and if the boundary points are not marked, supplementing the boundary points; and finally, outputting the boundary points in a reverse time needle sorting mode, setting the initial value of the boundary mark array to be zero (Y (i, j) ═ 0, i ═ 1, M, j ═ 1, N), and setting the initial value l ═ 0 in the boundary point queue.

Longitudinal scanning method

1) Taking out the point X (1, j) of the jth column and row 1, if the point is not zero, it is directly the boundary point, and the mark Y (1, j) is 1, l +1, line (l) is 1, cdp (l) is j;

2) rows i-2 to i-M-1, if X (i, j) >0 and X (i +1, j) >0 or X (i-1, j) ═ 0, then (i, j) is a boundary point, let Y (i, j) ═ 1, l ═ l +1, line (l) ═ i, cdp (l) ═ j, allowing multiple boundary points to appear in a column;

3) and taking out the point X (M, j) of the jth column and the Mth row, if the point value is not zero, directly indicating a boundary point, wherein the mark Y (M, j) is 1, l is l +1, LINE (l) is M, and CDP (l) is j.

② transverse scanning method

1) Taking out the point X (i,1) on the ith row and the 1 st column, if not marked and the point value is not zero, directly the boundary point, marking Y (i,1) ═ 1, l ═ l +1, line (l) ═ i, cdp (l) ═ 1;

2) columns from j 2 to j N-1, if unlabeled and if X (i, j) >0 and X (i, j +1) ═ 0 or X (i, j-1) ═ 0, (i, j) is a boundary point, let Y (i, j) ═ 1, l ═ l +1, line (l) ═ i, cdp (l) ═ j, allowing multiple boundary points to appear on a row;

3) the point X (i, N) on the ith row and nth column is taken, and if not marked and if the point value is not zero, it is directly the boundary point, and the mark Y (i, N) is 1, l +1, line (l) i, cdp (l) N.

Finally, a boundary point queue (line (l), cdp (l)) is obtained, where l is 1,2, …, LP. LP is the result of the previous boundary point count.

Algorithm for centroid calculation and counterclockwise boundary point sorting

1) And calculating the centroid o of the connected domain polygon, and taking the centroid as the central point of the anticlockwise rotation. The centroid calculation formula is as follows:

2) calculating the included angle between each boundary point p and the centroid o;

3) the bubble method is used for sorting the included angles from small to large, if the included angles are the same, the distance between the included angles and the centroid is calculated, and the small person ranks in front.

FIG. 7 is a diagram of a theoretical model for boundary labeling, centroid calculation and area calculation, with a connected domain in the model consisting of 38 elements. The model boundary polygon is designed to be convex and concave, so that the model boundary polygon is consistent with the actual reservoir. Boundary points (18) that are easily identified longitudinally by the aforementioned scanning method: (2,2), (1,3), (4,3), (3,4), (6,4), (3,5), (7,5), (2,6), (7,6), (2,7), (7,7), (3,8), (6,8), (2,9), (5,9), (2,10), (3,10), (2, 11); lateral complement identified boundary points (4): (2,3),(3,3),(4,9),(5,4). Fig. 8 shows the results after longitudinal and transverse scanning.

And S5 calculation of the area of the connected domain. For n points (x)₁,y₁),(x₂,y₂),...,(x_n,y_n) The points of the formed polygon are sequentially ordered in a counterclockwise way, and the area calculation formula is as follows:

from the previous formula, the centroid can be found, which for this model is (4, 6). The angle of each boundary point can be calculated, and the human-computer interaction can be further corrected according to the concave-convex characteristics of the boundary to obtain an ideal boundary sequence. Finally, the connected domain area is 25, and the result is completely correct, as shown in FIG. 9.

The invention is further illustrated by the following specific application examples:

FIG. 10 shows the results of the eagle mountain group hole detection in a certain actual work area. Wherein (a) 213 holes are found for the result of 10ms identification under the stratum O3 l; (b) for the result of the 25ms identification, a total of 206 slots were found. In the time period, the seam holes are most developed, and the detection and characterization results are completely consistent with the known wells, thereby showing the good application prospect of the method.

For the 205 th via extending 10ms (total 320 points) in fig. 10O3l, the centroid point was calculated as (518,167), and the via area was 272 × 25m — 0.17Km²And the result is correct, and as shown in fig. 11, the method provides great convenience for the calculation of the reserves of the oil field.

In addition, other three-dimensional work areas are tested, the effect is very obvious, and the universality and the stability of the method are fully demonstrated.

Claims (9)

1. A method for detecting and characterizing a slot and hole connected domain is characterized by comprising the following steps:

s1 extracting the coherence of the three-dimensional seismic data and calculating a first characteristic value of the coherence to construct a first characteristic value data body;

s2 selecting five neighborhoods as connected domain analysis point neighborhoods;

s3 connected domain marking, wherein the marking algorithm adopts a one-time scanning method of double-queue recursive search, and comprises the following steps:

s31, searching all non-zero value positions to form a large queue;

s32, establishing and marking a small queue of five neighborhoods and connected domains in the large queue;

s33, starting from the first point, judging whether the large queue data point is communicated with the five neighborhoods, if so, marking, and if not, reducing the length of the large queue by one, and recursing until the length of the large queue is zero;

s34 removing abnormal values, outputting marked slices and recording the serial numbers of connected domains;

s4, marking the connected domain boundary through longitudinal scanning and transverse scanning, calculating the centroid, and calculating the included angle and the distance between the centroid and the boundary point by taking the centroid as the central point to realize the anticlockwise sequencing of the boundary point;

s5 calculates the connected component area.

2. The method for detecting and characterizing a slot and hole connected domain according to claim 1, wherein the step S1 is as follows:

for a certain earthquake three-dimensional data volume, N sampling points of adjacent J channels are taken to form an NxJ earthquake sub-volume, and a matrix D is used_N×JTo express, the matrix C is composed of D_J×J：

Calculating a first eigenvalue λ of C₁And constructing a first characteristic value data body.

3. The method for detecting and characterizing a slot and hole connected domain according to claim 1, wherein the step S31 is as follows:

extracting first characteristic value data according to horizontal slices or stratum slices, setting a threshold coefficient, carrying out threshold processing, and setting zeros smaller than a threshold value;

firstly, according to row i and then according to column j, finding out all non-zero value positions to form a large queue; line number is defined as LINE (k) and track number is defined as CDP (k); k ═ 1,2, …, NP; k is the sequence number of the searched non-zero value data, and NP is the total number of the non-zero value data.

4. The method for detecting and characterizing a slot and hole connected domain according to claim 1, wherein the step S32 is as follows:

starting from the first point of the large queue, i is LINE (1), and j is CDP (1), judging whether the five neighborhood points are communicated, and establishing a corresponding small queue with the longest point of 5 points; setting the first connected domain as F as 1, setting the initial value of the point number contained in the F-th connected domain as NL (F) as 0, setting the head position L of the neighborhood small queue as 1, and setting the mark Y (i, j) as F;

(1) judging and identifying a left side point; if the value X (i, j-1) of the point (i, j-1) is greater than zero and is unmarked, i.e., X (i, j-1) >0& Y (i, j-1) ═ 0, then set Y (i, j-1) ═ F, and find out the point from the queue, set the neighborhood small queue location point L ═ L +1, and set the neighborhood point connectivity flag G1 ═ 1, nl (F) ═ nl (F) + 1; otherwise, setting the neighborhood point non-communication mark G1 as 0, and executing the next step (2);

(2) judging and identifying a right side point; if X (i, j +1) >0& Y (i, j +1) ═ 0, then Y (i, j +1) ═ F, and then the neighborhood small queue location point L ═ L +1, and then the neighborhood point connectivity flag G2 ═ 1, nl (F) ═ nl (F) +1, because this point immediately follows point (i, j), it is not necessary to search for and replace it; otherwise, setting the neighborhood point non-communication mark G2 as 0, and executing the next step (3);

(3) judging and identifying a lower point on the left side; if X (i +1, j-1) >0& Y (i +1, j-1) ═ 0, then set Y (i +1, j-1) ═ F, find out the point from the queue, replace it to neighborhood small queue location point L ═ L +1, set the neighborhood point connectivity flag G3 ═ 1, nl (F) ═ nl (F) + 1; otherwise, setting the neighborhood point non-communication mark G3 as 0, and executing the next step (4);

(4) judging and identifying a lower point; if X (i +1, j) >0& Y (i +1, j) ═ 0, then set Y (i +1, j) ═ F, find out the point from the queue, set neighborhood small queue location point L ═ L +1, set the neighborhood point connectivity flag G4 ═ 1, nl (F) ═ nl (F) + 1; otherwise, setting the neighborhood point non-communication mark G4 as 0, and executing the next step (5);

(5) judging and identifying a lower right point; if X (i +1, j +1) >0& Y (i +1, j +1) ═ 0, then set Y (i +1, j +1) ═ F, find out the point from the queue, set neighborhood small queue location point L ═ L +1, set the neighborhood point connectivity flag G5 ═ 1, nl (F) ═ nl (F) + 1; otherwise, if the neighborhood point non-communication flag G5 is set to 0, step S33 is executed.

5. The method for detecting and characterizing a slot and hole connected domain according to claim 1, wherein the step S33 is as follows:

large queue length minus 1: NP-1;

(1) if NP is 0, go to step S34;

(2) otherwise, finishing and dequeuing the 1 st point of the large queue, and reordering the queue; the queue is shifted left 1 bit from 2 to NP;

1) if G1+ G2+ G3+ G4+ G5 is 0, then F is F +1, indicating the end of the first connected domain;

2) otherwise, the first connected domain mark is not finished;

cases 1) and 2) both return to step S32, continuing the determination and identification.

6. A method for detecting and characterizing a slot and hole connected domain according to claim 1, wherein in step S3: the connected body identification can record the number and the number of connected points of each connected domain, and unreasonable values are removed during output; the user can set the maximum and minimum connection points by himself, so that small connecting bodies without mining value and unreasonable connection areas can be removed.

7. The method for detecting and characterizing a slot and hole connected domain according to claim 1, wherein the content of the longitudinal and transverse scanning in step S4 is as follows:

longitudinal scanning is firstly carried out to obtain longitudinal boundary points, and boundary marking is carried out; then transversely scanning, and if the boundary points are not marked, supplementing the boundary points; finally, the boundary points are output according to a reverse time needle sequence, the initial value of the boundary mark array is set to be zero (Y (i, j) ═ 0, i ═ 1, M, j ═ 1, N), and the initial value l ═ 0 of the boundary point queue;

longitudinal scanning method

1) Taking out the point X (1, j) of the jth column and row 1, if the point is not zero, it is directly the boundary point, and the mark Y (1, j) is 1, l +1, line (l) is 1, cdp (l) is j;

2) rows i-2 to i-M-1, if X (i, j) >0 and X (i +1, j) >0 or X (i-1, j) ═ 0, then (i, j) is a boundary point, let Y (i, j) ═ 1, l ═ l +1, line (l) ═ i, cdp (l) ═ j, allowing multiple boundary points to appear in a column;

3) taking out a point X (M, j) in the jth column and mth row, if the point is not zero, the point is directly a boundary point, and the mark Y (M, j) is 1, l +1, line (l) is M, cdp (l) is j;

② transverse scanning method

1) Taking out the point X (i,1) on the ith row and the 1 st column, if not marked and the point value is not zero, directly the boundary point, marking Y (i,1) ═ 1, l ═ l +1, line (l) ═ i, cdp (l) ═ 1;

2) columns from j 2 to j N-1, if unlabeled and if X (i, j) >0 and X (i, j +1) ═ 0 or X (i, j-1) ═ 0, (i, j) is a boundary point, let Y (i, j) ═ 1, l ═ l +1, line (l) ═ i, cdp (l) ═ j, allowing multiple boundary points to appear on a row;

3) taking out the point X (i, N) on the ith row and nth column, if not marked and if the point value is not zero, directly being a boundary point, marking Y (i, N) 1, l +1, line (l) i, cdp (l) N;

finally, a boundary point queue (line (l), cdp (l)) is obtained, wherein l is 1,2, …, LP; LP is the result of the previous boundary point count.

8. The method for detecting and characterizing a slot-and-hole connected domain according to claim 1, wherein the centroid calculation and boundary point reverse-time-needle ordering algorithm in step S4 comprises the following steps:

1) calculating the centroid o of the connected domain polygon, and taking the centroid as the central point of the anticlockwise rotation; centroid calculation formula:

2) calculating the included angle between each boundary point p and the centroid o;

wherein, LINE (p) represents the line number of the boundary point p, CDP (p) represents the track number of the boundary point p;

3) sorting the included angles from small to large by a bubbling method, if the included angles are the same, calculating the distance between the included angles and the centroid, and arranging the small in front of the centroid;

9. a method for detecting and characterizing a connected region of a slot and a hole as claimed in claim 1, wherein the area of the connected region in step S5 is calculated as follows:

for n points (x)₁,y₁),(x₂,y₂),...,(x_n,y_n) The points of the formed polygon are sequentially ordered in a counterclockwise way, and the area calculation formula is as follows:

Images