CN115639600A - Automatic Seismic Facies Identification Method Based on Region Growth Technique - Google Patents

Abstract

The invention provides a seismic facies automatic identification method based on a region growing technology, which comprises the following steps: step 1, converting original seismic data into stratum domain seismic data through Wheeler domain conversion; step 2, determining seismic attributes sensitive to seismic facies classification; step 3, selecting seed points; step 4, setting seismic facies plane and vertical cut-off conditions; step 5, executing the growth process of the seed region according to the cutoff condition; step 6, forcibly fusing the regions without categories into the adjacent most similar and classified regions; step 7, acquiring seismic facies recognition results of slices in the target layer every 1ms and the like based on the sensitive attributes; and 8, synthesizing the identification results of the slices into a three-dimensional seismic facies boundary. The method for automatically identifying the seismic facies based on the region growing technology can quickly and accurately identify the seismic facies boundary in actual seismic data, is simple to operate, has high three-dimensional seismic facies picking efficiency and accuracy, and has practical popularization and application significance.

Description

Automatic Seismic Facies Identification Method Based on Region Growth Technology

technical field

The invention relates to the technical field of oil and gas exploration, in particular to an automatic identification method of seismic facies based on region growth technology.

Background technique

In geological research, the definition of sedimentary facies is the sum of the environment, conditions and characteristics of sediments, and the analysis of sedimentary facies is also an important basic work in the field of geological research and oil and gas exploration, and its reliability even directly determines the The success or failure of oil and gas exploration. In the period when 3D seismic data were not popularized, the overall understanding of sedimentary facies in a region was often based on the analysis of well logging facies, and the development of sedimentary facies between wells was inferred based on the geological understanding of the region. "One-hole view" lacks actual evidence, and there are great problems in the analysis of sedimentary facies between wells.

With the wide application of 3D seismic data, the understanding of regional sedimentary facies can be obtained through the analysis and transformation of seismic facies on the basis of fine calibration of well seismic, and seismic facies can be understood as the appearance of sedimentary facies on seismic sections. In sum, the spatial distribution characteristics of sedimentary facies can be established by analyzing the seismic facies in 3D seismic data, combined with information such as drilling and provenance direction.

The current seismic facies analysis methods mainly include waveform classification method, seismic attribute feature mapping method, seismic geomorphology facies division method and so on. Among them, the waveform classification method is based on various attributes of the arrangement and combination of events in the seismic reflection interface, such as disorder, wave shape, and composite waveform, and adopts the method of multivariate statistics to classify and use in the analysis of seismic phases. This method has a calculation method It is simple and has the advantages of fast calculation speed, but it is easily affected by the noise in the seismic data, which makes the distribution of the obtained seismic phase map messy, which is quite different from the existing geological laws; the seismic attribute feature mapping method is the basis for the calculation of a large number of seismic attributes In terms of well-seismic combination, multiple dominant seismic attributes that are sensitive to sedimentary facies are selected, and seismic facies are generated based on the characteristics of different dominant seismic attributes. The advantage of this method is that it accepts the soft constraints of the well and has higher accuracy. The influence of lithology, faults, and fluids has multiple solutions for the interpretation of sedimentary facies; the seismic geomorphology method combines high-precision 3D seismic exploration technology with sedimentology and geomorphology, and uses high-resolution seismic data. This method also requires the use of a large number of seismic attributes for calculation but is not restricted by wells. The prediction accuracy is high, but the calculation is relatively complicated and difficult to achieve compared with other methods.

In addition, there are two common problems in the above three commonly used methods: First, a large number of manual interpretations are needed to realize the analysis of 3D seismic facies, and the nonlinear representation ability of human beings is limited, and the interpretation experience varies from person to person, which inevitably leads to The reliability and efficiency of the interpretation results are reduced; secondly, they only stay at the stage of two-dimensional planar seismic phase analysis, and cannot realize the analysis and mapping of three-dimensional seismic phases.

Therefore, how to obtain seismic facies features faster and with higher precision is an urgent need for researchers, especially in the face of various geological phenomena overlapping in a short time and in a small area in the later stage of exploration, this problem becomes more and more serious. It is urgent to introduce new data analysis techniques.

In the Chinese patent application with application number: CN201611198747.8, it involves a quantitative characterization method for the activity intensity of growth faults. Among them, the method includes: constructing a high-precision sequence stratigraphic framework in the target interval; performing spatial data rotation based on the fracture polygon and sequence interface data to obtain the rotated fracture polygon and sequence interface data; The polygon and sequence interface data, combined with the spatial geometric relationship, determines the total vector slip distance of the fault; the spatial multiple cross-correlation algorithm is used to realize the three-dimensional spatial positioning of the same fault at the top and bottom interfaces of the same sequence; according to the top and bottom interfaces of the same sequence Based on the 3D spatial positioning results of the same faults, the total vector slip distances at the corresponding breakpoints of the faults are matched, and the matching difference operation is performed to obtain the sequence vector slip distances of the synsedimentary fault system during the sequence formation process.

In the Chinese patent application with application number: CN201510696919.3, it relates to a method for describing the growth of reverse faults in small and medium-sized extensional fault basins, which belongs to the field of geological exploration technology. The method includes: collecting and processing data in the area to be identified; performing sequence division under the guidance of sequence stratigraphy theory, and establishing an isochronous sequence stratigraphic framework in the area to be identified; constraining the isochronous sequence stratigraphic framework in the area to be identified Next, perform structural-stratigraphic linkage interpretation, identify and divide normal and reverse faults; perform stratigraphic-sedimentary linkage interpretation on the divided normal and reverse faults, eliminate false faults that have been misidentified, and establish the fault framework of the area to be identified; Perform structural-sedimentary linkage interpretation on the normal and reverse faults in the fault framework of the area to be identified, establish the structural framework of growth faults, and determine the growth reverse faults; construct the structural geological model of growth reverse faults and its derived structural geological models, analyze The origin of growth reverse faults in each structural geological model guides oil and gas exploration in areas to be identified.

In the Chinese patent application with application number: CN202110092388.2, it involves a method for identifying hidden faults based on fault growth mechanism and seismic attribute prediction. The method comprises the following steps: (a) interpreting the structure of seismic data, counting the growth and development characteristics of regional faults, and determining the main fault; (b) according to the relationship between the main fault throw and distance position change determined in step (a), Make the "fault distance-distance" curve; (c) find the point where the fault distance value of the "fault distance-distance" curve decreases, and preliminarily predict the location of hidden fault development; (d) scan and extract the fault attributes for the post-stack space 3D seismic body Feature map, judging the fault path according to the abnormal response in the map; (e) when the predicted position in step (c) also meets the fault path response of step (d) fault attribute map identification, it is interpreted as a concealed fault.

The above existing technologies are quite different from the present invention, and cannot solve the technical problems we want to solve. Therefore, we have invented a new method for automatic identification of seismic facies based on region growing technology.

Contents of the invention

The purpose of the present invention is to provide a seismic facies automatic identification method based on the region growing technology, which has good application effect on the identification and division of three-dimensional seismic facies.

The object of the present invention can be realized by following technical measure: the seismic facies automatic recognition method based on the regional growth technology, this seismic facies automatic recognition method based on the regional growth technology comprises:

Step 1, convert the original seismic data into stratigraphic domain seismic data through Wheeler domain conversion;

Step 2, determine the seismic attributes that are sensitive to seismic facies classification;

Step 3, select the seed point;

Step 4, setting the seismic phase plane and vertical cut-off conditions;

Step 5, execute the seed region growth process according to the cut-off condition;

Step 6. Forcibly fuse the regions that still have no category into the most similar and category-like regions in the vicinity;

Step 7, based on the sensitivity attribute, obtain the seismic facies identification result of every 1ms isochronous slice in the target layer;

In step 8, the recognition results of each slice are synthesized into a three-dimensional seismic phase boundary.

The purpose of the present invention can also be achieved through the following technical measures:

Step 1 includes:

(1) Establish the isochronous stratigraphic framework of the work area on the basis of tracking the main layers of the target interval through automatic interpretation or manual interpretation in the seismic data;

(2) Carry out fine tracking of sub-layers inside the above-mentioned stratigraphic framework, and further divide each period of sedimentary events in the target interval;

(3) Horizontal leveling is carried out according to a certain geological horizon above the target interval, so as to obtain stratigraphic domain seismic data.

In step 2, the well-drilled log facies are compared with different seismic attributes to select seismic attributes that are sensitive to seismic facies classification.

In step 3, the selection of seed points is based on the following two basic principles:

(1) The well point on each layer slice must be used as the seed point, because on the basis of fine calibration, the well logging facies on the well can assign sedimentary facies labels to the seismic facies next to the well;

(2) Non-well points that can reflect the typical morphology of sedimentary facies can also be used as seed points.

In step 4, the seismic phase plane cut-off condition is:

(1) Calculate the gray mean value m of the seed point area, set the seed point image area R, where the number of pixels is N, then the gray mean value of the seed point area is expressed as:

(2) Compare the gray mean value of the neighborhood area with the gray value of the seed point area, if the absolute value of the two is less than the threshold, then merge them:

max|f(x, y)-m| _{(x, y)∈0} <K

Among them, f(x, y) is the average gray value of the neighborhood area, m is the average gray value of the seed point area, and K is the set threshold.

In step 4, the vertical cut-off condition of the seismic phase is:

(1) Carry out well-seismic fine calibration, convert the well logging data from the depth domain to the time domain, and make a good match with the seismic;

(2) Determine the cut-off condition for the vertical growth of the seismic facies next to the well by using the well logging facies interpretation data on the well.

Step 5 includes:

(1) Use the seed point selected in step 3 as the starting point of growth, and according to the growth cut-off condition provided in step 4, judge whether there are pixels that meet the growth criteria in the neighborhood of the current region, and if so, merge them into the grown region , so as to complete an iteration;

(2) Start the second iteration according to the principle and method of the first iteration, until no neighboring pixels satisfying the conditions are divided into the grown region, the region growing algorithm ends.

In step 6, for the pixels that are missed or over-segmented in the merging process, the threshold K in step 4 is appropriately reduced, and the merging calculation is performed on the above-mentioned pixels, so as to force fusion to the adjacent most similar and classified area middle.

Step 7 includes:

(1) Obtain a large number of isochronous slices at intervals of 1 ms for the stratigraphic domain seismic data obtained through Wheeler domain conversion in step 1;

(2) Use the seed points provided in step 3, the cut-off conditions provided in step 4, and the region growing method provided in step 5 to identify the sensitive attributes in each isochronous slice.

In step 8, the identification results of seismic facies boundaries in different slices are integrated, and synthesized into a three-dimensional seismic facies boundary in space.

The seismic facies automatic identification method based on the region growing technology in the present invention, on the basis of making full use of the well data as a hard constraint, gives the plane cut-off condition and the vertical cut-off condition of the growth according to the data characteristics of the actual work area, through the advanced region The growth algorithm realizes automatic identification and division of 3D seismic facies. This method can quickly and accurately identify the seismic phase boundary in the actual seismic data only by simply giving the seed point, growth criterion and growth stop condition. The operation is simple, and the 3D seismic phase picking efficiency and accuracy are high. , which greatly reduces the workload and has practical significance for popularization and application.

Description of drawings

Fig. 1 is the flow chart of a specific embodiment of the seismic phase automatic identification method based on the region growing technology of the present invention;

Fig. 2 is a schematic diagram of the distribution of 23 seed points in a specific embodiment of the present invention;

Fig. 3 is a schematic diagram of the vertical cut-off conditions of central shoal microfacies and underwater distributary channel microfacies of a specific embodiment of the present invention;

Fig. 4 is a three-dimensional seismic facies distribution map automatically constructed by the region growing technique in a specific embodiment of the present invention.

Detailed ways

It should be noted that the following detailed description is exemplary and intended to provide further explanation of the present invention. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

It should be noted that the terminology used here is only for describing specific embodiments, and is not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly dictates otherwise, the singular is intended to include the plural, and it should also be understood that when the terms "comprising" and/or "comprising" are used in this specification, they mean There are features, steps, operations and/or combinations thereof.

The seismic phase automatic recognition method based on the region growing technique of the present invention comprises the following steps:

Step 1, transform the original seismic data into stratigraphic domain seismic data through Wheeler domain conversion; Step 2, determine the seismic attributes sensitive to seismic facies classification; Step 3, select the seed point; Step 4, set the seismic phase plane and vertical cut-off conditions ; Step 5, execute the seed region growth process according to the cut-off condition; Step 6, forcefully merge the regions that still have no category into the adjacent most similar and category-oriented regions; Seismic facies recognition results of time slices; step 8, synthesize the recognition results of each slice into a three-dimensional seismic facies boundary.

The invention can quickly and accurately identify the three-dimensional seismic phase boundary in actual seismic data, greatly improves the precision and efficiency of three-dimensional seismic phase analysis, and has practical popularization and application significance.

The following are several specific embodiments of the application of the present invention. Different embodiments have adopted the technical process provided by the present invention, but the specific calculation parameters are different with reference to the actual conditions of different work areas.

Example 1:

In the specific embodiment 1 of the application of the present invention, as shown in Figure 1, Figure 1 is a flow chart of a method for automatic identification of seismic phases based on the region growing technology of the present invention, which includes the following eight steps in this embodiment And related parameter settings:

The first step is to transform the original seismic data into stratigraphic domain seismic data through Wheeler domain conversion. In this case, the seismic facies identification of a 3D seismic work area in western China is taken as an example. The main target layer in this block is the Jurassic Sangonghe Formation, and K ₁ q, J ₂ x, J ₁ s ₃ , J ₁ s ₂ ^2-2 , J ₁ s ₂ ^2-1 , and J ₁ s ₂ ¹ are constructed on six interpretation horizons. K ₁ q performs stratigraphic leveling to obtain stratigraphic domain seismic data in the work area.

In the second step, the seismic attributes that are sensitive to seismic facies classification are identified. In this case, the sensitive seismic attributes of seismic facies classification selected by comparing the logging facies of actual drilling with different seismic attributes include: root mean square amplitude and arc length.

The third step is to select the seed point. According to the two principles of seed point selection: (1) the well point on each slice must be used as the seed point; (2) the non-well point that can reflect the typical shape of the sedimentary facies can also be used as the seed point, and Fig. 2 is selected in this case In addition to the 20 drilling well points in the work area, 3 non-well point seed points were selected for channel sand bars, sheet sand and coastal shallow lakes.

The fourth step is to set the seismic phase plane and vertical cut-off conditions. The cut-off condition of the plane is: by calculating whether the difference between the gray mean value in the seed point area and the gray level mean value in the neighborhood area is less than the set threshold K to determine whether to merge with the seed point, in this case, the threshold K is set to 0.02; Figure 3 is a schematic diagram of the vertical cut-off conditions of channel microfacies and subaqueous distributary channel microfacies in this embodiment. On the basis of well-seismic fine calibration, the logging facies interpretation data of the drilled well is used as the seismic facies beside the well. Vertical growth determines cut-off conditions.

The fifth step is to execute the seed region growing process according to the cut-off condition. In this case, the 20 well point locations selected in the third step and 3 typical non-well point locations, a total of 23 seed points are used as the growth starting point, and the neighbors in the current area are judged according to the growth cut-off conditions provided in the fourth step. Whether there is a pixel in the domain that meets the growth criterion, through multiple iterations, until no neighborhood pixel that meets the condition is divided into the growing area, the region growing algorithm ends.

In the sixth step, the areas that still have no category are forced to be fused into the most similar adjacent area that has a category. This step is mainly used to deal with pixels that are missed or over-segmented during the merging process. In this case, the threshold K in the fourth step is adjusted to 0.015, and the missed pixels are re-merged.

The seventh step is to obtain the seismic facies identification results of every 1 ms isochronous slice in the target layer based on the sensitive attributes. In this case, we obtained a total of 322 isochronous slices in the stratigraphic domain data, and each slice contains two sensitive attributes of root mean square amplitude and arc length. Based on the above sensitive attributes, 23 seed points in the third step are used , according to the growth cut-off condition provided in the fourth step, the seismic facies identification results of every 1ms isochronous slice are obtained.

In the eighth step, the recognition results of each slice are synthesized into a three-dimensional seismic phase boundary. The area of the 3D work area is 280Km ² , excluding data preparation and other work. The whole identification process of seismic facies takes 3 hours, which greatly improves the efficiency of seismic facies analysis. Fig. 4 is the 3D seismic facies distribution map automatically constructed by the region growing The results are basically consistent with the existing geological knowledge in this area, and also strongly confirm the accuracy of the seismic facies automatic identification method based on the region growing technology of the present invention.

Example 2:

In embodiment 2 of the application of the present invention, similar to embodiment 1, the following eight implementation steps are also included, but different parameter settings are adopted:

The first step is to transform the original seismic data into stratigraphic domain seismic data through Wheeler domain conversion. In this case, the seismic facies identification of a 3D seismic work area in western China is taken as an example. The main target layer in this block is the Jurassic Xishanyao Formation, and K ₁ q, J ₂ x ₂ ² , J ₂ x ₂ ¹ , J ₂ x ₁ ² , J ₂ x ₁ ¹ , and J _{1 s 3} six interpretation levels are constructed. Layer leveling was carried out to obtain the stratigraphic domain seismic data in the work area.

In the second step, the seismic attributes that are sensitive to seismic facies classification are determined. In this case, by comparing the logging facies of the actual well with different seismic attributes, the selected seismic facies classification sensitive seismic attribute is the maximum amplitude.

The third step is to select the seed point. According to the two principles of seed point selection: (1) well points on each slice must be used as seed points; (2) non-well points that can reflect the typical shape of sedimentary facies can also be used as seed points, and 13 well points are selected in this example. seed points, including 10 drilling well points in the work area, and 3 non-well point seed points selected for sheet sand, peat swamp and coastal shallow lake.

The fourth step is to set the seismic phase plane and vertical cut-off conditions. The plane cut-off condition is: by calculating whether the difference between the gray mean value in the seed point area and the gray level mean value in the neighborhood area is less than the set threshold K to determine whether to merge with the seed point, in this embodiment, the threshold K is set to 0.03 ; The vertical cut-off condition is: on the basis of well-seismic fine calibration, the cut-off condition is determined for the vertical growth of seismic facies next to the well by using the well-drilled logging facies interpretation data.

The fifth step is to execute the seed region growing process according to the cut-off condition. In this case, the 10 well point locations selected in the third step and 3 typical non-well point locations, a total of 13 seed points are used as the starting point of growth, and according to the growth cut-off conditions provided in the fourth step, the neighbors of the current area are judged. Whether there is a pixel in the domain that meets the growth criterion, through multiple iterations, until no neighborhood pixel that meets the condition is divided into the growing area, the region growing algorithm ends.

In the sixth step, the areas that still have no category are forced to be fused into the most similar adjacent area that has a category. This step is mainly used to deal with pixels that are missed or over-segmented during the merging process. In this case, the threshold K in the fourth step is adjusted to 0.035, and the missed pixels are re-merged.

The seventh step is to obtain the seismic facies identification results of every 1 ms isochronous slice in the target layer based on the sensitive attributes. In this example, we obtained a total of 256 isochronous slices in the stratigraphic domain data, and each slice contains the maximum amplitude of the sensitive attribute. Based on the above-mentioned sensitive attributes, we use the 13 seed points in the third step, and provide according to the fourth step The growth cut-off conditions of each phase are obtained, and the seismic facies identification results of each 1ms isochronous slice are obtained.

In the eighth step, the recognition results of each slice are synthesized into a three-dimensional seismic phase boundary. The area of the 3D work area is 398Km ² , excluding data preparation and other work. The whole identification process took 6.2 hours. Due to the larger amount of data, the calculation speed has been reduced, but compared with the manual picking of seismic phases, the speed still meets expectations. A good technical implementation effect has been achieved.

Example 3:

In embodiment 3 of the application of the present invention, similar to embodiment 1, the following eight implementation steps are also included, but different parameter settings are adopted:

The first step is to transform the original seismic data into stratigraphic domain seismic data through Wheeler domain conversion. In this case, the seismic facies identification of a 3D seismic work area in eastern China is taken as an example. The main target layer in this block is the Paleogene Shahejie Formation, and Es3z, Es3x, and Es4s ¹ are selected for the establishment of the isochronous stratigraphic framework of the work area. , Es4s ² , and Es4s ³ five interpretation horizons were constructed, and on the basis of fine tracing of the small layers inside the grid, layer leveling was carried out along Es3z, so as to obtain the stratigraphic domain seismic data of the work area.

In the second step, the seismic attributes that are sensitive to seismic facies classification are identified. In this case, by comparing the logging facies of the actual well with different seismic attributes, the sensitive seismic attributes of the seismic facies classification selected are the maximum absolute amplitude and arc length.

The third step is to select the seed point. According to the two principles of seed point selection: (1) well points on each layer slice must be used as seed points; (2) non-well points that can reflect the typical shape of sedimentary facies can also be used as seed points. In this example, a total of 35 well points were selected. 3 seed points, including 32 drilling well points in the work area, and 3 non-well point seed points selected for the fan delta plain water distributary channel, delta plain natural embankment and lake bottom fan.

The fourth step is to set the seismic phase plane and vertical cut-off conditions. The plane cut-off condition is: by calculating whether the difference between the gray mean value in the seed point area and the gray level mean value in the neighborhood area is less than the set threshold K to determine whether to merge with the seed point, in this embodiment, the threshold K is set to 0.015 ; The vertical cut-off condition is: on the basis of well-seismic fine calibration, the cut-off condition is determined for the vertical growth of seismic facies next to the well by using the well-drilled logging facies interpretation data.

The fifth step is to execute the seed region growing process according to the cut-off condition. In this case, the 32 well point locations selected in the third step and 3 typical non-well point locations, a total of 35 seed points are used as the starting point of growth, and according to the growth cut-off conditions provided in the fourth step, the neighbor Whether there is a pixel in the domain that meets the growth criterion, through multiple iterations, until no neighborhood pixel that meets the condition is divided into the growing area, the region growing algorithm ends.

In the sixth step, the areas that still have no category are forced to be fused into the most similar adjacent area that has a category. This step is mainly used to deal with pixels that are missed or over-segmented in the merging process. In this case, the threshold K in the fourth step is adjusted to 0.01, and the missed pixels are re-merged.

The seventh step is to obtain the seismic facies identification results of every 1 ms isochronous slice in the target layer based on the sensitive attributes. In this example, we obtained a total of 302 isochronous slices in the stratigraphic domain data, and each slice contains the maximum absolute amplitude and arc length of sensitive attributes. Based on the above sensitive attributes, we use 35 seed points in the third step, according to The growth cut-off condition provided in the fourth step obtains the seismic facies identification result of every 1ms isochronous slice.

In the eighth step, the recognition results of each slice are synthesized into a three-dimensional seismic phase boundary. The area of the 3D work area is 223Km ² , excluding data preparation and other work. The whole identification process takes 5.8 hours. Since the 3D data is high-density seismic data, the bins are small and the amount of data is large, the calculation speed is affected. However, compared with manual picking In the seismic phase, the speed still meets expectations, and at the same time, good technical implementation results have been achieved.

The present invention can be applied to the identification and analysis of seismic facies in the field of oil and gas exploration. On the basis of making full use of well data as hard constraints, the method specifies the plane cut-off conditions and vertical cut-off conditions for growth according to the data characteristics of the actual work area. Advanced region growing algorithm realizes automatic identification and division of 3D seismic facies. This method can quickly and accurately identify the seismic phase boundary in the actual seismic data only by simply giving the seed point, growth criterion and growth stop condition. The operation is simple, and the 3D seismic phase picking efficiency and accuracy are high. , which greatly reduces the workload and has practical significance for popularization and application.

Finally, it should be noted that: the above is only a preferred embodiment of the present invention, and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, for those skilled in the art, it can still The technical solutions described in the foregoing embodiments are modified, or some of the technical features are equivalently replaced. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Except for the technical features described in the instructions, all are known technologies by those skilled in the art.

Claims (10)

1. The method for automatic identification of seismic facies based on regional growth technology is characterized in that the method for automatic identification of seismic facies based on regional growth technology comprises: Step 1, convert the original seismic data into stratigraphic domain seismic data through Wheeler domain conversion; Step 2, determine the seismic attributes that are sensitive to seismic facies classification; Step 3, select the seed point; Step 4, setting the seismic phase plane and vertical cut-off conditions; Step 5, execute the seed region growth process according to the cut-off condition; Step 6. Forcibly fuse the regions that still have no category into the most similar and category-like regions in the vicinity; Step 7, based on the sensitivity attribute, obtain the seismic facies identification result of every 1ms isochronous slice in the target layer; In step 8, the recognition results of each slice are synthesized into a three-dimensional seismic phase boundary. 2. the method for automatic identification of seismic phases based on region growing technology according to claim 1, is characterized in that, step 1 comprises: (1) Establish the isochronous stratigraphic framework of the work area on the basis of tracking the main layers of the target interval through automatic interpretation or manual interpretation in the seismic data; (2) Carry out fine tracking of sub-layers inside the above-mentioned stratigraphic framework, and further divide each period of sedimentary events in the target interval; (3) Horizontal leveling is carried out according to a certain geological horizon above the target interval, so as to obtain stratigraphic domain seismic data. 3. The method for automatic identification of seismic facies based on region growing technology according to claim 1, characterized in that, in step 2, the logging phase of the real drilling well is compared with different seismic attributes to select earthquakes sensitive to seismic facies classification Attributes. 4. the method for automatic identification of seismic phases based on region growing technology according to claim 1, is characterized in that, in step 3, selects seed point based on following two basic principles: (1) The well point on each layer slice must be used as the seed point, because on the basis of fine calibration, the well logging facies on the well can assign sedimentary facies labels to the seismic facies next to the well; (2) Non-well points that can reflect the typical morphology of sedimentary facies can also be used as seed points. 5. the method for automatically identifying seismic phases based on region growing technology according to claim 1, is characterized in that, in step 4, the seismic phase plane cut-off condition is: (1) Calculate the gray mean value m of the seed point area, set the seed point image area R, where the number of pixels is N, then the gray mean value of the seed point area is expressed as:

(2) Compare the gray mean value of the neighborhood area with the gray value of the seed point area, if the absolute value of the two is less than the threshold, then merge them: max|f(x,y)=-m| _(x,y)∈0 <K Among them, f(x, y) is the average gray value of the neighborhood area, m is the average gray value of the seed point area, and K is the set threshold. 6. the method for automatic identification of seismic phases based on region growing technology according to claim 5, is characterized in that, in step 4, the vertical cut-off condition of seismic phases is: (1) Carry out well-seismic fine calibration, convert the well logging data from the depth domain to the time domain, and make a good match with the seismic; (2) Determine the cut-off condition for the vertical growth of the seismic facies next to the well by using the well logging facies interpretation data on the well. 7. the method for automatic identification of seismic phases based on region growing technology according to claim 1, is characterized in that, step 5 comprises: (1) Use the seed point selected in step 3 as the starting point of growth, and according to the growth cut-off condition provided in step 4, judge whether there are pixels that meet the growth criteria in the neighborhood of the current region, and if so, merge them into the grown region , so as to complete an iteration; (2) Start the second iteration according to the principle and method of the first iteration, until no neighboring pixels satisfying the conditions are divided into the grown region, the region growing algorithm ends. 8. The method for automatic identification of seismic phases based on region growing technology according to claim 1, characterized in that in step 6, for pixels that are omitted or over-segmented in the merging process, the threshold K in step 4 is appropriately Reduce, re-combine and calculate the above pixels, so as to force fusion into the adjacent most similar and classified area. 9. the method for automatic identification of seismic phases based on region growing technology according to claim 1, is characterized in that, step 7 comprises: (1) Obtain a large number of isochronous slices at intervals of 1 ms for the stratigraphic domain seismic data obtained through Wheeler domain conversion in step 1; (2) Use the seed points provided in step 3, the cut-off conditions provided in step 4, and the region growing method provided in step 5 to identify the sensitive attributes in each isochronous slice. 10. The method for automatic identification of seismic facies based on region growing technology according to claim 1, characterized in that in step 8, the identification results of seismic facies boundaries in different slices are synthesized into three-dimensional seismic facies boundaries in space.

Images