Shale oil and gas well site classification method and shale oil and gas development activity monitoring method

Abstract

The invention provides a shale oil and gas well site classification method and a shale oil and gas development activity monitoring method, and belongs to the technical field of remote sensing geoscience application, and the classification method comprises the steps: determining the position information of each well site in a target region; according to the position information, extracting a first spectral feature and a second spectral feature of the well site from each target remote sensing image; sorting the first spectral features and the second spectral features of each well site, and correspondingly constructing a first spectral feature time sequence and a second spectral feature time sequence of each well site; based on the first spectral feature time sequence and the second spectral feature time sequence, respectively constructing a total spectral feature time sequence of each well site; respectively carrying out clustering analysis on the total spectral characteristic time group of each total spectral characteristic time sequence to obtain a clustering result of each well site; and based on a clustering result, obtaining a well site state category of each well site. According to the shale oil and gas well field classification method, high-accuracy classification of the shale oil and gas well field is achieved, and effective monitoring of shale oil and gas development activities is facilitated.

Description

Technical Field

[0001] The present application belongs to the field of remote sensing geoscience application technology, and in particular, relates to a shale oil and gas well site classification and shale oil and gas development activity monitoring method. Background Art

[0002] At present, it is necessary to classify the well sites in the region based on the survival status of shale oil and gas well sites, so as to take corresponding prevention and control measures for the areas where the well sites of each category are located.

[0003] Optical remote sensing images contain surface information. With the development of deep learning technology, deep learning technology is used in the prior art to identify shale oil and gas well sites based on long-term optical remote sensing images, so as to quickly obtain the spatiotemporal distribution of shale oil and gas well sites. Based on the spatiotemporal distribution of shale oil and gas well sites, the existence status of each shale oil and gas well site in the target time period is obtained. However, due to the small land area occupied by shale oil and gas well sites and the complex land background, the background of shale oil and gas well sites is diverse, such as shale oil and gas well sites located in mountainous areas, bare land, blocks, woodlands, farmlands and / or rivers and lakes. At the same time, due to the high similarity between the spectra of various ground features such as bare land and impermeable surfaces and the spectra of shale oil and gas well sites, a lot of noise interference is caused to the spectra of shale oil and gas well sites, resulting in low accuracy of the prior art in obtaining shale oil and gas well sites in a large range and long time period, and it is difficult to accurately obtain the existence status of shale oil and gas well sites in the target time period, and the classification accuracy of each shale oil and gas well site in the target area is limited.

[0004] Based on this, how to accurately extract the survival status of shale oil and gas well sites over a long period of time and over a large area and improve the classification accuracy of shale oil and gas well sites are important issues that need to be solved urgently. Summary of the invention

[0005] In view of the above-mentioned shortcomings of the prior art, the purpose of the present application is to provide a shale oil and gas well site classification method to solve the problems of the existing shale oil and gas well site classification method, such as low accuracy in obtaining the survival status of shale oil and gas well sites for a long time and a large range, and difficulty in accurately classifying shale oil and gas well sites based on the survival status of shale oil and gas well sites.

[0006] To achieve the above-mentioned and other related purposes, the present invention provides a shale oil and gas well site classification method, comprising the following steps: Determine the location information of each well site in the target area; Extracting, according to the location information of the well site, a first spectral feature and a second spectral feature corresponding to the well site in each target remote sensing image of the target area within a target time period; The first spectral features and the second spectral features of each well site are respectively sorted according to the acquisition time of the target remote sensing image, and the first spectral feature time series and the second spectral feature time series of each well site are correspondingly constructed; based on the first spectral feature time series and the second spectral feature time series of each well site, a total spectral feature time series of each well site is respectively constructed; the total spectral feature time series includes a total spectral feature time group; Performing cluster analysis on the total spectral feature time groups of each total spectral feature time series respectively to obtain clustering results of each well site; classifying each well site based on the clustering results to obtain a well site status category of each well site; The first spectral feature and the second spectral feature are pre-constructed spectral features that characterize changes in the well site before and after its existence.

[0007] In one embodiment of the present invention, the method for acquiring the location information of each well site in the target area includes: Acquiring remote sensing images of each of the targets; Using the trained wellsite extraction model, perform wellsite extraction on each of the target remote sensing images to obtain a first wellsite extraction result corresponding to each of the target remote sensing images; Based on the image geographic location of the target remote sensing image, geographic coordinate conversion is performed on the image coordinates of each of the first well site extraction results to obtain the spatial position distribution corresponding to each of the first well site extraction results; based on the spatial position distribution corresponding to each of the first well site extraction results, the location information of each of the well sites is obtained accordingly.

[0008] In one embodiment of the present invention, in order to improve the accuracy of obtaining the location information of the well site, after obtaining the spatial location distribution corresponding to each of the first well site extraction results, the determining the location information of each well site in the target area further includes: Acquire the building land data of the target area within the target time period; construct building area masks corresponding to the acquisition time and image geographical location of each of the target remote sensing images based on the building land data; match each of the building area masks with each of the target remote sensing images according to the principle of time-space correspondence matching; use each of the building area masks to perform mask processing on the first well site extraction result of the target remote sensing image matched therewith, so as to remove the building land mistakenly extracted as the well site in the first well site extraction result, and obtain each of the well sites and their location information based on the remaining first well site extraction results.

[0009] In one embodiment of the present invention, the method for acquiring the first spectral feature and the second spectral feature includes: Acquire the spectral band change characteristics of the well site; construct a well site extraction index based on two spectral bands in the spectral band change characteristics as the first spectral characteristic of the well site; extract a spectral band from the remaining spectral band change characteristics as the second spectral characteristic of the well site; The spectral band change characteristics include: in the same remote sensing image product, the top N spectral bands with the largest degree of change before and after the well site exists; N is greater than or equal to 3.

[0010] In one embodiment of the present invention, cluster analysis is performed on the total spectral feature time groups of each of the total spectral feature time series to obtain clustering results of each of the well fields. When cluster analysis is performed on the total spectral feature time groups in a single total spectral feature time series, the method includes: Using the acquisition time, the first spectral feature and the second spectral feature as clustering variables, respectively, performing three-dimensional clustering on the total spectral feature time group in the total spectral feature time series to obtain cluster clusters; Detecting whether the number of the clusters is 1, if so, obtaining the clusters as the clustering results of the well site; if not, dividing each cluster into a high mean cluster and a low mean cluster according to the size of the first spectral feature mean or the second spectral feature mean; The three-dimensional clustering is performed again on the total spectral feature time group in the high-mean clustering cluster to obtain the high-mean clustering subcluster of the high-mean clustering cluster; and the three-dimensional clustering is performed again on the total spectral feature time group in the low-mean clustering cluster to obtain the low-mean clustering subcluster of the low-mean clustering cluster, and the high-mean clustering cluster, the low-mean clustering cluster, the high-mean clustering subcluster and the low-mean clustering subcluster are used as the clustering results of the well site.

[0011] In one embodiment of the present invention, the well sites are classified based on the clustering result to obtain the well site status category of each well site. When obtaining the well site status category of a single well site, the method includes: It is detected whether there is only one cluster in the clustering result, or whether the number of high-mean cluster subclusters and the number of low-mean cluster subclusters in the clustering result are both less than or equal to 1. If so, the well site is determined to be a persistent well site; if not, the well site is determined to be a non-persistent well site.

[0012] In one embodiment of the present invention, after obtaining the well site status category of each of the well sites, the shale oil and gas well site classification method further includes: performing a persistent existence category detection on each of the persistent well sites, and when performing the persistent existence category detection on a single persistent well site, it includes: Detect whether there is only one cluster in the clustering result of the persistent well field, whether the persistent well field meets the preset spectral feature requirements, or whether the start time and end time of a cluster sub-cluster in the clustering result of the persistent well field are respectively within the start time period and the end time period of the target time period. If so, determine that the persistent existence category of the persistent well field is long-term persistent existence in the target time period; if not, It is then detected whether the start time of the low-mean cluster of the persistent well field is earlier than the start time of its high-mean cluster. If so, the persistent category of the persistent well field is determined to be persistent in the later period of the target time period; if not, the persistent category of the persistent well field is determined to be persistent in the early period of the target time period.

[0013] The present invention also provides a method for monitoring shale oil and gas development activities, comprising: using the shale oil and gas well site classification method as described above to obtain a non-persistent well site, and performing construction time node extraction and end time node extraction on the non-persistent well site; wherein, The construction time node extraction includes: obtaining a low-mean cluster subcluster located in a starting time period as a starting cluster cluster; based on a preset first sliding step size and a first time window size, starting from the latest total spectral feature time group, gradually moving the first time window on the starting cluster cluster in descending time order until, in the current first time window, the number of the total spectral feature time groups of the low-mean cluster subcluster is greater than the number of the total spectral feature time groups of the high-mean cluster cluster, obtaining the end time of the low-mean cluster subcluster in the current first time window as the construction time node; The end time node extraction includes: obtaining the low mean cluster sub-cluster located in the end time period as the end cluster cluster; based on the preset second sliding step size and the second time window size, starting from the earliest total spectral feature time group, gradually moving the second time window on the end cluster cluster in ascending time order until the number of the total spectral feature time groups of the low mean cluster sub-cluster in the current second time window is greater than the number of the total spectral feature time groups of the high mean cluster cluster, and obtaining the start time of the low mean cluster sub-cluster in the current second time window as the end time node; Among them, the starting time period is a time period in the target time period, including the starting moment of the target time period and having a preset first time length; the ending time period is a time period in the target time period, including the ending moment of the target time period and having a preset second time length.

[0014] In one embodiment of the present invention, after obtaining the well sites that continue to exist in the later period of the target time period and the well sites that continue to exist in the early period of the target time period, the shale oil and gas development activity monitoring method further includes: for the well sites that continue to exist in the later period of the target time period, executing the construction time node extraction to obtain the construction time node of the well sites; for the well sites that continue to exist in the early period of the target time period, executing the end time node extraction to obtain the end time node of the well sites.

[0015] Correspondingly, the present invention provides a shale oil and gas well site classification system, comprising: A location determination module, used to determine the location information of each well site in the target area; A feature extraction module is used to extract a first spectral feature and a second spectral feature corresponding to the well site in each target remote sensing image of the target area within a target time period according to the location information of the well site; the first spectral feature and the second spectral feature are pre-constructed spectral features that characterize changes in the well site before and after its existence; A feature time series construction module is used to sort the first spectral features and the second spectral features of each well site according to the acquisition time of the target remote sensing image, and correspondingly construct the first spectral feature time series and the second spectral feature time series of each well site; based on the first spectral feature time series and the second spectral feature time series of each well site, respectively construct a total spectral feature time series of each well site; the total spectral feature time series includes a total spectral feature time group; The classification module is used to perform cluster analysis on the total spectral feature time groups of each total spectral feature time series to obtain clustering results of each well site; based on the clustering results, classify each well site to obtain a well site status category of each well site.

[0016] As described above, the shale oil and gas well site classification method provided in this application has at least the following beneficial effects: The method comprises the following steps: determining the location information of each well site in the target area; extracting the first spectral feature and the second spectral feature corresponding to the well site in each target remote sensing image of the target area within the target time period according to the location information of the well site; sorting the first spectral feature and the second spectral feature of each well site according to the acquisition time of the target remote sensing image, and correspondingly constructing the first spectral feature time series and the second spectral feature time series of each well site; constructing the total spectral feature time series of each well site based on the first spectral feature time series and the second spectral feature time series of each well site; clustering analysis is performed on the total spectral feature time groups in each total spectral feature time series based on the first spectral feature and the second spectral feature that can characterize the changes of the well site before and after the existence of the well site, and obtaining the clustering results of each well site; and classifying each well site based on the clustering results to obtain the well site status category of each well site. The method accurately obtains the well site existence status in the target time period in the geographical location area where the well site is located based on the pre-constructed first spectral feature and the second spectral feature, and improves the classification accuracy of a large range of shale oil and gas well sites based on the well site existence status. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A schematic diagram of a process flow of a shale oil and gas well site classification method provided in the present application in one embodiment is shown.

[0018] Figure 2 A schematic diagram showing the flow of the shale oil and gas well site classification method provided in the present application in another embodiment.

[0019] Figure 3 A schematic diagram showing a flow chart of the shale oil and gas well site classification method provided in the present application in another embodiment.

[0020] Figure 4 A flow chart of a shale oil and gas development activity monitoring method provided in the present application in one embodiment is shown.

[0021] Figure 5 A module schematic diagram of a shale oil and gas well site classification system provided by the present application in one embodiment is shown.

[0022] Description of Reference Numerals S1~S5, steps; S41~S43, steps; S41'~S43', steps; 200, shale oil and gas well site classification system; 201, location determination module; 202, feature extraction module; 203, feature time series construction module; 204, classification module. DETAILED DESCRIPTION

[0023] The following describes the embodiments of the present application through specific examples, and those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. The present application can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0024] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application, and thus the drawings only show components related to the present application rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed at will, and the component layout may also be more complicated.

[0025] To facilitate the understanding of the technical solutions provided by the present application, the relevant terms in the present application are explained before the specific embodiments, as follows: Medium-resolution remote sensing image: refers to remote sensing image with medium resolution, whose spatial resolution is between a few meters and tens of meters, and can provide macroscopic surface information.

[0026] High-resolution remote sensing images: refers to high-resolution remote sensing images with a spatial resolution of meters or even sub-meters (below 0.5 meters), which can provide finer surface details.

[0027] The Database of Global Administrative Areas, GADM for short, provides administrative boundary data at multiple levels, including national boundaries, provincial boundaries, municipal boundaries, and district boundaries, for countries and regions around the world.

[0028] Global Human Settlement Layer: Global Human Settlement Layer, referred to as GHSL, is a series of high-precision datasets that provide human settlement patterns around the world.

[0029] Geospatial Data Abstraction Library: Geospatial Data Abstraction Library, referred to as GDAL, is an open source geospatial data processing library that provides a series of functions for reading, writing and processing various geospatial data formats.

[0030] Geographic transformation parameters: A parameter that describes the mathematical relationship between each pixel on a remote sensing image and the geographic coordinate system. It usually includes a tuple of six elements: upper left corner x coordinate, pixel width, row rotation, upper left corner y coordinate, column rotation and pixel height.

[0031] Short-wave infrared band: refers to light with a wavelength range of 0.9-1.7μm.

[0032] Shortwave infrared band 1: refers to the first band of shortwave infrared. In the Landsat 8 satellite, shortwave infrared band 1 corresponds to band 6, and its wavelength range is 1 610.4~1 613.7nm.

[0033] Shortwave infrared 2 band: refers to the second band of shortwave infrared. In the Landsat 8 satellite, shortwave infrared 2 band corresponds to band 7, and its wavelength range is 2185.7~2202.4 nm.

[0034] Thermal infrared band: refers to infrared radiation with a wavelength range of 8 to 14μm.

[0035] Red light band: refers to light waves in visible light with wavelengths between 620nm and 770nm.

[0036] The embodiments of the present application are described in detail below in conjunction with the accompanying drawings. In the absence of conflict, the features of the following embodiments and implementations can be combined with each other.

[0037] The following embodiment of the present application provides a shale oil and gas well site classification method, which determines the location information of each well site in a target area; extracts the first spectral feature and the second spectral feature corresponding to the well site in each target remote sensing image of the target area within a target time period according to the location information of the well site; sorts the first spectral feature and the second spectral feature of each well site according to the acquisition time of the target remote sensing image, and correspondingly constructs the first spectral feature time series and the second spectral feature time series of each well site; constructs the total spectral feature time series of each well site based on the first spectral feature time series and the second spectral feature time series of each well site; sequence; based on the first spectral feature and the second spectral feature that can characterize the changes of the well site before and after its existence, cluster analysis is performed on the total spectral feature time groups in each of the total spectral feature time series to obtain clustering results of each of the well sites; based on the clustering results, each of the well sites is classified to obtain a well site status category of each of the well sites. The method accurately obtains the geographical location area where the well site is located and the well site existence status within the target time period based on the pre-constructed first spectral feature and the second spectral feature, thereby improving the classification accuracy of a large range of shale oil and gas well sites based on the well site existence status.

[0038] See also Figure 1 , showing a flow chart of a shale oil and gas well site classification method provided by the present invention in one embodiment.

[0039] like Figure 1As shown, in this embodiment, the shale oil and gas well site classification method provided by the present invention includes the following steps: Step S1, obtaining the geographical location of each well site in the target area; Specifically, the target area and each target remote sensing image within the target time period are obtained; the well field extraction is performed on each target remote sensing image using the trained well field extraction model to obtain the first well field extraction result corresponding to each target remote sensing image; based on the image geographic location of the target remote sensing image, the image coordinates of each first well field extraction result in the corresponding target remote sensing image are respectively converted into geographic coordinates to obtain the spatial position distribution corresponding to each first well field extraction result; based on the spatial position distribution corresponding to each first well field extraction result, each well field and its position information are correspondingly obtained; The target remote sensing image includes: a medium-resolution remote sensing image or a high-resolution remote sensing image; both the medium-resolution remote sensing image and the high-resolution remote sensing image are optical remote sensing images, which have corresponding acquisition time and image geographical location; the image geographical location of each of the optical remote sensing images is located in the target area; The first well site extraction result includes: the well site prediction probability output by the trained well site extraction model, and the extraction result that is greater than a preset probability threshold.

[0040] Optionally, the preset probability threshold includes: 85%.

[0041] Optionally, the method for obtaining the first well site extraction result includes: Each of the target remote sensing images is input into the trained well site extraction model to obtain the well site prediction probability corresponding to each of the target remote sensing images, and the extraction results whose well site prediction probability is greater than the preset probability threshold are respectively used as the first well site extraction results of the corresponding target remote sensing image.

[0042] Optionally, the implementation of the geographic coordinate conversion includes: Based on the image coordinates of the first well site extraction results in the corresponding target remote sensing image, and the image geographic location of the corresponding target remote sensing image, according to geographic transformation parameters, the image coordinates corresponding to each of the first well site extraction results are converted into geographic coordinates to obtain the spatial position distribution corresponding to each of the first well site extraction results.

[0043] Optionally, obtaining each well site and its geographical location based on the spatial position distribution corresponding to each first well site extraction result includes: It is detected whether the spatial position distributions corresponding to the first well site extraction results at different acquisition times are the same, and the first well site extraction results with the same spatial position distribution are determined to correspond to the same well site, so as to obtain the well sites and their position information.

[0044] Optionally, based on the geographic location of the image, use the imagexy2geo function in the geospatial data abstraction library to convert the image coordinates corresponding to the first well site extraction result into geographic coordinates according to the geographic transformation parameters to obtain the spatial location distribution of the first well site extraction result.

[0045] Optionally, after obtaining the spatial position distribution corresponding to each of the first well site extraction results, determining the position information of each well site in the target area further includes: performing duplicate well site screening on each of the first well site extraction results, including: In each of the target remote sensing images, extract each of the target remote sensing images with overlapping areas at the same acquisition time; respectively use the target remote sensing images with overlapping areas between each other at the same acquisition time as a group of images to be screened; perform identification result repeatability screening on each of the image groups to be screened to obtain repeatability screening results of each of the image groups to be screened; the repeatability screening results include: repeatable identification well sites and non-repeatable identification well sites; perform repeatable identification well site elimination on each of the image groups to be screened, so that each of the image groups to be screened contains non-repeatable identification well sites, and obtain the first well site extraction results after elimination as the second well site extraction results; Among them, the repeatedly identified well field includes: the first well field extraction results whose central geographical distance is less than or equal to the preset distance threshold; the central geographical distance is the distance between the geographical locations of the centers of the identification boxes corresponding to the two first well field extraction results; the non-repeatedly identified well field includes: other well fields except the repeatedly identified well field.

[0046] Optionally, the implementation method of eliminating the duplicate identified well sites includes: In the corresponding image group to be screened, the first well field extraction result corresponding to one of the corresponding two repeatedly identified well fields is deleted until all the image groups to be screened are the non-repeatedly identified well fields.

[0047] Optionally, the preset distance threshold includes 30 meters.

[0048] In this embodiment, taking into account the possible overlapping areas between multi-source remote sensing images, duplicate well fields are screened out from the first well field extraction result to avoid erroneous extraction of the same well field in the overlapping area of ​​the multi-source remote sensing images as different well fields, thereby improving the extraction accuracy of a large range of well fields.

[0049] It should be noted that since the present application monitors a large range of shale oil and gas development activities, and a remote sensing image only contains surface information of a small range, the target remote sensing image in the present application includes multi-source remote sensing images, and there will be overlapping areas between the multi-source remote sensing images.

[0050] Optionally, obtaining each corresponding well site and its geographical location based on the spatial position distribution corresponding to each first well site extraction result includes: The duplicate well site screening is performed on each of the first well site extraction results to obtain each of the second well site extraction results and their corresponding spatial position distribution; it is detected whether the geographical locations corresponding to each of the second well site extraction results at different acquisition times are the same, and if so, it is determined that each of the second well site extraction results with the same geographical location corresponds to the same well site, and each of the well sites and their geographical locations are obtained.

[0051] Optionally, after obtaining the spatial position distribution corresponding to each of the first well site extraction results, determining the position information of each well site in the target area further includes: performing noise removal on each of the first well site extraction results, including: Acquire the building land data of the target area within the target time period; construct building area masks corresponding to the acquisition time and image geographical location of each of the target remote sensing images based on the building land data; the image size of the building area mask is the same as the image size of the target remote sensing image; match each of the building area masks with each of the target remote sensing images according to the time-space correspondence matching principle; use each of the building area masks to perform mask processing on the first well site extraction result of the target remote sensing image matched therewith, and obtain each of the first well site extraction results after mask processing as each of the third well site extraction results; The time-space correspondence matching principle includes: the time and geographical location of the building area mask correspond to the acquisition time and geographical location of the target remote sensing image; When the time attribute of the building land data is a time period, the time of the building area mask corresponds to the time of the target remote sensing image, including: the acquisition time of the target remote sensing image is within the time period corresponding to the building land data.

[0052] In this embodiment, by acquiring the building land data, a building area mask is constructed based on the building land data, and the building area mask is used to perform mask processing on the first well field extraction result of the target remote sensing image that corresponds to the building area mask in time and space, so as to remove the building land that is mistakenly extracted as a well field in the first well field extraction result of the target remote sensing image, thereby improving the accuracy of the extracted well field, and is particularly suitable for the well field extraction model obtained based on the training of the center-resolution remote sensing image.

[0053] It should be noted that the resolution of the medium-resolution remote sensing image is lower than that of the high-resolution remote sensing image, but the acquisition frequency of the medium-resolution remote sensing image is higher than that of the high-resolution remote sensing image. Therefore, the data volume of the high-resolution remote sensing image is less than that of the medium-resolution remote sensing image. In practical applications, it is difficult to meet the training requirements of large data volumes for deep learning; while the data volume of the medium-resolution remote sensing image can meet the training requirements of large data volumes for deep learning. However, due to the low resolution of the medium-resolution remote sensing image, the well site extraction accuracy of the network model obtained by training based on the medium-resolution remote sensing image is lower than that of the network model obtained by training based on the high-resolution remote sensing image. Moreover, since the spectrum and other features of the building land are highly similar to the spectrum and other features of the well site, in the present application, after obtaining the spatial position distribution corresponding to each first well site extraction result, noise removal is performed on each first well site extraction result to remove the building land that is mistakenly extracted as the well site, thereby improving the extraction accuracy of the well site, and is particularly suitable for the well site extraction model obtained by training based on the medium-resolution remote sensing image.

[0054] Optionally, when the method does not perform the noise removal before obtaining each second well site extraction result, the method for obtaining the third well site extraction result includes: Acquire the building land data of the target area within the target time period; construct building area masks corresponding to the acquisition time and image geographical location of each of the target remote sensing images based on the building land data; match each of the building area masks with each of the target remote sensing images according to the principle of time-space correspondence; use each of the building area masks to perform mask processing on the second well site extraction results of the target remote sensing images matched therewith, and obtain each of the second well site extraction results after mask processing as each of the third well site extraction results.

[0055] Optionally, the building land data is obtained based on earth observation data of global human settlement spatial information of a global administrative division database and / or global human settlement layer data.

[0056] Optionally, obtaining each corresponding well site and its geographical location based on the spatial position distribution corresponding to each first well site extraction result further includes: The duplicate well site screening is performed on each of the first well site extraction results to obtain each of the second well site extraction results and their corresponding spatial position distribution; the noise removal is performed on each of the second well site extraction results to obtain each of the third well site extraction results and their spatial position distribution; it is detected whether the geographical locations corresponding to the third well site extraction results at different acquisition times are the same, and if so, it is determined that the third well site extraction results with the same geographical location correspond to the same well site, and each of the well sites and their geographical locations are obtained.

[0057] Optionally, the image size of the target remote sensing image includes: 300*300 pixels.

[0058] Optionally, the color channels of the target remote sensing image are three channels of RGB.

[0059] Optionally, the well site extraction model includes an existing neural network model such as a Faster R-CNN model using VGG-16 as a feature extraction backbone network.

[0060] Optionally, the trained wellsite extraction model is obtained by training and testing using a pre-built training set and test set.

[0061] Optionally, the method for obtaining the training set and the test set includes: An original optical remote sensing image is obtained, and the original optical remote sensing image is cropped to obtain a sample image of a preset image size containing a shale oil and gas well site; the shale oil and gas well site in the sample image is labeled to obtain a sample set; and the sample set is divided into a training set and a test set according to a preset division ratio.

[0062] Optionally, the original optical remote sensing image includes: Sentinel-2 MSI optical remote sensing image, etc.

[0063] Optionally, the preset image size includes: 300*300 pixels.

[0064] Optionally, the preset division ratio includes: 9:1.

[0065] Optionally, after obtaining the training set, the method for obtaining the training set further includes: The sample images in the training set are subjected to operations such as flipping, translating, rotating and twisting to expand the sample images in the training set and increase the detail variation features of the sample images.

[0066] Exemplarily, the target time period includes: 1999~2019.

[0067] Step S2, extracting a first spectral feature and a second spectral feature corresponding to the well site from each target remote sensing image of the target area within a target time period according to the location information of the well site; The first spectral feature and the second spectral feature are pre-constructed spectral features that characterize changes in the well site before and after its existence.

[0068] In one embodiment, the first spectral feature and the second spectral feature are obtained by: Acquire the spectral band change characteristics of the well site; construct a well site extraction index based on two spectral bands in the spectral band change characteristics as the first spectral characteristic of the well site; extract a spectral band from the remaining spectral band change characteristics as the second spectral characteristic of the well site; The spectral band change characteristics include: the top N spectral bands with the largest change before and after the well site exists in the same remote sensing image product; N is greater than or equal to 3; The remaining spectral band variation characteristics are spectral bands other than the spectral bands used to construct the well site extraction index in the spectral band variation characteristics; The well site extraction index represents the degree of differentiation between the region and the well site, and its value range is between 0 and 1. The closer the value is to 0, the higher the possibility that the region is a non-well site, and the closer the value is to 1, the higher the possibility that the region is a well site.

[0069] Optionally, the remote sensing image products include: Landsat-5 TM, Landsat-7 ETM+, Landsat-8 OLI and / or Sentinel-2 MSI and other remote sensing image products.

[0070] Optionally, the method for acquiring the spectral band change characteristics includes: Acquire the spectral data of each well site within the target time period; based on the spectral data, compare and analyze the spectral bands of the well site before and after construction, and obtain the top N spectral bands with the largest degree of change between the well site before and after construction as the spectral band change characteristics.

[0071] Exemplarily, the spectral band change characteristics include: short-wave infrared band, thermal infrared band and red light band, etc.; the short infrared band includes short infrared band 1 band and short infrared band 2 band.

[0072] Optionally, the well site extraction index is constructed by: Among the spectral band change characteristics, the first two spectral bands with the largest degree of change are selected, and a normalization formula for the difference between the first two spectral bands is constructed. Based on the normalization formula, the well site extraction index is obtained, which is the value calculated by the normalization formula.

[0073] Optionally, the well site extraction index is a normalized ratio of the difference between the reflectivity of the short-wave infrared band and the reflectivity of the thermal infrared band.

[0074] Exemplarily, the method for obtaining the well site extraction index is as follows:

[0075] In the above formula, WPI represents the well site extraction index; SWIR Indicates short infrared band; TIR Represents the thermal infrared band.

[0076] Exemplarily, the first spectral feature is obtained based on the average reflectivity of the short-wave infrared band and the average reflectivity of the thermal infrared band in the geographical location area where the well site is located at the corresponding acquisition time.

[0077] Exemplarily, the second spectral feature is obtained based on the average reflectivity of the red light band in the geographical location area where the well site is located at the corresponding acquisition time.

[0078] Step S3, respectively sorting the first spectral features and the second spectral features of each well site according to the acquisition time of the target remote sensing image, and correspondingly constructing the first spectral feature time series and the second spectral feature time series of each well site; based on the first spectral feature time series and the second spectral feature time series of each well site, respectively constructing the total spectral feature time series of each well site; The total spectral feature time series includes: each total spectral feature time group within the target time period; the total spectral feature time group includes: the first spectral feature and the second spectral feature at the acquisition time and the corresponding acquisition time.

[0079] Specifically, the first spectral features and the second spectral features of each well site are sorted according to the acquisition time of the target remote sensing image, and the first spectral features of the time series of each well site and the second spectral features of the time series are obtained as the first spectral feature time series and the second spectral feature time series corresponding to each well site; the first spectral features and the second spectral features at the acquisition time and the corresponding acquisition time are taken as a total spectral feature time group, and the total spectral feature groups corresponding to each well site are constructed as the total spectral feature time series corresponding to each well site.

[0080] Step S4, respectively performing cluster analysis on the total spectral feature time groups of each total spectral feature time series to obtain clustering results of each well site; based on the clustering results, classifying each well site to obtain a well site status category of each well site.

[0081] Specifically, cluster analysis is performed on each of the total spectral feature time groups corresponding to each of the total spectral feature time series to obtain clustering results corresponding to each of the well sites; based on the clustering results, each of the well sites is classified to obtain a well site status category of each of the well sites.

[0082] Optional, such as Figure 2 As shown, the cluster analysis is performed on the total spectral feature time groups of each total spectral feature time series to obtain the clustering results of each well field. When the cluster analysis is performed on the total spectral feature time groups in a single total spectral feature time series, it includes: Step S41, using the acquisition time, the first spectral feature and the second spectral feature as clustering variables respectively, performing three-dimensional clustering on the total spectral feature time group in the total spectral feature time series to obtain cluster clusters; Step S42, detecting whether the number of the clusters is 1, if so, obtaining the clusters as the clustering result of the well site; if not, dividing each cluster into a high mean cluster and a low mean cluster according to the size of the first spectral feature mean or the second spectral feature mean; Step S43, re-perform the three-dimensional clustering on the total spectral feature time group in the high-mean cluster cluster to obtain the high-mean cluster sub-cluster of the high-mean cluster cluster; and re-perform the three-dimensional clustering on the total spectral feature time group in the low-mean cluster cluster to obtain the low-mean cluster sub-cluster of the low-mean cluster cluster, and use the high-mean cluster cluster, low-mean cluster cluster, high-mean cluster sub-cluster and low-mean cluster sub-cluster as the clustering results of the well site.

[0083] Optionally, the three-dimensional clustering can be implemented using a Gaussian mixture model or other existing clustering models.

[0084] Optionally, the dividing each cluster into a high mean cluster and a low mean cluster according to the size of the first spectral feature mean or the second spectral feature mean includes: According to the size of the first spectral feature mean or the second spectral feature mean corresponding to each cluster, the cluster with the smallest mean is defined as a low-mean cluster; the cluster with a mean greater than the corresponding mean of the low-mean cluster is defined as a high-mean cluster.

[0085] Optionally, based on the clustering result, each well site is classified to obtain a well site status category of each well site, and when obtaining the well site status category of a single well site, the method includes: It is detected whether there is only one cluster in the clustering result, or whether the number of high-mean cluster subclusters and the number of low-mean cluster subclusters in the clustering result are both less than or equal to 1. If so, the well site is determined to be a persistent well site; if not, the well site is determined to be a non-persistent well site.

[0086] Optionally, after obtaining the well site status category of each of the well sites, the shale oil and gas well site classification method further includes: performing a persistent existence category detection on each of the persistent well sites, and when performing the persistent existence category detection on a single persistent well site, it includes: Detect whether there is only one cluster in the clustering result of the persistent well field, whether the persistent well field meets the preset spectral feature requirements, or whether the start time and end time of a cluster sub-cluster in the clustering result of the persistent well field are respectively within the start time period and the end time period of the target time period. If so, determine that the persistent existence category of the persistent well field is long-term persistent existence in the target time period; if not, Then, it is detected whether the starting time of the low-mean cluster of the persistent well field is earlier than the starting time of its high-mean cluster. If so, the persistent category of the persistent well field is determined to be persistent in the late period of the target time period; if not, the persistent category of the persistent well field is determined to be persistent in the early period of the target time period; Wherein, the cluster sub-cluster includes: the high mean cluster sub-cluster or the low mean cluster sub-cluster; The starting time period is a time period in the target time period that includes the starting moment of the target time period and has a preset first time length; the ending time period is a time period in the target time period that includes the ending moment of the target time period and has a preset second time length.

[0087] Optionally, the detecting whether the persistent well field meets a preset spectral characteristic requirement includes: Based on the high-mean clusters and low-mean clusters in the clustering results, the first spectral characteristic difference and the second spectral characteristic difference of the persistent well field are obtained accordingly, and it is determined whether the first spectral characteristic difference is less than a preset first difference threshold, and whether the second spectral characteristic difference is less than a preset second difference threshold. If so, it is determined that the persistent well field meets the spectral characteristic requirements.

[0088] Exemplarily, the preset first difference threshold is 0.14; the preset second difference threshold is 0.08.

[0089] Optional, such as Figure 3 As shown, the method for obtaining the first spectral feature difference and the second spectral feature difference includes: Step S41', obtaining the first spectral feature mean and the second spectral feature mean of the high mean cluster, and the first spectral feature mean and the second spectral feature mean of the low mean cluster; Step S42', based on the mean of the first spectral characteristic value of the high mean clustering cluster and the mean of the first spectral characteristic value of the low mean clustering cluster, obtain the first spectral characteristic difference of the well site, which is the difference between the mean of the first spectral characteristic value of the high mean clustering cluster and the mean of the first spectral characteristic value of the low mean clustering cluster; Step S43', based on the second spectral feature mean of the high mean clustering cluster and the second spectral feature mean of the low mean clustering cluster, obtain the second spectral feature difference of the well site, which is the difference between the second spectral feature mean of the high mean clustering cluster and the second spectral feature mean of the low mean clustering cluster.

[0090] It should be noted that, considering that the spectral bands of the geographical location area where the well site is located show obvious changes in the time series before and after the existence of the well site, that is, the mean values ​​of the spectral band change characteristics of the geographical location area where the well site is located before the existence of the well site are lower than the mean values ​​of the spectral band change characteristics after the existence of the well site (such as the mean short infrared band reflectivity, the mean thermal infrared band reflectivity and the mean red light band reflectivity of the geographical location area where the well site is located before the existence of the well site are lower than the mean short infrared band reflectivity, the mean thermal infrared band reflectivity and the mean red light band reflectivity of the geographical location area where the well site is located after the existence of the well site, respectively); therefore, in the present application, the existence status of the well site within the target time period is determined based on the first spectral feature and the second spectral feature constructed in advance based on the spectral band change characteristics.

[0091] A shale oil and gas well site classification method provided in the above embodiment determines the location information of each well site in a target area; based on the location information of the well site, extracts the first spectral feature and the second spectral feature corresponding to the well site in each target remote sensing image of the target area within a target time period; sorts the first spectral feature and the second spectral feature of each well site according to the acquisition time of the target remote sensing image, and correspondingly constructs the first spectral feature time series and the second spectral feature time series of each well site; and constructs the total spectral feature time series of each well site based on the first spectral feature time series and the second spectral feature time series of each well site. ; Based on the first spectral feature and the second spectral feature that can characterize the changes in the well site before and after its existence, cluster analysis is performed on the total spectral feature time groups in each of the total spectral feature time series to obtain clustering results for each of the well sites; based on the clustering results, each of the well sites is classified to obtain a well site status category for each of the well sites. The method accurately obtains the geographical location area where the well site is located and the well site existence status within the target time period based on the pre-constructed first spectral feature and the second spectral feature, thereby improving the classification accuracy of a large range of shale oil and gas well sites based on the well site existence status.

[0092] See also Figure 4 , showing a flow chart of a shale oil and gas development activity monitoring method provided by the present invention in one embodiment.

[0093] like Figure 4 As shown, in this embodiment, the shale oil and gas development activity monitoring method provided by the present invention adopts the shale oil and gas well site classification method as described above, and after obtaining the non-persistent well site, the shale oil and gas development activity monitoring method further includes: Step S5, performing construction time node extraction and end time node extraction on the non-persistent well site; wherein, The construction time node extraction includes: obtaining the low mean cluster sub-cluster located in the starting time period as the starting cluster cluster; based on the preset first sliding step size and the first time window size, starting from the latest total spectral feature time group, gradually moving the first time window on the starting cluster cluster in descending time order until the number of the total spectral feature time groups of the low mean cluster sub-cluster in the current first time window is greater than the number of the total spectral feature time groups of the high mean cluster cluster, and obtaining the end time of the low mean cluster sub-cluster in the current first time window as the construction time node; The end time node extraction includes: obtaining the low mean cluster sub-cluster located in the end time period as the end cluster cluster; based on the preset second sliding step size and the second time window size, starting from the earliest total spectral feature time group, gradually moving the second time window on the end cluster cluster in ascending time order until the number of the total spectral feature time groups of the low mean cluster sub-cluster in the current second time window is greater than the number of the total spectral feature time groups of the high mean cluster cluster, and obtaining the start time of the low mean cluster sub-cluster in the current second time window as the end time node;

[0094] The low mean cluster sub-cluster located within the starting time period or the ending time period is the low mean cluster sub-cluster whose start time is within the starting time period or whose end time is within the ending time period.

[0095] Optionally, the shale oil and gas development activity monitoring method further includes: for a well site that continues to exist in the later period within the target time period, performing the construction time node extraction to obtain the construction time node of the well site.

[0096] Optionally, the shale oil and gas development activity monitoring method further includes: for a well site that has existed continuously in the early stage within the target time period, performing the end time node extraction to obtain the end time node of the well site.

[0097] It should be noted that since the construction time of the well site is a continuous process, the first spectral feature and the second spectral feature may be unstable during the construction of the well site. Therefore, directly using the start time or end time of the low-mean cluster subcluster as the corresponding construction time node or end time node will result in an error with the actual time node of the well site. Based on the total number of spectral feature time groups of the low-mean cluster subcluster and the total number of spectral feature time groups of the high-mean cluster subcluster in the current time window, the well site is judged before or after its existence, which can reduce the influence of the instability of the first spectral feature and the second spectral feature during the construction process, improve the accuracy of the judgment of the well site before or after its existence, and accurately obtain the time node of the well site.

[0098] In this embodiment, by comparing the total number of spectral feature time groups of low-mean clustering subclusters and high-mean clustering subclusters within the time window, a time window is obtained in which the total number of spectral feature time groups of the low-mean clustering subclusters is greater than the total number of spectral feature time groups of the high-mean clustering subclusters. Based on the time window, the construction time node or the end time node of the well site is obtained, thereby improving the accuracy of extracting each time node of the well site and achieving high-accuracy monitoring of the shale oil and gas development activities.

[0099] like Figure 5 As shown, in this embodiment, the present invention provides a shale oil and gas well site classification system 200, including: A location determination module 201 is used to determine the location information of each well site in the target area; The feature extraction module 202 is used to extract the first spectral feature and the second spectral feature corresponding to the well site in each target remote sensing image of the target area within the target time period according to the location information of the well site; the first spectral feature and the second spectral feature are pre-constructed spectral features that characterize the changes of the well site before and after its existence; The feature time series construction module 203 is used to sort the first spectral feature and the second spectral feature of each well site according to the acquisition time of the target remote sensing image, and construct the first spectral feature time series and the second spectral feature time series of each well site accordingly; based on the first spectral feature time series and the second spectral feature time series of each well site, respectively construct the total spectral feature time series of each well site; the total spectral feature time series includes a total spectral feature time group; the total spectral feature time group includes: the first spectral feature and the second spectral feature at the acquisition time and the corresponding acquisition time; The classification module 204 performs cluster analysis on the total spectral feature time groups of each total spectral feature time series to obtain clustering results of each well site; and classifies each well site based on the clustering results to obtain a well site status category of each well site.

[0100] The embodiment of the present application also provides a computer-readable storage medium. A person of ordinary skill in the art can understand that all or part of the steps in the method for implementing the above embodiment can be completed by instructing the processor through a program, and the program can be stored in a computer-readable storage medium, and the storage medium is a non-transitory medium, such as a random access memory, a read-only memory, a flash memory, a hard disk, a solid-state hard disk, a magnetic tape, a floppy disk, an optical disc, and any combination thereof. The above storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more available media integrated. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a digital video disc (DVD)), or a semiconductor medium (e.g., a solid-state disk (SSD)), etc.

[0101] The embodiment of the present application may also provide a computer program product, which includes one or more computer instructions. When the computer instructions are loaded and executed on a computing device, the process or function described in the embodiment of the present application is generated in whole or in part. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer or data center to another website, computer or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means.

[0102] When the computer program product is executed by a computer, the computer executes the method described in the above method embodiment. The computer program product may be a software installation package, and when the above method is required, the computer program product may be downloaded and executed on a computer.

[0103] The descriptions of the processes or structures corresponding to the above-mentioned figures have different emphases. For parts that are not described in detail in a certain process or structure, please refer to the relevant descriptions of other processes or structures.

[0104] The above embodiments are merely illustrative of the principles and effects of the present application and are not intended to limit the present application. Anyone familiar with the technology may modify or change the above embodiments without violating the spirit and scope of the present application. Therefore, all equivalent modifications or changes made by a person of ordinary skill in the art without departing from the spirit and technical ideas disclosed in the present application shall still be covered by the claims of the present application.

Claims

1. A shale oil and gas well site classification method, characterized in that: include: Determine the location information of each well site in the target area; Extracting, according to the location information of the well site, a first spectral feature and a second spectral feature corresponding to the well site in each target remote sensing image of the target area within a target time period; The first spectral features and the second spectral features of each well site are respectively sorted according to the acquisition time of the target remote sensing image, and the first spectral feature time series and the second spectral feature time series of each well site are correspondingly constructed; based on the first spectral feature time series and the second spectral feature time series of each well site, the total spectral feature time series of each well site is respectively constructed; The total spectrum feature time series includes a total spectrum feature time group; Performing cluster analysis on the total spectral feature time groups of each total spectral feature time series respectively to obtain clustering results of each well site; classifying each well site based on the clustering results to obtain a well site status category of each well site; The first spectral feature and the second spectral feature are pre-constructed spectral features that characterize changes in the well site before and after its existence.

2. The method according to claim 1, characterized in that The method for acquiring the location information of each well site in the target area includes: Acquiring remote sensing images of each of the targets; Using the trained wellsite extraction model, perform wellsite extraction on each of the target remote sensing images to obtain a first wellsite extraction result corresponding to each of the target remote sensing images; Based on the image geographic location of the target remote sensing image, geographic coordinate conversion is performed on the image coordinates of each of the first well site extraction results to obtain the spatial position distribution corresponding to each of the first well site extraction results; based on the spatial position distribution corresponding to each of the first well site extraction results, the location information of each of the well sites is obtained accordingly.

3. The method according to claim 2, characterized in that In order to improve the accuracy of obtaining the location information of the well site, after obtaining the spatial location distribution corresponding to each of the first well site extraction results, the determination of the location information of each well site in the target area further includes: Acquire the building land data of the target area within the target time period; construct building area masks corresponding to the acquisition time and image geographical location of each of the target remote sensing images based on the building land data; match each of the building area masks with each of the target remote sensing images according to the principle of time-space correspondence matching; use each of the building area masks to perform mask processing on the first well site extraction result of the target remote sensing image matched therewith, so as to remove the building land mistakenly extracted as the well site in the first well site extraction result, and obtain each of the well sites and their location information based on the remaining first well site extraction results.

4. The method according to claim 1, characterized in that: The method for extracting the first spectral feature and the second spectral feature includes: Acquire the spectral band change characteristics of the well site; construct a well site extraction index based on two spectral bands in the spectral band change characteristics as the first spectral characteristic of the well site; extract a spectral band from the remaining spectral band change characteristics as the second spectral characteristic of the well site; The spectral band change characteristics include: in the same remote sensing image product, the top N spectral bands with the largest degree of change before and after the well site exists; N is greater than or equal to 3.

5. The method according to claim 1, characterized in that The cluster analysis is performed on the total spectral feature time groups of each total spectral feature time series to obtain the clustering results of each well field. When the cluster analysis is performed on the total spectral feature time groups in a single total spectral feature time series, it includes: Using the acquisition time, the first spectral feature and the second spectral feature as clustering variables, respectively, performing three-dimensional clustering on the total spectral feature time group in the total spectral feature time series to obtain cluster clusters; Detecting whether the number of the clusters is 1, if so, obtaining the clusters as the clustering results of the well site; if not, dividing each cluster into a high mean cluster and a low mean cluster according to the size of the first spectral feature mean or the second spectral feature mean; The three-dimensional clustering is performed again on the total spectral feature time group in the high-mean clustering cluster to obtain the high-mean clustering subcluster of the high-mean clustering cluster; and the three-dimensional clustering is performed again on the total spectral feature time group in the low-mean clustering cluster to obtain the low-mean clustering subcluster of the low-mean clustering cluster, and the high-mean clustering cluster, the low-mean clustering cluster, the high-mean clustering subcluster and the low-mean clustering subcluster are used as the clustering results of the well site.

6. The method according to claim 5, characterized in that Based on the clustering result, each well site is classified to obtain the well site status category of each well site. When obtaining the well site status category of a single well site, it includes: It is detected whether there is only one cluster in the clustering result, or whether the number of high-mean cluster subclusters and the number of low-mean cluster subclusters in the clustering result are both less than or equal to 1. If so, the well site is determined to be a persistent well site; if not, the well site is determined to be a non-persistent well site.

7. The method according to claim 6, characterized in that After obtaining the well site status category of each well site, the shale oil and gas well site classification method further includes: performing a persistent existence category detection on each of the persistent well sites, and when performing the persistent existence category detection on a single persistent well site, it includes: Detect whether there is only one cluster in the clustering result of the persistent well field, whether the persistent well field meets the preset spectral feature requirements, or whether the start time and end time of a cluster sub-cluster in the clustering result of the persistent well field are respectively within the start time period and the end time period of the target time period. If so, determine that the persistent existence category of the persistent well field is long-term persistent existence in the target time period; if not, It is then detected whether the start time of the low-mean cluster of the persistent well field is earlier than the start time of its high-mean cluster. If so, the persistent category of the persistent well field is determined to be persistent in the later period of the target time period; if not, the persistent category of the persistent well field is determined to be persistent in the early period of the target time period.

8. A method for monitoring shale oil and gas development activities, characterized in that: include: The shale oil and gas well site classification method according to claim 6 or 7 is adopted to obtain a non-persistent well site, and construction time node extraction and end time node extraction are performed on the non-persistent well site; wherein, The construction time node extraction includes: obtaining a low-mean cluster subcluster located in a starting time period as a starting cluster cluster; based on a preset first sliding step size and a first time window size, starting from the latest total spectral feature time group, gradually moving the first time window on the starting cluster cluster in descending time order until, in the current first time window, the number of the total spectral feature time groups of the low-mean cluster subcluster is greater than the number of the total spectral feature time groups of the high-mean cluster cluster, obtaining the end time of the low-mean cluster subcluster in the current first time window as the construction time node; The end time node extraction includes: obtaining the low mean cluster sub-cluster located in the end time period as the end cluster cluster; based on the preset second sliding step size and the second time window size, starting from the earliest total spectral feature time group, gradually moving the second time window on the end cluster cluster in ascending time order until the number of the total spectral feature time groups of the low mean cluster sub-cluster in the current second time window is greater than the number of the total spectral feature time groups of the high mean cluster cluster, and obtaining the start time of the low mean cluster sub-cluster in the current second time window as the end time node; Among them, the starting time period is a time period in the target time period, including the starting moment of the target time period and having a preset first time length; the ending time period is a time period in the target time period, including the ending moment of the target time period and having a preset second time length.

9. The method according to claim 8, characterized in that After obtaining the well sites that continue to exist in the later period of the target time period and the well sites that continue to exist in the early period of the target time period, the shale oil and gas development activity monitoring method further includes: For the well field that continues to exist in the later period of the target time period, the construction time node extraction is performed to obtain the construction time node of the well field; for the well field that continues to exist in the early period of the target time period, the end time node extraction is performed to obtain the end time node of the well field.

10. A shale oil and gas well site classification system, characterized in that: include: A location determination module, used to determine the location information of each well site in the target area; A feature extraction module is used to extract a first spectral feature and a second spectral feature corresponding to the well site in each target remote sensing image of the target area within a target time period according to the location information of the well site; the first spectral feature and the second spectral feature are pre-constructed spectral features that characterize changes in the well site before and after its existence; A feature time series construction module is used to sort the first spectral features and the second spectral features of each well site according to the acquisition time of the target remote sensing image, and correspondingly construct the first spectral feature time series and the second spectral feature time series of each well site; based on the first spectral feature time series and the second spectral feature time series of each well site, respectively construct the total spectral feature time series of each well site; The total spectrum feature time series includes a total spectrum feature time group; The classification module is used to perform cluster analysis on the total spectral feature time groups of each total spectral feature time series to obtain clustering results of each well site; based on the clustering results, classify each well site to obtain a well site status category of each well site.

Images