An oil spill event timing remote sensing image automatic generation method, system and device

Abstract

The application discloses an oil spill event timing remote sensing image automatic generation method, system and equipment, relates to the technical field of water transportation and environmental protection, and comprises the following steps: obtaining structured oil spill event metadata based on original public opinion text data sets, and determining the spatial range and time window of a target observation area; concurrently searching a plurality of open source satellite data platforms to download satellite image data; cutting a sub-image covering a set size observation window from the satellite image data to obtain a preprocessed standardized sub-image; generating a timing index file based on the imaging time of the sub-image; sorting and packaging the preprocessed standardized sub-image based on the timing index file to obtain a multidimensional space-time data cube; and integrating the timing index file, the multidimensional space-time data cube and the oil spill event metadata to generate an oil spill event timing remote sensing monitoring product package and a core metadata file. The application improves the timeliness, accuracy and continuity of oil spill emergency event monitoring.

Description

Technical Field

[0001] This application relates to the field of water transportation and environmental protection technology, and in particular to a method, system and equipment for automatically generating time-series remote sensing images of oil spill events. Background Technology

[0002] Oil spills, as a sudden marine environmental disaster, require rapid and accurate monitoring for pollution control and ecological protection. Currently, remote sensing monitoring of oil spills mainly relies on the following technological approaches: (1) Regular satellite surveys and manual interpretation Currently, most operational oil spill monitoring systems rely on fixed-orbit remote sensing satellites, which periodically scan predetermined sea areas according to their inherent revisit cycles. Oil slicks are identified through manual or semi-automated interpretation of the acquired images by professionals. While this method achieves wide coverage, it is essentially a passive and indiscriminate survey. The monitoring gap between two revisits is several days to several weeks, preventing immediate monitoring after an oil spill, resulting in a lack of crucial initial pollution diffusion information. Furthermore, processing massive amounts of imagery from unrelated areas consumes enormous computing and storage resources, leading to low monitoring efficiency and a lack of specificity.

[0003] (2) Manually driven emergency observation after the incident After learning of an oil spill through ship reports and aerial reconnaissance, monitoring personnel manually plan observation tasks, including determining the observation area, querying available satellite data, issuing data ordering and download instructions, and performing subsequent data preprocessing and analysis. While this method is somewhat targeted, the entire process is slow and highly dependent on manual experience and operation. The long lag between the occurrence of the incident and the acquisition of the first batch of effective images makes it difficult to meet the time-sensitive requirements of emergency response. Furthermore, the coordinated retrieval and processing of multi-source data (such as SAR and optical data) is cumbersome, making it difficult to quickly generate time-series observation products that can be used for analysis.

[0004] (3) Automated data push system based on fixed area or sensor Some research or commercial platforms have implemented automated pipelines for subscribing to, downloading, and preprocessing data from specific regions or satellites. However, these systems are designed around "fixed areas" or "fixed data sources," and their observation tasks are pre-set and static. They cannot intelligently detect when and where sudden oil spill events requiring special attention occur, thus failing to achieve "on-demand observation." When events occur outside the pre-set areas, or when it is necessary to integrate other satellite data to compensate for insufficient spatiotemporal coverage, the system cannot automatically adapt and adjust, lacking flexibility and event-oriented capabilities.

[0005] (4) Simple reminder service based on traditional Internet information Some systems crawl online news containing specific keywords and send simple text alerts to users. However, these services only provide information alerts and do not transform the event descriptions (such as location and time) in the text information into structured, spatial parameters that can drive remote sensing observation systems. The alerts are completely disconnected from subsequent data acquisition and processing, failing to form a closed loop from "information perception" to "data product generation," thus limiting their practical value.

[0006] In summary, the existing remote sensing monitoring technology system for oil spills has significant shortcomings in terms of proactiveness, timeliness, and the degree of automation and intelligence in event response. Regular surveys are time-consuming and wasteful of resources; post-event manual intervention is slow and inefficient; fixed automated systems lack the ability to perceive and adapt to sudden events; and simple network alerts are disconnected from remote sensing observations. The core problem lies in the lack of a technical solution capable of automatically and in real-time capturing event clues from open internet information, intelligently triggering and coordinating multi-source remote sensing data acquisition and processing, and ultimately automatically generating spatiotemporal sequence monitoring products for specific events. This results in oil spill event monitoring often being delayed, discontinuous, and fragmented, failing to provide rapid, intuitive, and complete decision support for emergency command.

[0007] In summary, existing technologies generally suffer from problems such as delayed monitoring initiation, insufficient data source coordination, fragmented processing procedures and reliance on manual intervention, and difficulties in integrating event information with spatial observation data. An integrated technology chain of "intelligent event perception - precise data service - automatic product generation" has not yet been established. These shortcomings make it difficult to achieve rapid, accurate, and continuous three-dimensional monitoring in the face of sudden oil spill events, thus restricting the effectiveness of emergency response.

[0008] Therefore, it is necessary to provide a method for automatically generating time-series remote sensing images of oil spill events to solve the above problems. Summary of the Invention

[0009] The purpose of this application is to provide a method, system, and device for automatically generating time-series remote sensing images of oil spill events, so as to improve the timeliness, accuracy, and continuity of oil spill emergency monitoring.

[0010] To achieve the above objectives, this application provides the following solution: In a first aspect, this application provides a method for automatically generating time-series remote sensing images of oil spill events, the method comprising: Obtain the original public opinion text dataset and process it to obtain structured oil spill event metadata; Based on structured oil spill event metadata, the spatial extent and time window of the target observation area are determined; Concurrently retrieve data from multiple open-source satellite data platforms, generate corresponding structured data lists, and download satellite imagery data covering the spatial range within the time window based on the structured data lists; Based on the oil spill event metadata, sub-images covering a set size observation window are cropped from the satellite image data, and each sub-image is preprocessed to obtain multiple preprocessed standardized sub-images. The imaging timestamps of each sub-image are sorted in chronological order to generate a time-series index file; Based on the time-series index file, the preprocessed standardized sub-images are sorted, and the sorted sub-image sequence is encapsulated into a multi-dimensional spatiotemporal data cube. By integrating the time-series index file, the multidimensional spatiotemporal data cube, and the oil spill event metadata, a time-series remote sensing monitoring product package and core metadata file for oil spill events are generated, thus completing the automatic generation of time-series remote sensing images.

[0011] In one embodiment, the original public opinion text dataset is acquired and processed to obtain structured oil spill event metadata, specifically including: Based on a pre-defined keyword combination for the oil spill event, high-frequency scanning and text extraction are performed on internet information sources to form an original public opinion text dataset. Natural language processing was used to extract the geographic location names, event occurrence times or event reporting times, and descriptions of the oil spill scale from the original public opinion text dataset. The geographic location name is converted into latitude and longitude coordinates, and the latitude and longitude coordinates, the time of occurrence of the event or the time of event reporting, and the text describing the scale of the oil spill are encapsulated into structured oil spill event metadata.

[0012] In one implementation, the spatial extent and time window of the target observation area are determined based on structured oil spill event metadata, specifically including: Centered on the latitude and longitude coordinates in the oil spill event metadata, a spatial buffer is defined as the spatial range of the target observation area; Based on the event reporting time in the oil spill event metadata, a forward tracing and backward extension time window is defined as the time range for data retrieval in the target observation area.

[0013] In one implementation, multiple open-source satellite data platforms are retrieved concurrently to generate corresponding structured data lists, and satellite imagery data covering the spatial range within the time window is downloaded based on the structured data lists. Specifically, this includes: Simultaneously access the API interfaces of multiple open-source satellite data platforms, and submit data query requests to each open-source satellite data platform concurrently according to the time window and the spatial range to retrieve satellite image data covering the target observation area; The retrieved satellite imagery data is deduplicated and sorted to generate a structured data list; the structured data list includes: a unique identifier for the data product, the satellite platform source, the sensor type, the imaging time, and cloud cover information; Based on the structured data list, a download task is automatically initiated to download the satellite imagery data in the structured data list to the storage system.

[0014] In one embodiment, the satellite image data includes SAR image data and optical image data; Based on the oil spill event metadata, sub-images covering a set-size observation window are cropped from the satellite imagery data, and each sub-image is preprocessed to obtain multiple preprocessed standardized sub-images, specifically including: Using the latitude and longitude coordinates in the oil spill event metadata as the center, multiple sub-images covering the length of the set-size observation window are cropped from the satellite image data to obtain multiple cropped sub-images; Each cropped sub-image is subjected to radiometric and geometrical normalization to obtain multiple standardized sub-images; the standardized sub-images include standardized SAR sub-images and standardized optical sub-images. Generate cloud and cloud shadow masks using cloud quality identification bands or preset cloud detection algorithms; Based on cloud and cloud shadow masks, cloud-covered areas in each standardized optical sub-image are marked as invalid data; Standardized SAR sub-images and standardized optical sub-images marked with invalid data are identified as preprocessed standardized sub-images.

[0015] In one embodiment, the cropped sub-image includes a cropped optical sub-image and a cropped SAR sub-image; Radiometric and geometrical normalization are performed on each cropped sub-image to obtain multiple normalized sub-images, specifically including: Radiometric calibration and atmospheric correction are performed on each cropped optical sub-image to obtain multiple radiometrically normalized optical sub-images; Radiometric calibration and speckle noise filtering are performed on each cropped SAR sub-image to obtain multiple radiometrically normalized SAR sub-images; All radiometrically standardized optical sub-images and radiometrically standardized SAR sub-images are uniformly registered to a preset geographic coordinate system, and each registered optical sub-image and registered SAR sub-image is resampled according to a preset uniform spatial resolution to complete geometric standardization and obtain multiple standardized sub-images.

[0016] In one embodiment, the sorted sub-image sequence is encapsulated into a multi-dimensional spatiotemporal data cube, specifically including: Define the dimensions of a multidimensional spatiotemporal data cube, where the spatial dimension is defined by uniform geographic grid coordinates, the temporal dimension is defined by image layers arranged by timestamps, and the spectral dimension or polarization dimension is defined by the bands or polarization channels of each sub-image. The sorted sub-image sequences are organized and encapsulated according to the spatial dimension, the temporal dimension, and the spectral dimension or the polarization dimension to obtain a multidimensional spatiotemporal data cube.

[0017] In one implementation, the core metadata file includes event identification information, data summary information, data quality description, and usage recommendations.

[0018] Secondly, this application provides an automatic generation system for time-series remote sensing images of oil spill events. This system is used to implement the aforementioned automatic generation method for time-series remote sensing images of oil spill events. The automatic generation system includes: The oil spill event metadata generation unit is used to acquire the original public opinion text dataset and process the original public opinion text dataset to obtain structured oil spill event metadata. The spatiotemporal range determination unit is used to determine the spatial range and time window of the target observation area based on structured oil spill event metadata; The satellite imagery data retrieval unit is used to concurrently retrieve data from multiple open-source satellite data platforms, generate corresponding structured data lists, and download satellite imagery data covering the spatial range within the time window based on the structured data lists. The preprocessing unit is used to crop out sub-images covering a set size observation window from the satellite image data based on the oil spill event metadata, and to preprocess each sub-image to obtain multiple preprocessed standardized sub-images. The time-series index file construction unit is used to sort the imaging timestamps of each sub-image in chronological order to generate a time-series index file; The multidimensional spatiotemporal data cube generation unit is used to sort each preprocessed standardized sub-image based on the time-series index file, and encapsulate the sorted sub-image sequence into a multidimensional spatiotemporal data cube. The product packaging and output unit is used to integrate the time-series index file, the multi-dimensional spatiotemporal data cube, and the oil spill event metadata to generate an oil spill event time-series remote sensing monitoring product package and core metadata file, thereby completing the automatic generation of time-series remote sensing images.

[0019] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described method for automatically generating time-series remote sensing images of oil spill events.

[0020] According to the specific embodiments provided in this application, this application has the following technical effects: This application discloses a method, system, and device for automatically generating time-series remote sensing images of oil spill events, constructing a fully automated technical system from event perception to product delivery. By introducing internet public opinion as a monitoring trigger signal, and through real-time acquisition and intelligent analysis of online text, it automatically extracts key spatiotemporal information of oil spill events, replacing the traditional passive response mode that relies on manual reporting, thus improving monitoring timeliness. Based on a multi-source open-source satellite data platform for concurrent retrieval of event information, it automatically completes accurate retrieval and automatic download of remote sensing data. Finally, multi-temporal images are automatically organized into structured multi-dimensional data cubes according to the time dimension, and packaged into a time-series monitoring product package containing complete metadata, forming a continuous observation dataset that can be directly used for dynamic process analysis. From event network exposure to monitoring product generation, the entire process is automated and intelligent, thereby systematically improving the timeliness, accuracy, and continuity of oil spill emergency monitoring. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 A schematic flowchart of a method for automatically generating time-series remote sensing images of an oil spill event according to an embodiment of this application; Figure 2 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0023] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] The purpose of this application is to address the problems in existing oil spill remote sensing monitoring technologies, such as delayed event detection, passive monitoring initiation, cumbersome and fragmented data processing procedures, and difficulties in multi-source information fusion. This application provides a method for automatically generating time-series remote sensing images of oil spill events. This method automatically extracts the spatiotemporal information of the event by collecting and analyzing oil spill event reports from internet public opinion in real time. Based on this, it intelligently triggers and drives the accurate retrieval, automatic acquisition, rapid preprocessing, and standardized time-series organization of multi-source open-source satellite remote sensing data. Finally, it automatically generates a spatiotemporally continuous multi-source remote sensing monitoring image sequence product for a specific oil spill event.

[0025] This application aims to realize an automated generation process from "public opinion perception" to "data products," thereby improving the timeliness, pertinence, and continuity of emergency monitoring of oil spill incidents, reducing manual intervention, saving computing and storage resources, and providing intuitive, complete, and timely decision support information for emergency command of oil spill incidents, as well as providing core technical support for building an intelligent marine environmental disaster monitoring system.

[0026] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0027] In one exemplary embodiment, such as Figure 1 As shown, a method for automatically generating time-series remote sensing images of oil spill events is provided, including the following steps: Wherein: Step S1: Obtain the original public opinion text dataset and process the original public opinion text dataset to obtain structured oil spill event metadata.

[0028] As an optional implementation method, step S1 specifically includes: Step S11: Based on the preset keyword combination of the oil spill event, perform high-frequency scanning and text capture of Internet information sources to form the original public opinion text dataset.

[0029] Step S12: Extract the geographical location name, event occurrence time or event reporting time, and oil spill scale description text from the original public opinion text dataset through natural language processing.

[0030] Step S13: Convert the geographic location name into latitude and longitude coordinates, and encapsulate the latitude and longitude coordinates, the time of occurrence of the event or the time of event reporting, and the text describing the scale of the oil spill into structured oil spill event metadata.

[0031] Specifically, step S1 aims to promptly detect oil spill events from internet information streams and automatically extract key parameters that can drive remote sensing observations, including: Based on a pre-defined combination of keywords related to oil spill events (such as "oil spill", "marine pollution", "oil leakage", etc.), automated retrieval technology is used to perform high-frequency (e.g., once every 5 minutes) scanning and text capture of selected internet sources such as international news media, official websites of environmental monitoring agencies, and maritime safety notices, forming an original public opinion text dataset.

[0032] Natural language processing was performed on the text data in the original public opinion text dataset. First, the geographical location names mentioned in the text were extracted using named entity recognition technology. Then, geocoding services (such as calling the APIs of GeoNames or OpenStreetMap) were used to convert the geographical location names into precise latitude and longitude coordinates. At the same time, key information such as the time of the event or the time of the event reporting, the description of the scale of the oil spill (such as "large area", "several kilometers long"), and possible causes were parsed from the text and encapsulated into structured oil spill event metadata.

[0033] Step S2: Based on structured oil spill event metadata, determine the spatial range and time window of the target observation area.

[0034] As an optional implementation method, step S2 specifically includes: Step S21: Using the latitude and longitude coordinates in the oil spill event metadata as the center, define a spatial buffer (e.g., a circular area with a radius of 50 kilometers) as the spatial range of the target observation area.

[0035] Step S22: Based on the event reporting time in the oil spill event metadata, define a forward tracing and backward extension time window (e.g., all time from 3 days before the event to after the event) as the time range for data retrieval in the target observation area.

[0036] Step S3: Concurrently search multiple open-source satellite data platforms, generate corresponding structured data lists, and download satellite imagery data covering the spatial range within the time window based on the structured data lists.

[0037] As an optional implementation, step S3 specifically includes: Step S31: Simultaneously access the API interfaces of multiple open-source satellite data platforms, and submit data query requests concurrently to each open-source satellite data platform according to the time window and the spatial range to retrieve satellite image data covering the target observation area.

[0038] This involves building an automated program that simultaneously accesses the APIs of multiple open-source satellite data platforms (such as the ESA Copernicus Open Access Centre and the U.S. Geological Survey EarthExplorer). Based on the spatiotemporal scope, data query requests are submitted concurrently to each platform to retrieve available satellite imagery covering the target observation area within that spatiotemporal window. The search scope includes at least: Sentinel-1 SAR imagery data (for all-weather monitoring) and Sentinel-2 or Landsat series optical imagery data (for hyperspectral information acquisition).

[0039] Step S32: The retrieved satellite imagery data is deduplicated and sorted to generate a structured data list; the structured data list includes: unique identifier of data product, satellite platform source, sensor type, imaging time, and cloud cover information (for optical data).

[0040] Step S33: Based on the structured data list, automatically start the download task to download the satellite imagery data in the structured data list to the storage system.

[0041] Specifically, the system automatically initiates a download task, downloading all data products in the list to local or cloud storage systems in parallel. Step S4: Based on the oil spill event metadata, crop out sub-images covering a set size observation window from the satellite image data, and preprocess each sub-image to obtain multiple preprocessed standardized sub-images.

[0042] As an optional implementation, if the satellite image data includes SAR image data and optical image data, then step S4 specifically includes: Step S41: Using the latitude and longitude coordinates in the oil spill event metadata as the center, crop multiple sub-images from the satellite image data that cover the length of the set observation window to obtain multiple cropped sub-images.

[0043] Specifically, a uniform, fixed-size observation window (e.g., a square area with sides of 20 kilometers) is set as the center point, centered on the core latitude and longitude coordinates of the oil spill event. The system automatically reads the downloaded satellite imagery and its geographic metadata for each scene, and uses a spatial coordinate matching algorithm to precisely crop out sub-images covering the fixed window area from each large-scale image. Simultaneously, the unique identifier of each sub-image is recorded, especially its precise imaging timestamp, which serves as the core basis for subsequent time-series sequencing. The cropping operation significantly reduces the data volume of a single image, focuses on the event-related area, and improves subsequent processing efficiency.

[0044] Step S42 involves performing radiometric and geometric normalization on each cropped sub-image to obtain multiple normalized sub-images. The normalized sub-images include normalized SAR sub-images and normalized optical sub-images.

[0045] As an optional implementation, the cropped sub-image includes a cropped optical sub-image and a cropped SAR sub-image; then step S42 specifically includes: Step S421: Radiometric calibration and atmospheric correction are performed on each cropped optical sub-image to obtain multiple radiometrically normalized optical sub-images.

[0046] Step S422: Radiometric calibration and speckle noise filtering are performed on each cropped SAR sub-image to obtain multiple radiometrically normalized SAR sub-images.

[0047] Step S423: Register all radiometrically standardized optical sub-images and radiometrically standardized SAR sub-images to a preset geographic coordinate system, and resample all registered optical sub-images and registered SAR sub-images according to a preset uniform spatial resolution to complete geometric standardization and obtain multiple standardized sub-images.

[0048] Specifically, radiometric normalization involves radiometric calibration of cropped optical sub-images (such as Sentinel-2, Landsat), converting digital quantization values ​​into surface reflectance, and performing atmospheric correction to eliminate the effects of atmospheric scattering and absorption. For cropped SAR sub-images (such as Sentinel-1), radiometric calibration is performed, normalizing the backscattering coefficients and applying speckle noise filtering.

[0049] Geometric normalization: All sub-images are uniformly registered to a specified geographic coordinate system (such as WGS84 UTM). Subsequently, based on a preset uniform spatial resolution (e.g., 10 meters), all images are resampled to the same pixel size using a resampling algorithm (such as bilinear or cubic convolution) to ensure consistent spatial scale.

[0050] Step S43: Generate cloud and cloud shadow masks using cloud quality identification bands or preset cloud detection algorithms.

[0051] Step S44: Based on cloud and cloud shadow masks, mark the cloud-covered areas in each standardized optical sub-image as invalid data.

[0052] Step S45: The standardized SAR sub-image and the standardized optical sub-image marked with invalid data are identified as the preprocessed standardized sub-image.

[0053] Specifically, for optical images, clouds and cloud shadow masks are automatically generated using their attached cloud quality labeling bands or dedicated cloud detection algorithms. In standardized optical sub-images, cloud-covered areas are marked as invalid data to reduce the interference of clouds on subsequent time-series analysis (such as oil film color recognition).

[0054] Step S5: Sort the imaging timestamps of each sub-image in chronological order to generate a time-series index file.

[0055] Specifically, the system reads the imaging timestamps carried by all the preprocessed standardized sub-images, strictly sorts them in chronological order, and generates a time-series index file. This index file associates and records key metadata such as the file path, imaging time, sensor type, satellite platform, cloud coverage (optical), and processing version of each image, forming a directory of the entire image set.

[0056] Step S6: Based on the time-series index file, sort the preprocessed standardized sub-images and encapsulate the sorted sub-image sequence into a multi-dimensional spatiotemporal data cube.

[0057] As an optional implementation, in step S6, the sorted sub-image sequence is encapsulated into a multi-dimensional spatiotemporal data cube, specifically including: Step S61: Define the dimensions of the multidimensional spatiotemporal data cube, wherein the spatial dimension is defined by uniform geographic grid coordinates, the temporal dimension is defined by image layers arranged by timestamps, and the spectral dimension or polarization dimension is defined by the bands or polarization channels of each sub-image.

[0058] Step S62: The sorted sub-image sequence is organized and encapsulated according to the spatial dimension, the temporal dimension, and the spectral dimension or the polarization dimension to obtain a multidimensional spatiotemporal data cube.

[0059] Specifically, the spatial dimension (X, Y) corresponds to uniform geographic grid coordinates; the temporal dimension (T) corresponds to image layers arranged by timestamps; and the spectral / band dimension (or polarization dimension) (B) is organized according to data type: for example, all Sentinel-2 images are assigned the same band order (e.g., blue, green, red, near-infrared, etc.), and Sentinel-1 images are assigned VV and VH polarization channels. Encapsulation formats (such as using NetCDF, Zarr, or an organized GeoTIFF stack) allow spatial data of any time and any band to be accessed quickly and consistently, greatly facilitating time-series analysis.

[0060] Steps S5 and S6 generate a time-series index file and a multidimensional spatiotemporal data cube, organizing the preprocessed standardized sub-images into a structured time-series remote sensing image collection product that can be directly used for oil spill dynamic process analysis.

[0061] Furthermore, to ensure the continuity of timing, the system can perform consistency checks.

[0062] Step S7: Integrate the time-series index file, the multidimensional spatiotemporal data cube, and the oil spill event metadata to generate an oil spill event time-series remote sensing monitoring product package and core metadata file, thus completing the automatic generation of time-series remote sensing images.

[0063] As an optional implementation, the core metadata file includes event identification information, data summary information, data quality description, and usage recommendations.

[0064] Specifically, step S7, as the final step in the process, aims to encapsulate the time-series remote sensing image set constructed in S4 into a self-describing standardized product that can be directly delivered to downstream applications or manually analyzed, thus completing the transformation from raw data to end products.

[0065] The time-series index file generated in step S5, the multidimensional data cube constructed in step S6, and the oil spill event metadata (including original links that triggered public opinion, parsed location, time, scale descriptions, geographic coordinates, etc.) collected throughout step S1 are integrated and output in a unified manner. An independent data entity named "Oil Spill Event Time-Series Remote Sensing Monitoring Product Package" is generated. This product package uses an open, standard compression and archiving format (such as organization within a ZIP package following specific directory specifications) to ensure its portability and integrity.

[0066] In addition, it also includes the generation of product metadata and instruction manuals: A core metadata file (such as manifest.json in XML or JSON format) is automatically generated, serving as the product package's "identity card" and "instruction manual." This file details the following: 1) Event identification information: unique event ID, source of public opinion, and trigger time.

[0067] 2) Data summary information: Total number of images contained in the product, time range, spatial range, and a list of satellite platforms and sensors involved.

[0068] 3) Data quality description: In particular, for optical images, overall cloud coverage statistics are provided.

[0069] 4) Usage recommendations: Recommended data access methods and visualization tools.

[0070] Finally, product output is also included: The completed oil spill event time-series remote sensing monitoring product package and core metadata files are output through channels to complete the closed loop. The output is stored in a designated local or network storage directory using a file system, and named according to the "Event ID_Date" rule for easy archiving and management.

[0071] Message notification push: After the oil spill event time-series remote sensing monitoring product package is successfully generated, a notification is automatically sent to a preset message queue, email or collaborative office platform. The notification content includes an event summary, product access link and key metadata summary, realizing proactive service.

[0072] Based on the same inventive concept, this application also provides an automatic oil spill event time-series remote sensing image generation system for implementing the above-mentioned automatic oil spill event time-series remote sensing image generation method. The solution provided by this system is similar to the implementation scheme described in the above method. Therefore, the specific limitations of one or more embodiments of the automatic oil spill event time-series remote sensing image generation system provided below can be found in the limitations of the automatic oil spill event time-series remote sensing image generation method described above, and will not be repeated here.

[0073] In one exemplary embodiment, an automatic generation system for time-series remote sensing images of oil spill events is provided, comprising: The oil spill event metadata generation unit is used to acquire the original public opinion text dataset and process the original public opinion text dataset to obtain structured oil spill event metadata.

[0074] The spatiotemporal range determination unit is used to determine the spatial range and time window of the target observation area based on structured oil spill event metadata.

[0075] The satellite imagery data retrieval unit is used to concurrently retrieve data from multiple open-source satellite data platforms, generate corresponding structured data lists, and download satellite imagery data covering the spatial range within the time window based on the structured data lists.

[0076] The preprocessing unit is used to crop sub-images covering a set size observation window from the satellite image data based on the oil spill event metadata, and to preprocess each sub-image to obtain multiple preprocessed standardized sub-images.

[0077] The time-series index file construction unit is used to sort the imaging timestamps of each sub-image in chronological order to generate a time-series index file.

[0078] The multidimensional spatiotemporal data cube generation unit is used to sort each preprocessed standardized sub-image based on the time-series index file, and encapsulate the sorted sub-image sequence into a multidimensional spatiotemporal data cube.

[0079] The product packaging and output unit is used to integrate the time-series index file, the multi-dimensional spatiotemporal data cube, and the oil spill event metadata to generate an oil spill event time-series remote sensing monitoring product package and core metadata file, thereby completing the automatic generation of time-series remote sensing images.

[0080] Technical effects: 1) By innovatively integrating real-time internet public opinion perception with intelligent processing of multi-source remote sensing data, this method achieves full-chain automation and intelligence in the monitoring of sudden oil spill events, from "discovery" to "observation" and then to "analysis." This approach not only reconstructs the response paradigm of oil spill emergency monitoring but also significantly improves the timeliness, accuracy, and continuity of monitoring, providing an unprecedentedly efficient technological tool for marine environmental protection and emergency decision-making.

[0081] 2) By scanning news and environmental bulletins at high frequency, the system can capture and locate the oil spill within a very short time (e.g., within 15 minutes) after the incident is reported. This changes the passive situation of relying on periodic satellite transits or manual reporting after the event, and significantly advances the starting point of monitoring and response from "after data acquisition" to "after the event occurs".

[0082] 3) Through full-process automation, monitoring efficiency has been improved without the need for manual intervention, solving the pain points of traditional emergency observation, such as process breaks, cumbersome operation, and high dependence on the experience of professional personnel.

[0083] 4) The resulting "time-series remote sensing monitoring product package" is a spatiotemporally aligned, multi-source fusion, and ready-to-use analysis dataset that directly supports accurate and continuous quantitative analysis of the oil spill diffusion range, drift path, and rate of change, overcoming the limitations of analysis based on a single time phase or a single data source. The accompanying automatic briefing generation function can further transform the data into intuitive decision-making information, greatly assisting commanders in judging and assessing the development of the situation.

[0084] In one exemplary embodiment, a computer device is provided, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a method for automatically generating time-series remote sensing images of an oil spill event.

[0085] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 2 As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for automatically generating time-series remote sensing images of an oil spill event.

[0086] Those skilled in the art will understand that Figure 2 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0087] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0088] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0089] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0090] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0091] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods, systems, and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for automatically generating time-series remote sensing images of an oil spill event, characterized in that, The method for automatically generating time-series remote sensing images of the oil spill event includes: Obtain the original public opinion text dataset and process it to obtain structured oil spill event metadata; Based on structured oil spill event metadata, the spatial extent and time window of the target observation area are determined; Concurrently retrieve data from multiple open-source satellite data platforms, generate corresponding structured data lists, and download satellite imagery data covering the spatial range within the time window based on the structured data lists; Based on the oil spill event metadata, sub-images covering a set size observation window are cropped from the satellite image data, and each sub-image is preprocessed to obtain multiple preprocessed standardized sub-images. The imaging timestamps of each sub-image are sorted in chronological order to generate a time-series index file; Based on the time-series index file, the preprocessed standardized sub-images are sorted, and the sorted sub-image sequence is encapsulated into a multi-dimensional spatiotemporal data cube. By integrating the time-series index file, the multidimensional spatiotemporal data cube, and the oil spill event metadata, a time-series remote sensing monitoring product package and core metadata file for oil spill events are generated, thus completing the automatic generation of time-series remote sensing images.

2. The method for automatically generating time-series remote sensing images of oil spill events according to claim 1, characterized in that, Obtain the original public opinion text dataset and process it to obtain structured oil spill event metadata, specifically including: Based on a pre-defined keyword combination for the oil spill event, high-frequency scanning and text extraction are performed on internet information sources to form an original public opinion text dataset. Natural language processing was used to extract the geographic location names, event occurrence times or event reporting times, and descriptions of the oil spill scale from the original public opinion text dataset. The geographic location name is converted into latitude and longitude coordinates, and the latitude and longitude coordinates, the time of occurrence of the event or the time of event reporting, and the text describing the scale of the oil spill are encapsulated into structured oil spill event metadata.

3. The method for automatically generating time-series remote sensing images of oil spill events according to claim 2, characterized in that, Based on structured oil spill event metadata, the spatial extent and time window of the target observation area are determined, specifically including: Centered on the latitude and longitude coordinates in the oil spill event metadata, a spatial buffer is defined as the spatial range of the target observation area; Based on the event reporting time in the oil spill event metadata, a forward tracing and backward extension time window is defined as the time range for data retrieval in the target observation area.

4. The method for automatically generating time-series remote sensing images of oil spill events according to claim 3, characterized in that, Concurrently retrieve data from multiple open-source satellite data platforms, generate corresponding structured data lists, and download satellite imagery data covering the spatial range within the time window based on the structured data lists. Specifically, this includes: Simultaneously access the API interfaces of multiple open-source satellite data platforms, and submit data query requests to each open-source satellite data platform concurrently according to the time window and the spatial range to retrieve satellite image data covering the target observation area; The retrieved satellite imagery data is deduplicated and sorted to generate a structured data list; the structured data list includes: a unique identifier for the data product, the satellite platform source, the sensor type, the imaging time, and cloud cover information; Based on the structured data list, a download task is automatically initiated to download the satellite imagery data in the structured data list to the storage system.

5. The method for automatically generating time-series remote sensing images of oil spill events according to claim 1, characterized in that, The satellite imagery data includes SAR imagery data and optical imagery data; Based on the oil spill event metadata, sub-images covering a set-size observation window are cropped from the satellite imagery data, and each sub-image is preprocessed to obtain multiple preprocessed standardized sub-images, specifically including: Using the latitude and longitude coordinates in the oil spill event metadata as the center, multiple sub-images covering the length of the set-size observation window are cropped from the satellite image data to obtain multiple cropped sub-images; Each cropped sub-image is subjected to radiometric and geometrical normalization to obtain multiple standardized sub-images; the standardized sub-images include standardized SAR sub-images and standardized optical sub-images. Generate cloud and cloud shadow masks using cloud quality identification bands or preset cloud detection algorithms; Based on cloud and cloud shadow masks, cloud-covered areas in each standardized optical sub-image are marked as invalid data; Standardized SAR sub-images and standardized optical sub-images marked with invalid data are identified as preprocessed standardized sub-images.

6. The method for automatically generating time-series remote sensing images of oil spill events according to claim 5, characterized in that, The cropped sub-image includes the cropped optical sub-image and the cropped SAR sub-image; Radiometric and geometrical normalization are performed on each cropped sub-image to obtain multiple normalized sub-images, specifically including: Radiometric calibration and atmospheric correction are performed on each cropped optical sub-image to obtain multiple radiometrically normalized optical sub-images; Radiometric calibration and speckle noise filtering are performed on each cropped SAR sub-image to obtain multiple radiometrically normalized SAR sub-images; All radiometrically standardized optical sub-images and radiometrically standardized SAR sub-images are uniformly registered to a preset geographic coordinate system. Then, each registered optical sub-image and registered SAR sub-image is resampled according to a preset uniform spatial resolution to complete geometric standardization and obtain multiple standardized sub-images.

7. The method for automatically generating time-series remote sensing images of oil spill events according to claim 6, characterized in that, The sorted sub-image sequences are encapsulated into a multi-dimensional spatiotemporal data cube, specifically including: Define the dimensions of a multidimensional spatiotemporal data cube, where the spatial dimension is defined by uniform geographic grid coordinates, the temporal dimension is defined by image layers arranged by timestamps, and the spectral dimension or polarization dimension is defined by the bands or polarization channels of each sub-image. The sorted sub-image sequences are organized and encapsulated according to the spatial dimension, the temporal dimension, and the spectral dimension or the polarization dimension to obtain a multidimensional spatiotemporal data cube.

8. The method for automatically generating time-series remote sensing images of oil spill events according to claim 1, characterized in that, The core metadata file includes event identification information, data summary information, data quality descriptions, and usage recommendations.

9. An automatic generation system for time-series remote sensing images of oil spill events, characterized in that, The automatic generation system for time-series remote sensing images of oil spill events is used to implement the automatic generation method for time-series remote sensing images of oil spill events according to any one of claims 1-8, and the automatic generation system for time-series remote sensing images of oil spill events includes: The oil spill event metadata generation unit is used to acquire the original public opinion text dataset and process the original public opinion text dataset to obtain structured oil spill event metadata. The spatiotemporal range determination unit is used to determine the spatial range and time window of the target observation area based on structured oil spill event metadata; The satellite imagery data retrieval unit is used to concurrently retrieve data from multiple open-source satellite data platforms, generate corresponding structured data lists, and download satellite imagery data covering the spatial range within the time window based on the structured data lists. The preprocessing unit is used to crop out sub-images covering a set size observation window from the satellite image data based on the oil spill event metadata, and to preprocess each sub-image to obtain multiple preprocessed standardized sub-images. The time-series index file construction unit is used to sort the imaging timestamps of each sub-image in chronological order to generate a time-series index file; The multidimensional spatiotemporal data cube generation unit is used to sort each preprocessed standardized sub-image based on the time-series index file, and encapsulate the sorted sub-image sequence into a multidimensional spatiotemporal data cube. The product packaging and output unit is used to integrate the time-series index file, the multi-dimensional spatiotemporal data cube, and the oil spill event metadata to generate an oil spill event time-series remote sensing monitoring product package and core metadata file, and to complete the automatic generation of time-series remote sensing images.

10. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the method for automatically generating time-series remote sensing images of an oil spill event according to any one of claims 1-8.

Images