Offshore oil and gas platform monitoring method and system based on multi-source remote sensing data

Abstract

The invention provides an offshore oil and gas platform monitoring method and system based on multi-source remote sensing data, and the method comprises the following steps: S1, obtaining image data from a multi-source remote sensing platform, storing the image data in a database, and carrying out the preprocessing of the image data; s2, marking the position and the type of an oil and gas platform in the optical image to obtain a training set and a verification set; s3, taking the coordinates of the general survey result of the offshore oil and gas platform as candidate areas; s4, using the trained offshore oil and gas platform identification model to identify the oil and gas platform in the high-resolution optical image containing the candidate area to obtain the accurate position and category of the platform; and S5, calculating the accurate position of the platform and infrared image features or noctilucent image features of the category, and judging the operation state of the oil and gas platform to be monitored in combination with a preset threshold value. According to the invention, through multi-modal data fusion, accurate identification and operation state monitoring of the oil and gas platform are realized, and technical support is provided for safety supervision of the offshore oil and gas platform.

Description

Technical Field

[0001] The present invention relates to the field of smart maritime technology, and in particular to a method and system for monitoring an offshore oil and gas platform using multi-source remote sensing data. Background Art

[0002] As important facilities for offshore energy extraction, the distribution and operation status of oil and gas platforms have a significant impact on the marine environment and marine economy. In order to ensure the safe operation and environmental protection of offshore oil and gas platforms, it is essential to carry out monitoring of oil and gas platforms. Traditional oil and gas platform monitoring mainly relies on manual inspections and ship cruises, but these methods are not only costly and inefficient, but also difficult to implement in harsh sea conditions and have great limitations. With the development of remote sensing technology, synthetic aperture radar (SAR) and optical remote sensing have been widely used in offshore oil and gas platform monitoring. SAR has all-day and all-weather imaging capabilities, can penetrate clouds and rainfall, and is suitable for large-scale survey monitoring; high-resolution optical remote sensing provides detailed visible light images, which can be used to finely identify the structure and type of oil and gas platforms. However, monitoring with a single data source has certain limitations. For example, SAR images are easily interfered by speckle noise, and optical images are affected by clouds and light conditions. Therefore, it has become a trend to use multi-source remote sensing data fusion to carry out oil and gas platform monitoring.

[0003] In existing multi-source remote sensing monitoring methods, SAR is generally used for preliminary detection of the platform, and then combined with optical images for accurate identification. However, these methods still face challenges in multimodal data alignment and fusion. In particular, due to the differences in the imaging mechanisms of SAR and optical images, it is difficult to directly align the spatial features of the two, resulting in poor multimodal data fusion. In addition, with the development of infrared remote sensing and night light remote sensing technology, it has become possible to monitor the operating status of oil and gas platforms using infrared or night light images, but how to integrate this information with SAR and optical data has not been fully studied.

[0004] Most of the current oil and gas platform monitoring methods focus on static detection, lack the ability to dynamically monitor the platform's operating status and change trends, and cannot effectively identify abnormal platform operations. Therefore, building a comprehensive monitoring method based on multi-source remote sensing data can not only improve the accuracy of oil and gas platform identification, but also realize real-time monitoring of the platform's operating status, which has important practical significance. Summary of the invention

[0005] In view of this, the purpose of the present invention is to propose a method and system for monitoring offshore oil and gas platforms using multi-source remote sensing data, which realizes accurate identification and real-time status monitoring of offshore oil and gas platforms by integrating SAR, optical, infrared and night light remote sensing data. By using large-scale surveys of SAR images and refined monitoring of optical images, combined with the judgment of the operating status of infrared and night light images, the platform type can be accurately identified and its dynamic changes can be continuously tracked. This method improves the accuracy and comprehensiveness of monitoring, reduces manual intervention, and provides efficient and reliable technical support for the safety supervision of offshore oil and gas platforms.

[0006] The technical means adopted by the present invention are as follows:

[0007] A method for monitoring an offshore oil and gas platform using multi-source remote sensing data comprises the following steps:

[0008] S1. Acquire image data from a multi-source remote sensing platform and store it in a database, and pre-process the image data; the image data includes SAR images, optical images, infrared images, and night-light images;

[0009] S2. Mark the location and type of the oil and gas platform in the optical image to generate initial annotation data; retrieve the corresponding area of ​​the initial annotation data in the SAR image and mark it to obtain a training set and a validation set;

[0010] S3, inputting the training set into the offshore oil and gas platform recognition model for training, using the trained offshore oil and gas platform recognition model to recognize the oil and gas platforms in the SAR image to be monitored, obtaining the offshore oil and gas platform survey results, and using the coordinates of the offshore oil and gas platform survey results as candidate areas;

[0011] S4. Filter out high-resolution optical images containing the candidate area from the image data, and use the trained offshore oil and gas platform recognition model to recognize the oil and gas platform in the high-resolution optical images containing the candidate area to obtain the precise location and category of the platform;

[0012] S5. Calculate the precise location and category of the platform using infrared image features or night-light image features, and determine the operating status of the oil and gas platform to be monitored using a preset threshold.

[0013] Furthermore, the specific steps of S2 are as follows:

[0014] S21. marking the oil and gas platform in the optical image, and marking the location, outline and type of the platform using the visual information of the optical image;

[0015] S22. Retrieve the corresponding image from the SAR image library using the position, outline and type of the annotated platform extracted from the optical image; align the platform position in the optical image, project these positions into the retrieved SAR image, accurately annotate the platform in the SAR image, and keep the annotations of the optical and SAR data consistent through geometric position calibration.

[0016] Furthermore, in S3, the offshore oil and gas platform identification model includes an input layer, a convolutional layer, a fully connected layer, a feature fusion layer and an output layer.

[0017] Furthermore, the specific steps of S3 are as follows:

[0018] S31. Input the training set into the convolution layer through the input layer of the offshore oil and gas platform recognition model for feature extraction to obtain a feature map. The convolution layer formula is:

[0019] f(x)=ReLU(W1*x+b1)

[0020] Among them, f(x) is the convolution output, W1 is the convolution kernel, * represents the convolution operation, b1 is the bias, and ReLU is the activation function;

[0021] S32, map the feature map to a feature vector of fixed length through a fully connected layer. The formula of the fully connected layer is:

[0022] z=W2z+b2

[0023] Among them, z is the output feature vector, W2 is the weight matrix, and b2 is the bias;

[0024] The characteristic vectors of the two modes of S33, SAR image and optical image are F SAR 、F Optical , then the fused features are expressed as:

[0025] F fusion =[F SAR ; F Optical ]

[0026] S34. Add a fully connected layer to the concatenated comprehensive feature vector to perform feature fusion and dimensionality reduction to extract comprehensive features:

[0027] F f ' usion =ReLU(W fusion ·F fusion +b fusion )

[0028] Among them, W fusion and b fusion are the weights and biases of the fusion layer;

[0029] Assuming there are N types of platforms, the output is the probability distribution of each type of platform:

[0030] p type =softmax(W type ·F f ' usion +b type )

[0031] Among them, W type and b type are the weights and biases of the classifier, and the softmax function is used to normalize the output probability;

[0032] S35. Use the validation set to optimize the offshore oil and gas platform identification model, use the adaptive optimization algorithm to minimize the loss function, and obtain the optimized offshore oil and gas platform identification model; the loss function uses the cross entropy loss function:

[0033]

[0034] in, is the one-hot encoding of the actual type, is the predicted class probability.

[0035] Furthermore, the specific steps of S5 are as follows:

[0036] S51, extract the average temperature characteristics of the platform area, the calculation process is as follows:

[0037] Calculate the average temperature using the formula:

[0038]

[0039] Among them, T i is the temperature value of the i-th pixel in the platform area, and N is the total number of pixels in the platform area;

[0040] S52, extracting the average brightness feature of the platform area, the calculation process is as follows:

[0041] Calculate the average brightness using the formula:

[0042]

[0043] Among them, L i is the brightness value of the i-th pixel in the platform area, and N is the total number of pixels in the platform area;

[0044] S53. Set the threshold of the infrared image feature or the night light image feature, and distinguish the operation status of the platform by the average temperature, average brightness and the threshold:

[0045]

[0046] Among them, T threshold and L threshold is the threshold value.

[0047] The present invention also provides an offshore oil and gas platform monitoring system using multi-source remote sensing data, which is used to implement any of the above-mentioned offshore oil and gas platform monitoring methods using multi-source remote sensing data, including:

[0048] Data acquisition module: used to obtain SAR, optical, infrared and night light image data from multi-source remote sensing platforms and store the data in the database;

[0049] Data preprocessing module: used to preprocess image data, including radiation correction, geometric correction and denoising;

[0050] Sample annotation and model training module: used to annotate the location, outline and type of oil and gas platforms in high-resolution optical images, and then use these geometric positions to retrieve corresponding images in the SAR image library; by aligning the platform positions in the optical images, projecting these positions into the SAR images and accurately annotating them, the annotation consistency of optical and SAR data is maintained;

[0051] SAR survey module: used to use SAR images for preliminary detection of oil and gas platforms and to extract candidate areas of the platforms using detection algorithms;

[0052] Optical inspection module: used to identify the type of platform in the optical image and obtain the precise location and type of the platform, including jack-up drilling platform, semi-submersible drilling platform, and semi-submersible production storage and offloading platform;

[0053] Infrared / night light monitoring module: used to analyze platform features in infrared or night light images and determine the platform's operating status.

[0054] The present invention also provides a storage medium, which includes a stored program, wherein when the program is run, any of the above-mentioned methods for monitoring offshore oil and gas platforms using multi-source remote sensing data is executed.

[0055] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes any of the above-mentioned methods for monitoring offshore oil and gas platforms using multi-source remote sensing data through the computer program.

[0056] Compared with the prior art, the present invention has the following advantages:

[0057] The present invention effectively integrates the features of different remote sensing data through multimodal data alignment and fusion technology. The oil and gas platforms in the optical image are first marked, and then the corresponding areas are retrieved and marked in the SAR image library to achieve the alignment of SAR and optical samples. Multimodal fusion not only improves the accuracy of oil and gas platform detection, but also enhances the ability to identify platform types.

[0058] The present invention utilizes infrared and night-light remote sensing images, and can monitor the operation of oil and gas platforms and determine whether the platforms are in operation. This function provides an important reference for the safety monitoring of offshore platforms, helps to detect abnormal situations in a timely manner, and ensures the safety of platform operations.

[0059] The present invention reduces the reliance on manual inspections and manual annotations through automatic processing and analysis of multi-source remote sensing data, improves monitoring efficiency, and reduces labor costs. This highly automated monitoring method is more suitable for large-scale, long-term oil and gas platform supervision tasks. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0061] Figure 1 Schematic diagram of the monitoring method of the present invention.

[0062] Figure 2 This is a schematic diagram of the offshore oil and gas platform identification model of the present invention.

[0063] Figure 3 Schematic diagram of the monitoring system of the present invention. DETAILED DESCRIPTION

[0064] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0065] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0066] like Figure 1 As shown, the present invention provides a method for monitoring an offshore oil and gas platform using multi-source remote sensing data, comprising the following steps:

[0067] S1, obtain SAR, optical, infrared and night light image data from multi-source remote sensing platforms, store the data in a database, and perform preprocessing;

[0068] Specifically, remote sensing images from January to December in the Bohai Sea were selected from the image library. SAR images used the strip mode of the GF-3 satellite to perform radiation correction, geometric correction, multi-view processing and speckle noise suppression; high-resolution optical images used GF-2 images to perform radiation correction, geometric correction and atmospheric correction; infrared images used the TIRS of the Landsat 8 satellite to perform temperature correction and convert it into brightness temperature; night light images used NPP-VIIRS to perform spectral correction on night light data.

[0069] S2. Mark the location and type of the oil and gas platform in the optical image to generate initial annotation data. Then, based on the annotation results in the optical image, retrieve the corresponding area in the SAR image and mark it to obtain the training set and the validation set; the training set and the validation set are randomly divided according to the preset ratio;

[0070] S21. Annotate oil and gas platforms in high-resolution optical images, and use their detailed visualization information to manually annotate the location, outline, and type of the platforms;

[0071] S22. Using the geometric positions extracted from the optical image, the corresponding image is retrieved from the SAR image library. By aligning the platform positions in the optical image, these positions are projected into the retrieved SAR image, and the platform is manually accurately annotated in the SAR image. The annotations of the optical and SAR data are kept consistent through geometric position calibration.

[0072] S3, inputting the training set into the offshore oil and gas platform recognition model for training, using the trained model to identify the oil and gas platforms in the SAR image to be monitored, obtaining the offshore oil and gas platform survey results, and using their coordinates as candidate areas;

[0073] Please refer to Figure 2 , Figure 2 The offshore oil and gas platform recognition model in S3 includes an input layer, a convolution layer, a fully connected layer, a feature fusion layer, and an output layer. The training process is as follows:

[0074] S31, input the training set obtained in S2 into the convolution layer through the input layer of the offshore oil and gas platform recognition model for feature extraction. The convolution layer formula is:

[0075] f(x)=ReLU(W1*x+b1)

[0076] Among them, f(x) is the convolution output, W1 is the convolution kernel, * represents the convolution operation, b1 is the bias, and ReLU is the activation function;

[0077] S32, map the feature map to a feature vector of fixed length through a fully connected layer. The formula of the fully connected layer is:

[0078] z=W2z+b2

[0079] Among them, z is the output feature vector, W2 is the weight matrix, and b2 is the bias;

[0080] The eigenvectors of the S33, SAR and optical modes are F SAR 、F Optical , then the fused features are expressed as:

[0081] F fusion =[F SAR ; F Optical ]

[0082] S34. Add a fully connected layer to the concatenated comprehensive feature vector to perform feature fusion and dimensionality reduction to extract comprehensive features:

[0083] F f ' usion =ReLU(W fusion ·F fusion +b fusion )

[0084] Among them, W fusion and b fusion are the weights and biases of the fusion layer;

[0085] Assuming there are N types of platforms, the output is the probability distribution of each type of platform:

[0086] p type =softmax(W type ·F f ' usion +b type )

[0087] Among them, W type and b type are the weights and biases of the classifier, and the softmax function is used to normalize the output probability;

[0088] S35. Use the validation set to optimize the offshore oil and gas platform identification model, use the adaptive optimization algorithm to minimize the loss function, and obtain the optimized offshore oil and gas platform identification model; the loss function uses the cross entropy loss function:

[0089]

[0090] in, is the one-hot encoding of the actual type, is the predicted class probability.

[0091] S4. Based on the candidate areas provided by the SAR images, high-resolution optical images containing the candidate areas are screened out, and the oil and gas platforms in the images are identified using the model to obtain the precise location and category of the platforms;

[0092] S5. Determine the operating status of the oil and gas platform to be monitored by calculating the features of the infrared / night light image and presetting the threshold.

[0093] S5 calculates the features of infrared / night light images and presets thresholds to determine the operating status of the oil and gas platform to be monitored. The determination method is:

[0094] When the S51 oil and gas platform is in operation, the platform temperature will be high due to the operation and heating of the equipment. The average temperature characteristics of the platform area are extracted, and the calculation process is as follows:

[0095] Calculate the average temperature using the formula:

[0096]

[0097] Among them, T i is the temperature value of the i-th pixel in the platform area, and N is the total number of pixels in the platform area.

[0098] S52. The platform in operation usually has a higher brightness, such as lighting and other facilities. The average brightness of the platform area is extracted, and the calculation process is as follows:

[0099] Calculate the average brightness using the formula:

[0100]

[0101] Among them, L i is the brightness value of the i-th pixel in the platform area, and N is the total number of pixels in the platform area.

[0102] S53. Setting a threshold of infrared or night light characteristics. The set threshold is used to distinguish the operation status of the platform:

[0103]

[0104] Among them, T threshold and L threshold is the threshold value.

[0105] Reference Figure 3 The present invention also provides an offshore oil and gas platform monitoring system based on multi-source remote sensing data, which mainly includes the following modules:

[0106] Data acquisition module: responsible for acquiring SAR, optical, infrared and night light image data from multi-source remote sensing platforms and storing the data in the database;

[0107] Data preprocessing module: preprocess the image data, including radiation correction, geometric correction, denoising, etc., to provide high-quality data for subsequent processing;

[0108] Sample annotation and model training module: Manually annotate the location, outline and type of oil and gas platforms in high-resolution optical images, and then use these geometric positions to retrieve corresponding images in the SAR image library. By aligning the platform positions in the optical images, projecting these positions into the SAR images, and manually annotating them accurately, the annotation consistency of optical and SAR data is maintained;

[0109] SAR survey module: Use SAR images to conduct preliminary detection of oil and gas platforms and use detection algorithms to extract candidate areas of the platforms;

[0110] Optical inspection module: Identify the type of platform in the optical image to obtain the precise location and category of the platform, including jack-up drilling platforms, semi-submersible drilling platforms, and semi-submersible production storage and offloading platforms;

[0111] Infrared / night light monitoring module: Analyzes platform features in infrared or night light images to determine the platform's operating status.

[0112] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring an offshore oil and gas platform using multi-source remote sensing data, characterized in that: The steps include: S1. Acquire image data from a multi-source remote sensing platform and store it in a database, and pre-process the image data; the image data includes SAR images, optical images, infrared images, and night-light images; S2. Mark the location and type of the oil and gas platform in the optical image to generate initial annotation data; retrieve the corresponding area of ​​the initial annotation data in the SAR image and mark it to obtain a training set and a validation set; S3, inputting the training set into the offshore oil and gas platform recognition model for training, using the trained offshore oil and gas platform recognition model to recognize the oil and gas platforms in the SAR image to be monitored, obtaining the offshore oil and gas platform survey results, and using the coordinates of the offshore oil and gas platform survey results as candidate areas; S4. Filter out high-resolution optical images containing the candidate area from the image data, and use the trained offshore oil and gas platform recognition model to recognize the oil and gas platform in the high-resolution optical images containing the candidate area to obtain the precise location and category of the platform; S5. Calculate the precise location and category of the platform using infrared image features or night-light image features, and determine the operating status of the oil and gas platform to be monitored using a preset threshold.

2. The offshore oil and gas platform monitoring method based on multi-source remote sensing data according to claim 1, characterized in that: The specific steps of S2 are as follows: S21. marking the oil and gas platform in the optical image, and marking the location, outline and type of the platform using the visual information of the optical image; S22. Retrieve the corresponding image from the SAR image library using the position, outline and type of the annotated platform extracted from the optical image; align the platform position in the optical image, project these positions into the retrieved SAR image, accurately annotate the platform in the SAR image, and keep the annotations of the optical and SAR data consistent through geometric position calibration.

3. The offshore oil and gas platform monitoring method based on multi-source remote sensing data according to claim 1, characterized in that: In S3, the offshore oil and gas platform recognition model includes an input layer, a convolutional layer, a fully connected layer, a feature fusion layer and an output layer.

4. The offshore oil and gas platform monitoring method based on multi-source remote sensing data according to claim 3 is characterized in that: The specific steps of S3 are as follows: S31. Input the training set into the convolution layer through the input layer of the offshore oil and gas platform recognition model for feature extraction to obtain a feature map. The convolution layer formula is: f(x)=ReLU(W1*x+b1) Among them, f(x) is the convolution output, W1 is the convolution kernel, * represents the convolution operation, b1 is the bias, and ReLU is the activation function; S32, map the feature map to a feature vector of fixed length through a fully connected layer. The formula of the fully connected layer is: z=W2z+b2 Among them, z is the output feature vector, W2 is the weight matrix, and b2 is the bias; The characteristic vectors of the two modes of S33, SAR image and optical image are F SAR 、F Optical , then the fused features are expressed as: F fusion =[F SAR ;F Optical ] S34. Add a fully connected layer to the concatenated comprehensive feature vector to perform feature fusion and dimensionality reduction to extract comprehensive features: F′ fusion =ReLU(W fusion ·F fusion +b fusion ) Among them, W fusion and b fusion are the weights and biases of the fusion layer; Assuming there are N types of platforms, the output is the probability distribution of each type of platform: p type =softmax(W type ·F f ′ usion +b type ) Among them, W type and b type are the weights and biases of the classifier, and the softmax function is used to normalize the output probability; S35. Use the validation set to optimize the offshore oil and gas platform identification model, use the adaptive optimization algorithm to minimize the loss function, and obtain the optimized offshore oil and gas platform identification model; the loss function uses the cross entropy loss function: in, is the one-hot encoding of the actual type, is the predicted class probability.

5. The offshore oil and gas platform monitoring method based on multi-source remote sensing data according to claim 1, characterized in that: The specific steps of S5 are as follows: S51, extract the average temperature characteristics of the platform area, the calculation process is as follows: Calculate the average temperature using the formula: Among them, T i is the temperature value of the i-th pixel in the platform area, and N is the total number of pixels in the platform area; S52, extracting the average brightness feature of the platform area, the calculation process is as follows: Calculate the average brightness using the formula: Among them, L i is the brightness value of the i-th pixel in the platform area, and N is the total number of pixels in the platform area; S53. Set the threshold of the infrared image feature or the night light image feature, and distinguish the operation status of the platform by the average temperature, average brightness and the threshold: Among them, T threshold and L threshold is the threshold value.

6. A multi-source remote sensing data offshore oil and gas platform monitoring system, used to implement the multi-source remote sensing data offshore oil and gas platform monitoring method according to any one of claims 1 to 5, characterized in that: include: Data acquisition module: used to obtain SAR, optical, infrared and night light image data from multi-source remote sensing platforms and store the data in the database; Data preprocessing module: used to preprocess image data, including radiation correction, geometric correction and denoising; Sample annotation and model training module: used to annotate the location, outline and type of oil and gas platforms in high-resolution optical images, and then use these geometric positions to retrieve corresponding images in the SAR image library; by aligning the platform positions in the optical images, projecting these positions into the SAR images and accurately annotating them, the annotation consistency of optical and SAR data is maintained; SAR survey module: used to use SAR images for preliminary detection of oil and gas platforms and to extract candidate areas of the platforms using detection algorithms; Optical inspection module: used to identify the type of platform in the optical image and obtain the precise location and type of the platform, including jack-up drilling platform, semi-submersible drilling platform, and semi-submersible production storage and offloading platform; Infrared / night light monitoring module: used to analyze platform features in infrared or night light images and determine the platform's operating status.

7. A storage medium, characterized in that: The storage medium includes a stored program, wherein when the program is run, the offshore oil and gas platform monitoring method of multi-source remote sensing data described in any one of claims 1 to 5 is executed.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: The processor executes the offshore oil and gas platform monitoring method using multi-source remote sensing data as claimed in any one of claims 1 to 5 by running the computer program.

Images