A task processing method, apparatus, equipment, and medium based on the consistency characteristics of ocean polarization direction.

Abstract

This invention relates to the field of marine remote sensing technology, and discloses a task processing method, apparatus, equipment, and medium based on marine polarization direction consistency characteristics. Responding to processing commands for marine observation tasks, it performs radiometric calibration, correction, neighborhood window selection, and feature parameter setting on raw polarization remote sensing observation data of the target sea area, obtaining effective image data, selected neighborhood window parameters, and feature parameters adapted to the requirements of PDCF (Polypolarization Direction Consistency Feature) extraction. Based on the neighborhood window parameters, feature parameters, and the created marine PDCF index definition formula, PDCF is extracted from the effective image data to obtain a PDCF map. A corresponding multi-polarization feature fused image is generated based on the PDCF map. The multi-polarization feature fused image is input into a trained marine observation task processing model for task processing, obtaining the corresponding task processing results. This effectively improves the processing speed of marine observation tasks, thereby enhancing the real-time performance of marine observation task processing.

Description

Technical Field

[0001] This invention relates to the field of marine remote sensing technology, and in particular to a task processing method, apparatus, equipment and medium based on the consistency characteristics of marine polarization direction. Background Technology

[0002] With the development of science and technology, marine remote sensing technology is constantly improving.

[0003] Currently, polarization remote sensing, as an important means to improve target identification capabilities and enhance observation dimensions, has been widely introduced into marine remote sensing systems. By capturing the polarization state of water surface reflections, related technologies can effectively reflect the physical structure of the sea surface, interface disturbances, the distribution of floating bodies, and their interaction with the atmosphere, significantly enhancing the stability and accuracy of remote sensing methods under various complex sea conditions and enabling the processing of different marine observation tasks.

[0004] However, the relevant technologies are slow to process ocean observation tasks and have weak real-time processing capabilities. Summary of the Invention

[0005] This invention provides a task processing method, apparatus, equipment, and medium based on the consistency characteristics of ocean polarization direction, in order to solve the defects of slow processing speed and weak real-time processing capability of ocean observation tasks in related technologies, improve the processing speed of ocean observation tasks, and enhance the real-time performance of ocean observation task processing.

[0006] In a first aspect, the present invention provides a task processing method based on the consistency characteristics of ocean polarization direction, comprising:

[0007] In response to the processing instructions for the ocean observation mission, radiometric calibration, correction, neighborhood window selection, and feature parameter setting are performed on the raw polarization remote sensing observation data of the target sea area to obtain effective image data, well-selected neighborhood window parameters, and feature parameters for adapting to the requirements of PDCF feature extraction of ocean polarization direction consistency features.

[0008] Based on the neighborhood window parameters, the feature parameters, and the created marine PDCF index definition formula, PDCF is extracted from the effective image data to obtain a PDCF map;

[0009] Generate a corresponding multi-polarization feature fusion image based on the PDCF map;

[0010] The multi-polarization feature fused image is input into a trained ocean observation task processing model for task processing to obtain the corresponding task processing results.

[0011] Optionally, the formula for defining the marine PDCF index is:

[0012] ;

[0013] ;

[0014] in, and These are the cell row order and cell column order in the valid image data, respectively, used to uniquely identify a cell; For the pixels in the effective pixel data of ; This represents the total number of pixels in the neighborhood. For pixels The polarization angle; the PDCF range is A higher value indicates a more consistent direction.

[0015] Optionally, the process of radiometric calibration, correction, neighborhood window selection, and feature parameter setting of the raw polarization remote sensing data of the target sea area to obtain effective image data, selected neighborhood window parameters, and feature parameters adapted to the requirements of PDCF feature extraction for ocean polarization direction consistency includes:

[0016] Radiometric calibration was performed on the raw polarization remote sensing observation data of the target sea area to obtain radiometrically calibrated Stokes vector data;

[0017] The Stokes vector data is corrected to obtain the effective image data.

[0018] Neighborhood windows are selected from the valid image data to obtain the neighborhood window parameters;

[0019] Feature parameters are set in the effective image data and the neighborhood window parameters to obtain feature parameters that can be adapted to the needs of marine PDCF extraction.

[0020] Optionally, the correction of the Stokes vector data to obtain the effective image data includes:

[0021] The Stokes vector data is polarized atmospherically corrected to obtain surface polarized Stokes vector data;

[0022] Geometric correction is performed on the surface polarization Stokes vector data to obtain a standard map projection image;

[0023] The standard map projection image is subjected to cloud masking to obtain the effective image data.

[0024] Optionally, generating the corresponding multi-polarization feature fusion image based on the PDCF map includes:

[0025] Obtain the polarization comparison map and angular polarization map corresponding to the original polarization remote sensing observation data;

[0026] The multi-polarization feature fusion image is generated by combining the polarization comparison map, the angular polarization map, and the PDCF map.

[0027] Optionally, when the ocean observation task is ship monitoring and trajectory tracking, the task processing result is the geographical location information of the target ship at different timestamps;

[0028] When the ocean observation task is to identify disturbance boundaries, the task processing result is the disturbance boundary delineation result;

[0029] When the ocean observation task is water body classification, the task processing result is a water body classification result that distinguishes the image areas corresponding to sea ice, sediment areas, and clean seawater in the target sea area using different colors or symbols.

[0030] Secondly, the present invention provides a task processing device based on the consistency characteristics of ocean polarization direction, comprising:

[0031] The first processing unit is used to respond to the processing instructions of the ocean observation mission, and to perform radiometric calibration, correction, neighborhood window selection and feature parameter setting on the polarization remote sensing raw observation data of the target sea area, so as to obtain effective image data, selected neighborhood window parameters and feature parameters for adapting to the requirements of PDCF feature extraction of ocean polarization direction consistency features.

[0032] The feature extraction unit is used to extract PDCF from the effective image data according to the neighborhood window parameters, the feature parameters, and the created marine PDCF index definition formula to obtain a PDCF map.

[0033] A fusion unit is used to generate a corresponding multi-polarization feature fused image based on the PDCF map;

[0034] The second processing unit is used to input the multi-polarization feature fused image into the trained marine observation task processing model for task processing, and obtain the corresponding task processing results.

[0035] Optionally, the formula for defining the marine PDCF index is:

[0036] ;

[0037] ;

[0038] in, and These are the cell row order and cell column order in the valid image data, respectively, used to uniquely identify a cell; For the pixels in the effective pixel data of ; This represents the total number of pixels in the neighborhood. For pixels The polarization angle; the PDCF range is A higher value indicates a more consistent direction.

[0039] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the task processing method based on the ocean polarization direction consistency characteristics described in the first aspect or any corresponding embodiment.

[0040] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the task processing method based on ocean polarization direction consistency characteristics described in the first aspect or any corresponding embodiment thereof.

[0041] The present invention provides a task processing method, apparatus, device, and medium based on ocean polarization direction consistency characteristics. Responding to processing instructions for ocean observation tasks, it performs radiometric calibration, correction, neighborhood window selection, and feature parameter setting on raw polarization remote sensing observation data of the target sea area to obtain effective image data, selected neighborhood window parameters, and feature parameters adapted to the requirements of ocean polarization direction consistency feature (PDCF) extraction. Based on the neighborhood window parameters, feature parameters, and the created ocean PDCF index definition formula, PDCF is extracted from the effective image data to obtain a PDCF map. A corresponding multi-polarization feature fused image is generated based on the PDCF map. The multi-polarization feature fused image is input into a trained ocean observation task processing model for task processing to obtain the corresponding task processing results, effectively improving the processing speed of ocean observation tasks and thus enhancing the real-time performance of ocean observation task processing. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in this invention or related technologies, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 A flowchart illustrating a task processing method based on the consistency characteristics of ocean polarization direction provided in an embodiment of the present invention;

[0044] Figure 2 A task processing flowchart based on a remote sensing solution system is provided in this embodiment of the invention;

[0045] Figure 3 A schematic diagram illustrating a software-to-hardware implementation according to an embodiment of the present invention;

[0046] Figure 4 This is a schematic diagram of real-time remote sensing calculation and software / hardware application of an unmanned aerial vehicle (UAV) platform provided in an embodiment of the present invention.

[0047] Figure 5 This is a schematic diagram comparing the time consumption of different schemes for remote sensing calculation on an unmanned aerial vehicle platform, provided by an embodiment of the present invention.

[0048] Figure 6 This is a schematic diagram of a task processing device based on the consistency characteristics of ocean polarization direction provided in an embodiment of the present invention;

[0049] Figure 7 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0051] The following is combined Figures 1-5 This invention describes a task processing method based on the consistency characteristics of ocean polarization direction.

[0052] like Figure 1 As shown, this embodiment proposes a first task processing method based on the consistency characteristics of ocean polarization direction. This method may include the following steps:

[0053] S101. In response to the processing instructions for the ocean observation mission, perform radiometric calibration, correction, neighborhood window selection, and feature parameter setting on the raw polarization remote sensing observation data of the target sea area to obtain effective image data, selected neighborhood window parameters, and feature parameters for adapting to the requirements of ocean polarized direction coherence feature (PDCF) extraction.

[0054] Among them, ocean observation tasks can include ship monitoring and tracking, disturbance boundary detection, and water body classification in target sea areas.

[0055] Optionally, step S101 includes:

[0056] Radiometric calibration was performed on the raw polarization remote sensing observation data of the target sea area to obtain radiometrically calibrated Stokes vector data;

[0057] The Stokes vector data is corrected to obtain effective image metadata;

[0058] Neighborhood window selection is performed on valid image metadata to obtain neighborhood window parameters;

[0059] Feature parameters are set in the effective image metadata and neighborhood window parameters to obtain feature parameters that can be adapted to the needs of marine PDCF extraction.

[0060] Specifically, the raw polarization remote sensing data for the target sea area includes pre-calibrated grayscale values, which have no physical meaning and correspond to the radiance flux under polarizers at various bands, angles, and angles. This embodiment radiometrically calibrates the raw polarization remote sensing data, obtaining calibration coefficients through a calibration field / target. This yields radiometrically calibrated atmospheric top-of-atmosphere sensor observation data, i.e., radiometrically calibrated Stokes vector data, which has practical physical meaning and is identical to the physical meaning of radiance, thus eliminating errors caused by the sensor's own response characteristics.

[0061] Specifically, this embodiment can perform physical correction on the radiometrically calibrated Stokes vector data to obtain the corresponding effective image data.

[0062] Specifically, in this embodiment, a neighborhood window can be selected from the effective image data. A neighborhood window of appropriate size is selected for each pixel to obtain the selected neighborhood window parameters. Then, this embodiment can set feature parameters based on the neighborhood window parameters and the effective image data to obtain feature parameters suitable for extracting ocean polarization direction consistency features (PDCF).

[0063] It should be noted that this embodiment can be applied to marine observation tasks such as ship detection, water pollution (e.g., lakes, rivers, tidal flats, oceans), and boundary identification. It introduces the PDCF index for joint analysis with other polarization layers, and sets feature parameters such as sliding window scale, angular bandwidth, and spatial constraints to achieve adaptive configuration of feature extraction strategies. This embodiment can achieve dynamic optimization and scenario customization according to the application requirements of marine observation tasks, demonstrating task-oriented flexibility.

[0064] Optionally, the above correction of the Stokes vector data to obtain effective image metadata includes:

[0065] Polarization atmospheric correction is performed on the Stokes vector data to obtain surface polarization Stokes vector data;

[0066] Geometric correction is performed on the surface polarization Stokes vector data to obtain a standard map projection image;

[0067] Cloud masking is performed on standard map projection images to obtain effective image data.

[0068] Specifically, this embodiment can construct a polarization radiation transfer model based on sea-atmosphere interaction, correct atmospheric interference with polarization signals, and restore the true polarization state of the sea surface, including linear polarization degree, polarization angle, and multi-angle polarization reflectivity.

[0069] Specifically, this embodiment can perform polarization atmospheric correction on the radiometrically calibrated Stokes vector data through vector radiative transfer calculation and atmospheric extinction characteristic inversion to obtain surface polarized Stokes vector data after removing atmospheric influence, thus realizing the process from the top of the atmosphere to the surface.

[0070] Specifically, this embodiment can combine attitude and orbit data, ephemeris information, and digital elevation model to eliminate distortion and misalignment of images during the imaging process through precise geometric position adjustment. This eliminates the geometric distortion of sea surface polarization state data after polarization atmospheric correction, achieves geometric correction of surface polarization Stokes vector data, and obtains a standard map projection image, i.e., sea surface polarization state data under a unified spatial reference system, providing spatial consistency assurance for subsequent feature extraction and analysis.

[0071] Specifically, this embodiment can combine polarization texture features and radiation consistency index to construct a cloud detection and recognition model. The cloud detection and recognition model is used to automatically mark and remove cloud-affected areas in the standard map projection image. The cloud masking of the standard map projection image by the line of sight obtains effective image data, providing a reliable observation basis for subsequent polarization direction consistency feature extraction.

[0072] S102. Based on the neighborhood window parameters, feature parameters, and the created PDCF definition formula, extract the ocean polarization direction consistency feature PDCF from the effective image data to obtain the PDCF map.

[0073] Specifically, this embodiment uses neighborhood window parameters, set feature parameters, and a created marine PDCF index definition formula to extract PDCF from valid image data to obtain a PDCF map.

[0074] Optionally, the formula for defining the marine PDCF index is:

[0075] ;

[0076] ;

[0077] in, and These are the cell row order and cell column order in the valid pixel data, respectively, used to uniquely identify a cell; For effective pixel data of ; This represents the total number of pixels in the neighborhood. For pixels The polarization angle; the PDCF range is A higher value indicates a more consistent direction.

[0078] S103. Generate the corresponding multi-polarization feature fusion image based on the PDCF map.

[0079] Specifically, in this embodiment, after obtaining the PDCF image, a multi-polarization feature fusion image can be generated based on the PDCF image.

[0080] Optionally, step S103 includes:

[0081] Obtain polarization comparison maps and angular polarization maps corresponding to the raw polarization remote sensing observation data;

[0082] A multi-polarization feature fusion image is generated by combining polarization contrast map, angular polarization map and PDCF map.

[0083] Specifically, in this embodiment, after generating the PDCF map, an enhanced multi-polarization feature fusion image can be generated by simultaneously combining the polarization contrast map and the angular polarization map.

[0084] It should be noted that this embodiment can achieve complementary fusion of multi-dimensional polarization features, which can not only enhance the salience of weak targets (such as distant ships and buoys), but also improve the ability to identify complex disturbances (such as oil film diffusion and wave interference).

[0085] S104. Input the multi-polarization feature fused image into the trained marine observation task processing model for task processing to obtain the corresponding task processing results.

[0086] Among them, the ocean observation task processing model is a trained ocean observation task processing model obtained by training a pre-trained task processing model using training data.

[0087] Specifically, the processing model for ocean observation tasks can be an artificial intelligence model such as vector machine, convolutional neural network, and deep neural network.

[0088] Optionally, when the ocean observation task is ship monitoring and trajectory tracking, the task processing result is the geographical location information of the target ship at different timestamps;

[0089] When the ocean observation task is to identify disturbance boundaries, the task processing result is the disturbance boundary delineation result;

[0090] When the ocean observation task is water body classification, the task processing result is the water body classification result that uses different colors or symbols to distinguish the image areas corresponding to sea ice, sediment areas, and clean seawater in the target sea area.

[0091] Specifically, in this embodiment, multi-polarization feature fused images can be input into a trained marine observation task processing model, so that the marine observation task processing model can perform task processing by comprehensively utilizing multi-scale spatial features and polarization direction consistency features, and obtain the target location, disturbance boundary or thematic classification map output by the model.

[0092] Here, target location refers to the specific coordinate information of a particular target in the ocean, such as a ship or buoy, on the image, determined through model analysis. For example, in a remote sensing image of a certain area, the model identifies a ship, and its target location might be expressed in pixel coordinates (…). x , y This indicates that the coordinates correspond to a geographical location in the actual ocean and can be used to track the real-time movement or historical trajectory of ships. In this case, this embodiment can be used to accurately identify the geographical location of specific targets such as ships and buoys, and can achieve dynamic monitoring and trajectory tracking by combining time series data.

[0093] The disturbance boundary refers to the dividing line between the area of ​​the sea surface disturbed by factors such as oil slick diffusion and ship navigation, and the surrounding normal sea surface area. For example, in oil slick detection, the disturbance boundary clearly delineates the extent of oil slick coverage, helping staff quickly determine the size and shape of the contaminated area and providing accurate spatial reference for subsequent cleanup and control work. In this case, this embodiment can detect the boundary area of ​​the sea surface caused by oil slicks, navigation tracks, or other abnormal disturbances, forming a spatially continuous disturbance profile.

[0094] Thematic classification maps refer to images that are divided into different thematic categories and presented visually according to different marine observation needs. For example, in water body classification, thematic classification maps mark different types of water bodies such as sea ice, sedimentary areas, and clean seawater with different colors or symbols, enabling researchers or relevant departments to intuitively understand the water body distribution in the area and providing a clear classification basis for marine resource surveys and environmental assessments. In this embodiment, multi-source polarization features and machine learning classifiers can be combined to finely classify different water body categories such as sea ice, sedimentary areas, and clean seawater, and output them in a visual form.

[0095] It should be noted that this embodiment introduces PDCF as a core indicator to reveal sea surface physical processes from the perspective of polarization direction consistency, overcoming the limitations of related technologies based on single radiation or spectral features. This embodiment achieves joint modeling of Stokes vectors, polarization contrast maps, angular polarization maps, and PDCF, significantly improving feature discrimination. This embodiment can also achieve intelligent scene adaptation, adaptively adjusting parameter configuration and feature fusion strategies through a learning model, enabling the system to maintain efficiency and robustness in different application scenarios. The application-oriented solution output of this embodiment not only provides basic imagery but can also directly generate target detection results, disturbance monitoring results, and thematic classification products that can be used for decision support.

[0096] The task processing method based on ocean polarization direction consistency features proposed in this embodiment can respond to processing instructions for ocean observation tasks by performing radiometric calibration, correction, neighborhood window selection, and feature parameter setting on the raw polarization remote sensing observation data of the target sea area to obtain effective image data, selected neighborhood window parameters, and feature parameters adapted to the requirements of ocean polarization direction consistency feature (PDCF) extraction. Based on the neighborhood window parameters, feature parameters, and the created ocean PDCF index definition formula, PDCF is extracted from the effective image data to obtain a PDCF map. A corresponding multi-polarization feature fused image is generated based on the PDCF map. The multi-polarization feature fused image is input into a trained ocean observation task processing model for task processing to obtain the corresponding task processing results, effectively improving the processing speed of ocean observation tasks and thus enhancing the real-time performance of ocean observation task processing.

[0097] based on Figure 1 ,like Figure 2 As shown, this embodiment proposes a second task processing method based on the consistency characteristics of ocean polarization direction. In this method, the radiometric calibration, correction, neighborhood window selection, feature parameter setting, PCDF extraction and task processing of the original polarization remote sensing observation data can be packaged into modules to construct a remote sensing solution system.

[0098] Specifically, the remote sensing solution system includes a data preprocessing and physical correction module, as well as a polarization direction consistency feature extraction and application solution module.

[0099] Among them, the data preprocessing module is a radiometric calibration module, which is used to radiometrically calibrate the raw polarization remote sensing observation data of the target sea area to obtain radiometrically calibrated Stokes vector data.

[0100] The physical correction module includes a polarization atmospheric correction module, a geometric correction module, and a cloud masking module. The polarization atmospheric correction module performs polarization atmospheric correction on the radiometrically calibrated Stokes vector data to obtain surface polarization Stokes vector data. The geometric correction module performs geometric correction on the surface polarization Stokes vector data to obtain a standard map projection image. The cloud masking module applies a cloud mask to the standard map projection image to obtain valid image data.

[0101] Specifically, the polarization direction consistency feature extraction and application solution module includes a neighborhood window selection module, a feature parameter setting module, a PCDF feature extraction module, and a marine observation application module.

[0102] Specifically, in this embodiment, valid image data can be input into the neighborhood window parameters to select a neighborhood window, thus obtaining the neighborhood window parameters. The valid image data, neighborhood window parameters, and valid image data are then input into the feature parameter setting module for parameter setting, resulting in the configured feature parameters.

[0103] In this embodiment, effective image data, neighborhood window parameters, and feature parameters can be input into the PCDF feature extraction module for PCDF extraction to obtain a PCDF image.

[0104] Specifically, the marine observation application module includes a fusion module and a trained marine observation task processing model. In this embodiment, the PCDF image is input into the fusion module so that the fusion module can combine the polarization contrast image, angular polarization image, and PDCF image to generate a multi-polarization feature fused image.

[0105] Specifically, in this embodiment, multi-polarization feature fused images can be input into a trained ocean observation task processing model for task processing to obtain the corresponding task processing results.

[0106] It should also be noted that, given the limitations of software processing in terms of processing speed and deployment flexibility, and the exponential growth of high-resolution remote sensing data, in order to achieve platform-level real-time computing capabilities and low-power operation capabilities, this embodiment further proposes a software-hardware implementation architecture. Through software-hardware collaborative deployment and function mapping design, key computing links are migrated from software to dedicated acceleration units in the hardware platform, achieving overall optimization from the algorithm level to the platform level.

[0107] To achieve a breakthrough in system-level computational performance, this embodiment introduces a modular and layered design concept, constructing a three-layer integrated architecture of control scheduling, feature computation, communication feedback, and task management.

[0108] In this embodiment, a heterogeneous computing architecture is adopted in the control and scheduling layer, including logical modules such as data reception, task division, parameter configuration, and instruction scheduling, which are executed by a general processing unit (such as the main control MCU).

[0109] The feature computation layer is the core of hardware acceleration. It supports efficient input and output of image data and processing results, employs standardized communication protocols and a unified intermediate data format, ensuring high compatibility and stability of the system's external interfaces.

[0110] In practical applications, test results show that this software-hardware collaborative platform can process a 4000×4000 pixel ocean polarization remote sensing image and extract ocean polarization direction consistency features in less than 1.5 seconds. This represents a 5-10 times improvement in processing efficiency compared to traditional platforms, with power consumption below 15W. This fully meets the lightweight embedded deployment requirements of edge devices, shipborne platforms, and UAV processors, fully validating its engineering practicality. The overall software-hardware implementation architecture is as follows: Figure 3 As shown.

[0111] This embodiment can achieve the following technical effects in typical application scenarios. In marine target detection scenarios, this embodiment enhances the robustness of traditional grayscale image recognition by identifying ships, buoys, and residual targets on the water surface through polarization angle comparison. In water body classification and pollution identification, this embodiment can accurately identify different types of water bodies such as sea ice, sediment, and oil films and extract their boundaries based on polarization texture differences and spectral consistency. This embodiment can also perform physical parameter inversion, combining a multi-angle polarization reflection model to invert parameters such as sea surface wind speed, wave height, roughness, and chlorophyll concentration.

[0112] In the scenario of automated generation of thematic products, this embodiment has developed an intelligent product generation system that can automatically output various types of products such as angular polarization maps, target location information maps, pollution monitoring thematic maps, and time series change maps, and supports format standardization and platform integration.

[0113] In practical applications, the software and hardware implementation methods of this embodiment have been initially applied to UAV remote sensing platforms, achieving end-to-end real-time processing from data acquisition to product generation. The application effects and processes are as follows: Figure 4 As shown.

[0114] This embodiment quantitatively evaluates the verification results on the UAV remote sensing platform, such as... Figure 5 As shown. This embodiment, while maintaining consistency in algorithm, code, and features, only considers the advantages brought by software-hardware integration. Considering the combined improvement in computing speed compared to traditional solutions across three scenarios, the efficiency improvement meets theoretical expectations of 5-10 times. Experiments quantitatively demonstrate the efficiency advantage of this patented solution. On satellite platforms, due to the greater complexity of atmospheric correction and the higher algorithm complexity, the improvement in computing speed will be even more significant.

[0115] This embodiment introduces polarization direction consistency features for the first time in the field of polarized ocean remote sensing, and combines them with a software and hardware solution framework to achieve the following technical effects: effectively enhance the sensitivity and stability of identifying sea surface disturbance targets; introduce a sliding window consistency calculation index to improve the efficiency of polarization angle signal utilization; the software and hardware collaborative structure significantly reduces processing latency and power consumption; the features are highly scalable, adaptable to various remote sensing tasks, and have broad engineering application prospects.

[0116] Specifically, this embodiment achieves accurate identification of sea surface disturbance anomalies (such as ships, oil slicks, and floating bodies) by fusing the spatial continuity of polarization direction information and constructing a quantitative consistency index. Combined with a hardware and software collaborative deployment architecture, this invention realizes real-time extraction, dynamic analysis, and efficient application output of polarization features. It is applicable to various platforms such as edge computing, airborne, and shipborne systems, and features low power consumption, high real-time performance, and strong environmental adaptability, meeting the stringent timeliness requirements for data processing in scenarios such as marine environmental monitoring and emergency response.

[0117] like Figure 6 As shown, this embodiment proposes a task processing device based on the consistency characteristics of ocean polarization direction. The device may include:

[0118] The first processing unit 601 is used to respond to the processing instructions for the ocean observation mission, and to perform radiometric calibration, correction, neighborhood window selection and feature parameter setting on the polarization remote sensing raw observation data of the target sea area, so as to obtain effective image data, selected neighborhood window parameters and feature parameters for adapting to the requirements of PDCF feature extraction of ocean polarization direction consistency features.

[0119] The feature extraction unit 602 is used to extract PDCF from effective image data according to the neighborhood window parameters, feature parameters and the created marine PDCF index definition formula to obtain a PDCF map.

[0120] The fusion unit 603 is used to generate a corresponding multi-polarization feature fused image based on the PDCF map;

[0121] The second processing unit 604 is used to input multi-polarization feature fused images into a trained marine observation task processing model for task processing, and obtain the corresponding task processing results.

[0122] It should be noted that the processing procedures of the first processing unit 601, the feature extraction unit 602, the fusion unit 603, and the second processing unit 604, as well as their beneficial effects, can be referred to respectively. Figure 1 Steps S101 to S104 in the process will not be described again.

[0123] Optionally, the formula for defining the marine PDCF index is:

[0124] ;

[0125] ;

[0126] in, and These are the cell row order and cell column order in the valid pixel data, respectively, used to uniquely identify a cell; For effective pixel data of ; This represents the total number of pixels in the neighborhood. For pixels The polarization angle; the PDCF range is A higher value indicates a more consistent direction.

[0127] Optionally, the first processing unit 601 is also used for:

[0128] Radiometric calibration was performed on the raw polarization remote sensing observation data of the target sea area to obtain radiometrically calibrated Stokes vector data;

[0129] The Stokes vector data is corrected to obtain effective image metadata;

[0130] Neighborhood window selection is performed on valid image metadata to obtain neighborhood window parameters;

[0131] Feature parameters are set in the effective image metadata and neighborhood window parameters to obtain feature parameters that can be adapted to the needs of marine PDCF extraction.

[0132] Optionally, the first processing unit 601 is also used for:

[0133] Polarization atmospheric correction is performed on the Stokes vector data to obtain surface polarization Stokes vector data;

[0134] Geometric correction is performed on the surface polarization Stokes vector data to obtain a standard map projection image;

[0135] Cloud masking is performed on standard map projection images to obtain effective image data.

[0136] Optionally, the fusion unit 603 is also used for:

[0137] Obtain polarization comparison maps and angular polarization maps corresponding to the raw polarization remote sensing observation data;

[0138] A multi-polarization feature fusion image is generated by combining polarization contrast map, angular polarization map and PDCF map.

[0139] Optionally, when the ocean observation task is ship monitoring and trajectory tracking, the task processing result is the geographical location information of the target ship at different timestamps;

[0140] When the ocean observation task is to identify disturbance boundaries, the task processing result is the disturbance boundary delineation result;

[0141] When the ocean observation task is water body classification, the task processing result is the water body classification result that distinguishes the image areas corresponding to sea ice, sediment areas and clean seawater in the target sea area using different colors or symbols.

[0142] The task processing device based on ocean polarization direction consistency characteristics proposed in this embodiment can respond to processing instructions for ocean observation tasks by performing radiometric calibration, correction, neighborhood window selection, and feature parameter setting on raw polarization remote sensing observation data of the target sea area. This yields effective image data, selected neighborhood window parameters, and feature parameters adapted to the requirements of PDCF (Position Direction Consistency Function) extraction. Based on the neighborhood window parameters, feature parameters, and the created ocean PDCF index definition formula, PDCF is extracted from the effective image data to obtain a PDCF map. A corresponding multi-polarization feature fusion image is generated based on the PDCF map. This multi-polarization feature fusion image is then input into a trained ocean observation task processing model for task processing, yielding the corresponding task processing results. This effectively improves the processing speed of ocean observation tasks, thereby enhancing the real-time performance of ocean observation task processing.

[0143] In this embodiment, the task processing device based on the ocean polarization direction consistency characteristic is presented in the form of a functional unit. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0144] This invention also provides a computer device having the above-described features. Figure 6 The shown is a task processing device based on the consistency characteristics of ocean polarization direction.

[0145] Please see Figure 7The present invention provides a schematic diagram of the structure of a computer device according to an optional embodiment. The computer device includes one or more processors 10, a memory 20, and interfaces for connecting the various components, including high-speed interfaces and low-speed interfaces. The various components are interconnected via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, multiple processors and / or multiple buses can be used with multiple memories, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 7 Take a processor 10 as an example.

[0146] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0147] The memory 20 stores instructions executable by at least one processor 10 to cause at least one processor 10 to perform the method shown in the above embodiments.

[0148] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function. The data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, which can be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0149] Memory 20 may include volatile memory, such as random access memory. Memory may also include non-volatile memory, such as flash memory, hard disk, or solid-state drive. Memory 20 may also include combinations of the above types of memory.

[0150] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0151] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0152] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A task processing method based on the consistency characteristics of ocean polarization direction, characterized in that, include: In response to the processing instructions for the ocean observation mission, radiometric calibration, correction, neighborhood window selection, and feature parameter setting are performed on the raw polarization remote sensing observation data of the target sea area to obtain effective image data, well-selected neighborhood window parameters, and feature parameters for adapting to the requirements of PDCF feature extraction of ocean polarization direction consistency features. Based on the neighborhood window parameters, the feature parameters, and the created marine PDCF index definition formula, PDCF is extracted from the effective image data to obtain a PDCF map; Generate a corresponding multi-polarization feature fusion image based on the PDCF map; The multi-polarization feature fused image is input into a trained marine observation task processing model for task processing to obtain the corresponding task processing results. The formula for defining the marine PDCF index is as follows: ； ； in, and These are the cell row order and cell column order in the valid image data, respectively, used to uniquely identify a cell; For the pixels in the effective pixel data of ; This represents the total number of pixels in the neighborhood. For pixels The polarization angle; the PDCF range is A higher value indicates a more consistent direction.

2. The method according to claim 1, characterized in that, The process involves radiometric calibration, correction, neighborhood window selection, and feature parameter setting of the raw polarization remote sensing data for the target sea area to obtain effective image data, well-selected neighborhood window parameters, and feature parameters adapted to the requirements of PDCF (Polypolarization Direction Consistency Feature) extraction for ocean polarization direction consistency. This includes: Radiometric calibration was performed on the raw polarization remote sensing observation data of the target sea area to obtain radiometrically calibrated Stokes vector data; The Stokes vector data is corrected to obtain the effective image data. Neighborhood windows are selected from the valid image data to obtain the neighborhood window parameters; Feature parameters are set in the effective image data and the neighborhood window parameters to obtain feature parameters that can be adapted to the needs of marine PDCF extraction.

3. The method according to claim 2, characterized in that, The correction of the Stokes vector data to obtain the effective image data includes: The Stokes vector data is polarized atmospherically corrected to obtain surface polarized Stokes vector data; Geometric correction is performed on the surface polarization Stokes vector data to obtain a standard map projection image; The standard map projection image is subjected to cloud masking to obtain the effective image data.

4. The method according to claim 1, characterized in that, The step of generating a corresponding multi-polarization feature fusion image based on the PDCF map includes: Obtain the polarization comparison map and angular polarization map corresponding to the original polarization remote sensing observation data; The multi-polarization feature fusion image is generated by combining the polarization comparison map, the angular polarization map, and the PDCF map.

5. The method according to any one of claims 1 to 4, characterized in that, When the ocean observation task is ship monitoring and trajectory tracking, the task processing result is the geographical location information of the target ship at different timestamps; When the ocean observation task is to identify disturbance boundaries, the task processing result is the disturbance boundary delineation result; When the ocean observation task is water body classification, the task processing result is a water body classification result that distinguishes the image areas corresponding to sea ice, sediment areas, and clean seawater in the target sea area using different colors or symbols.

6. A task processing device based on the consistency characteristics of ocean polarization direction, characterized in that, include: The first processing unit is used to respond to the processing instructions of the ocean observation mission, and to perform radiometric calibration, correction, neighborhood window selection and feature parameter setting on the polarization remote sensing raw observation data of the target sea area, so as to obtain effective image data, selected neighborhood window parameters and feature parameters for adapting to the requirements of PDCF feature extraction of ocean polarization direction consistency features. The feature extraction unit is used to extract PDCF from the effective image data according to the neighborhood window parameters, the feature parameters, and the created marine PDCF index definition formula to obtain a PDCF map. A fusion unit is used to generate a corresponding multi-polarization feature fused image based on the PDCF map; The second processing unit is used to input the multi-polarization feature fused image into the trained marine observation task processing model for task processing, and obtain the corresponding task processing results. The formula for defining the marine PDCF index is as follows: ； ； in, and These are the cell row order and cell column order in the valid image data, respectively, used to uniquely identify a cell; For the pixels in the effective pixel data of ; This represents the total number of pixels in the neighborhood. For pixels The polarization angle; the PDCF range is A higher value indicates a more consistent direction.

7. A computer device, characterized in that, include: A memory and a processor are communicatively connected, the memory stores computer instructions, and the processor executes the task processing method based on ocean polarization direction consistency characteristics as described in any one of claims 1 to 5 by executing the computer instructions.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the task processing method based on the ocean polarization direction consistency feature as described in any one of claims 1 to 5.

Images