Marine search and rescue method and system based on remote sensing target search and water fall trajectory prediction
By obtaining marine parameters and using drifting algorithm models to predict the trajectory of maritime target objects, and combining remote sensing images to confirm the location, the commonality and timeliness of existing maritime search and rescue methods are solved, and fast and accurate search and rescue results are achieved, and through co-construction, The platform supports global use and data optimization, improving search and rescue efficiency.
Patent Information
- Application Number
- CN202111000442.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-08-27
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2041-08-27
AI Technical Summary
The existing maritime search and rescue methods lack universality and timeliness, and cannot continuously improve the overall search and rescue efficiency. They mainly rely on experience and local water trace meteorological data, making it difficult to ensure the correctness and accuracy of the results.
By obtaining the ocean parameters of the fallen site, selecting appropriate drifting algorithm models, combining water mark meteorological data, float coefficients and geographical information to predict the target object trajectory, and using remote sensing images to confirm the location, using search and rescue sharing and co-construction platform to achieve flexible import of the algorithm model and the accuracy of the calculation results.
The target position confirmation is achieved quickly and accurately, ensuring the referenceability and effectiveness of search and rescue results. By adjusting the granularity and time of deduction, the search and rescue efficiency is improved, and the co-construction platform supports global universality and diversity. Use of data source.
Smart Images

Figure CN113963244B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a search and rescue method and system, and in particular to a search and rescue method and system at sea based on remote sensing target search and water fall trajectory prediction. Background Art
[0002] With the expansion of the scale of marine development and the frequent maritime activities, maritime search and rescue work has received more and more attention from coastal countries. For the rapidly developing maritime transportation and fishing industries, maritime search and rescue operations can provide irreplaceable protection for the safety of Shanghai operators and property.
[0003] my country has vast sea areas, and the water patterns in various sea areas are complex and vary greatly, resulting in different search and rescue methods. At present, my country's maritime search and rescue command and coordination work mainly relies on the experience and judgment of search and rescue commanders to implement search and rescue. Some areas improve search and rescue efficiency by introducing local water pattern meteorological data and training drift algorithm models to predict the drift trajectory of falling into the water, but this method is not universal and cannot continuously improve the overall search and rescue efficiency. A method that can be widely used and has timeliness, can ensure the correctness of the results, and improve the search and rescue efficiency is needed. Summary of the invention
[0004] The technical problem to be solved by the present invention is to provide a maritime search and rescue method and system based on remote sensing target search and water-falling trajectory prediction, determine the ocean parameters according to the water-falling position of the target object, predict the position of the target object based on the ocean parameters, and confirm the position of the target object according to the remote sensing image near the predicted position, so as to quickly obtain the results and ensure the reference and accuracy of the results; and realize the flexible introduction of the algorithm model through the search and rescue sharing and co-construction platform to ensure the advancement of the algorithm model and the accuracy of the calculation results.
[0005] The technical solution adopted by the present invention to solve the above-mentioned technical problems is to provide a marine search and rescue method based on remote sensing target search and water-falling trajectory prediction, comprising the following steps: S1: obtaining the location where the target falls into the water; selecting a drift algorithm model according to the access parameters of the water-falling location; S2: calculating the target trajectory according to the access parameters and the drift algorithm model to obtain the predicted position of the target; S3: obtaining remote sensing data around the predicted position of the target, selecting a possible range according to the remote sensing data, and extracting a set of marine targets through target detection; S4: screening the set of marine targets, confirming the location of the target, and launching search and rescue.
[0006] Furthermore, the parameters accessed according to the falling point in step S1 include water pattern meteorological data near the falling point, the floating coefficient and / or wind drag coefficient of the target object near the falling point, and geographic information data near the falling point.
[0007] Further, the step S2 includes: setting an integration time interval ΔT, setting a prediction time, according to the target floating coefficient, according to the time integration interval, using water ripple meteorological data, floating coefficient and / or wind drag coefficient and geographic information to obtain the target drift position in the time interval ΔT through the drift algorithm model, repeatedly iterating until the prediction time is set, and obtaining the target object prediction position of the set prediction time, the steps are as follows: S21: The water ripple meteorological data is discrete water ripple data, and the discrete water ripple data is superimposed on the vector direction according to the flow velocity, and the data is linked and smoothed by using a Gaussian function, and then the data is derived to obtain the direction and speed value of the flow velocity at each position, that is, the fluid field data; S22: According to the geographic information data, the fluid field data is introduced for fitting, so that the ocean current velocity and direction within the data range have longitude and latitude data; S23: According to ΔS=V*ΔT, the predicted position of the target object is obtained by integral calculation:
[0008]
[0009] Furthermore, the step S3 includes: obtaining remote sensing images near the predicted position of the target object, slicing the remote sensing images into image tiles, cutting the remote sensing images into multiple small pieces for image recognition, and obtaining a set of marine targets through an image target detection algorithm.
[0010] Another technical solution adopted by the present invention to solve the above-mentioned technical problems is to provide a maritime search and rescue system based on remote sensing target search and falling-water trajectory prediction, including a search and rescue sharing and co-construction platform, which implements the steps of the method described in any one of claims 1 to 4, and the search and rescue sharing and co-construction platform includes a WEB interaction terminal and a server terminal, and the WEB interaction terminal realizes user data input and parameter and model import; the server terminal calculates the target object position according to the input data and imported parameters and models, and evaluates the model and solidifies the model according to user feedback; there are multiple WEB interaction terminals, and the multiple WEB interaction terminals interact with the server terminal by establishing a long link, and the server terminal actively pushes the calculated data to the WEB interaction terminal according to the parameters and models, and presents the target object status in real time.
[0011] Furthermore, the WEB interactive terminal includes a user interaction module, which is used for the user to input the location where the target object falls into the water, set the integration time interval ΔT and the prediction duration; a data source import module, which is used to import water ripple meteorological data near the falling point and geographic information data near the falling point; a floating coefficient / wind drag coefficient setting module, which is used to import the floating coefficient and / or wind drag coefficient of the target object near the falling point; a drift algorithm model import module, which is used to import the drift algorithm model; a remote sensing detection model import module, which is used to import the remote sensing detection model; a service interaction module, which is used to establish a long link with the server through WEBSOCKET to realize data interaction.
[0012] Furthermore, the data source importing module imports water pattern meteorological data near the water impact point and geographic information data near the water impact point based on remote sensing geographic information.
[0013] Furthermore, the server includes a geographic information module for loading global geographic GIS information; a water ripple data processing module for loading water ripple meteorological data; a floatation coefficient / wind drag coefficient processing module for loading the floatation coefficient and / or wind drag coefficient; a drift algorithm module for loading the drift algorithm model; a remote sensing target detection module for loading the remote sensing target detection algorithm model; a parallel computing module for performing parallel computing based on the loaded data and model; a service support module for connecting to multiple WEB interactive terminals; a front-end interactive control module for controlling the service module to connect to multiple WEB interactive terminals at the same time; and a user parameter cache module for caching user parameters.
[0014] Furthermore, the server also includes a hardware resource management module for managing the hardware resources of the server where the server is located; a performance evaluation and model optimization module for evaluating and optimizing the model based on user feedback on search and rescue results; a data solidification module for solidifying the current optimal model; and a log module for recording work logs.
[0015] Compared with the prior art, the present invention has the following beneficial effects: the maritime search and rescue method and system based on remote sensing target search and falling-water trajectory prediction provided by the present invention determine ocean parameters according to the falling-water position of the target object, predict the position of the target object based on the ocean parameters and algorithm models, and confirm the position of the target object according to the remote sensing images near the predicted position, so as to quickly obtain the results and ensure the reference and accuracy of the results; simulate the trajectory deduction after falling into the water through ocean parameters and algorithm models, and ensure the effectiveness by adjusting the deduction time granularity and deduction time; realize the flexible import of algorithm models through the search and rescue sharing and co-construction platform, ensure the advancement of the algorithm model and the accuracy of the calculation results, and the search and rescue sharing and co-construction platform is based on the Web to ensure ease of use; the search and rescue sharing and co-construction platform is universal and supports global use, supports multiple data sources, and ensures practicality; the search and rescue sharing and co-construction platform has an entry that supports extensive participation in jointly improving data and optimizing algorithm models, so as to ensure the accuracy of the method results. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] Figure 1 A flow chart of a method for searching and rescuing at sea based on remote sensing target search and prediction of falling-into-water trajectory according to an embodiment of the present invention;
[0017] Figure 2 This is an architecture diagram of a maritime search and rescue method based on remote sensing target search and water fall trajectory prediction according to an embodiment of the present invention;
[0018] Figure 3This is an architecture diagram of the search and rescue sharing and co-building platform according to an embodiment of the present invention. DETAILED DESCRIPTION
[0019] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.
[0020] Figure 1 A flow chart of a method for searching and rescuing at sea based on remote sensing target search and prediction of falling-into-water trajectory according to an embodiment of the present invention; Figure 2 This is an architecture diagram of a maritime search and rescue method based on remote sensing target search and water fall trajectory prediction according to an embodiment of the present invention.
[0021] See also Figure 1 and Figure 2 The marine search and rescue method based on remote sensing target search and water fall trajectory prediction of an embodiment of the present invention comprises the following steps:
[0022] S1: Obtain the location where the target object falls into the water; select the drift algorithm model according to the access parameters of the falling point; the parameters accessed according to the falling point include water ripple meteorological data near the falling point, the floating coefficient and / or wind drag coefficient of the target object near the falling point, and the geographic information data near the falling point.
[0023] S2: Calculate the target trajectory according to the access parameters and the drift algorithm model to obtain the predicted position of the target; set the integration time interval ΔT, set the prediction time, and according to the target floating coefficient, according to the time integration interval, use the water pattern meteorological data, floating coefficient and / or wind drag coefficient and geographic information through the drift algorithm model to obtain the target drift position in the time interval ΔT, and iterate repeatedly until the prediction time is set to obtain the predicted position of the target with the set prediction time. The steps are as follows:
[0024] S21: The water ripple meteorological data is discrete water ripple data. The discrete water ripple data is superimposed on the vector direction according to the flow velocity. After data linking and smoothing by using Gaussian function, the data is derived to obtain the direction and velocity value of the flow velocity at each position, i.e., the fluid field data.
[0025] S22: According to the geographic information data, the fluid field data is introduced for fitting, so that the velocity and direction of the ocean current within the data range have longitude and latitude data;
[0026] S23: According to ΔS=V*ΔT, the predicted position of the target object is obtained by integral calculation:
[0027]
[0028] S3: Acquire remote sensing data around the predicted position of the target, select a possible range based on the remote sensing data, and extract a set of offshore targets through target detection; acquire remote sensing images near the predicted position of the target, tile the remote sensing images, cut the remote sensing images into multiple small pieces for image recognition, and obtain a set of offshore targets through an image target detection algorithm.
[0029] S4: Screen the target set at sea, confirm the location of the target, and start search and rescue.
[0030] Please also see Figure 3 The maritime search and rescue system based on remote sensing target search and falling-water trajectory prediction of the embodiment of the present invention includes a search and rescue sharing and co-construction platform, which includes a WEB interaction terminal and a server terminal. The WEB interaction terminal realizes user data input and parameter and model import; the server terminal calculates the target object position according to the input data and the imported parameters and models, and evaluates the model and solidifies the model according to user feedback; there are multiple WEB interaction terminals, and the multiple WEB interaction terminals interact with the server terminal by establishing a long link. The server terminal actively pushes the calculated data to the WEB interaction terminal according to the parameters and the model, and presents the target object status in real time.
[0031] Compared with the traditional front-end making service requests to the back-end and the back-end service returning data, the server actively pushes the calculated data to the WEB interaction end to ensure the real-time nature of the calculation data and the initiative of the calculation.
[0032] Specifically, the WEB interactive terminal includes a user interaction module, which is used for the user to input the location where the target object falls into the water, set the integration time interval ΔT and the prediction duration; a data source import module, which is used to import the water ripple meteorological data near the water drop point and the geographic information data near the water drop point; a floating coefficient / wind drag coefficient setting module, which is used to import the floating coefficient and / or wind drag coefficient of the target object near the water drop point; a drift algorithm model import module, which is used to import the drift algorithm model; a remote sensing detection model import module, which is used to import the remote sensing detection model; and a service interaction module, which is used to establish a long link with the server through WEBSOCKET to realize data interaction. Among them, the data source import module imports the water ripple meteorological data near the water drop point and the geographic information data near the water drop point based on the remote sensing geographic information.
[0033] The drifting algorithm model import module and the remote sensing detection model import module allow users to import the created drifting algorithm model and remote sensing detection model to ensure the applicability of the model when the original drifting algorithm model import module and remote sensing detection model are not applicable. At the same time, the remote sensing detection model import module has an open interface, which encourages and supports the use of third-party target detection models, so that the search and rescue sharing and construction platform can use the most advanced target detection algorithms to ensure its advancement.
[0034] Specifically, the server includes a geographic information module for loading global geographic GIS information; a water ripple data processing module for loading water ripple meteorological data; a floatation coefficient / wind drag coefficient processing module for loading floatation coefficient and / or wind drag coefficient; a drift algorithm module for loading drift algorithm models; a remote sensing target detection module for loading remote sensing target detection algorithm models; a parallel computing module for performing parallel computing based on the loaded data and models; a service support module for connecting to multiple WEB interactive terminals; a front-end interactive control module for controlling the service module to connect to multiple WEB interactive terminals at the same time; and a user parameter cache module for caching user parameters.
[0035] Preferably, the server also includes a hardware resource management module for managing the hardware resources of the server where the server is located; a performance evaluation and model optimization module for evaluating and optimizing the model based on user feedback on search and rescue results; a data solidification module for solidifying the current optimal model; and a log module for recording work logs.
[0036] In summary, the maritime search and rescue method and system based on remote sensing target search and falling-water trajectory prediction of the embodiments of the present invention determine ocean parameters according to the falling-water location of the target object, predict the location of the target object based on the ocean parameters and algorithm models, and confirm the location of the target object based on remote sensing images near the predicted location, so as to quickly obtain results and ensure that the results are referenceable and accurate; simulate the trajectory deduction after falling into the water through ocean parameters and algorithm models, and ensure effectiveness by adjusting the deduction time granularity and deduction time; realize flexible import of algorithm models through the search and rescue sharing and co-construction platform, ensure the advancement of the algorithm model and the accuracy of the calculation results, and the search and rescue sharing and co-construction platform is based on the Web to ensure ease of use; the search and rescue sharing and co-construction platform is universal and supports global use, supports multiple data sources, and ensures practicality; the search and rescue sharing and co-construction platform has an entry that supports extensive participation in jointly improving data and optimizing algorithm models, so as to ensure the accuracy of the method results.
[0037] Although the present invention has been disclosed as above with preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art may make some modifications and improvements without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be based on the definition of the claims.
Claims
1. A method for marine search and rescue based on remote sensing target search and water fall trajectory prediction, characterized in that: The steps include: S1: Obtain the location where the target falls into the water; select the drift algorithm model according to the access parameters of the location where the target falls into the water; S2: Calculate the target trajectory based on the access parameters and the drift algorithm model to obtain the predicted position of the target; S3: Acquire remote sensing data around the predicted location of the target object, select a possible range based on the remote sensing data, and extract a set of marine targets through target detection; S4: Screen the target set at sea, confirm the location of the target, and start search and rescue; The parameters accessed according to the water-falling point in step S1 include water pattern meteorological data near the water-falling point, floating coefficient and wind drag coefficient of the target object near the water-falling point, and geographic information data near the water-falling point; The step S2 includes: setting an integration time interval ΔT, setting a prediction time, according to the target floating coefficient, according to the time integration interval, using water pattern meteorological data, floating coefficient, wind drag coefficient and geographic information data through a drift algorithm model to obtain the target drift position in the time interval ΔT, repeatedly iterating until the prediction time is set, and obtaining the target object prediction position of the set prediction time, the steps are as follows: S21: The water ripple meteorological data is discrete water ripple data. The discrete water ripple data is superimposed on the vector direction according to the flow velocity, and the data is linked and smoothed using a Gaussian function, and then the data is derived to obtain the direction and velocity value of the flow velocity at each position, that is, the fluid field data; S22: According to the geographic information data, the fluid field data is introduced for fitting, so that the velocity and direction of the ocean current within the data range have longitude and latitude data; S23: According to ΔS=V*ΔT, the predicted position of the target object is obtained by integral calculation:
2. The method for searching and rescuing at sea based on remote sensing target search and prediction of falling-into-water trajectory according to claim 1, characterized in that: The step S3 includes: obtaining a remote sensing image near the predicted position of the target object, slicing the remote sensing image into image tiles, cutting the remote sensing image into multiple small pieces for image recognition, and obtaining a set of marine targets through an image target detection algorithm.
3. A marine search and rescue system based on remote sensing target search and water trajectory prediction, characterized in that: It includes a search and rescue sharing and co-construction platform, which implements the steps of the method as described in any one of claims 1-2, and includes a WEB interaction terminal and a server terminal, and the WEB interaction terminal realizes user data input and parameter and model import; the server terminal calculates the target object position according to the input data and imported parameters and models, and evaluates the model and solidifies the model according to user feedback; there are multiple WEB interaction terminals, and the multiple WEB interaction terminals interact with the server terminal by establishing a long link, and the server terminal actively pushes the calculated data to the WEB interaction terminal according to the parameters and models, and presents the target object status in real time.
4. The marine search and rescue system based on remote sensing target search and water fall trajectory prediction as claimed in claim 3, characterized in that: The WEB interactive terminal includes a user interactive module for the user to input the location where the target object falls into the water, set the integration time interval ΔT and the prediction time length; The data source import module is used to import water ripple meteorological data and geographic information data near the water drop point; the floating coefficient / wind drag coefficient setting module is used to import the floating coefficient and / or wind drag coefficient of the target object near the water drop point; the drift algorithm model import module is used to import the drift algorithm model; the remote sensing detection model import module is used to import the remote sensing detection model; the service interaction module is used to establish a long link with the server through WEBSOCKET to realize data interaction.
5. The marine search and rescue system based on remote sensing target search and water fall trajectory prediction as claimed in claim 4, characterized in that: The data source import module imports water pattern meteorological data near the water drop point and geographic information data near the water drop point based on remote sensing geographic information.
6. The marine search and rescue system based on remote sensing target search and water fall trajectory prediction as claimed in claim 3, characterized in that: The server includes a geographic information module for loading global geographic GIS information; Water ripple data processing module, used to load water ripple meteorological data; floating coefficient / wind drag coefficient processing module, used to load floating coefficient and / or wind drag coefficient; drift algorithm module, used to load drift algorithm model; remote sensing target detection module, used to load remote sensing target detection algorithm model; parallel operation module, used to perform parallel calculation according to loaded data and model; service support module, used to connect with multiple WEB interactive terminals; front-end interactive control module, used to control the service module to connect with multiple WEB interactive terminals at the same time; User parameter cache module, used to cache user parameters.
7. The marine search and rescue system based on remote sensing target search and water fall trajectory prediction as claimed in claim 3, characterized in that: The server also includes a hardware resource management module for managing the hardware resources of the server where the server is located; a performance evaluation and model optimization module for evaluating and optimizing the model based on the feedback of the user's search and rescue results; The data solidification module is used to solidify the current optimal model; the log module is used to record the work log.
