Method and System for Wounded Personnel Location Based on Neural Network Prediction

Abstract

The present invention discloses a method and system for positioning wounded personnel based on neural network prediction. The method includes: acquiring sensing data of multiple wounded personnel in the accident area; inputting the sensing data into a first prediction network to obtain the injury conditions of each wounded personnel; the first prediction network is arranged in the edge device in the accident area; screening out multiple wounded personnel to be positioned whose injury conditions are greater than a preset condition threshold from all the wounded personnel; inputting the sensing data of the wounded personnel to be positioned into a second prediction network arranged in the cloud device to obtain the position information and injury information corresponding to the wounded personnel to be positioned. It can be seen that the present invention can achieve efficient and accurate screening and positioning of wounded personnel, ensure that resources are preferentially allocated to critically wounded personnel in subsequent rescue operations, optimize the rescue scheduling efficiency, and improve the overall rescue effect.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method and system for locating a wounded person based on neural network prediction. Background Art

[0002] In existing accident emergency rescue technologies, the assessment and location of the injured usually rely on manual judgment or traditional monitoring equipment. In traditional methods, rescuers need to check the status of the injured one by one, or rely on fixed monitoring equipment to obtain injury information. However, these methods are difficult to quickly cover the entire accident area in the early stages after the accident, and are prone to omissions or misjudgments. In addition, due to the lack of an efficient data prediction mechanism, the existing injured positioning technology is difficult to quickly screen the location of the injured who need priority treatment, which in turn affects the timeliness and accuracy of the overall rescue. It can be seen that the existing technology has defects that need to be solved urgently. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a method and system for locating the wounded based on neural network prediction, which can realize efficient and accurate screening and locating of the wounded, ensure that resources are allocated preferentially to critically injured persons in subsequent rescue, optimize rescue scheduling efficiency, and improve the overall rescue effect.

[0004] In order to solve the above technical problems, the first aspect of the present invention discloses a method for locating a wounded person based on neural network prediction, the method comprising:

[0005] Acquire sensor data of multiple casualties in the accident area;

[0006] Inputting the sensor data into a first prediction network to obtain the injury status of each of the injured persons; the first prediction network is arranged in an edge device in the accident area;

[0007] Screening out a plurality of injured persons to be located whose injury conditions are greater than a preset threshold value from all the injured persons;

[0008] The sensor data of the injured person to be located is input into a second prediction network arranged in a cloud device to obtain the position information and injury information corresponding to the injured person to be located.

[0009] As an optional embodiment, in the first aspect of the present invention, the sensor data is obtained by a drone device equipped with a sensor module installed in the accident area and / or a sensor module installed on the injured person.

[0010] As an alternative embodiment, in the first aspect of the present invention, the sensing data includes at least one of physiological data, motion data, environmental sensing data, image data, and sound data; the physiological data includes at least one of heart rate, blood oxygen saturation, blood pressure, body temperature, respiratory rate, electrocardiogram data, blood glucose level, and bleeding detection data; the motion data includes at least one of acceleration, angular velocity, displacement information, and limb mobility; and the environmental sensing data includes at least one of temperature, humidity, air quality, dust concentration, and light intensity.

[0011] As an alternative embodiment, in the first aspect of the present invention, the first prediction network is trained using a training data set including a plurality of training casualty sensing data and corresponding injury annotations; the second prediction network is trained using a training data set including a plurality of training casualty sensing data and corresponding location annotations and injury annotations; and the model parameter scale of the second prediction network is larger than that of the first prediction network.

[0012] As an alternative embodiment, in the first aspect of the present invention, the first prediction network and the second prediction network are trained through the following steps:

[0013] Obtain a common training data set including a plurality of training casualty sensing data and corresponding injury annotations;

[0014] Based on the common training data set, train a preset basic prediction network model until convergence to obtain a trained common prediction network model;

[0015] Based on preset distillation target parameters, distill the common prediction network model to obtain the first prediction network;

[0016] Obtain an advanced training data set including a plurality of training casualty sensing data and corresponding location annotations and injury annotations;

[0017] According to the advanced training data set, train the common prediction network model until convergence to obtain the second prediction network.

[0018] As an alternative embodiment, in the first aspect of the present invention, the step of distilling the common prediction network model based on preset distillation target parameters to obtain the first prediction network includes:

[0019] Obtain the device parameters of the edge device in the accident area;

[0020] Based on a preset parameter correspondence analysis algorithm, determine the model carrying capacity parameters corresponding to the device parameters;

[0021] Input the model carrying capacity parameter into the model scale prediction network corresponding to the network architecture of the prediction network base model to obtain the predicted model scale; the network architecture includes at least one of a CNN architecture, an RNN architecture, an LSTM architecture, and a random forest architecture;

[0022] According to the predicted model scale, the current model scale of the common prediction network model, and the preset corresponding relationship between the scale and the distillation parameter, determine the distillation target parameter; the distillation target parameter includes a distillation target function value, a distillation constraint condition function, and a distillation operation duration;

[0023] Distill the common prediction network model according to the distillation target parameter to obtain the first prediction network.

[0024] As an optional implementation manner, in the first aspect of the present invention, the inputting the sensing data of the to-be-located wounded into the second prediction network set in the cloud device to obtain the position information and injury information corresponding to the to-be-located wounded includes:

[0025] For each to-be-located wounded, transmit the sensing data of the to-be-located wounded to the cloud device, and record the acquisition location, acquisition time of the sensing data, and the receiving time of the cloud device;

[0026] Input the sensing data into the second prediction network set in the cloud device to obtain the first predicted position and predicted injury degree corresponding to the to-be-located wounded;

[0027] Input the acquisition time and the receiving time into the trained position prediction network to obtain the second predicted position; the position prediction network is trained through a training data set including a plurality of training transmission time information and corresponding position annotations;

[0028] According to the acquisition location, the first predicted position, and the second predicted position, determine the position information corresponding to the to-be-located wounded.

[0029] As an optional implementation manner, in the first aspect of the present invention, the determining the position information corresponding to the to-be-located wounded according to the acquisition location, the first predicted position, and the second predicted position includes:

[0030] Calculate a position correction weight proportional to the predicted injury degree;

[0031] Calculate the product of the position correction weight and the acquisition location, and the average value between the product, the first predicted position, and the second predicted position to obtain the position information corresponding to the to-be-located wounded.

[0032] In the second aspect of the embodiments of the present invention, a casualty positioning system based on neural network prediction is disclosed. The system includes:

[0033] An acquisition module, configured to acquire sensing data of multiple casualties in the accident area;

[0034] A first prediction module, configured to input the sensing data into a first prediction network to obtain the injury condition of each casualty; the first prediction network is set in the edge device in the accident area;

[0035] A screening module, configured to screen out multiple casualties to be located whose injury condition is greater than a preset condition threshold from all the casualties;

[0036] A second prediction module, configured to input the sensing data of the casualties to be located into a second prediction network set in the cloud device to obtain the position information and injury information corresponding to the casualties to be located.

[0037] As an optional implementation manner, in the second aspect of the present invention, the sensing data is obtained by a drone device equipped with a sensor module set in the accident area and / or a sensor module set on the casualty.

[0038] As an optional implementation manner, in the second aspect of the present invention, the sensing data includes at least one of physiological data, motion data, environmental sensing data, image data, and sound data; the physiological data includes at least one of heart rate, blood oxygen saturation, blood pressure, body temperature, respiratory rate, electrocardiogram data, blood glucose level, and bleeding detection data; the motion data includes at least one of acceleration, angular velocity, displacement information, and limb mobility; the environmental sensing data includes at least one of temperature, humidity, air quality, dust concentration, and light intensity.

[0039] As an optional implementation manner, in the second aspect of the present invention, the first prediction network is trained by a training data set including multiple training casualty sensing data and corresponding injury annotations; the second prediction network is trained by a training data set including multiple training casualty sensing data and corresponding position annotations and injury annotations; the model parameter scale of the second prediction network is larger than that of the first prediction network.

[0040] As an optional implementation manner, in the second aspect of the present invention, the first prediction network and the second prediction network are trained through the following steps:

[0041] Obtain a common training data set including multiple training casualty sensing data and corresponding injury annotations;

[0042] Based on the common training data set, train a preset basic prediction network model until convergence to obtain a trained common prediction network model;

[0043] Based on preset distillation target parameters, distill the common prediction network model to obtain the first prediction network;

[0044] Obtain an advanced training data set including multiple training casualty sensing data and corresponding location annotations and injury condition annotations;

[0045] According to the advanced training data set, train the common prediction network model until convergence to obtain the second prediction network.

[0046] As an optional implementation manner, in the second aspect of the present invention, the distilling the common prediction network model based on preset distillation target parameters to obtain the first prediction network includes:

[0047] Obtain the device parameters of the edge devices in the accident area;

[0048] Based on a preset parameter correspondence analysis algorithm, determine the model carrying capacity parameters corresponding to the device parameters;

[0049] Input the model carrying capacity parameters into a model scale prediction network corresponding to the network architecture of the basic prediction network model, and obtain a predicted model scale; the network architecture includes at least one of a CNN architecture, an RNN architecture, an LSTM architecture, and a random forest architecture;

[0050] According to the predicted model scale, the current model scale of the common prediction network model, and the preset correspondence between the scale and the distillation parameters, determine the distillation target parameters; the distillation target parameters include a distillation target function value, a distillation constraint condition function, and a distillation operation duration;

[0051] According to the distillation target parameters, distill the common prediction network model to obtain the first prediction network.

[0052] As an optional implementation manner, in the second aspect of the present invention, the specific manner in which the second prediction module inputs the sensing data of the to-be-located casualty into the second prediction network set in the cloud device to obtain the location information and injury condition information corresponding to the to-be-located casualty includes:

[0053] For each to-be-located casualty, transmit the sensing data of the to-be-located casualty to the cloud device, and record the acquisition location, acquisition time of the sensing data, and the reception time of the cloud device;

[0054] Input the sensing data into a second prediction network set in the cloud device to obtain a first predicted position and a predicted injury degree corresponding to the casualty to be located;

[0055] Input the acquisition time and the reception time into a trained position prediction network to obtain a second predicted position; the position prediction network is trained by a training data set including a plurality of training transmission time information and corresponding position annotations;

[0056] Determine the position information corresponding to the casualty to be located according to the acquired position, the first predicted position, and the second predicted position.

[0057] As an optional implementation manner, in the second aspect of the present invention, the specific manner in which the second prediction module determines the position information corresponding to the casualty to be located according to the acquired position, the first predicted position, and the second predicted position includes:

[0058] Calculate a position correction weight proportional to the predicted injury degree;

[0059] Calculate the product of the position correction weight and the acquired position, and the average value among the product, the first predicted position, and the second predicted position to obtain the position information corresponding to the casualty to be located.

[0060] The third aspect of the present invention discloses another casualty positioning system based on neural network prediction, and the system includes:

[0061] A memory storing executable program code;

[0062] A processor coupled to the memory;

[0063] The processor calls the executable program code stored in the memory and executes some or all of the steps in the casualty positioning method based on neural network prediction disclosed in the first aspect of the present invention.

[0064] The fourth aspect of the present invention discloses a computer storage medium, and the computer storage medium stores computer instructions, which are used to execute some or all of the steps in the casualty positioning method based on neural network prediction disclosed in the first aspect of the present invention when being called.

[0065] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0066] The present invention can quickly evaluate the injury conditions of each injured person based on the sensing data of multiple injured persons in the accident area through the injury prediction network in the edge device, and screen out the injured persons to be located with serious injuries. Then, the positioning prediction network of the cloud device is used to accurately determine the location information of these injured persons, so as to achieve efficient and accurate screening and positioning of the injured persons, ensure that resources are preferentially allocated to critically injured persons in subsequent rescues, optimize the rescue scheduling efficiency, and improve the overall rescue effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0068] Figure 1 It is a schematic flowchart of a method for positioning injured persons based on neural network prediction disclosed in an embodiment of the present invention.

[0069] Figure 2 It is a schematic structural diagram of a system for positioning injured persons based on neural network prediction disclosed in an embodiment of the present invention.

[0070] Figure 3 It is a schematic structural diagram of another system for positioning injured persons based on neural network prediction disclosed in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0071] In order to enable those skilled in the art to better understand the solution of the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0072] The terms "first", "second", etc. in the specification and claims of the present invention and the above drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, device, product or equipment that includes a series of steps or units is not limited to the listed steps or units, but may optionally further include steps or units not listed, or may optionally further include other steps or units inherent to these processes, methods, products or equipment.

[0073] References to "embodiments" in this specification mean that specific features, structures, or characteristics described in connection with the embodiments can be included in at least one embodiment of the invention. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0074] The present invention discloses a method and system for locating injured persons based on neural network prediction. It can quickly evaluate the injury conditions of each injured person through an injury prediction network in an edge device based on the sensing data of multiple injured persons in an accident area, and screen out the injured persons to be located with serious injuries. Then, the positioning prediction network of the cloud device is used to accurately determine the location information of these injured persons, so as to achieve efficient and accurate screening and positioning of injured persons, ensure that resources are preferentially allocated to critically injured persons in subsequent rescues, optimize the rescue scheduling efficiency, and improve the overall rescue effect. The following will be described in detail respectively.

[0075] Embodiment 1

[0076] Please refer to Figure 1 , Figure 1 which is a schematic flowchart of a method for locating injured persons based on neural network prediction disclosed in an embodiment of the present invention. Among them, Figure 1 the described method for locating injured persons based on neural network prediction can be applied to a data processing system / data processing device / data processing server (wherein, the server includes a local processing server or a cloud processing server). As Figure 1 shown, the method for locating injured persons based on neural network prediction can include the following operations:

[0077] 101. Obtain the sensing data of multiple injured persons in the accident area.

[0078] 102. Input the sensing data into the first prediction network to obtain the injury condition of each injured person.

[0079] Optionally, the first prediction network is set in an edge device in the accident area.

[0080] 103. Screen out multiple injured persons to be located whose injury conditions are greater than a preset condition threshold from all injured persons.

[0081] 104. Input the sensing data of the injured persons to be located into the second prediction network set in the cloud device to obtain the location information and injury information corresponding to the injured persons to be located.

[0082] It can be seen that the above-described embodiments of the invention can, based on the sensing data of multiple wounded in the accident area, quickly evaluate the injury conditions of each wounded through the injury prediction network in the edge device, screen out the wounded to be located with serious injuries, and then use the positioning prediction network of the cloud device to accurately determine the location information of these wounded, so as to achieve efficient and accurate screening and positioning of the wounded, ensure that resources are preferentially allocated to critically wounded in subsequent rescue operations, optimize the rescue scheduling efficiency, and improve the overall rescue effect.

[0083] As an alternative embodiment, in the above steps, the sensing data is obtained by a drone device equipped with a sensor module set in the accident area and / or a sensor module set on the wounded.

[0084] It can be seen that through the above alternative embodiment, the sensing data is collected by the drone device equipped with a sensor module in the accident area or the sensor module worn by the wounded, so as to achieve comprehensive data perception of the accident scene, improve the coverage of injury monitoring and the real-time nature of data acquisition, provide more accurate data support for subsequent injury analysis and rescue decision-making, assist in achieving efficient and accurate screening and positioning of the wounded, ensure that resources are preferentially allocated to critically wounded in subsequent rescue operations, optimize the rescue scheduling efficiency, and improve the overall rescue effect.

[0085] As an alternative embodiment, in the above steps, the sensing data includes at least one of physiological data, motion data, environmental sensing data, image data, and sound data; the physiological data includes at least one of heart rate, blood oxygen saturation, blood pressure, body temperature, respiratory rate, electrocardiogram data, blood glucose level, and bleeding detection data; the motion data includes at least one of acceleration, angular velocity, displacement information, and limb mobility; the environmental sensing data includes at least one of temperature, humidity, air quality, dust concentration, and light intensity.

[0086] It can be seen that through the above alternative embodiment, the content of the sensing data is defined to comprehensively perceive the status of the wounded in the accident area and the surrounding environment, so as to facilitate subsequent efficient and accurate screening and positioning of the wounded, ensure that resources are preferentially allocated to critically wounded in subsequent rescue operations, optimize the rescue scheduling efficiency, and improve the overall rescue effect.

[0087] As an alternative embodiment, in the above steps, the first prediction network is trained by a training data set including a plurality of training wounded sensing data and corresponding injury annotations; the second prediction network is trained by a training data set including a plurality of training wounded sensing data and corresponding location annotations and injury annotations; the model parameter scale of the second prediction network is larger than that of the first prediction network.

[0088] It can be seen that through the above optional embodiments, the training details and model parameter details of the first prediction network and the second prediction network are defined, so that the simpler first prediction network set on the edge device can quickly realize the identification of the injuries of the wounded, while the second prediction network needs to process more complex spatial information and injury analysis, and the scale of its model parameters is larger than that of the first prediction network, thereby improving the comprehensive prediction ability of the state and position information of the wounded, assisting in realizing efficient and accurate screening and positioning of the wounded, ensuring that resources in subsequent rescue are preferentially allocated to critically wounded patients, optimizing the rescue scheduling efficiency, and improving the overall rescue effect.

[0089] As an optional embodiment, in the above steps, the first prediction network and the second prediction network are trained through the following steps:

[0090] Obtain a common training data set including multiple training wounded sensing data and corresponding injury annotations;

[0091] Based on the common training data set, train the preset prediction network basic model until convergence to obtain a trained common prediction network model;

[0092] Based on the preset distillation target parameters, distill the common prediction network model to obtain the first prediction network;

[0093] Obtain an advanced training data set including multiple training wounded sensing data and corresponding position annotations and injury annotations;

[0094] Train the common prediction network model according to the advanced training data set until convergence to obtain the second prediction network.

[0095] It can be seen that through the above optional embodiments, the technical details of training two prediction networks by defining the steps of common training and separate distillation optimization and training optimization are defined, so that the two networks can respectively achieve more reasonable and accurate identification, assist in realizing efficient and accurate screening and positioning of the wounded, ensure that resources in subsequent rescue are preferentially allocated to critically wounded patients, optimize the rescue scheduling efficiency, and improve the overall rescue effect.

[0096] As an optional embodiment, in the above steps, based on the preset distillation target parameters, distill the common prediction network model to obtain the first prediction network, including:

[0097] Obtain the device parameters of the edge device in the accident area;

[0098] Based on the preset parameter correspondence analysis algorithm, determine the model carrying capacity parameters corresponding to the device parameters;

[0099] Input the model carrying capacity parameter into the model scale prediction network corresponding to the network architecture of the prediction network base model to obtain the predicted model scale. Optionally, the network architecture includes at least one of the CNN architecture, RNN architecture, LSTM architecture, and random forest architecture;

[0100] According to the predicted model scale, the current model scale of the common prediction network model, and the preset corresponding relationship between the scale and the distillation parameter, determine the distillation target parameter. Optionally, the distillation target parameter includes the distillation target function value, the distillation constraint condition function, and the distillation operation duration;

[0101] Distill the common prediction network model according to the distillation target parameter to obtain the first prediction network.

[0102] It can be seen that through the above optional embodiments, based on the edge device parameters in the accident area, the model carrying capacity parameter of the device is determined by using the parameter correspondence analysis algorithm, and the adapted predicted model scale is calculated through the model scale prediction network. Thus, according to the scale of the current common prediction network model and the preset scale-distillation parameter correspondence relationship, the distillation target parameter is reasonably determined, and under the constraints of the distillation target function, the distillation constraint condition, and the distillation operation duration, the common prediction network model is distilled and optimized to obtain the first prediction network adapted to the edge device, so as to improve the deployment efficiency and inference performance of the prediction model in a resource-constrained environment, assist in achieving efficient and accurate screening and positioning of the wounded, ensure that resources are preferentially allocated to critically wounded in subsequent rescues, optimize the rescue scheduling efficiency, and improve the overall rescue effect.

[0103] As an optional embodiment, in the above steps, input the sensing data of the wounded to be located into the second prediction network set in the cloud device to obtain the position information and injury information corresponding to the wounded to be located, including:

[0104] For each wounded to be located, transmit the sensing data of the wounded to be located to the cloud device, and record the acquisition location, acquisition time of the sensing data, and the receiving time of the cloud device;

[0105] Input the sensing data into the second prediction network set in the cloud device to obtain the first predicted position and predicted injury degree corresponding to the wounded to be located;

[0106] Input the acquisition time and the receiving time into the trained position prediction network to obtain the second predicted position. Optionally, the position prediction network is trained through a training data set including multiple training transmission time information and corresponding position annotations;

[0107] Determine the position information corresponding to the wounded to be located according to the acquisition location, the first predicted position, and the second predicted position.

[0108] It can be seen that through the above optional embodiments, for the acquisition location, acquisition time, and reception time of the sensing data of each casualty to be located, the second prediction network of the cloud device analyzes the sensing data to obtain the first predicted location and the predicted injury degree. At the same time, by combining the acquisition location, the first predicted location, and the second predicted location, the location information of the casualty to be located is comprehensively determined, thereby improving the accuracy and real-time performance of casualty location, realizing efficient and accurate casualty screening and location, ensuring that resources are preferentially allocated to critically injured casualties in subsequent rescue operations, optimizing the rescue dispatch efficiency, and enhancing the overall rescue effect.

[0109] As an optional embodiment, in the above steps, determining the location information corresponding to the casualty to be located according to the acquisition location, the first predicted location, and the second predicted location includes:

[0110] Calculating a location correction weight proportional to the predicted injury degree;

[0111] Calculating the product of the location correction weight and the acquisition location, and the average value among the product, the first predicted location, and the second predicted location to obtain the location information corresponding to the casualty to be located.

[0112] It can be seen that through the above optional embodiments, a correction weight proportional to the injury condition is defined, so that the weight of the initial sensing position of the casualty with a serious injury is increased. This is because the possibility of the location movement of such casualties is reduced, and the location comprehensively determined based on this is more accurate, realizing efficient and accurate casualty screening and location, ensuring that resources are preferentially allocated to critically injured casualties in subsequent rescue operations, optimizing the rescue dispatch efficiency, and enhancing the overall rescue effect.

[0113] Embodiment 2

[0114] Please refer to Figure 2 , Figure 2 which is a schematic structural diagram of a casualty location system based on neural network prediction disclosed in an embodiment of the present invention. Among them, Figure 2 The described casualty location system based on neural network prediction can be applied to a data processing system / data processing device / data processing server (wherein, the server includes a local processing server or a cloud processing server). As Figure 2 shown, the casualty location system based on neural network prediction can include:

[0115] An acquisition module 201, configured to acquire sensing data of multiple casualties in the accident area.

[0116] A first prediction module 202, configured to input the sensing data into the first prediction network to obtain the injury condition of each casualty.

[0117] Optionally, the first prediction network is set in an edge device in the accident area.

[0118] A screening module 203, configured to screen out multiple to-be-located wounded from all the wounded, where the injury conditions of the to-be-located wounded are greater than a preset condition threshold.

[0119] A second prediction module 204, configured to input the sensing data of the to-be-located wounded into a second prediction network set in a cloud device, so as to obtain the location information and injury information corresponding to the to-be-located wounded.

[0120] It can be seen that the above-mentioned invention embodiments can, based on the sensing data of multiple wounded in the accident area, quickly evaluate the injury conditions of each wounded through an injury prediction network in an edge device, screen out the to-be-located wounded with serious injuries, and then accurately determine the location information of these wounded by using the location prediction network of the cloud device, so as to achieve efficient and accurate screening and positioning of the wounded, ensure that resources are preferentially allocated to critical wounded in subsequent rescue operations, optimize the rescue dispatch efficiency, and improve the overall rescue effect.

[0121] As an alternative embodiment, the sensing data is obtained by a drone device equipped with a sensor module set in the accident area and / or a sensor module set on the wounded.

[0122] It can be seen that through the above alternative embodiment, the sensing data is collected by a drone device equipped with a sensor module in the accident area or a sensor module worn by the wounded, so as to achieve comprehensive data perception of the accident scene, improve the coverage of injury monitoring and the real-time nature of data acquisition, provide more accurate data support for subsequent injury analysis and rescue decision-making, assist in achieving efficient and accurate screening and positioning of the wounded, ensure that resources are preferentially allocated to critical wounded in subsequent rescue operations, optimize the rescue dispatch efficiency, and improve the overall rescue effect.

[0123] As an alternative embodiment, the sensing data includes at least one of physiological data, motion data, environmental sensing data, image data, and sound data; the physiological data includes at least one of heart rate, blood oxygen saturation, blood pressure, body temperature, respiratory rate, electrocardiogram data, blood glucose level, and bleeding detection data; the motion data includes at least one of acceleration, angular velocity, displacement information, and limb mobility; the environmental sensing data includes at least one of temperature, humidity, air quality, dust concentration, and light intensity.

[0124] It can be seen that through the above alternative embodiment, the content of the sensing data is defined to comprehensively perceive the status of the wounded in the accident area and the surrounding environment, so as to facilitate subsequent efficient and accurate screening and positioning of the wounded, ensure that resources are preferentially allocated to critical wounded in subsequent rescue operations, optimize the rescue dispatch efficiency, and improve the overall rescue effect.

[0125] As an optional embodiment, the first prediction network is trained by a training data set including a plurality of training casualty sensing data and corresponding injury annotations; the second prediction network is trained by a training data set including a plurality of training casualty sensing data and corresponding location annotations and injury annotations; the model parameter scale of the second prediction network is larger than that of the first prediction network.

[0126] It can be seen that through the above optional embodiment, the training details and model parameter details of the first prediction network and the second prediction network are defined, so that the simpler first prediction network set on the edge device can quickly realize the identification of the casualty's injury, while the second prediction network needs to process more complex spatial information and injury analysis, and its model parameter scale is larger than that of the first prediction network, thereby improving the comprehensive prediction ability of the casualty's state and location information, assisting in realizing efficient and accurate casualty screening and positioning, ensuring that resources are preferentially allocated to critical casualties in subsequent rescue, optimizing the rescue scheduling efficiency, and improving the overall rescue effect.

[0127] As an optional embodiment, the first prediction network and the second prediction network are trained through the following steps:

[0128] Obtain a common training data set including a plurality of training casualty sensing data and corresponding injury annotations;

[0129] Based on the common training data set, train a preset prediction network basic model until convergence to obtain a trained common prediction network model;

[0130] Based on the preset distillation target parameters, distill the common prediction network model to obtain the first prediction network;

[0131] Obtain an advanced training data set including a plurality of training casualty sensing data and corresponding location annotations and injury annotations;

[0132] According to the advanced training data set, train the common prediction network model until convergence to obtain the second prediction network.

[0133] It can be seen that through the above optional embodiment, the technical details of training the two prediction networks based on common training and separate distillation optimization and training optimization are defined, so that the two networks can respectively achieve more reasonable and accurate identification, assist in realizing efficient and accurate casualty screening and positioning, ensure that resources are preferentially allocated to critical casualties in subsequent rescue, optimize the rescue scheduling efficiency, and improve the overall rescue effect.

[0134] As an optional embodiment, based on the preset distillation target parameters, distilling the common prediction network model to obtain the first prediction network includes:

[0135] Obtain the device parameters of the edge devices in the accident area;

[0136] Based on a preset parameter correspondence analysis algorithm, determine the model carrying capacity parameters corresponding to the device parameters;

[0137] Input the model carrying capacity parameters into the model scale prediction network corresponding to the network architecture of the prediction network basic model to obtain the predicted model scale; Optionally, the network architecture includes at least one of the CNN architecture, RNN architecture, LSTM architecture, and random forest architecture;

[0138] According to the predicted model scale, the current model scale of the common prediction network model, and the preset correspondence between the scale and the distillation parameters, determine the distillation target parameters; Optionally, the distillation target parameters include the distillation target function value, the distillation constraint condition function, and the distillation operation duration;

[0139] According to the distillation target parameters, distill the common prediction network model to obtain the first prediction network.

[0140] It can be seen that through the above optional embodiments, based on the edge device parameters in the accident area, the model carrying capacity parameters of the device are determined using the parameter correspondence analysis algorithm, and the appropriate predicted model scale is calculated through the model scale prediction network. Thus, according to the scale of the current common prediction network model and the preset scale-distillation parameter correspondence, the distillation target parameters are reasonably determined, and under the constraints of the distillation target function, the distillation constraint conditions, and the distillation operation duration, the common prediction network model is distilled and optimized to obtain the first prediction network adapted to the edge device, so as to improve the deployment efficiency and inference performance of the prediction model in a resource-constrained environment, assist in achieving efficient and accurate screening and positioning of the wounded, ensure that resources are preferentially allocated to critically wounded in subsequent rescues, optimize the rescue scheduling efficiency, and improve the overall rescue effect.

[0141] As an optional embodiment, the specific manner in which the second prediction module inputs the sensing data of the wounded to be located into the second prediction network set in the cloud device to obtain the position information and injury information corresponding to the wounded to be located includes:

[0142] For each wounded to be located, transmit the sensing data of the wounded to be located to the cloud device, and record the acquisition location, acquisition time of the sensing data, and the reception time of the cloud device;

[0143] Input the sensing data into the second prediction network set in the cloud device to obtain the first predicted position and predicted injury degree corresponding to the wounded to be located;

[0144] Input the acquisition time and reception time into the trained location prediction network to obtain a second predicted location. Optionally, the location prediction network is trained using a training dataset including multiple pieces of training transmission time information and corresponding location annotations.

[0145] Determine the location information corresponding to the casualty to be located based on the acquired location, the first predicted location, and the second predicted location.

[0146] It can be seen that through the above optional embodiments, for the acquisition location, acquisition time, and reception time of the sensing data of each casualty to be located, the second prediction network of the cloud device analyzes the sensing data to obtain the first predicted location and the predicted injury degree. At the same time, by combining the acquired location, the first predicted location, and the second predicted location, the location information of the casualty to be located is comprehensively determined, thereby improving the accuracy and timeliness of casualty location, realizing efficient and accurate casualty screening and location, ensuring that resources are preferentially allocated to critical casualties in subsequent rescues, optimizing the rescue dispatch efficiency, and enhancing the overall rescue effect.

[0147] As an optional embodiment, the specific manner in which the second prediction module determines the location information corresponding to the casualty to be located based on the acquired location, the first predicted location, and the second predicted location includes:

[0148] Calculate a location correction weight proportional to the predicted injury degree.

[0149] Calculate the product of the location correction weight and the acquired location, and the average value among the product, the first predicted location, and the second predicted location to obtain the location information corresponding to the casualty to be located.

[0150] It can be seen that through the above optional embodiments, a correction weight proportional to the injury condition is defined, so that the weight of the initial sensing position of the casualty with a serious injury is increased because the possibility of the location movement of such casualties is reduced. Based on this, the comprehensively determined location is more accurate, realizing efficient and accurate casualty screening and location, ensuring that resources are preferentially allocated to critical casualties in subsequent rescues, optimizing the rescue dispatch efficiency, and enhancing the overall rescue effect.

[0151] Embodiment III

[0152] Please refer to Figure 3 , Figure 3 which is another casualty location system based on neural network prediction disclosed in the embodiments of the present invention. Figure 3 The described casualty location system based on neural network prediction is applied in a data processing system / data processing device / data processing server (wherein, the server includes a local processing server or a cloud processing server). As Figure 3 shown, the casualty location system based on neural network prediction may include:

[0153] A memory 301 storing executable program code;

[0154] A processor 302 coupled to the memory 301;

[0155] Wherein, the processor 302 calls the executable program code stored in the memory 301 to execute the steps of the method for locating wounded personnel based on neural network prediction described in the first embodiment.

[0156] Embodiment Four

[0157] An embodiment of the present invention discloses a computer-readable storage medium storing a computer program for electronic data exchange, wherein the computer program causes a computer to execute the steps of the method for locating wounded personnel based on neural network prediction described in the first embodiment.

[0158] Embodiment Five

[0159] An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to cause a computer to execute the steps of the method for locating wounded personnel based on neural network prediction described in the first embodiment.

[0160] The above describes specific embodiments of this specification, and other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve the desired result. Additionally, the processes depicted in the drawings do not necessarily have to be performed in the particular order or continuous order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0161] The systems, devices, modules or units illustrated in the above embodiments can be specifically implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer can be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or any combination of these devices.

[0162] For convenience of description, when describing the above devices, various units are described separately according to their functions. Of course, when implementing this specification, the functions of each unit can be implemented in one or more software and / or hardware.

[0163] Those skilled in the art should understand that the embodiments of this specification can be provided as a method, a system, or a computer program product. Therefore, the embodiments of this specification can take the form of an all-hardware embodiment, an all-software embodiment, or an embodiment combining software and hardware aspects. Moreover, the embodiments of this specification can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0164] This specification is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of this specification. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one or more flows and / or blocks Figure 1 one or more blocks.

[0165] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means that implement the functions specified in Figure 1 one or more flows and / or blocks Figure 1 one or more blocks.

[0166] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one or more flows and / or blocks Figure 1 one or more blocks.

[0167] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.

[0168] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM), and / or non-volatile memory such as read-only memory (ROM) or flash memory (flash RAM). The memory is an example of computer-readable media.

[0169] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0170] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0171] This specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. This specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0172] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0173] Finally, it should be noted that what is disclosed by a method and system for positioning the wounded based on neural network prediction disclosed in the embodiments of the present invention is only the preferred embodiments of the present invention, which are only used to illustrate the technical solutions of the present invention, rather than limiting it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for locating wounded personnel based on neural network prediction, characterized in that The method includes: Obtaining sensing data of multiple wounded in the accident area; Inputting the sensing data into a first prediction network to obtain the injury condition of each wounded; the first prediction network is set in the edge device in the accident area; Screening out multiple wounded to be located with injury conditions greater than a preset condition threshold from all the wounded; Inputting the sensing data of the wounded to be located into a second prediction network set in the cloud device to obtain the position information and injury information corresponding to the wounded to be located; the first prediction network is trained through a training data set including multiple training wounded sensing data and corresponding injury annotations; the second prediction network is trained through a training data set including multiple training wounded sensing data and corresponding position annotations and injury annotations; the model parameter scale of the second prediction network is larger than that of the first prediction network, and the first prediction network and the second prediction network are trained through the following steps: Obtaining a common training data set including multiple training wounded sensing data and corresponding injury annotations; Based on the common training data set, training a preset prediction network basic model until convergence to obtain a trained common prediction network model; Obtaining the device parameters of the edge device in the accident area; Based on a preset parameter correspondence analysis algorithm, determining the model carrying capacity parameters corresponding to the device parameters; Inputting the model carrying capacity parameters into a model scale prediction network corresponding to the network architecture of the prediction network basic model to obtain a predicted model scale; the network architecture includes at least one of a CNN architecture, an RNN architecture, an LSTM architecture, and a random forest architecture; According to the predicted model scale, the current model scale of the common prediction network model, and a preset correspondence between the scale and distillation parameters, determining distillation target parameters; the distillation target parameters include a distillation target function value, a distillation constraint function, and a distillation operation duration; According to the distillation target parameters, distilling the common prediction network model to obtain the first prediction network; Obtaining an advanced training data set including multiple training wounded sensing data and corresponding position annotations and injury annotations; Training the common prediction network model according to the advanced training data set until convergence to obtain the second prediction network.

2. The method for positioning the wounded based on neural network prediction according to claim 1, wherein The sensing data is obtained by a drone device equipped with a sensor module set in the accident area and / or a sensor module set on the wounded.

3. The method for positioning the wounded based on neural network prediction according to claim 1, wherein The sensing data includes at least one of physiological data, motion data, environmental sensing data, image data, and sound data; the physiological data includes at least one of heart rate, blood oxygen saturation, blood pressure, body temperature, respiratory rate, electrocardiogram data, blood glucose level, and bleeding detection data; the motion data includes at least one of acceleration, angular velocity, displacement information, and limb mobility; the environmental sensing data includes at least one of temperature, humidity, air quality, dust concentration, and light intensity.

4. The method for locating the wounded based on neural network prediction according to claim 1, wherein, Inputting the sensing data of the casualty to be located into a second prediction network set in a cloud device to obtain the position information and injury condition information corresponding to the casualty to be located, includes: For each casualty to be located, transmitting the sensing data of the casualty to be located to the cloud device, and recording the acquisition position, acquisition time of the sensing data, and the reception time of the cloud device; Inputting the sensing data into a second prediction network set in the cloud device to obtain a first predicted position and a predicted injury degree corresponding to the casualty to be located; Inputting the acquisition time and the reception time into a trained position prediction network to obtain a second predicted position; the position prediction network is trained by a training data set including a plurality of training transmission time information and corresponding position annotations; Determining the position information corresponding to the casualty to be located according to the acquisition position, the first predicted position, and the second predicted position.

5. The method for positioning the wounded based on neural network prediction according to claim 4, characterized in that, The determining the position information corresponding to the casualty to be located according to the acquisition position, the first predicted position, and the second predicted position, includes: Calculating a position correction weight proportional to the predicted injury degree; Calculating the product of the position correction weight and the acquisition position, and the average value between the product, the first predicted position, and the second predicted position to obtain the position information corresponding to the casualty to be located.

6. A wounded personnel positioning system based on neural network prediction, characterized in that, The system includes: An acquisition module, configured to acquire sensing data of multiple casualties in an accident area; A first prediction module, configured to input the sensing data into a first prediction network to obtain the injury condition of each casualty; the first prediction network is set in an edge device in the accident area; A screening module, configured to screen out multiple casualties to be located whose injury condition is greater than a preset condition threshold from all the casualties; A second prediction module, configured to input the sensing data of the casualties to be located into a second prediction network set in a cloud device to obtain the position information and injury condition information corresponding to the casualties to be located; the first prediction network is trained by a training data set including a plurality of training casualty sensing data and corresponding injury annotations; the second prediction network is trained by a training data set including a plurality of training casualty sensing data and corresponding position annotations and injury annotations; the model parameter scale of the second prediction network is larger than that of the first prediction network, and the first prediction network and the second prediction network are trained through the following steps: Obtaining a common training data set including a plurality of training casualty sensing data and corresponding injury annotations; Based on the common training data set, training a preset prediction network basic model until convergence to obtain a trained common prediction network model; Obtaining the device parameters of the edge device in the accident area; Based on a preset parameter correspondence analysis algorithm, determining the model carrying capacity parameters corresponding to the device parameters; Input the model carrying capacity parameter into the model scale prediction network corresponding to the network architecture of the prediction network basic model, and obtain the predicted model scale; the network architecture includes at least one of the CNN architecture, RNN architecture, LSTM architecture, and random forest architecture; Determine the distillation target parameter according to the predicted model scale, the current model scale of the common prediction network model, and the preset corresponding relationship between the scale and the distillation parameter; the distillation target parameter includes the distillation target function value, the distillation constraint condition function, and the distillation operation duration; Perform distillation on the common prediction network model according to the distillation target parameter to obtain the first prediction network; Obtain an advanced training data set including a plurality of training casualty sensing data and corresponding location annotations and injury annotations; Train the common prediction network model according to the advanced training data set until convergence to obtain the second prediction network.

7. A casualty positioning system based on neural network prediction, characterized in that The system includes: A memory storing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory and executes the casualty positioning method based on neural network prediction according to any one of claims 1-5.

Images