Nursing processing method and device, storage medium and electronic equipment

Abstract

The embodiment of the invention discloses a nursing processing method and device, a storage medium and electronic equipment, and the method comprises the steps: monitoring a target scene video of a nursing scene for a nursing object based on an object nursing task which comprises a nursing triggering condition and a nursing triggering action, based on the target scene, using a nursing large model to determine nursing object situation data, and based on the nursing object situation data and the object nursing task, performing task trigger detection to obtain a task trigger detection result, and performing nursing response processing on the nursing object based on the task trigger detection result and the trigger nursing action.

Description

Technical Field

[0001] This specification relates to the field of computer technology, and in particular to a nursing processing method, device, storage medium, and electronic device. Background Art

[0002] In recent years, with the rapid development of computer vision, artificial intelligence, and multimodal data processing technologies, the smart home and home care sectors are undergoing unprecedented transformation. Existing technologies primarily rely on real-time video surveillance and environmental perception, requiring remote monitoring to detect and provide early warnings of object behavior and scene conditions within the home. Therefore, how to more efficiently and accurately extract meaningful information from complex environmental and behavioral data, and implement intelligent security warnings and interventions, has become a key technical challenge urgently needed in the industry. Summary of the Invention

[0003] The embodiments of this specification provide a nursing processing method, device, storage medium, and electronic device, and the technical solutions are as follows:

[0004] In a first aspect, an embodiment of the present specification provides a nursing treatment method, the method comprising:

[0005] Based on the object care task, monitoring the target scene video of the care scene for the care object, the object care task includes a care triggering condition and a triggering care action;

[0006] Determine care object context data using a large care model based on the target scenario, and perform task trigger detection based on the care object context data and the care object task to obtain a task trigger detection result;

[0007] Performing care response processing on the care object based on the task trigger detection result and the triggered care action.

[0008] In a feasible implementation, determining the care object scenario data using the care model based on the target scenario includes:

[0009] Based on the target scene, a large nursing model is used to extract object environment information, object behavior information, and scene potential risk hotspot information for the nursing object;

[0010] Based on the object environment information, object behavior information and scene potential risk hotspot information, care situation data is generated.

[0011] In a feasible implementation, the generating of care situation data based on the object environment information, the object behavior information, and the scene potential risk hotspot information includes:

[0012] Multimodal contextual semantic features that fuse the object environment information, object behavior information, and scene potential risk hotspot information;

[0013] The multimodal situational semantic features are subjected to situational semantic modeling in a preset semantic space to obtain nursing situation data.

[0014] In a feasible implementation manner, the performing task trigger detection based on the care object context data and the care object task to obtain a task trigger detection result includes:

[0015] Mapping the care object context data and the care object task to the same semantic vector space to obtain a care object context semantic vector and a care object task semantic vector;

[0016] Vector matching is performed on the care object context semantic vector and the object care task semantic vector in the semantic vector space to obtain a task trigger detection result.

[0017] In a feasible implementation manner, performing vector matching on the care object context semantic vector and the object care task semantic vector in the semantic vector space to obtain a task trigger detection result includes:

[0018] Calculating the semantic matching degree between the care object context semantic vector and the care object task semantic vector in the semantic vector space;

[0019] If the semantic matching degree is greater than the care risk matching threshold, a first task trigger detection result of the care object abnormality type is generated;

[0020] If the semantic matching degree is less than or equal to the care risk matching threshold, a second task triggering detection result of a normal type of the care object is generated.

[0021] In a feasible implementation manner, performing care response processing on the care object based on the task trigger detection result and the triggered care action includes:

[0022] If the task trigger detection result is an abnormal type of the care object, determining a target response device based on the triggered care action, and calling the target response device to perform the triggered care action on the care object;

[0023] If the task trigger detection result is that the care object is of a normal type, the step of monitoring the care scene for the target scene video of the care object is executed.

[0024] In a feasible embodiment, the method further includes:

[0025] Obtaining a target object associated video for the care object in the target scene video, and predicting a reference risk behavior pattern for the target object based on the care object situational data and the task trigger detection result;

[0026] Performing future care risk prediction processing based on the reference risk behavior pattern and the target object-associated video to obtain future care risk hotspot distribution information and a potential risk time series behavior model;

[0027] Building a potential risk warning reminder based on the future care risk hotspot distribution information and the potential risk time series behavior model;

[0028] Based on the potential risk warning reminder, risk warning monitoring and processing are performed on the care object.

[0029] In a feasible implementation, the constructing of a potential risk warning reminder based on the future care risk hotspot distribution information and the potential risk time series behavior model includes:

[0030] Based on the future care risk hotspot distribution information and the potential risk time series behavior model, potential risk interpretation information of the care object is generated, and based on the potential risk interpretation information of the care object, a risk warning reminder is generated.

[0031] In a second aspect, an embodiment of the present specification provides a nursing processing device, the device comprising:

[0032] A monitoring module, configured to monitor a target scene video of a care object in a care scene based on a care object task, wherein the care object task includes a care triggering condition and a triggering care action;

[0033] a care module, configured to determine care object context data using a care model based on the target scenario, and perform task trigger detection based on the care object context data and the care object care task to obtain a task trigger detection result;

[0034] A processing module is used to perform care response processing on the care object based on the task trigger detection result and the triggered care action.

[0035] In a feasible implementation, determining the care object scenario data using the care model based on the target scenario includes:

[0036] Based on the target scene, a large nursing model is used to extract object environment information, object behavior information, and scene potential risk hotspot information for the nursing object;

[0037] Based on the object environment information, object behavior information and scene potential risk hotspot information, care situation data is generated.

[0038] In a feasible implementation, the generating of care situation data based on the object environment information, the object behavior information, and the scene potential risk hotspot information includes:

[0039] Multimodal contextual semantic features that fuse the object environment information, object behavior information, and scene potential risk hotspot information;

[0040] The multimodal situational semantic features are subjected to situational semantic modeling in a preset semantic space to obtain nursing situation data.

[0041] In a feasible implementation manner, the performing task trigger detection based on the care object context data and the care object task to obtain a task trigger detection result includes:

[0042] Mapping the care object context data and the care object task to the same semantic vector space to obtain a care object context semantic vector and a care object task semantic vector;

[0043] Vector matching is performed on the care object context semantic vector and the object care task semantic vector in the semantic vector space to obtain a task trigger detection result.

[0044] In a feasible implementation manner, performing vector matching on the care object context semantic vector and the object care task semantic vector in the semantic vector space to obtain a task trigger detection result includes:

[0045] Calculating the semantic matching degree between the care object context semantic vector and the care object task semantic vector in the semantic vector space;

[0046] If the semantic matching degree is greater than the care risk matching threshold, a first task trigger detection result of the care object abnormality type is generated;

[0047] If the semantic matching degree is less than or equal to the care risk matching threshold, a second task triggering detection result of a normal type of the care object is generated.

[0048] In a feasible implementation manner, performing care response processing on the care object based on the task trigger detection result and the triggered care action includes:

[0049] If the task trigger detection result is an abnormal type of the care object, determining a target response device based on the triggered care action, and calling the target response device to perform the triggered care action on the care object;

[0050] If the task trigger detection result is that the care object is of a normal type, the step of monitoring the care scene for the target scene video of the care object is executed.

[0051] In a feasible embodiment, the device further includes:

[0052] Obtaining a target object associated video for the care object in the target scene video, and predicting a reference risk behavior pattern for the target object based on the care object situational data and the task trigger detection result;

[0053] Performing future care risk prediction processing based on the reference risk behavior pattern and the target object-associated video to obtain future care risk hotspot distribution information and a potential risk time series behavior model;

[0054] Building a potential risk warning reminder based on the future care risk hotspot distribution information and the potential risk time series behavior model;

[0055] Based on the potential risk warning reminder, risk warning monitoring and processing are performed on the care object.

[0056] In a feasible implementation, the constructing of a potential risk warning reminder based on the future care risk hotspot distribution information and the potential risk time series behavior model includes:

[0057] Based on the future care risk hotspot distribution information and the potential risk time series behavior model, potential risk interpretation information of the care object is generated, and based on the potential risk interpretation information of the care object, a risk warning reminder is generated.

[0058] In a third aspect, an embodiment of this specification provides a computer storage medium, wherein the computer storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded by a processor and executing the above-mentioned method steps.

[0059] In a fourth aspect, an embodiment of this specification provides an electronic device, which may include: a processor and a memory; wherein the memory stores a computer program, and the computer program is suitable for being loaded by the processor and executing the above-mentioned method steps.

[0060] The beneficial effects of the technical solutions provided by some embodiments of this specification include at least:

[0061] In one or more embodiments of the present specification, through real-time video monitoring and user preset task loading, raw data and task basis are provided for subsequent processing, and the video data is deeply analyzed using the large-scale care model to generate situational data of the care object, and match it with the preset trigger conditions, and output the task trigger detection result. Based on the detection results and the preset trigger care action, the care object is intervened and protected in real time to ensure that risks are effectively controlled before the development of the event, forming a closed-loop task triggering and response system, realizing early risk detection and immediate intervention of the care object, significantly improving the accuracy, real-time and personalized effect of the care response, and continuously optimizing the system performance through the feedback mechanism, providing an efficient, intelligent and sustainable solution for family safety care. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the embodiments of this specification or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0063] Figure 1 This is a flowchart of a nursing treatment method provided in an embodiment of this specification;

[0064] Figure 2 This is a schematic diagram of an interface for interactively setting an object care task provided by an embodiment of this specification;

[0065] Figure 3 This is a flow chart of a method of determining the scenario data of a care object by using a large care model provided in an embodiment of this specification;

[0066] Figure 4 This is a flowchart of a task trigger detection provided by an embodiment of this specification;

[0067] Figure 5 This is a schematic diagram of a vector matching process provided by an embodiment of this specification;

[0068] Figure 6 This is a flow chart of risk early warning monitoring provided by the embodiments of this specification;

[0069] Figure 7 This is a structural diagram of a nursing treatment device provided in an embodiment of this specification;

[0070] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this specification;

[0071] Figure 9 This is a schematic diagram of the structure of the operating system and user space provided in the embodiments of this specification;

[0072] Figure 10 yes Figure 9 The architecture diagram of the Android operating system;

[0073] Figure 11 yes Figure 9 Architecture diagram of the IOS operating system. DETAILED DESCRIPTION

[0074] The following will be combined with the drawings in the embodiments of this specification to clearly and completely describe the technical solutions in the embodiments of this specification. Obviously, the embodiments described are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of this specification.

[0075] In the description of this specification, it should be understood that the terms "first", "second", etc. are used for descriptive purposes only and should not be understood as indicating or implying relative importance. In the description of this specification, it should be noted that, unless otherwise expressly specified and limited, "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or devices. For those of ordinary skill in the art, the specific meanings of the above terms in this specification can be understood according to the specific circumstances. In addition, in the description of this specification, unless otherwise specified, "multiple" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the associated objects before and after are in an "or" relationship.

[0076] The present specification is described in detail below with reference to specific embodiments.

[0077] In one embodiment, Figure 1As shown, a monitoring processing method is proposed. This method can be implemented using a computer program and can be run on a monitoring processing device based on the von Neumann architecture. The computer program can be integrated into an application or run as a standalone tool application. The monitoring processing device can be an electronic device, including but not limited to: a smart camera, a personal computer, a tablet computer, a handheld device, an in-vehicle device, a wearable device, a computing device, or other processing device connected to a wireless modem. Terminal devices in different networks can be called different names, such as user equipment, access terminal, subscriber unit, subscriber station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal, wireless communication device, user agent or user device, cellular phone, cordless phone, device in 5G network or future evolution network, etc.

[0078] Specifically, the nursing treatment method includes:

[0079] S102: Monitoring a target scene video of a care scene for a care object based on a care object task, wherein the care object task includes a care triggering condition and a triggering care action;

[0080] Object care tasks: These are user-defined tasks that require the system to perform specific care tasks (such as children, elderly people, pets, etc.) in specific scenarios. These tasks typically consist of two core components:

[0081] Caregiving triggers: describe the circumstances under which the task should be activated, such as "when a child is near the kitchen stove" or "when a pet is on the table";

[0082] Triggering care action: defines the response measures that the system needs to perform when the trigger conditions are met, such as making a phone call, voice broadcasting, or sounding an alarm.

[0083] Target scene video: refers to the video stream captured by the device camera and contains all visual information in the care scene.

[0084] For example, a pre-set object care task is initiated, with the trigger conditions and corresponding actions within the task serving as the basis for subsequent judgment. In the conversational care scenario, cameras installed within the care environment collect video data in real time, ensuring continuous monitoring of the entire care scene. The system initially monitors the video stream to ensure it captures key areas and object behavior relevant to the task.

[0085] For example, suppose the task is set as "Call me when the child approaches the kitchen stove." In this case, the camera continuously monitors the video data of the kitchen area, paying attention to the child's movement trajectory and location distribution in real time, preparing to determine whether to trigger the task.

[0086] In a feasible implementation, the electronic device provides an intelligent agent service to the user, and the user can interact with the intelligent agent service to set an object care task, such as Figure 2 As shown, Figure 2 It is a schematic diagram of the interface for interactively setting up an object care task through an intelligent service. The user initiates a care start request in the intelligent service "Xiao An" dialogue interface → the system parses the request and completes the configuration of the object care task → finally returns a prompt of "Task setting successful", indicating that the object care task has been activated and entered the monitoring state. At this time, the monitoring care scene is the target scene video of the care object.

[0087] S104: Determine care object context data using a large care model based on the target scenario, and perform task trigger detection based on the care object context data and the care object task to obtain a task trigger detection result;

[0088] Large care model: refers to the large care model obtained by adapting the pre-trained basic large language model to the object care scene. It can analyze multimodal video data, extract object environment information, object behavior information and scene potential risk hotspot information to form comprehensive care object situational data, and perform task trigger detection with the object care task to obtain the task trigger detection result; in some embodiments, the care response processing can also be performed on the care object based on the task trigger detection result and the triggered care action.

[0089] Care object situational data: a structured description of the current care scene. Care object situational data may include but is not limited to: environmental information: the spatial layout and object distribution of each object in the scene (such as kitchen stoves, furniture, dangerous equipment, etc.); behavioral information: the real-time actions, movement trajectories and abnormal behaviors of the care object (such as running, approaching dangerous areas); potential risk hotspots: areas or objects in the scene that may induce accidents (such as high-temperature stoves, sharp knives, etc.).

[0090] Task trigger detection: The system matches the extracted situational data with the care trigger conditions set by the user, and determines whether the requirements for triggering the care action are met by comparing the semantic or feature similarity between the two.

[0091] Schematically, the large-scale care model analyzes the target scene video to extract environmental, behavioral, and potential risk information. This information, along with other sources, is then integrated to generate unified contextual data for the care recipient. For example, image recognition technology is used to determine the location of the kitchen stove, while behavior recognition captures the child's movement trajectory. This is then combined with a risk assessment algorithm to determine the risk score for high-temperature areas.

[0092] Furthermore, the caregiving model performs semantic encoding or feature extraction on both the contextual data and the trigger conditions in the task, mapping them into the same feature space for comparison. This comparison calculates the similarity or match between the contextual data and the preset trigger conditions. When the match score exceeds a preset threshold, the trigger condition is considered met. Conversely, when the match score does not exceed the preset threshold, the condition is considered unmet. This generates a task trigger detection result, which typically includes information such as the match score, trigger time, and a description of the associated risk.

[0093] S106: Performing care response processing on the care object based on the task trigger detection result and the triggered care action.

[0094] Triggering nursing action: refers to the system's preset response measures when the task trigger conditions are met, such as phone notification, voice broadcast, alarm or control instructions of other linked devices.

[0095] Care response processing: refers to the system calling the preset trigger care action according to the task trigger detection results, executing actual intervention measures, protecting and guiding the care object, and preventing potential accidents.

[0096] Schematically, based on the task trigger detection result generated in S104, it is determined whether the current scenario has actually met the trigger condition. When it is determined that the current scenario has met the trigger condition, the system calls the corresponding tool or device and performs the response operation corresponding to the trigger care action according to the preset trigger care action. During the action execution process, the system continuously monitors the response effect (such as whether the call is dialed or the voice prompt is effective) to confirm the successful execution of the task.

[0097] Optionally, S106: performing care response processing on the care object based on the task trigger detection result and the triggered care action can be executed by controlling a large care model.

[0098] For example, if the detection results indicate that a child is approaching the kitchen stove and poses a high risk, the system immediately initiates the "Call Me" trigger action based on the pre-set task. The system automatically dials the pre-set phone number through the communication module and plays a voice prompt: "Child approaching the kitchen, please be careful." The system also records the intervention and awaits user confirmation or further response.

[0099] In the embodiments of this specification, through real-time video monitoring and user preset task loading, raw data and task basis are provided for subsequent processing, and the large-scale care model is used to deeply analyze the video data, generate the care object situational data, and match it with the preset trigger conditions, output the task trigger detection result, and then based on the detection results and the preset trigger care action, intervene and protect the care object in real time to ensure that risks are effectively controlled before the development of the event, forming a closed-loop task triggering and response system, realizing early risk detection and immediate intervention of the care object, significantly improving the accuracy, real-time and personalized effect of the care response, and continuously optimizing the system performance through the feedback mechanism, providing an efficient, intelligent and sustainable solution for family safety care.

[0100] Optional, see Figure 3 , Figure 3 This is a flow chart of a method for determining the scenario data of a care recipient using a large care model proposed in this specification. Specifically, the method of determining the scenario data of a care recipient using a large care model based on the target scenario includes:

[0101] S202: extracting object environment information, object behavior information, and scene potential risk hotspot information for the care object based on the target scene using a large care model;

[0102] Object environment information: refers to factors related to the surrounding environment of the care object, such as indoor layout, furniture location, temperature, humidity, lighting conditions, etc.

[0103] Specifically, the large-scale model can be used to extract environmental layout recognition to obtain object environmental information: image segmentation and object detection can be performed on the target scene video to identify the indoor layout, furniture distribution, and the spatial location of fixed facilities (such as sofas, tables and chairs, doors and windows, etc.). Furthermore, object status perception is performed, and by combining sensor or image recognition results, the status information of important objects in the scene (for example, the temperature of a hot water cup, whether electrical appliances are powered on, whether doors and windows are open, etc.) is obtained. At the same time, environmental condition parameters are determined, and information such as ambient temperature, humidity, and illumination is obtained through sensors or visual estimation.

[0104] Object behavior information: describes the posture, movements, movement trajectory, emotional expression, and other information of the care object (such as children, the elderly, and pets).

[0105] Specifically, the caregiver model can be used to identify and track the caregiver based on the model input, applying target tracking and action recognition mechanisms to capture their motion trajectory, posture changes, and behavior patterns (for example, detecting that the child is crawling or moving to a specific area);

[0106] Scene potential risk hotspot information: refers to areas or items in the target scene that may pose safety hazards, such as high-temperature equipment, sharp objects, and areas prone to falling. The danger of this information can be determined through risk assessment algorithms.

[0107] Specifically, the big data model can be used to identify potential risk hotspots in a scene. For example, by detecting the temperature of a hot water cup and the behavior of approaching a dangerous area, it can determine whether there is a risk of burns or collision.

[0108] S204: Generate care situation data based on the object environment information, object behavior information and scene potential risk hotspot information.

[0109] Care scenario data, formed through multimodal fusion, temporal correlation, and semantic modeling of object environment information, object behavior information, and potential risk hotspots in the scenario, comprehensively reflects the overall state of the current care scenario. This care scenario data includes static environment, dynamic behavior, and risk assessment results, providing an intuitive description of the safety status of the care scenario.

[0110] In a feasible implementation, the generation of care situation data based on the object environment information, the object behavior information, and the scene potential risk hotspot information may be performed in the following manner:

[0111] A2: Multimodal contextual semantic features that fuse the object environment information, object behavior information, and scene potential risk hotspot information;

[0112] Multimodal situational semantic features: Object environment information, object behavior information, and scene potential risk hotspot information are mapped into a unified semantic representation space through feature extraction and fusion technology, resulting in a feature vector with high-level semantic expression capabilities. This vector can comprehensively reflect the status and potential risks of the current scene.

[0113] For example, the information in the object environment information is parsed, and the layout, object location and environmental parameters of the static scene corresponding to the object environment information are converted into the object environment feature vector; the object behavior information is identified to capture the movement trajectory, behavior pattern and abnormal movements of the object being cared for, and generate the corresponding object behavior dynamic features; the scene potential risk hotspot information is identified to extract the scene potential risk hotspot features. The object environment feature vector, object behavior dynamic features and scene potential risk hotspot features are normalized and dimensionally aligned, and then a fusion network (such as a Transformer based on an attention mechanism, a fusion convolutional neural network or a fusion layer) is used to deeply fuse the features of each modality to form a unified multimodal situational semantic feature.

[0114] Example scenario:

[0115] In the family living room, the camera surveillance found:

[0116] Environmental information: There is a coffee table in the living room, its location is clear, and the indoor temperature is 28°C;

[0117] Behavior information: A child is detected crawling and moving towards the coffee table;

[0118] Risk hotspot information: There is a cup of hot water on the coffee table, its temperature is about 60℃, and it is assessed as a high-risk hotspot.

[0119] Execute step A2: The system first preprocesses the scene image, using object detection and semantic segmentation to extract the position and status of the coffee table, child, and hot water cup, generating environmental and risk features, respectively. Simultaneously, the motion recognition and target tracking modules obtain the child's motion trajectory and behavioral characteristics. Next, these environmental, behavioral, and risk features are normalized and integrated into a unified contextual semantic feature vector using a multimodal fusion network. This vector comprehensively expresses the information that "the child is rapidly approaching the coffee table, and there is hot water on the coffee table."

[0120] The fused feature vector not only retains the key information of each modality, but also strengthens the overall semantic expression through cross-modal association, providing an accurate data basis for subsequent task matching.

[0121] A4: Performing situational semantic modeling on the multimodal situational semantic features in a preset semantic space to obtain nursing situation data.

[0122] Preset semantic space: This is a high-level semantic representation space derived from prior knowledge, domain rules, and large-scale corpus training. In this space, different text, image, or fusion features are mapped into semantically meaningful vectors, whose distribution reflects the semantic similarity and hierarchical relationships between concepts.

[0123] Contextual semantic modeling: This involves mapping multimodal contextual semantic features into a pre-defined semantic space and extracting high-level contextual semantic representations through dimensionality reduction, clustering, classification, or regression. This process automatically summarizes and summarizes the current scenario's risk status, behavioral trends, and environmental characteristics, forming a structured contextual description.

[0124] Care scenario data: This is structured data, including a comprehensive description of environmental conditions, object behaviors, and risk hotspots, along with high-level semantic information mapped from a pre-defined semantic space. This data can be expressed as a numerical risk score or accompanied by a natural language description, such as a high-risk burn hazard in the current scenario.

[0125] Schematically, the resulting multimodal contextual semantic features are semantically embedded. Using a pre-set semantic space, the fused features are mapped into a high-level semantic vector representation, preserving the correlation and semantic structure between the different modal information. The mapped high-dimensional semantic vectors are then subjected to dimensionality reduction (e.g., using PCA or t-SNE) to more intuitively visualize the data distribution. The data is then clustered and grouped to extract high-level semantic labels for similar contexts. Based on pre-set caregiving rules and an expert knowledge base, the caregiving processing model automatically assigns semantic labels to different clustering results, such as "low-risk scenario," "high-risk area," and "abnormal behavior warning." In addition to static labels, the caregiving processing model also performs contextual reasoning based on temporal information and behavioral trends to generate a comprehensive description of the current situation, such as "the child is rapidly approaching high-temperature equipment, posing a serious risk of burns." Finally, the results of this mapping and modeling are organized into structured caregiving scenario data.

[0126] In the embodiments of this specification, the system generates a unified contextual semantic feature vector by extracting, normalizing, and multimodally fusing information about the subject's environment, behavior, and potential risk hotspots. The system then maps these multimodal contextual semantic features to a pre-defined semantic space and performs contextual semantic modeling to generate structured care scenario data. This data, in the form of high-level semantic vectors and natural language descriptions, comprehensively reflects the current care scenario's environment, behavior, and risk status, providing a precise data foundation for subsequent task triggering and intervention responses.

[0127] Optional, see Figure 4 , Figure 4 This is a flowchart of a task trigger detection process. Specifically, the task trigger detection based on the care object context data and the care object care task is performed to obtain the task trigger detection result. The following method can be referred to:

[0128] S302: Mapping the care object context data and the care object task to the same semantic vector space to obtain a care object context semantic vector and a care object task semantic vector;

[0129] Care object context semantic vector: The vector representation obtained by mapping the care object context data into the semantic vector space captures the high-level semantic features of the environment, behavior and risks in the current scenario.

[0130] Object care task semantic vector: The vector representation obtained by mapping the user's preset care task description into the same semantic vector space, which contains the semantic information of the task triggering conditions and expected response measures.

[0131] Schematically, the text description of the care object situation data and the description of the object care task are converted into vectors of fixed dimension using a semantic embedding mechanism (such as a semantic embedding network). The obtained vectors are respectively called "care object situation semantic vector" and "object care task semantic vector".

[0132] S304: performing vector matching on the care object context semantic vector and the object care task semantic vector in the semantic vector space to obtain a task trigger detection result.

[0133] Task trigger detection result: It is a judgment result based on vector matching, indicating whether the current care situation meets the trigger conditions of the user's preset task, and may include information such as matching score, trigger mark and recommended response measures.

[0134] Schematically, in the same semantic vector space, vector matching methods such as cosine similarity or Euclidean distance are used to calculate the semantic matching degree between the semantic vector of the care object situation and the semantic vector of the care object task. According to a preset threshold (which can be set by historical data and expert experience), it is judged whether the semantic matching degree meets the requirements for triggering the task. If the semantic matching degree exceeds the threshold, it is considered that the task trigger condition has been met, and the task trigger detection result is output; otherwise, it is not triggered. The task trigger detection result is then generated, and the task trigger detection result usually includes a matching score, a trigger flag (such as trigger / not trigger) and possible recommended response measures;

[0135] In one possible implementation, see Figure 5 , Figure 5 This is a flow chart of vector matching, specifically performing vector matching on the care object context semantic vector and the object care task semantic vector in the semantic vector space to obtain the task trigger detection result, which can be referred to the following method:

[0136] S402: Calculating the semantic matching degree between the care object context semantic vector and the care object task semantic vector in the semantic vector space;

[0137] Semantic matching degree: refers to the value obtained by calculating the similarity (such as cosine similarity) between the semantic vector of the care object context and the semantic vector of the care object task. The higher the value, the closer the semantics of the two are. The matching degree reflects whether the current scene meets the preset task triggering conditions.

[0138] Care Risk Matching Threshold: This is a preset numerical threshold used to distinguish between abnormal and normal types of care recipients. When the semantic match exceeds this threshold, it indicates that the current scenario is risky and requires triggering an abnormal task; otherwise, it is considered a normal scenario.

[0139] S406: If the semantic matching degree is greater than the care risk matching threshold, generating a first task trigger detection result of the care object abnormality type;

[0140] S408: If the semantic matching degree is less than or equal to the care risk matching threshold, a second task triggering detection result of a normal type of the care object is generated.

[0141] Task trigger detection results: Based on the semantic matching degree and care risk matching threshold, the trigger detection results generated by the system are divided into:

[0142] The first task triggers the detection result (abnormal type): when the matching degree is greater than the care risk matching threshold, it indicates that the care recipient is in an abnormal or high-risk state;

[0143] The second task triggers the detection result (normal type): when the matching degree is less than or equal to the care risk matching threshold, it indicates that the care object is in a normal state and does not require immediate intervention.

[0144] In this specification, through the processing flow of S402, S406 and S408, the system matches the care object situation semantic vector and the object care task semantic vector in the same semantic vector space, calculates the semantic matching degree, and judges whether the current scene is in an abnormal state based on the preset care risk matching threshold, thereby generating corresponding task trigger detection results, providing accurate basis for subsequent intervention response.

[0145] Optional, see Figure 6 , Figure 6 This is a flow chart of risk early warning monitoring. When executing the nursing treatment method of one or more embodiments, the following methods may also be referred to:

[0146] S502: Obtain a target object associated video for the care object in the target scene video, and predict a reference risk behavior pattern for the target object based on the care object situation data and the task trigger detection result;

[0147] Target object-related video: refers to extracting sub-videos related to specific care objects from the target scene video, usually focusing on the care object and its surrounding key objects, such as toys, furniture, equipment, etc.

[0148] Reference risk behavior patterns: This refers to the analysis of the target object's movement, interactive behavior, and historical data in the video to predict the target object's possible future risk behavior characteristic patterns, such as abnormal proximity to dangerous objects, sudden running, or repeated high-risk behaviors.

[0149] Schematically, the large nursing processing model uses the existing nursing object situational data and task trigger detection results to perform time series association processing, that is, to integrate these two parts of data in chronological order to form a time series data sequence containing historical behavior, environmental status and risk trigger signals. Then, based on the time series data sequence, key time series features are extracted, such as the position, speed, acceleration, behavior change rate of the target object, and the risk signals (such as high matching degree, abnormal mark, etc.) indicated in the task trigger detection results; based on the extracted key time series features, a time series prediction network is used to build a behavior prediction model to predict the behavior state changes and risk index of the target object in the future period of time, and obtain the predicted future behavior trend and predicted risk score. Then, the predicted future behavior trend and predicted risk score are combined to form a comprehensive future behavior prediction description. Based on the preset pattern generation rules, the comprehensive future behavior prediction description is mapped to a specific risk behavior pattern description, such as "In the next 5 seconds, the target object has a high probability of approaching the high-temperature equipment area at a higher speed, and the risk level reaches a high level."

[0150] Example: Assume that in a home scenario, the system obtains the following information through video surveillance:

[0151] Caregiver situational data: Records show that "the child was moving from the living room to the kitchen area at a speed of approximately 0.8 m / s. There was a cup of hot water in the surrounding environment. The current risk matching score was 0.85."

[0152] Task trigger detection result: Indicates that the current scene is abnormal and there is a potential risk of burns.

[0153] The execution process of the large model is as follows:

[0154] The large-scale care processing model integrates data: The system integrates the situational data and test results within the past 10 seconds into continuous time series data, such as recording the changes in the child's position, speed, and test results.

[0155] The large care processing model performs time series feature extraction: extracting children's movement trajectory, acceleration and risk signals (such as the changing trend of risk matching scores) from the integrated data.

[0156] The large care processing model performs behavior prediction: the extracted time series features are input into the trained LSTM model to predict the child's movement trend in the next 5 seconds. The model predicts that the child's speed may increase to about 0.9 meters per second, and the distance between the child and the hot water cup will decrease to 0.4 meters in the next 3 seconds. The risk score may rise to 0.92.

[0157] The large-scale care processing model constructs a model: Based on the prediction results and preset rules, the system constructs a reference risk behavior model: "It is predicted that the risk of a child rapidly approaching a hot area and touching a hot water cup within the next 5 seconds is extremely high, and the risk score is on an upward trend."

[0158] S504: performing future care risk prediction based on the reference risk behavior pattern and the target object-associated video to obtain future care risk hotspot distribution information and a potential risk time series behavior model;

[0159] Future care risk hotspot distribution information: It predicts the distribution of risk levels that may occur in various areas of the scenario within a period of time in the future, including risk probability, risk intensity, and the spatial location of high-risk areas.

[0160] Potential risk temporal behavior model: A temporal model that describes the behavioral evolution of a target object over a future time period. It reflects the trend of dynamic changes in the target object's behavior and risk status over time, and is used to predict when and where the risk will reach a trigger level.

[0161] Schematically, the large care processing model uses the reference risk behavior pattern generated in the previous step and the target object associated video extracted from the target scene video, and uses the reference risk behavior pattern as a priori risk signal, and aligns it with the spatiotemporal dynamic data in the target object associated video to form a time series data set containing historical and current states. The risk time series signal predicted in the reference risk behavior pattern (for example, the rising trend of the risk score) is fused into the time series data set as an additional feature; the time series prediction network is then used to predict the target object's behavior trend in the future based on the integrated time series data set to obtain the target object's future motion trajectory and risk score distribution, and the risk score is mapped to the spatial scene corresponding to the target object's future motion trajectory to obtain future care risk hotspot distribution information, that is, the future care risk hotspot distribution information characterizes the risk situation of each area in the current home space scene. At the same time, the predicted future behavior trend of the target object and the change of the risk score over time are constructed into a potential risk time series behavior model, which describes the dynamic evolution of the risk state.

[0162] S506: Constructing a potential risk warning reminder based on the future care risk hotspot distribution information and the potential risk time series behavior model;

[0163] Potential Risk Warning Alert: Automatically generated warning information is used to provide early warning of high-risk events that may occur in the future and provide recommended autonomous intervention measures. Warning alerts typically include the risk area, predicted risk level, time point, and autonomous response measures.

[0164] Schematically, the control care processing large model aligns the future care risk hotspot distribution information and the potential risk time series behavior model according to the time and space dimensions to form a comprehensive risk prediction input, and performs weighted fusion of the hotspot distribution information (for example, the risk score of each area) and the risk indicators predicted in the time series behavior model to generate a comprehensive risk indicator that changes with time and space. At the same time, based on the comprehensive risk indicator, an early warning level and autonomous intervention suggestions are generated according to the preset early warning strategy. Based on the early warning level and autonomous intervention suggestions, an early warning reminder message is generated according to the reminder content template;

[0165] Optional warning reminder information includes but is not limited to the following:

[0166] Risk Areas: Indicates which areas are about to become high-risk areas.

[0167] Risk Level: Gives a predicted risk score or grade.

[0168] Forecast time: describes the time period during which the risk may reach a critical point.

[0169] Autonomous intervention suggestions: Provide specific response measures, such as "play preset reminder audio", "dial emergency number", etc.

[0170] In a feasible implementation, the following method may be used to construct a potential risk warning reminder based on the future care risk hotspot distribution information and the potential risk time series behavior model:

[0171] Based on the future care risk hotspot distribution information and the potential risk time series behavior model, potential risk interpretation information of the care object is generated, and based on the potential risk interpretation information of the care object, a risk warning reminder is generated.

[0172] Interpretation information of potential risks for the care recipient: refers to the information output after a comprehensive interpretation of future risk hotspot distribution information and potential risk time-series behavior models. This interpretation converts the original numerical prediction into an easy-to-understand risk description (such as "high-risk area", "risk will rise to a high level in the next 3 seconds", etc.).

[0173] The risk warning reminder here can be understood as a warning message generated based on the potential risk interpretation information of the care recipient, which is used to notify caregivers of current or future high-risk conditions and provide corresponding intervention suggestions, such as making a phone call, voice alarm or other emergency measures.

[0174] Indicatively, the data of future care risk hotspot distribution information (such as risk scores and risk probabilities for each region) and potential risk time series behavior models (describing the target object's behavior and risk trends over time) are aligned and integrated; ensuring the consistency of the two parts of data in time and space dimensions, providing a complete basis for subsequent interpretation. Using the preset risk mapping rules or expert knowledge base, the integrated quantitative risk indicators are converted into qualitative risk descriptions - potential risk interpretation information for the care object. For example, if the risk score of a certain area exceeds the set threshold, it is interpreted as a "high-risk area", and combined with the time series model, it is pointed out that "the risk will increase significantly in the next 3 seconds";

[0175] Furthermore, the care processing model uses the risk level and predicted time revealed in the interpretation information and uses pre-set warning strategies (for example, if the risk is "high" and reaches the warning level within 3 seconds, an immediate warning is triggered) to determine whether the warning trigger conditions have been met and select appropriate intervention measures. Based on the risk interpretation information, a detailed potential risk warning is constructed, including risk description, predicted time, risk area, and recommended autonomous intervention measures.

[0176] S508: Perform risk warning monitoring on the care object based on the potential risk warning reminder.

[0177] Risk warning monitoring and processing: refers to real-time monitoring of the status and behavior of the care recipient after identifying a potential risk warning reminder, verifying the accuracy of the warning information, and triggering subsequent autonomous intervention measures based on real-time data, as well as status recording and feedback adjustments.

[0178] In this manual, based on previously generated potential risk warnings, real-time monitoring and data collection are performed on the care recipient. By comparing real-time data with the warning content, the presence of risk conditions is verified, triggering appropriate intervention measures. The entire monitoring process is also recorded to dynamically adjust the prediction model and warning strategy. This process ensures that warnings are not just informational reminders, but are truly implemented in real-time monitoring and intervention, providing a closed-loop risk management mechanism for the care system.

[0179] The following will be combined Figure 7 , the nursing treatment device provided in the embodiment of this specification is introduced in detail. It should be noted that, Figure 7 The nursing treatment device shown is used to execute this instruction Figures 1 to 6 For the convenience of explanation, only the part related to the embodiment of this specification is shown. For the specific technical details not disclosed, please refer to this specification. Figures 1 to 6 The embodiment shown.

[0180] See Figure 7, which shows a schematic diagram of the structure of the nursing processing device according to an embodiment of the present specification. The nursing processing device 1 can be implemented as all or part of the user terminal through software, hardware, or a combination of both. According to some embodiments, the nursing processing device 1 includes a monitoring module 11, a nursing module 12, and a processing module 13, which are specifically used to:

[0181] A monitoring module 11 is configured to monitor a target scene video of a nursing scene for a nursing object based on a nursing task, wherein the nursing task includes a nursing trigger condition and a triggering nursing action;

[0182] A care module 12 is configured to determine care object context data using a care model based on the target scenario, and perform task trigger detection based on the care object context data and the care object care task to obtain a task trigger detection result;

[0183] The processing module 13 is configured to perform a care response process on the care object based on the task trigger detection result and the triggered care action.

[0184] In a feasible implementation, determining the care object scenario data using the care model based on the target scenario includes:

[0185] Based on the target scene, a large nursing model is used to extract object environment information, object behavior information, and scene potential risk hotspot information for the nursing object;

[0186] Based on the object environment information, object behavior information and scene potential risk hotspot information, care situation data is generated.

[0187] In a feasible implementation, the generating of care situation data based on the object environment information, the object behavior information, and the scene potential risk hotspot information includes:

[0188] Multimodal contextual semantic features that fuse the object environment information, object behavior information, and scene potential risk hotspot information;

[0189] The multimodal situational semantic features are subjected to situational semantic modeling in a preset semantic space to obtain nursing situation data.

[0190] In a feasible implementation manner, the performing task trigger detection based on the care object context data and the care object task to obtain a task trigger detection result includes:

[0191] Mapping the care object context data and the care object task to the same semantic vector space to obtain a care object context semantic vector and a care object task semantic vector;

[0192] Vector matching is performed on the care object context semantic vector and the object care task semantic vector in the semantic vector space to obtain a task trigger detection result.

[0193] In a feasible implementation manner, performing vector matching on the care object context semantic vector and the object care task semantic vector in the semantic vector space to obtain a task trigger detection result includes:

[0194] Calculating the semantic matching degree between the care object context semantic vector and the care object task semantic vector in the semantic vector space;

[0195] If the semantic matching degree is greater than the care risk matching threshold, a first task trigger detection result of the care object abnormality type is generated;

[0196] If the semantic matching degree is less than or equal to the care risk matching threshold, a second task triggering detection result of a normal type of the care object is generated.

[0197] In a feasible implementation manner, performing care response processing on the care object based on the task trigger detection result and the triggered care action includes:

[0198] If the task trigger detection result is an abnormal type of the care object, determining a target response device based on the triggered care action, and calling the target response device to perform the triggered care action on the care object;

[0199] If the task trigger detection result is that the care object is of a normal type, the step of monitoring the care scene for the target scene video of the care object is executed.

[0200] In a feasible embodiment, the device further includes:

[0201] Obtaining a target object associated video for the care object in the target scene video, and predicting a reference risk behavior pattern for the target object based on the care object situational data and the task trigger detection result;

[0202] Performing future care risk prediction processing based on the reference risk behavior pattern and the target object-associated video to obtain future care risk hotspot distribution information and a potential risk time series behavior model;

[0203] Building a potential risk warning reminder based on the future care risk hotspot distribution information and the potential risk time series behavior model;

[0204] Based on the potential risk warning reminder, risk warning monitoring and processing are performed on the care object.

[0205] In a feasible implementation, the constructing of a potential risk warning reminder based on the future care risk hotspot distribution information and the potential risk time series behavior model includes:

[0206] Based on the future care risk hotspot distribution information and the potential risk time series behavior model, potential risk interpretation information of the care object is generated, and based on the potential risk interpretation information of the care object, a risk warning reminder is generated.

[0207] It should be noted that the aforementioned embodiments of the nursing care device, when executing the nursing care method, only illustrate the division of the aforementioned functional modules. In actual applications, the aforementioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the nursing care device and the nursing care method embodiments provided in the aforementioned embodiments are based on the same concept. The implementation process is detailed in the method embodiments and will not be repeated here.

[0208] The serial numbers of the embodiments in this specification are for description only and do not represent the advantages or disadvantages of the embodiments.

[0209] The embodiment of this specification also provides a computer storage medium, which can store multiple instructions, and the instructions are suitable for being loaded and executed by a processor as described above. Figures 1 to 6 The detailed implementation process of the nursing treatment method in the embodiment shown can be found in Figures 1 to 6 The detailed description of the illustrated embodiment will not be repeated here.

[0210] This specification also provides a computer program product, which stores at least one instruction, and the at least one instruction is loaded and executed by the processor as described above. Figures 1 to 6 The detailed implementation process of the nursing treatment method in the embodiment shown can be found in Figures 1 to 6 The detailed description of the illustrated embodiment will not be repeated here.

[0211] Please refer to Figure 8 , which shows a block diagram of the structure of an electronic device provided by an exemplary embodiment of this specification. The electronic device described in this specification may include one or more of the following components: a processor 110, a memory 120, an input device 130, an output device 140, and a bus 150. The processor 110, the memory 120, the input device 130, and the output device 140 may be connected via the bus 150.

[0212] The processor 110 may include one or more processing cores. The processor 110 utilizes various interfaces and circuits to connect various components within the electronic device. It executes instructions, programs, code sets, or instruction sets stored in the memory 120, as well as accesses data stored in the memory 120, to perform various functions of the electronic device and process data. Optionally, the processor 110 may be implemented using at least one of the following hardware forms: a digital signal processing (DSP), a field-programmable gate array (FPGA), or a programmable logic array (PLA). The processor 110 may integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU primarily handles the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing display content; and the modem handles wireless communications. It is understood that the modem may not be integrated into the processor 110 and may be implemented separately via a communications chip.

[0213] The memory 120 may include a random access memory (RAM) or a read-only memory (ROM). Optionally, the memory 120 includes a non-transitory computer-readable storage medium. The memory 120 may be used to store instructions, programs, codes, code sets, or instruction sets. The memory 120 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the following various method embodiments, etc. The operating system may be an Android system, including a system deeply developed based on the Android system, an IOS system developed by Apple, including a system deeply developed based on the IOS system or other systems. The data storage area may also store data created by the electronic device during use, such as a phone book, audio and video data, chat record data, etc.

[0214] See also Figure 9As shown, the memory 120 can be divided into operating system space and user space. The operating system runs in the operating system space, and native and third-party applications run in the user space. In order to ensure that different third-party applications can achieve better operating results, the operating system allocates corresponding system resources to different third-party applications. However, the requirements for system resources in different application scenarios in the same third-party application are also different. For example, in the local resource loading scenario, the third-party application has higher requirements for disk reading speed; in the animation rendering scenario, the third-party application has higher requirements for GPU performance. The operating system and the third-party application are independent of each other, and the operating system often cannot perceive the current application scenario of the third-party application in a timely manner, resulting in the operating system being unable to perform targeted system resource adaptation according to the specific application scenario of the third-party application.

[0215] In order for the operating system to distinguish the specific application scenarios of third-party applications, it is necessary to open up data communication between third-party applications and the operating system so that the operating system can obtain the current scenario information of third-party applications at any time, and then perform targeted system resource adaptation based on the current scenario.

[0216] Taking the Android operating system as an example, the programs and data stored in the memory 120 are as follows: Figure 10As shown, the memory 120 may store a Linux kernel layer 320, a system runtime library layer 340, an application framework layer 360, and an application layer 380. The Linux kernel layer 320, the system runtime library layer 340, and the application framework layer 360 belong to the operating system space, and the application layer 380 belongs to the user space. The Linux kernel layer 320 provides underlying drivers for various hardware components of electronic devices, such as display drivers, audio drivers, camera drivers, Bluetooth drivers, Wi-Fi drivers, power management, etc. The system runtime library layer 340 provides major feature support for the Android system through some C / C++ libraries. For example, the SQLite library provides database support, the OpenGL / ES library provides 3D drawing support, and the Webkit library provides browser kernel support. The system runtime library layer 340 also provides the Android runtime library (Android runtime), which mainly provides some core libraries that allow developers to write Android applications using the Java language. The application framework layer 360 provides various APIs that may be used when building applications. Developers can also use these APIs to build their own applications, such as activity management, window management, view management, notification management, content provider management, package management, call management, resource management, and location management. The application layer 380 runs at least one application. These applications can be native applications that come with the operating system, such as contacts, SMS, clock, and camera applications, or third-party applications developed by third-party developers, such as games, instant messaging programs, and photo enhancement programs.

[0217] Taking the operating system as the IOS system as an example, the programs and data stored in the memory 120 are as follows: Figure 9As shown, the IOS system includes: a core operating system layer 420 (Core OS layer), a core service layer 440 (Core Services layer), a media layer 460 (Media layer), and a touchable layer 480 (Cocoa Touch Layer). The core operating system layer 420 includes the operating system kernel, drivers, and underlying program frameworks. These underlying program frameworks provide functions closer to the hardware for use by the program framework located in the core service layer 440. The core service layer 440 provides system services and / or program frameworks required by applications, such as the foundation framework, account framework, advertising framework, data storage framework, network connection framework, geographic location framework, motion framework, etc. The media layer 460 provides applications with audio-visual interfaces, such as graphics and image-related interfaces, audio technology-related interfaces, video technology-related interfaces, and wireless playback (AirPlay) interfaces for audio and video transmission technologies. The touchable layer 480 provides various commonly used interface-related frameworks for application development. The touchable layer 480 is responsible for user touch interaction operations on electronic devices. For example, local notification service, remote push service, advertising framework, game tool framework, message user interface (UI) framework, user interface UIKit framework, map framework, etc.

[0218] exist Figure 11 Among the frameworks shown, those relevant to most applications include, but are not limited to, the Foundation framework in the core services layer 440 and the UIKit framework in the touchable layer 480. The Foundation framework provides many basic object classes and data types, offering fundamental system services for all applications and having nothing to do with the UI. The classes provided by the UIKit framework are the foundational UI class library for creating touch-based user interfaces. iOS applications can use the UIKit framework to provide their UIs, providing the application infrastructure for building user interfaces, drawing, handling user interaction events, responding to gestures, and so on.

[0219] Among them, the method and principle of implementing data communication between third-party applications and the operating system in the IOS system can be referred to the Android system, and this manual will not go into details here.

[0220] Among them, the input device 130 is used to receive input instructions or data, and the input device 130 includes but is not limited to a keyboard, a mouse, a camera, a microphone or a touch device. The output device 140 is used to output instructions or data, and the output device 140 includes but is not limited to a display device and a speaker. In one example, the input device 130 and the output device 140 can be combined, and the input device 130 and the output device 140 are touch screen displays, which are used to receive touch operations on or near the user using any suitable objects such as fingers and touch pens, and to display the user interface of each application. The touch screen display is usually provided on the front panel of the electronic device. The touch screen display can be designed as a full screen, a curved screen or a special-shaped screen. The touch screen display can also be designed as a combination of a full screen and a curved screen, or a combination of a special-shaped screen and a curved screen, which is not limited in the embodiments of this specification.

[0221] In addition, those skilled in the art will understand that the structures of the electronic devices shown in the above figures do not limit the electronic devices. The electronic devices may include more or fewer components than shown, or may combine certain components, or arrange the components differently. For example, the electronic devices may also include radio frequency circuits, input units, sensors, audio circuits, wireless fidelity (WiFi) modules, power supplies, Bluetooth modules, and other components, which are not described in detail here.

[0222] In the embodiments of this specification, the execution entity of each step can be the electronic device described above. Optionally, the execution entity of each step is the operating system of the electronic device. The operating system can be Android, iOS, or other operating systems, and this embodiment of this specification does not limit this.

[0223] The electronic device of the embodiment of this specification may further be equipped with a display device, and the display device may be any device capable of realizing a display function, such as a cathode ray tube display (CR), a light-emitting diode display (LED), an electronic ink screen, a liquid crystal display (LCD), a plasma display panel (PDP), etc. The user may use the display device on the electronic device to view displayed text, images, videos and other information. The electronic device may be a smart phone, a tablet computer, a gaming device, an AR (Augmented Reality) device, a car, a data storage device, an audio playback device, a video playback device, a notebook, a desktop computing device, a wearable device such as an electronic watch, electronic glasses, an electronic helmet, an electronic bracelet, an electronic necklace, electronic clothing and the like.

[0224] exist Figure 8 In the electronic device shown, which may be a terminal, the processor 110 may be configured to call an application program stored in the memory 120 and specifically execute the nursing processing method of one or more embodiments of this specification.

[0225] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing related hardware through a computer program. The program can be stored in a computer-readable storage medium, and when executed, the program can include the processes in the above-described method embodiments. The storage medium can be a magnetic disk, an optical disk, a read-only memory, or a random access memory.

[0226] The above disclosure is only a preferred embodiment of this specification, and certainly cannot be used to limit the scope of rights of this specification. Therefore, equivalent changes made according to the claims of this specification are still within the scope covered by this specification.

Claims

1. A nursing treatment method, characterized in that: The method comprises: Based on the object care task, monitoring the target scene video of the care scene for the care object, the object care task includes a care triggering condition and a triggering care action; Determine care object context data using a large care model based on the target scenario, and perform task trigger detection based on the care object context data and the care object task to obtain a task trigger detection result; Performing care response processing on the care object based on the task trigger detection result and the triggered care action.

2. The method according to claim 1, wherein determining the care object scenario data using the care model based on the target scenario comprises: Based on the target scene, a large nursing model is used to extract object environment information, object behavior information, and scene potential risk hotspot information for the nursing object; Based on the object environment information, object behavior information and scene potential risk hotspot information, care situation data is generated.

3. The method according to claim 2, wherein generating care situation data based on the subject's environmental information, subject's behavioral information, and scene potential risk hotspot information comprises: Multimodal contextual semantic features that fuse the object environment information, object behavior information, and scene potential risk hotspot information; The multimodal situational semantic features are subjected to situational semantic modeling in a preset semantic space to obtain nursing situation data.

4. The method according to claim 1, wherein the step of performing task trigger detection based on the care object context data and the care object task to obtain a task trigger detection result comprises: Mapping the care object context data and the care object task to the same semantic vector space to obtain a care object context semantic vector and a care object task semantic vector; Vector matching is performed on the care object context semantic vector and the object care task semantic vector in the semantic vector space to obtain a task trigger detection result.

5. The method according to claim 4, wherein the performing vector matching on the care object context semantic vector and the care object task semantic vector in the semantic vector space to obtain a task trigger detection result comprises: Calculating the semantic matching degree between the care object context semantic vector and the care object task semantic vector in the semantic vector space; If the semantic matching degree is greater than the care risk matching threshold, a first task trigger detection result of the care object abnormality type is generated; If the semantic matching degree is less than or equal to the care risk matching threshold, a second task triggering detection result of a normal type of the care object is generated.

6. The method according to claim 1, wherein performing a care response process on the care object based on the task trigger detection result and the triggered care action comprises: If the task trigger detection result is an abnormal type of the care object, determining a target response device based on the triggered care action, and calling the target response device to perform the triggered care action on the care object; If the task trigger detection result is that the care object is of a normal type, the step of monitoring the care scene for the target scene video of the care object is executed.

7. The method according to claim 6, further comprising: Obtaining a target object associated video for the care object in the target scene video, and predicting a reference risk behavior pattern for the target object based on the care object situational data and the task trigger detection result; Performing future care risk prediction processing based on the reference risk behavior pattern and the target object-associated video to obtain future care risk hotspot distribution information and a potential risk time series behavior model; Building a potential risk warning reminder based on the future care risk hotspot distribution information and the potential risk time series behavior model; Based on the potential risk warning reminder, risk warning monitoring and processing are performed on the care object.

8. The method according to claim 7, wherein the constructing of a potential risk warning reminder based on the future care risk hotspot distribution information and the potential risk time series behavior model comprises: Based on the future care risk hotspot distribution information and the potential risk time series behavior model, potential risk interpretation information of the care object is generated, and based on the potential risk interpretation information of the care object, a risk warning reminder is generated.

9. A nursing treatment device, characterized in that: The device comprises: A monitoring module, configured to monitor a target scene video of a care scene for a care object based on a care object task, wherein the care object task includes a care triggering condition and a triggering care action; a care module, configured to determine care object context data using a care model based on the target scenario, and perform task trigger detection based on the care object context data and the care object care task to obtain a task trigger detection result; A processing module is used to perform care response processing on the care object based on the task trigger detection result and the triggered care action.

10. An electronic device, characterized in that: include: A processor and a memory; wherein the memory stores a computer program, and the computer program is suitable for being loaded by the processor and executing the method steps according to any one of claims 1 to 8.

Images