Remote monitoring method and child watch

Abstract

The invention provides a remote monitoring method and a child watch, and relates to the technical field of monitoring, and the method comprises the steps: extracting motion features and sound features of a mobile terminal from various sensor data, and generating a current track; based on the motion features and the current track, identifying a behavior pattern of a user corresponding to the mobile terminal through a preset behavior pattern library to obtain a current behavior pattern; inputting the motion features and the sound features into a pre-trained prediction model, and judging whether a moving object is approaching the mobile terminal or not by using the prediction model; when it is judged that the moving object is approaching the mobile terminal, a preset distance threshold value is dynamically adjusted based on the motion feature and the sound feature at the current moment and the current track; whether the relative distance between the moving object and the mobile terminal is smaller than the dynamically adjusted distance threshold value or not is judged, and when the relative distance between the moving object and the mobile terminal is smaller than the dynamically adjusted distance threshold value, alarm information is generated and pushed to an associated terminal pre-paired with the mobile terminal, and the remote monitoring method is more intelligent and personalized.

Description

Technical Field

[0001] The present invention relates to the field of monitoring technology, and in particular to a remote monitoring method and a children's watch. Background Art

[0002] As society pays more and more attention to children's safety, technologies and products for children's safety monitoring are constantly emerging. Among them, portable terminals such as children's watches, as devices worn by children every day, not only have basic functions such as calling and positioning, but also gradually integrate remote monitoring technology to achieve more comprehensive and detailed safety monitoring of children.

[0003] Traditional child watch monitoring methods mainly rely on GPS positioning technology. By obtaining children's location information in real time, parents can keep track of their children's whereabouts at any time. However, this method is often severely affected in positioning accuracy in complex environments (such as indoors, high-rise buildings, etc.), and cannot provide parents with accurate and reliable child location information.

[0004] Therefore, it is necessary to provide a remote monitoring method and a children's watch to solve the above technical problems. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides a remote monitoring method and a children's watch, which provide a more intelligent and personalized remote monitoring method to meet parents' higher demands for child safety monitoring.

[0006] The present invention provides a remote monitoring method, which comprises the following steps:

[0007] Extracting motion features and sound features of the mobile terminal from a variety of sensor data, and generating a current trajectory based on the motion features, wherein the various sensor data are collected by a sensor component installed on the mobile terminal;

[0008] Based on the motion feature and the current trajectory, identifying the behavior pattern of the user corresponding to the mobile terminal through a preset behavior pattern library to obtain a current behavior pattern;

[0009] Inputting the motion feature and the sound feature into a pre-trained prediction model, and using the prediction model to determine whether a moving object is approaching the mobile terminal;

[0010] When it is determined that the moving object is approaching the mobile terminal, dynamically adjusting a preset distance threshold based on the motion characteristics, sound characteristics and current trajectory at the current moment, wherein the distance threshold represents a safe distance that should be maintained between the mobile terminal and the moving object in the current behavior mode;

[0011] It is determined whether the relative distance between the mobile object and the mobile terminal is less than the dynamically adjusted distance threshold, and if so, an alarm message is generated and pushed to an associated terminal pre-paired with the mobile terminal.

[0012] Preferably, the multiple sensor data include accelerometer data, gyroscope data, GPS data and sound data, wherein the accelerometer data, the gyroscope data and the GPS data are used to extract the motion features, and the sound data is used to extract the sound features, wherein the motion features include accelerometer data features, gyroscope data features and GPS data features corresponding to the accelerometer data, the gyroscope data and the GPS data, respectively.

[0013] Preferably, the step of generating the current trajectory is:

[0014] Based on the extracted accelerometer data features and gyroscope data features, combined with the position change information and timestamp data in the GPS data features, a map matching algorithm is used to generate the current trajectory of the mobile terminal.

[0015] Preferably, based on the motion feature and the current trajectory, identifying the behavior pattern of the user corresponding to the mobile terminal through a preset behavior pattern library to obtain the current behavior pattern includes:

[0016] Extracting a plurality of behavior pattern-related feature vectors from the accelerometer data features, the gyroscope data features, and the GPS data features, wherein the feature vectors include acceleration, angular velocity, speed, direction, and speed change rate;

[0017] Matching the feature vector with a plurality of behavior patterns in a preset behavior pattern library, wherein the behavior pattern library contains a plurality of known behavior patterns and their corresponding feature vectors;

[0018] The similarity between the feature vector and the feature vector of each known behavior pattern is calculated, and the known behavior pattern with the highest similarity is selected as the current behavior pattern.

[0019] Preferably, the step of inputting the motion feature and the sound feature into a pre-trained prediction model, and using the prediction model to determine whether a moving object is approaching the mobile terminal, comprises:

[0020] Fusing the extracted motion features and the extracted sound features to form a fused feature vector;

[0021] Inputting the fused feature vector into a pre-trained prediction model, wherein the prediction model is a neural network model that is trained to recognize and analyze the fused feature vector to determine whether a moving object is approaching the mobile terminal;

[0022] Output the judgment result. If the judgment result is that a moving object is approaching, the subsequent distance threshold dynamic adjustment step is triggered.

[0023] Preferably, the training process of the prediction model includes:

[0024] Collect annotated datasets containing motion features, sound features, and corresponding moving object proximity information;

[0025] Using the data set to train a neural network model, and adjusting model parameters during the training process to minimize the prediction error rate;

[0026] The performance of the neural network model is evaluated through the cross-validation method, and finally the trained model is saved as a pre-trained prediction model.

[0027] Preferably, when it is determined that the moving object is approaching the mobile terminal, dynamically adjusting the preset distance threshold based on the motion characteristics, sound characteristics and current trajectory at the current moment includes:

[0028] The preset basic distance threshold is adjusted to obtain a preliminary distance threshold, wherein the adjustment adopts the following formula:

[0029]

[0030] Among them, T p is the initial distance threshold, T0 is the basic distance threshold, and F v is the speed adjustment factor, F d is the direction adjustment factor, F n is the noise adjustment factor;

[0031] The behavior pattern adjustment factor is calculated, and the preliminary distance threshold is adjusted based on the behavior pattern adjustment factor to obtain a final dynamic threshold as the adjusted distance threshold, wherein the adjustment of the preliminary distance threshold adopts the following formula:

[0032]

[0033] Among them, T f is the final dynamic threshold, F b is the behavior mode adjustment factor, v is the speed of the mobile terminal at the current moment, and v max is the maximum speed of the mobile terminal in the current trajectory.

[0034] Preferably, in the adjustment of the basic distance threshold, the speed adjustment factor is calculated as follows:

[0035]

[0036] Among them, F v is the speed adjustment factor, v is the speed of the mobile terminal at the current moment, Δv is the speed change rate in the current trajectory, and α is the influence coefficient of the preset speed change rate;

[0037] The calculation formula of the direction adjustment factor is:

[0038]

[0039] Among them, F d is the direction adjustment factor, d is the direction of the mobile terminal at the current moment, Δd is the direction change rate in the current trajectory, and β is the influence coefficient of the preset direction change rate;

[0040] The calculation formula of the noise adjustment factor is:

[0041]

[0042] Among them, F n is the noise adjustment factor, n is the ambient noise level of the mobile terminal at the current moment, Δn is the noise change rate in the current trajectory, and γ is the influence coefficient of the preset noise change rate.

[0043] Preferably, the calculation formula of the behavior mode adjustment factor is:

[0044] F b =K b +δ×Δb

[0045] Among them, F b is the behavior pattern adjustment factor, Δb is the behavior pattern change rate, and is expressed as the ratio of the number of times the user's behavior pattern changes in the current trajectory to the total time, δ is the influence coefficient of the preset behavior pattern change rate, and K b Adjust coefficients for behavioral patterns.

[0046] The present invention also provides a children's watch, comprising a sensor component and a processor.

[0047] The sensor assembly is used for various sensor data;

[0048] The processor is used to extract motion features and sound features of the mobile terminal from multiple sensor data, and generate a current trajectory based on the motion features;

[0049] The processor is used to identify the behavior pattern of the user corresponding to the mobile terminal through a preset behavior pattern library based on the motion characteristics and the current trajectory to obtain the current behavior pattern;

[0050] The processor is used to input the motion feature and the sound feature into a pre-trained prediction model, and use the prediction model to determine whether a moving object is approaching the mobile terminal;

[0051] The processor is configured to dynamically adjust a preset distance threshold based on the motion characteristics, sound characteristics, and current trajectory at a current moment when it is determined that the mobile object is approaching the mobile terminal, wherein the distance threshold represents a safe distance that should be maintained between the mobile terminal and the mobile object in a current behavior mode;

[0052] The processor is used to determine whether the relative distance between the moving object and the mobile terminal is less than the dynamically adjusted distance threshold, and if so, generate an alarm message and push it to an associated terminal pre-paired with the mobile terminal.

[0053] Compared with the related art, the remote monitoring method and children's watch provided by the present invention have the following beneficial effects:

[0054] The present invention achieves accurate perception of user behavior patterns by extracting the motion and sound features of mobile terminals from a variety of sensor data and generating current trajectory and behavior pattern recognition results based on these features. At the same time, the method also uses a pre-trained prediction model to intelligently judge the approach of moving objects and dynamically adjusts the distance threshold according to the judgment results, thereby achieving personalized monitoring of safety distances under different behavior patterns. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A flow chart of the remote monitoring method provided by the present invention;

[0056] Figure 2 This is a module structure diagram of the children's watch provided by the present invention. DETAILED DESCRIPTION

[0057] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only the parts related to the present invention, rather than all structures, are shown in the accompanying drawings. In addition, the embodiments of the present invention and the features in the embodiments may be combined with each other without conflict.

[0058] It should also be noted that, for ease of description, only the part relevant to the present invention but not all content is shown in the accompanying drawings. It should be mentioned before discussing exemplary embodiments in more detail that some exemplary embodiments are described as processing or methods depicted as flow charts. Although the flow chart describes each operation (or step) as sequential processing, many operations therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of each operation can be rearranged. When its operation is completed, the processing can be terminated, but it can also have additional steps not included in the accompanying drawings. The processing can correspond to methods, functions, procedures, subroutines, subprograms, etc.

[0059] Embodiment 1

[0060] The present invention provides a remote monitoring method, referring to Figure 1 As shown, the monitoring method comprises the following steps:

[0061] S1: extracting motion features and sound features of a mobile terminal from a variety of sensor data, and generating a current trajectory based on the motion features, wherein the various sensor data are collected by a sensor component installed on the mobile terminal.

[0062] It should be noted that the multiple sensor data include accelerometer data, gyroscope data, GPS data and sound data, wherein the accelerometer data, the gyroscope data and the GPS data are used to extract the motion features, and the sound data is used to extract the sound features, wherein the motion features include accelerometer data features, gyroscope data features and GPS data features corresponding to the accelerometer data, the gyroscope data and the GPS data respectively.

[0063] In this embodiment, the accelerometer data is derived from an accelerometer installed on the mobile terminal, which can measure the linear acceleration on three axes (X, Y, Z). The accelerometer data is mainly used to extract the linear motion characteristics of the mobile terminal. The magnitude of the acceleration can be obtained by calculating the square root of the sum of the squares of the acceleration values ​​on the three axes. The direction of the acceleration is represented by a unit vector obtained by dividing the acceleration value on each axis by the magnitude of the acceleration. The rate of change of acceleration is obtained by calculating the difference in acceleration between two consecutive time points divided by the time interval.

[0064] Gyroscope data comes from the gyroscope installed on the mobile terminal, which can measure the angular velocity on three axes (X, Y, Z). Gyroscope data is mainly used to extract the rotational motion characteristics of the mobile terminal. The angular velocity magnitude can be obtained by calculating the square root of the sum of the squares of the angular velocity values ​​on the three axes. The angular velocity direction is represented by a unit vector obtained by dividing the angular velocity value on each axis by the magnitude of the angular velocity. The angular velocity change rate is obtained by calculating the angular velocity difference between two consecutive time points and dividing it by the time interval.

[0065] GPS data comes from the GPS receiver installed on the mobile terminal, which can provide accurate geographic location information. GPS data is mainly used to extract the location and speed characteristics of the mobile terminal. The location information includes the longitude, latitude and altitude of the mobile terminal. The speed information includes the instantaneous speed of the mobile terminal. The direction information includes the heading angle of the mobile terminal. The position change rate is obtained by calculating the difference between the longitude, latitude and altitude of two consecutive time points and dividing it by the time interval.

[0066] The sound data comes from the microphone installed on the mobile terminal, which can collect the audio signals of the surrounding environment. The sound data is mainly used to extract the acoustic environment characteristics around the mobile terminal. The ambient noise level can be obtained by calculating the average sound pressure level of the sound signal.

[0067] From the above detailed description, it can be seen that the accelerometer data provides the linear motion characteristics of the mobile terminal, including the magnitude, direction and rate of change of acceleration. The gyroscope data provides the rotational motion characteristics of the mobile terminal, including the magnitude, direction and rate of change of angular velocity. The GPS data provides the location and velocity characteristics of the mobile terminal, including longitude, latitude, altitude, speed and direction. The sound data provides the acoustic environment characteristics around the mobile terminal, that is, the ambient noise level. These characteristics together constitute the motion characteristics and sound characteristics of the mobile terminal, providing rich data support for subsequent behavior pattern recognition and environmental monitoring.

[0068] At the same time, in step S1, the steps of generating the current trajectory are:

[0069] Based on the extracted accelerometer data features and gyroscope data features, combined with the position change information and timestamp data in the GPS data features, a map matching algorithm is used to generate the current trajectory of the mobile terminal.

[0070] In this embodiment, the collected accelerometer data, gyroscope data and GPS data are filtered to remove noise and abnormal values ​​to ensure the accuracy and consistency of the data.

[0071] The accelerometer data features and the gyroscope data features are combined to form a comprehensive motion feature vector, which includes information such as acceleration, angular velocity, speed and direction.

[0072] Using the location and timestamp information in the GPS data, the integrated motion feature vector is matched with the map data. The specific method includes the following steps:

[0073] Dead Reckoning: Estimates the path of a mobile terminal between two GPS points based on acceleration and angular velocity data.

[0074] Map matching: Match the estimated path to the road network on the map to correct errors caused by GPS signal drift.

[0075] Trajectory Optimization: Combines speed and direction information to further optimize the trajectory to make it smoother and more accurate.

[0076] S2: Based on the motion feature and the current trajectory, the behavior pattern of the user corresponding to the mobile terminal is identified through a preset behavior pattern library to obtain a current behavior pattern.

[0077] Specifically, step S2 includes the following steps:

[0078] S21: extracting a plurality of behavior pattern-related feature vectors from the accelerometer data features, the gyroscope data features and the GPS data features, wherein the feature vectors include acceleration, angular velocity, speed, direction and speed change rate.

[0079] In this embodiment, a plurality of behavior pattern-related feature vectors are extracted from the accelerometer data features, the gyroscope data features, and the GPS data features.

[0080] Among them, accelerometer data provides linear motion characteristics, gyroscope data provides rotational motion characteristics, and GPS data provides position and velocity characteristics. These feature vectors include acceleration, angular velocity, speed, direction, and velocity change rate.

[0081] Exemplarily, the acceleration can be obtained by calculating the square root of the sum of the squares of the acceleration values ​​on the three axes, the angular velocity can be obtained by calculating the square root of the sum of the squares of the angular velocity values ​​on the three axes, the speed can be directly obtained from the GPS data, the direction can be calculated from the gyroscope data, and the speed change rate is obtained by calculating the speed difference between two consecutive time points and dividing it by the time interval.

[0082] S22: Matching the feature vector with a plurality of behavior patterns in a preset behavior pattern library, wherein the behavior pattern library includes a plurality of known behavior patterns and their corresponding feature vectors.

[0083] In this embodiment, the extracted feature vector is matched with multiple behavior patterns in a preset behavior pattern library. The behavior pattern library contains multiple known behavior patterns and their corresponding feature vectors. These known behavior patterns include but are not limited to common behavior patterns such as walking, running, cycling, and standing still. The matching process involves comparing the extracted feature vector with the feature vector of each known behavior pattern in the library to find the most similar known behavior pattern. Specifically, matching can be performed by calculating the Euclidean distance or cosine similarity between the feature vectors.

[0084] S23: Calculate the similarity between the feature vector and the feature vector of each known behavior pattern, and select the known behavior pattern with the highest similarity as the current behavior pattern.

[0085] In this embodiment, the similarity between the feature vector and the feature vector of each known behavior pattern is calculated, and the known behavior pattern with the highest similarity is selected as the current behavior pattern.

[0086] Exemplarily, the Euclidean distance between the extracted feature vector and the feature vector of each known behavior pattern in the library is calculated, and the known behavior pattern with the smallest distance is selected as the current behavior pattern. This ensures that the identified behavior pattern is the one that best matches the actual behavior of the current user.

[0087] S3: Inputting the motion feature and the sound feature into a pre-trained prediction model, and using the prediction model to determine whether a moving object is approaching the mobile terminal.

[0088] Specifically, step S3 includes the following steps:

[0089] S31: Fusing the extracted motion features and the sound features to form a fused feature vector.

[0090] In this embodiment, it is first necessary to fuse the motion features and sound features extracted from the various sensor data to form a comprehensive feature vector. Specifically:

[0091] Motion features: including acceleration, angular velocity, speed, direction, and rate of change of velocity, etc. These features are extracted from accelerometer data, gyroscope data, and GPS data.

[0092] Sound characteristics: including the ambient noise level, these characteristics are extracted from the sound data collected by the microphone.

[0093] The specific steps of fusion processing include:

[0094] Data preprocessing: Normalize the extracted motion features and sound features to ensure that each feature is at the same order of magnitude to facilitate subsequent model input.

[0095] Feature concatenation: concatenate the normalized motion features and sound features into a high-dimensional feature vector. For example, if the dimension of the motion feature vector is five (acceleration, angular velocity, speed, direction, rate of change of speed), and the dimension of the sound feature vector is one (ambient noise level), then the dimension of the final fused feature vector is six.

[0096] S32: Inputting the fused feature vector into a pre-trained prediction model, wherein the prediction model is a neural network model that has been trained to identify and analyze the fused feature vector to determine whether a moving object is approaching the mobile terminal.

[0097] In this embodiment, first ensure that the neural network model used has been fully trained and can effectively identify and analyze the fused feature vector. The model is trained on a large number of labeled data sets, which contain the motion characteristics and sound characteristics of mobile terminals in various environments, as well as label information on whether there is a moving object approaching.

[0098] The fused feature vector formed in step S31 is passed as input to the pre-trained neural network model. The fused feature vector contains comprehensive information extracted from multiple sensors such as accelerometer, gyroscope, GPS and microphone, and can fully reflect the current state and surrounding environment of the mobile terminal.

[0099] After receiving the fused feature vector, the neural network model will gradually process the input data through a series of forward propagation operations. This process includes:

[0100] Feature transformation: Perform nonlinear transformation on input features through multiple layers of neurons to capture complex relationships in the data.

[0101] Feature extraction: The hidden layers inside the model are responsible for extracting higher-level abstract features from the input features, which help distinguish different moving object approach situations.

[0102] Decision function: Finally, the model generates a value between 0 and 1 through the sigmoid activation function of the output layer, which represents the probability of a moving object approaching. Usually a threshold (such as 0.5) is set, and if it exceeds the threshold, it is considered that a moving object is approaching.

[0103] Result output: The result generated by the model will be interpreted as a judgment of whether there is a moving object approaching. If the judgment result is "a moving object is approaching", the subsequent distance threshold dynamic adjustment step will be triggered.

[0104] Specifically, the training process of the prediction model includes:

[0105] Collect annotated datasets containing motion features, sound features, and corresponding moving object proximity situations.

[0106] At the beginning of model training, we first need to collect a labeled dataset containing rich information. This dataset should cover motion features, sound features, and corresponding labels of moving objects approaching in various environments.

[0107] Motion characteristics: collected through accelerometers, gyroscopes, and GPS, including acceleration, angular velocity, speed, direction, and rate of change of velocity.

[0108] Sound characteristics: The microphone collects sound characteristics such as ambient noise level.

[0109] Annotation: Each data point needs a clear label indicating whether there is a moving object approaching.

[0110] After the data is collected, it needs to be preprocessed, including data cleaning, missing value processing, outlier detection, etc., to ensure the accuracy and consistency of the data. In addition, for motion features and sound features, normalization processing is also required to make them at the same order of magnitude to facilitate subsequent model input.

[0111] The neural network model is trained using the data set, and the model parameters are adjusted during the training process to minimize the prediction error rate.

[0112] In this embodiment, a neural network model is constructed to identify and analyze the fused feature vector. This model is a multi-layer perceptron model, including:

[0113] Input layer: receives the fused feature vector as input.

[0114] Hidden layer: contains multiple layers of neurons, which are used to perform nonlinear transformation and feature extraction on input features.

[0115] Output layer: Use the sigmoid activation function to generate a value between 0 and 1, indicating the probability of the moving object approaching.

[0116] After the model is built, the model is trained using the collected labeled dataset. During the training process, the model parameters are continuously adjusted to minimize the prediction error rate, which is usually achieved through the back propagation algorithm.

[0117] The performance of the neural network model is evaluated through the cross-validation method, and finally the trained model is saved as a pre-trained prediction model.

[0118] In this embodiment, in order to evaluate the performance of the model, a cross-validation method is used. Specifically, the data set is divided into k equal parts (such as k = 5), and each time k-1 parts are used as training sets, and the remaining parts are used as test sets. In this way, k evaluation results can be obtained, and the average value is taken as the final performance evaluation indicator.

[0119] During the cross-validation process, the performance of the model is evaluated using indicators including but not limited to accuracy, recall, F1 score, etc. These indicators can help understand the performance of the model in different situations and provide a basis for adjusting model parameters.

[0120] After sufficient training and evaluation, the trained model is saved as a pre-trained prediction model. This model can be loaded and used in actual applications to determine whether a moving object is approaching the mobile terminal.

[0121] S33: Output the judgment result. If the judgment result is that a moving object is approaching, the subsequent distance threshold dynamic adjustment step is triggered.

[0122] In this embodiment, the prediction model outputs a binary classification result, indicating whether a moving object is approaching the mobile terminal. If the judgment result is that a moving object is approaching, step S4 will be entered to dynamically adjust the preset distance threshold based on the current motion characteristics, sound characteristics and current trajectory to ensure the safety of the user.

[0123] S4: When it is determined that the moving object is approaching the mobile terminal, the preset distance threshold is dynamically adjusted based on the motion characteristics, sound characteristics and current trajectory at the current moment, wherein the distance threshold represents the safe distance that should be maintained between the mobile terminal and the moving object under the current behavior mode.

[0124] Specifically, step S4 includes the following steps:

[0125] S41: Adjust the preset basic distance threshold to obtain a preliminary distance threshold, wherein the adjustment adopts the following formula:

[0126]

[0127] Among them, T p is the initial distance threshold, T0 is the basic distance threshold, and F v is the speed adjustment factor, F d is the direction adjustment factor, F n is the noise adjustment factor.

[0128] In this embodiment, the basic distance threshold is a preset initial value, which indicates the safe distance that should be maintained between the mobile terminal and the moving object under normal circumstances. This value is usually determined based on experience and experimental data. In order to adapt to different environments and behavior patterns, speed adjustment factors, direction adjustment factors and noise adjustment factors are introduced, which are calculated based on the current motion characteristics and sound characteristics.

[0129] The speed adjustment factor reflects the impact of the current speed and speed change on the safety distance. The faster the speed of the wearer of the children's watch, the greater the required safety distance, because when moving at high speed, the reaction time and stopping distance will increase, and a larger buffer space is needed to avoid collision. In addition, the speed change rate reflects the changing trend of the speed. If the speed change rate is large, it means that the speed changes frequently or drastically, which increases uncertainty and requires a larger safety distance to deal with possible emergencies.

[0130] The direction adjustment factor reflects the impact of the current direction and direction changes on the safety distance. The direction of the wearer of the children's watch will affect its movement trajectory. For example, the safety distance required when going straight is different from that when turning. When going straight, the path is relatively stable, and the safety distance can be appropriately reduced; while when turning, the path changes greatly, and a larger safety distance is required. In addition, the direction change rate reflects the frequency and magnitude of the change in direction. If the direction change rate is large, it means that the child changes direction frequently, which increases the uncertainty and complexity of the path, and a larger safety distance is required to ensure safety.

[0131] The noise adjustment factor reflects the impact of current environmental noise on the safety distance. High environmental noise levels can interfere with children's attention and perception, increasing the risk of accidents. Therefore, a larger safety distance is required in a high-noise environment to compensate for this interference. In addition, the noise change rate reflects the changing trend of environmental noise. If the noise change rate is large, it means that the environmental noise is unstable, increasing uncertainty and potential risks, and a larger safety distance is required to deal with possible emergencies.

[0132] By combining the speed adjustment factor, direction adjustment factor, and noise adjustment factor, the preliminary distance threshold can more comprehensively reflect the safety distance requirements of the current environment and children's behavior. Specifically, the speed adjustment factor ensures that the safety distance is large enough in the case of high speed or frequent acceleration and deceleration to prevent collisions caused by insufficient reaction time. The direction adjustment factor ensures that the safety distance is large enough in the case of frequent changes in direction or turns to prevent accidents caused by path changes. The noise adjustment factor ensures that the safety distance is large enough in the environment of high noise or frequent noise changes to prevent dangers caused by distraction.

[0133] S42: Calculate the behavior pattern adjustment factor, and adjust the preliminary distance threshold based on the behavior pattern adjustment factor to obtain a final dynamic threshold as the adjusted distance threshold, wherein the adjustment of the preliminary distance threshold adopts the following formula:

[0134]

[0135] Among them, T f is the final dynamic threshold, F b is the behavior mode adjustment factor, v is the speed of the mobile terminal at the current moment, and v max is the maximum speed of the mobile terminal in the current trajectory.

[0136] In the remote monitoring method provided by the present invention, the behavior pattern influencing factor has an important impact on the final dynamic threshold because the user's behavior pattern directly affects their security needs and risk level in a specific environment. Different behavior patterns mean that the user may be in different activity states and environmental conditions. The changes in these states and conditions require the security system to be able to flexibly adjust the safety distance threshold to provide more effective protection. Therefore, it is necessary to further adjust the preliminary distance threshold through the behavior pattern adjustment factor.

[0137] Among them, in the adjustment of the basic distance threshold, the calculation formula of the speed adjustment factor is:

[0138]

[0139] Among them, F v is the speed adjustment factor, v is the speed of the mobile terminal at the current moment, Δv is the speed change rate in the current trajectory, and α is the influence coefficient of the preset speed change rate.

[0140] In this embodiment, the value 10 is used to normalize the terminal velocity so that it affects the threshold within a reasonable range.

[0141] The calculation formula of the direction adjustment factor is:

[0142]

[0143] Among them, F d is the direction adjustment factor, d is the direction of the mobile terminal at the current moment, Δd is the direction change rate in the current trajectory, and β is the influence coefficient of the preset direction change rate.

[0144] In this embodiment, the value 90 is also used to standardize the terminal direction so that it affects the threshold within a reasonable range.

[0145] The calculation formula of the noise adjustment factor is:

[0146]

[0147] Among them, F n is the noise adjustment factor, n is the ambient noise level of the mobile terminal at the current moment, Δn is the noise change rate in the current trajectory, and γ is the influence coefficient of the preset noise change rate.

[0148] In this embodiment, the value 100 is also used to normalize the environmental noise level so that it affects the threshold within a reasonable range.

[0149] Specifically, the calculation formula of the behavior mode adjustment factor is:

[0150] F b =K b +δ×Δb

[0151] Among them, F b is the behavior pattern adjustment factor, Δb is the behavior pattern change rate, and is expressed as the ratio of the number of times the user's behavior pattern changes in the current trajectory to the total time, δ is the influence coefficient of the preset behavior pattern change rate, and K b Adjust coefficients for behavioral patterns.

[0152] In this embodiment, the behavior mode adjustment coefficient is an adjustment coefficient under different behavior modes. The corresponding adjustment coefficient can be found through a preset mapping table. This mapping table defines the adjustment coefficients corresponding to different behavior modes so as to be used when calculating the behavior mode adjustment factor. The construction process of the mapping table includes:

[0153] Define behavior patterns: First define all possible behavior patterns, such as standing still, walking, jogging, sprinting, running, cycling, driving, etc.

[0154] Set adjustment coefficients: Set a reasonable adjustment coefficient for each behavior mode. These coefficients can be determined based on actual testing and experience to ensure that the safety distance threshold can be reasonably adjusted under different behavior modes.

[0155] S5: Determine whether the relative distance between the moving object and the mobile terminal is less than the dynamically adjusted distance threshold, and if so, generate an alarm message and push it to an associated terminal pre-paired with the mobile terminal.

[0156] In this embodiment, the relative distance is first obtained, which specifically includes:

[0157] Get the location information of the mobile terminal: Get the current location information of the mobile terminal through GPS data, including longitude, latitude and altitude.

[0158] Estimate the location information of moving objects: The sound data collected by the sound sensor is combined with the accelerometer, gyroscope and GPS data to estimate the location information of the moving object. The specific method is as follows:

[0159] Sound data processing: Calculate the average sound pressure level of sound signals and evaluate the ambient noise level.

[0160] Motion data processing: Use accelerometer and gyroscope data to estimate the motion trajectory and direction of the mobile terminal.

[0161] Position estimation: Combining sound data and motion data, the position of moving objects is estimated using a sound source localization algorithm (specifically a method based on sound intensity and arrival time difference).

[0162] Calculate relative distance: Calculate the relative distance between the mobile terminal and the moving object based on the estimated position information of the mobile terminal and the moving object, where the relative distance is calculated using the Euclidean distance formula.

[0163] Then, the relative distance is compared with the dynamic threshold, including:

[0164] Obtaining a dynamic threshold: Obtaining the final dynamic threshold after dynamic adjustment from step S4.

[0165] Compare the relative distance with the final dynamic threshold: Compare the calculated relative distance with the final dynamic threshold to determine whether it is less than the dynamic threshold.

[0166] Next, generate alarm information, including:

[0167] If the relative distance is less than the final dynamic threshold, it is considered that the moving object is approaching the mobile terminal and there is a potential security risk, and an alarm message is generated, including but not limited to the following: alarm time, location information of the moving object, location information of the mobile terminal, relative distance, dynamic threshold and the user's current behavior pattern.

[0168] Finally, push alarm information, including:

[0169] Associated terminal: Determine the associated terminal pre-paired with the mobile terminal. The associated terminal can be the user's guardian mobile phone, smart watch and other devices.

[0170] Push method: Push alarm information to associated terminals through wireless communication technology (including but not limited to Bluetooth, Wi-Fi, cellular network, etc.).

[0171] Display alarm information: Display alarm information on the associated terminal to remind users to pay attention to safety. The display method can be but not limited to pop-up notification, vibration prompt, sound alarm, etc.

[0172] Embodiment 2

[0173] The present invention also provides a children's watch, referring to Figure 2 As shown, a sensor component 100 and a processor 200 are included.

[0174] The sensor assembly 100 is used for various sensor data.

[0175] The processor 200 is used to extract motion features and sound features of the mobile terminal from various sensor data, and generate a current trajectory based on the motion features.

[0176] The processor 200 is used to identify the behavior pattern of the user corresponding to the mobile terminal through a preset behavior pattern library based on the motion characteristics and the current trajectory, and obtain the current behavior pattern.

[0177] The processor 200 is used to input the motion feature and the sound feature into a pre-trained prediction model, and use the prediction model to determine whether a moving object is approaching the mobile terminal.

[0178] The processor 200 is used to dynamically adjust a preset distance threshold based on the motion characteristics, sound characteristics and current trajectory at the current moment when it is determined that the mobile object is approaching the mobile terminal, wherein the distance threshold represents a safe distance that should be maintained between the mobile terminal and the mobile object under the current behavior mode.

[0179] The processor 200 is used to determine whether the relative distance between the mobile object and the mobile terminal is less than the dynamically adjusted distance threshold, and if so, generate an alarm message and push it to an associated terminal pre-paired with the mobile terminal.

[0180] The present application is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0181] Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0182] 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.

Claims

1. A remote monitoring method, characterized in that: The monitoring method comprises the following steps: Extracting motion features and sound features of the mobile terminal from a variety of sensor data, and generating a current trajectory based on the motion features, wherein the various sensor data are collected by a sensor component installed on the mobile terminal; Based on the motion feature and the current trajectory, identifying the behavior pattern of the user corresponding to the mobile terminal through a preset behavior pattern library to obtain a current behavior pattern; Inputting the motion feature and the sound feature into a pre-trained prediction model, and using the prediction model to determine whether a moving object is approaching the mobile terminal; When it is determined that the moving object is approaching the mobile terminal, dynamically adjusting a preset distance threshold based on the motion characteristics, sound characteristics and current trajectory at the current moment, wherein the distance threshold represents a safe distance that should be maintained between the mobile terminal and the moving object in the current behavior mode; It is determined whether the relative distance between the mobile object and the mobile terminal is less than the dynamically adjusted distance threshold, and if so, an alarm message is generated and pushed to an associated terminal pre-paired with the mobile terminal.

2. A remote monitoring method according to claim 1, characterized in that: The multiple sensor data include accelerometer data, gyroscope data, GPS data and sound data, wherein the accelerometer data, the gyroscope data and the GPS data are used to extract the motion features, and the sound data is used to extract the sound features, wherein the motion features include accelerometer data features, gyroscope data features and GPS data features corresponding to the accelerometer data, the gyroscope data and the GPS data respectively.

3. A remote monitoring method according to claim 2, characterized in that: The steps for generating the current trajectory are: Based on the extracted accelerometer data features and gyroscope data features, combined with the position change information and timestamp data in the GPS data features, a map matching algorithm is used to generate the current trajectory of the mobile terminal.

4. A remote monitoring method according to claim 3, characterized in that: The identifying, based on the motion feature and the current trajectory, the behavior pattern of the user corresponding to the mobile terminal through a preset behavior pattern library to obtain the current behavior pattern includes: Extracting a plurality of behavior pattern-related feature vectors from the accelerometer data features, the gyroscope data features, and the GPS data features, wherein the feature vectors include acceleration, angular velocity, speed, direction, and speed change rate; Matching the feature vector with a plurality of behavior patterns in a preset behavior pattern library, wherein the behavior pattern library contains a plurality of known behavior patterns and their corresponding feature vectors; The similarity between the feature vector and the feature vector of each known behavior pattern is calculated, and the known behavior pattern with the highest similarity is selected as the current behavior pattern.

5. A remote monitoring method according to claim 4, characterized in that: The step of inputting the motion feature and the sound feature into a pre-trained prediction model, and using the prediction model to determine whether a moving object is approaching the mobile terminal, includes: Fusing the extracted motion features and the extracted sound features to form a fused feature vector; Inputting the fused feature vector into a pre-trained prediction model, wherein the prediction model is a neural network model that is trained to recognize and analyze the fused feature vector to determine whether a moving object is approaching the mobile terminal; Output the judgment result. If the judgment result is that a moving object is approaching, the subsequent distance threshold dynamic adjustment step is triggered.

6. A remote monitoring method according to claim 5, characterized in that: The training process of the prediction model includes: Collect annotated datasets containing motion features, sound features, and corresponding moving object proximity information; Using the data set to train a neural network model, and adjusting model parameters during the training process to minimize the prediction error rate; The performance of the neural network model is evaluated through the cross-validation method, and finally the trained model is saved as a pre-trained prediction model.

7. A remote monitoring method according to claim 6, characterized in that: When it is determined that the moving object is approaching the mobile terminal, dynamically adjusting the preset distance threshold based on the motion characteristics, sound characteristics and current trajectory at the current moment includes: The preset basic distance threshold is adjusted to obtain a preliminary distance threshold, wherein the adjustment adopts the following formula: Among them, T p is the initial distance threshold, T0 is the basic distance threshold, and F v is the speed adjustment factor, F d is the direction adjustment factor, F n is the noise adjustment factor; The behavior pattern adjustment factor is calculated, and the preliminary distance threshold is adjusted based on the behavior pattern adjustment factor to obtain a final dynamic threshold as the adjusted distance threshold, wherein the adjustment of the preliminary distance threshold adopts the following formula: Among them, T f is the final dynamic threshold, F b is the behavior mode adjustment factor, v is the speed of the mobile terminal at the current moment, and v max is the maximum speed of the mobile terminal in the current trajectory.

8. A remote monitoring method according to claim 7, characterized in that: In the adjustment of the basic distance threshold, the calculation formula of the speed adjustment factor is: Among them, F v is the speed adjustment factor, v is the speed of the mobile terminal at the current moment, Δv is the speed change rate in the current trajectory, and α is the influence coefficient of the preset speed change rate; The calculation formula of the direction adjustment factor is: Among them, F d is the direction adjustment factor, d is the direction of the mobile terminal at the current moment, Δd is the direction change rate in the current trajectory, and β is the influence coefficient of the preset direction change rate; The calculation formula of the noise adjustment factor is: Among them, F n is the noise adjustment factor, n is the ambient noise level of the mobile terminal at the current moment, Δn is the noise change rate in the current trajectory, and γ is the influence coefficient of the preset noise change rate.

9. A remote monitoring method according to claim 8, characterized in that: The calculation formula of the behavior mode adjustment factor is: F b =K b +δ×Δb Among them, F b is the behavior pattern adjustment factor, Δb is the behavior pattern change rate, and is expressed as the ratio of the number of times the user's behavior pattern changes in the current trajectory to the total time, δ is the influence coefficient of the preset behavior pattern change rate, and K b Adjust coefficients for behavioral patterns.

10. A children's watch, characterized in that: Including sensor components and processors, The sensor assembly is used for various sensor data; The processor is used to extract motion features and sound features of the mobile terminal from multiple sensor data, and generate a current trajectory based on the motion features; The processor is used to identify the behavior pattern of the user corresponding to the mobile terminal through a preset behavior pattern library based on the motion characteristics and the current trajectory to obtain the current behavior pattern; The processor is used to input the motion feature and the sound feature into a pre-trained prediction model, and use the prediction model to determine whether a moving object is approaching the mobile terminal; The processor is configured to dynamically adjust a preset distance threshold based on the motion characteristics, sound characteristics, and current trajectory at a current moment when it is determined that the mobile object is approaching the mobile terminal, wherein the distance threshold represents a safe distance that should be maintained between the mobile terminal and the mobile object in a current behavior mode; The processor is used to determine whether the relative distance between the moving object and the mobile terminal is less than the dynamically adjusted distance threshold, and if so, generate an alarm message and push it to an associated terminal pre-paired with the mobile terminal.

Images