Intelligent wearable device for pet behavior monitoring and early warning and electric shock blocking wounding behavior

Abstract

The invention relates to the technical field of pets, and provides an intelligent wearable device for pet behavior monitoring and early warning and electric shock blocking of wounding behaviors. Comprising a high-definition camera, a microphone, a preprocessing module, a feature extraction module, a behavior classification module, a built-in accelerometer, a gyroscope, a microphone and other sensors, a time sequence analysis module, a threshold setting module, an intelligent intervention decision module, a safety protection module, an intelligent electric shock blocking module, a personalized setting module and a data recording and analysis module. A remote control module; and a battery endurance and charging module. According to the intelligent wearable device for pet behavior monitoring and early warning and electric shock blocking of hurt behaviors, for a pet owner, the safety guarantee is provided, the understanding of the pet is enhanced, and personalized management is achieved. From the perspective of the pet, the equipment takes into account the health and welfare of the pet while correcting bad behaviors.

Description

Technical Field

[0001] The present invention relates to the technical field of pets, and particularly to a wearable device for intelligent pet behavior monitoring, early warning and electroshock blocking of attacking behaviors. Background Art

[0002] With the improvement of people's living standards, pet raising has become increasingly common. However, pet attacking incidents also occur from time to time, which not only cause physical harm to the victims, but may also trigger a series of social problems, such as neighborhood disputes and potential public safety hazards.

[0003] Traditional pet management methods mainly rely on the experience and training of the owner, but these methods have certain limitations. On the one hand, the owner cannot always closely monitor the behavior of the pet, especially when the pet is alone or the owner is busy. For example, when the owner is at work, doing housework or sleeping, it is difficult to detect whether the pet is about to exhibit aggressive behavior. On the other hand, for some pets that are difficult to train or have a relatively stubborn personality, simply relying on traditional training methods to correct their aggressive behavior has poor effects.

[0004] In the field of pet behavior research, although there are already some technical means for monitoring pet behavior, the existing technologies still have many deficiencies. For example, some pet monitoring devices can only perform simple activity tracking, such as recording the steps or activity range of the pet, but lack in-depth analysis of the pet's behavior intention and cannot accurately determine whether the pet is about to perform an aggressive behavior. In addition, although some devices can perform behavior recognition, they lack effective intervention measures and cannot prevent the pet's aggressive behavior in time when it is detected, and thus cannot truly protect people and other animals from being attacked by the pet.

[0005] At the same time, in terms of pet behavior intervention, traditional devices such as electric shock collars have problems such as uncontrollable electric shock intensity or lack of personalized settings. These devices often use a fixed electric shock intensity without considering individual differences of the pet, such as factors like weight, age, and health status, which may cause excessive harm to the pet and affect its physical and mental health. Moreover, the existing devices lack a comprehensive safety protection mechanism, and when the device fails, such as in the case of overcurrent or short circuit in the circuit, it may pose a greater risk to the pet.

[0006] In summary, in order to better ensure the safety of pet owners, family members, visitors and other animals, while taking into account the health and welfare of the pet, there is a need for an intelligent wearable device that can monitor pet behavior in real time, accurately warn of aggressive behavior, and block aggressive behavior in a safe and personalized manner when necessary. This is the research and development background of the present invention. Summary of the Invention

[0007] To overcome the deficiencies of the prior art, the purpose of the present invention is to provide a wearable device for intelligent pet behavior monitoring and early warning and electric shock blocking of injury-causing behaviors.

[0008] To achieve the above object, the technical solution of the present invention is realized as follows: A wearable device for intelligent pet behavior monitoring and early warning and electric shock blocking of injury-causing behaviors, comprising: A high-definition camera and a microphone, used to capture video and audio data of the pet; A preprocessing module, which performs preprocessing such as denoising and enhancement on the video and audio data to improve the data quality; A feature extraction module, which uses a convolutional neural network (CNN) to extract the posture, movement and expression features of the pet from video frames, and uses audio processing technology to extract the barking features of the pet; A behavior classification module, which inputs the extracted features into a support vector machine (SVM) or a deep neural network (DNN) for classification to distinguish between normal behaviors and attack behaviors of the pet; Built-in sensors such as an accelerometer, a gyroscope, and a microphone to real-time monitor the movement state and sound features of the pet; A time series analysis module, which performs time series analysis on the sensor data, and combines with a behavior pattern recognition algorithm to perform real-time classification and prediction on the pet behavior; A threshold setting module, which sets different thresholds according to the individual characteristics and behavior habits of the pet to accurately judge whether the pet is in an attack state; An intelligent intervention decision-making module, which is built-in with an electric shock device. When detecting an attack behavior, it comprehensively considers factors such as the behavior characteristics, attack intensity, and historical behavior records of the pet, decides whether electric shock blocking is needed, and automatically adjusts the electric shock intensity according to the weight, age, and health status of the pet; A safety protection module, which is built-in with safety protection mechanisms such as overcurrent protection and short-circuit protection to ensure that no accidental injury is caused to the pet or the pet owner during the electric shock process; An intelligent electric shock blocking module, when it is confirmed that the pet is about to attack, it emits a low-voltage electric shock signal to block the attack behavior of the pet. The electric shock signal has an intelligent adjustment function and can be personalized according to the specific situation of the pet; A personalized setting module, which allows the pet owner to perform personalized settings according to factors such as the type, age, and personality of the pet, including electric shock intensity, threshold setting, behavior pattern recognition, etc.; A data recording and analysis module, which can record the behavior data of the pet, including the number of attack behaviors, time, intensity, etc., and sends it to the mobile application of the pet owner through wireless transmission technology. The pet owner can view and analyze the behavior pattern of the pet through the mobile application to optimize the training effect; Remote control module: Pet owners can remotely control the device through a mobile application, including turning on or off the shock function, adjusting the shock intensity, viewing behavior data, etc. Battery life and charging module: It uses a high-capacity battery to support long-term use and also supports wireless charging technology, making it convenient for pet owners to charge the device at any time.

[0009] Preferably, in the behavior classification module, the normal behaviors of pets are distinguished as playing, resting, and eating, and the aggressive behaviors include barking, biting, and scratching.

[0010] Preferably, in the feature extraction module, the structure of the convolutional neural network (CNN) includes multiple convolutional layers and pooling layers to extract richer pet features.

[0011] Preferably, the time series analysis module uses the autoregressive moving average model (ARMA) or the Kalman filter algorithm to improve the accuracy of real-time classification and prediction of pet behaviors.

[0012] Preferably, in the threshold setting module, the setting of the threshold also considers factors such as the breed and size of the pet.

[0013] Preferably, in the intelligent intervention decision module, the triggering conditions for shock blocking also include the attack intention and the attack target of the pet.

[0014] Preferably, the overcurrent protection mechanism in the safety protection module includes a current detection circuit and a fast cut-off circuit, and the short-circuit protection mechanism includes a short-circuit detection circuit and an automatic recovery circuit.

[0015] Preferably, in the data recording and analysis module, the recorded pet behavior data also includes the activity range and activity trajectory of the pet.

[0016] Preferably, in the remote control module, the mobile application also provides a pet behavior warning function. When the device detects a potential aggressive behavior, it will send a warning message to the pet owner in a timely manner.

[0017] Preferably, in the battery life and charging module, the high-capacity battery uses a lithium-ion battery, and the wireless charging technology complies with the Qi standard.

[0018] The beneficial effects of the present invention are as follows: I. Beneficial effects for pet owners Ensuring safety It can effectively warn of the aggressive behaviors of pets and send warning messages to the owner in advance, enabling the owner to take timely measures to avoid being injured by the pet himself, his family members or visitors. Especially for families with children or the elderly, this warning function provides additional safety protection.

[0019] When the pet is in a state where it may attack, the device makes an intelligent decision to determine whether to perform an electric shock block, further reducing the risk of the pet hurting people.

[0020] In-depth understanding of the pet The owner can view the pet's behavior data through the mobile application, including the activity range, activity trajectory, number of attack behaviors, time, and intensity, etc. This helps to understand the pet's habits and behavior patterns in depth. For example, to understand the peak activity period of the pet in a day and the favorite activity areas.

[0021] Based on this data, the owner can better meet the pet's needs, such as adjusting the pet's diet, activity arrangements, etc., and improving the pet's quality of life.

[0022] Personalized management of the pet The owner can remotely control the device through the mobile application, including turning on or off the electric shock function, adjusting the electric shock intensity, etc. This enables the owner to perform personalized management according to the specific situation of the pet. For example, for a pet that is being trained or is unwell, the device settings can be flexibly adjusted.

[0023] The personalized setting module also allows the owner to adjust the threshold setting, behavior pattern recognition, etc. according to factors such as the pet's species, age, and personality, better adapting to the characteristics of different pets.

[0024] II. Beneficial effects on the pet Moderate correction of behavior When the pet shows an attack tendency, the low-voltage electric shock signal can timely block the attack behavior. This method helps to correct the pet's bad behavior without causing long-term adverse effects, enabling the pet to gradually learn to control its behavior.

[0025] The electric shock intensity will be automatically adjusted according to the pet's weight, age, and health status, avoiding excessive harm to the pet, reflecting the concern for the pet's health and welfare.

[0026] Health monitoring The sensors in the device continuously monitor the pet's movement state. For example, the accelerometer and gyroscope monitor the pet's movement direction, speed, and acceleration, etc. To a certain extent, this can reflect the pet's health status. For example, if the pet's movement data is abnormal, it may indicate that the pet is unwell, and the owner can take the pet to the doctor in time.

[0027] In summary, this wearable device for intelligent pet behavior monitoring, warning, and electric shock blocking of harmful behaviors has various beneficial effects. For pet owners, it not only provides safety protection but also enhances their understanding of pets and enables personalized management. From the perspective of pets, the device takes into account the health and welfare of pets while correcting bad behaviors. In addition, multiple functional modules of the device, such as data acquisition, preprocessing, feature extraction, behavior classification, sensor data processing, threshold setting, intelligent intervention decision-making, safety protection mechanism, electric shock blocking operation, and data recording, analysis, and remote control, work together to form a complete system. This system can adapt to the individual differences of different pets and can operate stably in both indoor and outdoor environments, providing effective technical support for the harmonious coexistence between pets and their owners. It is an innovative and practical pet management device. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In the drawings: Figure 1 is a schematic structural diagram of the control system of the present invention; Figure 2 is a schematic diagram of the electric shock device of the present invention and its internal structure distribution; Figure 3 is a schematic diagram of the structure of the electric shock device of the present invention inherited on the collar.

[0029] Description of the reference numerals: 1, high-definition camera; 2, microphone; 3, built-in accelerometer; 4, gyroscope; 5, electric shock device. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0030] The present invention will be further described in detail below with reference to the drawings and embodiments. Obviously, the described embodiments are only a part of the embodiments of the invention, not all of them. Without conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the invention without creative efforts shall fall within the scope of protection of the invention.

[0031] Please refer to the attached Figures 1-3 description, the present invention provides a wearable device for intelligent pet behavior monitoring, warning, and electric shock blocking of harmful behaviors: I. Composition and Installation of Device Hardware Hardware Components The main body of the wearable device integrates components such as a high-definition camera 1, a microphone 2, a built-in accelerometer 3, a gyroscope 4, and a battery. The device is designed in a form suitable for pets to wear, such as a collar or a harness structure, to ensure that it will not cause too much interference to the pets during their daily activities.

[0032] The positions of the high-definition camera 1 and the microphone 2 are carefully designed to capture the video and audio data of the pet in all directions. For example, the camera is located at the front of the device and can clearly capture the situation in front of and around the pet, while the microphones 2 are distributed around the device to accurately receive sounds from different directions.

[0033] Battery Installation and Charging A high-capacity lithium-ion battery is used. There is a dedicated battery compartment inside the device, and the battery is connected to each module of the device through a safe and reliable circuit.

[0034] It supports the Qi-standard wireless charging technology. There is a wireless charging receiving area set at the bottom of the device. The pet owner can use a Qi-standard wireless charger and place the device on the charger to charge it, which is convenient and fast and can meet the long-term usage needs of the device.

[0035] II. Data Acquisition and Preprocessing Video and Audio Data Acquisition When the device is turned on, the high-definition camera 1 and the microphone 2 work in real time. The camera captures the video images of the pet at a certain frame rate (e.g., 30 frames per second), and the microphone 2 continuously receives the sound signals around the pet.

[0036] During the pet's activities, data can be stably acquired whether in an indoor or outdoor environment.

[0037] Data Preprocessing The acquired video and audio data are transmitted to the preprocessing module. For video data, the preprocessing module uses an image denoising algorithm (such as median filtering) to remove the noise points in the image, and at the same time improves the clarity of the image through a contrast enhancement algorithm.

[0038] For audio data, audio filtering technology is used to remove background noise and enhance the intensity of effective audio signals such as the pet's barking, thereby improving the data quality and preparing for subsequent feature extraction.

[0039] III. Feature Extraction Feature Extraction of Posture, Action, and Expression (Video Data) The convolutional neural network (CNN) in the feature extraction module starts to work. Multiple convolutional layers and pooling layers of the CNN process the video frames in sequence. For example, the convolutional layer performs convolutional operations by sliding the convolutional kernel on the video frame to extract features at different scales, and the pooling layer downsamples the output of the convolutional layer to reduce the data volume while retaining the key features.

[0040] In this way, the CNN can accurately extract the features of the pet's posture (such as standing, squatting, lying down, etc.), action (such as running, jumping, etc.), and expression (such as excited, nervous, etc.) from the video frames.

[0041] Call feature extraction (audio data) Using audio processing techniques, such as audio feature extraction algorithms (e.g., Mel Frequency Cepstral Coefficient (MFCC) algorithm), process the preprocessed audio data to extract the call features of the pet, including features such as the frequency, duration, and timbre of the call.

[0042] IV. Behavior classification Classification model training and initialization In the behavior classification module, first use a large amount of labeled pet behavior data (including normal behaviors such as playing, resting, eating, and aggressive behaviors such as barking, biting, scratching) to train a Support Vector Machine (SVM) or a Deep Neural Network (DNN).

[0043] When the device is initialized, load the trained classification model into the device.

[0044] Real-time classification After extracting the posture, movement, expression, and call features of the pet, input these features into the classification model. The SVM or DNN classifies the pet's behavior according to the parameters and algorithms in the model, and determines whether it is a normal behavior or an aggressive behavior.

[0045] V. Sensor data processing and behavior prediction Sensor data acquisition and transmission The built-in accelerometer and gyroscope 4 continuously monitor the pet's motion state, such as data on the pet's motion direction, speed, acceleration, etc., and the microphone 2 simultaneously continuously collects sound feature data.

[0046] These sensor data are transmitted to the time series analysis module in real time.

[0047] Time series analysis The time series analysis module uses an AutoRegressive Moving Average model (ARMA) or a Kalman filter algorithm to process the sensor data. For example, the ARMA model predicts future values based on past sensor data, and the Kalman filter algorithm improves the accuracy of real-time classification and prediction of the pet's behavior by fusing data from multiple sensors and considering noise factors.

[0048] Combined with behavior pattern recognition algorithms, perform more accurate real-time classification and prediction of the pet's behavior, such as predicting whether the pet is about to change from a normal behavior to an aggressive behavior.

[0049] VI. Threshold setting and aggressive behavior judgment Basis for threshold setting When setting the threshold value, the threshold setting module comprehensively considers the individual characteristics and behavior habits of the pet, including factors such as breed, body size, and age. For example, for dog breeds with a larger body size and a relatively fierce temperament, the threshold for their attack behavior may be relatively loose, while for small and docile pets, the threshold is set more strictly.

[0050] Judgment of Attack Behavior Based on the real-time behavior data of the pet and the set threshold, it is judged whether the pet is in an attack state. If the behavior characteristics of the pet (such as the intensity of barking, the fierceness of movements, etc.) exceed the set threshold, it is considered that the pet may be in an attack state.

[0051] VII. Intelligent Intervention Decision Comprehensive Consideration of Decision-making Factors When it is detected that the pet may be in an attack state, the intelligent intervention decision-making module starts to work. This module comprehensively considers factors such as the behavior characteristics of the pet (such as the attack posture, the type of barking), the attack intensity (such as the force of pouncing and biting, the speed of scratching), and the historical behavior record (such as the frequency and severity of past attack behaviors).

[0052] At the same time, it also considers the attack intention of the pet (such as whether it is for self-defense or unprovoked attack) and the attack object (such as whether it is towards strangers, other pets, or the owner) to decide whether to activate the electric shock device 5 for electric shock blocking.

[0053] Adjustment of Electric Shock Intensity Automatically adjust the electric shock intensity according to the weight, age, and health status of the pet. For example, for pets with a lighter weight, a younger age, or a relatively weaker body, a lower electric shock intensity is set to avoid excessive harm to the pet.

[0054] VIII. Safety Protection Mechanism Overcurrent Protection The overcurrent protection mechanism in the safety protection module is activated. The current detection circuit continuously monitors the magnitude of the current in the device. When the current exceeds the set safety threshold, the fast cut-off circuit immediately operates to cut off the power supply and prevent excessive current from damaging the pet or the device itself.

[0055] Short Circuit Protection The short circuit detection circuit continuously detects whether there is a short circuit in the device circuit. Once a short circuit is detected, the automatic recovery circuit will automatically restore the normal operation of the device after eliminating the short circuit fault to ensure the safety of the device under various conditions.

[0056] IX. Electric Shock Blocking Operation Intelligent Adjustment and Personalized Settings When the intelligent electric shock blocking module confirms that the pet is about to attack, it emits a low-voltage electric shock signal. The electric shock signal has an intelligent adjustment function and can be adjusted according to the personalized settings made by the pet owner in the personalized setting module based on factors such as the pet's species, age, and personality. For example, for a pet with a sensitive personality, the electric shock intensity can be set lower.

[0057] The personalized setting module also allows the pet owner to make personalized adjustments to threshold setting, behavior pattern recognition, etc. to adapt to the characteristics of different pets.

[0058] Blocking effect The low-voltage electric shock signal acts on the pet to block the pet's attack behavior. The electric shock intensity and duration are carefully designed to effectively prevent the pet's attack behavior without causing long-term adverse effects on the pet.

[0059] X. Data recording, analysis and remote control Data recording and transmission The data recording and analysis module records the pet's behavior data, including not only the number, time, and intensity of attack behaviors, but also the pet's activity range and activity trajectory. For example, by analyzing the data of the accelerometer and gyroscope (4), the activity range and trajectory of the pet indoors or outdoors are determined.

[0060] These data are sent to the pet owner's mobile application through wireless transmission technology (such as Bluetooth or Wi-Fi).

[0061] Mobile application functions The pet owner can view and analyze the pet's behavior pattern through the mobile application. The mobile application also provides a pet behavior warning function, and when the device detects a potential attack behavior, it sends a warning message to the pet owner in a timely manner.

[0062] The pet owner can remotely control the device through the mobile application, including operations such as turning on or off the electric shock function, adjusting the electric shock intensity, and viewing behavior data.

[0063] Working steps: The high-definition camera 1 and microphone 2 continuously capture the pet's video and audio data, and the built-in sensors such as the accelerometer 3 and gyroscope 4 real-time monitor the pet's motion state and sound characteristics.

[0064] The collected video and audio data are transmitted to the preprocessing module. The video data improves the image quality through operations such as denoising and enhancement, and the audio data enhances the effective signal through processing such as filtering.

[0065] The feature extraction module extracts the pet's posture, action, and expression features from the video frames using a convolutional neural network (CNN), and extracts the pet's barking features from the audio data using audio processing technology.

[0066] The behavior classification module inputs the extracted features into a support vector machine (SVM) or a deep neural network (DNN) for classification. At the same time, the time series analysis module performs time series analysis on the sensor data using an autoregressive moving average model (ARMA) or a Kalman filter algorithm, and combines with a behavior pattern recognition algorithm to classify and predict the pet's behavior in real time.

[0067] The threshold setting module determines whether the pet reaches the critical value of the attack state according to the threshold set based on the pet's individual characteristics (including breed, body size, age, behavior habits, etc.).

[0068] When the intelligent intervention decision module detects that the pet may be in an attack state, it comprehensively considers factors such as the pet's behavior characteristics (such as attack posture, type of barking), attack intensity (such as biting force, scratching speed), historical behavior records (frequency and severity of past attack behaviors), attack intention (self-defense or unprovoked attack), and attack object (stranger, other pet or owner), etc., to decide whether to perform an electric shock block. At the same time, the electric shock intensity is automatically adjusted according to the pet's weight, age and health status.

[0069] Before deciding to perform an electric shock block, safety protection mechanisms such as overcurrent protection and short-circuit protection in the safety protection module perform self-checks to ensure that the device is in a safe and operable state. For example, the current detection circuit checks for overcurrent risks, and the short-circuit detection circuit checks whether the circuit is normal.

[0070] If the intelligent intervention decision module determines that an electric shock block is needed, the intelligent electric shock block module emits a low-voltage electric shock signal. The electric shock signal is intelligently adjusted according to the parameters set by the pet owner in the personalized setting module according to factors such as pet species, age, and personality to block the pet's attack behavior.

[0071] The data recording and analysis module records the relevant data of the pet's approaching attack behavior this time, including the time when the attack trend appears, the pet's behavior characteristics, whether an electric shock block has been performed, etc., and sends it to the pet owner's mobile application through wireless transmission technology. At the same time, if the pet owner has enabled the warning function, the mobile application will receive a warning message of potential attack behavior. The pet owner can remotely view the device status, behavior data, etc. through the mobile application, and can adjust the device settings as needed, such as adjusting the electric shock intensity or turning off the electric shock function, etc.

[0072] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A wearable device for intelligent pet behavior monitoring and early warning and electric shock to block harmful behavior, characterized in that: include: A high-definition camera (1) and a microphone (2) for capturing video and audio data of the pet; A preprocessing module performs preprocessing such as denoising and enhancement on the video and audio data to improve data quality; The feature extraction module uses a convolutional neural network (CNN) to extract the pet’s posture, movement, and expression features from video frames, and uses audio processing technology to extract the pet’s call features; The behavior classification module inputs the extracted features into the support vector machine (SVM) or deep neural network (DNN) for classification to distinguish between normal and aggressive behaviors of pets; Built-in sensors such as an accelerometer (3), a gyroscope (4), and a microphone (2) monitor the pet's movement status and sound characteristics in real time; The time series analysis module performs time series analysis on sensor data and combines it with the behavior pattern recognition algorithm to classify and predict pet behavior in real time; The threshold setting module sets different thresholds according to the individual characteristics and behavioral habits of the pet to accurately determine whether the pet is in an attacking state; The intelligent intervention decision module, when detecting aggressive behavior, comprehensively considers factors such as the pet's behavioral characteristics, attack intensity, and historical behavior records to decide whether electric shock blocking is needed, and automatically adjusts the intensity of electric shock according to the pet's weight, age, and health status; Safety protection module, with built-in over-current protection, short-circuit protection and other safety protection mechanisms, to ensure that no accidental harm will be caused to pets or pet owners during the electric shock process; An intelligent electric shock blocking module has a built-in electric shock device (5). When it is confirmed that the pet is about to attack, it sends a low-voltage electric shock signal to block the pet's attack behavior. The electric shock signal has an intelligent adjustment function and can be personalized according to the specific situation of the pet; The personalized settings module allows pet owners to make personalized settings based on the pet's type, age, personality and other factors, including electric shock intensity, threshold setting, behavioral pattern recognition, etc.; The data recording and analysis module can record the pet's behavior data, including the number, time, and intensity of aggressive behaviors, and send it to the pet owner's mobile phone application through wireless transmission technology. The pet owner can view and analyze the pet's behavior patterns through the mobile phone application to optimize the training effect; Remote control module, pet owners can remotely control the device through a mobile phone application, including turning on or off the electric shock function, adjusting the intensity of the electric shock, and viewing behavioral data; The battery life and charging module uses a high-capacity battery to support long-term use. It also supports wireless charging technology, making it convenient for pet owners to charge the device at any time.

2. According to claim 1, a wearable device for monitoring and warning of intelligent pet behavior and blocking harmful behavior by electric shock, characterized in that: In the behavior classification module, the normal behaviors of the pet are distinguished to include playing, resting, and eating, and the aggressive behaviors include barking, biting, and scratching.

3. According to claim 1, a wearable device for intelligent pet behavior monitoring and early warning and electric shock to block harmful behavior is characterized in that: In the feature extraction module, the structure of the convolutional neural network (CNN) includes multiple convolutional layers and pooling layers to extract richer pet features.

4. According to claim 1, a wearable device for intelligent pet behavior monitoring and early warning and electric shock to block harmful behavior is characterized in that: The time series analysis module adopts an autoregressive moving average model (ARMA) or a Kalman filter algorithm to improve the accuracy of real-time classification and prediction of pet behavior.

5. According to claim 1, a wearable device for monitoring and warning of intelligent pet behavior and electric shock to block harmful behavior is characterized in that: In the threshold setting module, the setting of the threshold also takes into account the breed and body shape of the pet.

6. According to claim 1, a wearable device for monitoring and warning of intelligent pet behavior and blocking harmful behavior by electric shock, characterized in that: In the intelligent intervention decision module, the triggering conditions for electric shock blocking also include the pet's attack intention and attack target.

7. According to claim 1, a wearable device for monitoring and warning of intelligent pet behavior and blocking harmful behavior by electric shock, characterized in that: The overcurrent protection mechanism in the safety protection module includes a current detection circuit and a fast cut-off circuit, and the short-circuit protection mechanism includes a short-circuit detection circuit and an automatic recovery circuit.

8. According to claim 1, a wearable device for monitoring and warning of intelligent pet behavior and blocking harmful behavior by electric shock, characterized in that: In the data recording and analysis module, the recorded pet behavior data also includes the pet's activity range and activity trajectory.

9. According to claim 1, a wearable device for monitoring and warning of intelligent pet behavior and blocking harmful behavior by electric shock, characterized in that: In the remote control module, the mobile phone application also provides a pet behavior warning function. When the device detects potential attack behavior, it sends a warning message to the pet owner in time.

10. The wearable device for intelligent pet behavior monitoring and early warning and electric shock to block harmful behavior according to claim 1, characterized in that: In the battery life and charging module, the high-capacity battery uses a lithium-ion battery, and the wireless charging technology complies with the Qi standard.

Images