IOT based monitoring method and system for detecting separation anxiety of pet using support vector machine and complex event processing
AssigneeKorea University Research and Business Foundation, Sejong Campus
InventorsDai Hee Park, Huasang Wang
Life of this patent
Estimating the term…
Abstract
Disclosed is a method and system for monitoring a separation anxiety of a pet that may recognize an activity of the pet using a support vector machine (SVM) based on pet-related data acquired from a sensor, and may determine whether the pet is in a separation anxiety state by analyzing input data that is the recognized activity of the pet using a complex event processing (CEP).
Claims9
A system for monitoring a separation anxiety of a pet, the system comprising:
- an activity recognizer configured to recognize an activity of the pet using a support vector machine (SVM) based on pet-related data acquired from at least one sensor wherein the activity recognizer is configured to recognize continuous primary activities of the pet based on data that is continuously acquired from the at least one sensor;
- and
- a separation anxiety determiner configured to determine whether the pet is in a separation anxiety state by analyzing input data that is the activity of the pet recognized by the activity recognizer using a complex event processing (CEP), wherein the separation anxiety determiner is configured to:
- determine whether the pet is in the separation anxiety state by analyzing the continuous primary activities using a hierarchical model, wherein the primary activities comprise at least one of a posture activity of the pet comprising walking, standing, sitting, or lying;
- a head motion activity of the pet comprising head up, head down, or horizontal head movement;
- a voice activity of the pet comprising howling, barking, or whining;
- and a motion activity of the pet comprising digging or jumping;
- set each of the primary activities as an event of a first level, to determine an event of a second level of the pet comprising at least one of sniffing and head shake based on the event of the first level, to determine an event of a third level of the pet comprising at least one of an exploratory behavior, a play behavior, a destructive behavior, an escape behavior, and a voice behavior based on at least one of the event of the first level and the event of the second level, and to determine whether the pet is in the separation anxiety state based on at least one of sequence in which a behavior corresponding to the event of the third level occurs and frequency thereof;
- and
- determine that the pet is in the separation anxiety state when the voice behavior is determined to be present as the event of the third level before or after the destructive behavior or the escape behavior.
A computer-implemented method of monitoring a separation anxiety of a pet, the method comprising:
- acquiring pet-related data from at least one sensor;
- recognizing an activity of the pet using a support vector machine (SVM) based on the acquired pet-related data, wherein the recognizing comprises recognizing continuous primary activities of the pet based on data that is continuously acquired from the at least one sensor;
- and
- determining whether the pet is in a separation anxiety state by analyzing input data that is the recognized activity of the pet using a complex event processing (CEP), wherein the determining comprises:
- determining whether the pet is in the separation anxiety state by analyzing the continuous primary activities using a hierarchical model, wherein the primary activities comprise at least one of a posture activity of the pet comprising walking, standing, sitting, or lying;
- a head motion activity of the pet comprising head up, head down, or horizontal head movement a voice activity of the pet comprising howling, barking, or whining;
- and a motion activity of the pet comprising digging or jumping;
- setting each of the primary activities as an event of a first level;
- determining an event of a second level of the pet comprising at least one of sniffing and head shake based on the event of the first level;
- determining an event of a third level of the pet comprising at least one of an exploratory behavior, a play behavior, a destructive behavior, an escape behavior, and a voice behavior based on at least one of the event of the first level and the event of the second level;
- and
- determining whether the pet is in the separation anxiety state based on at least one of sequence in which a behavior corresponding to the event of the third level occurs and frequency thereof.
A system for monitoring a separation anxiety of a pet, the system comprising:
- a plurality of sensors configured to collect pet-related data, at least one of the sensors being worn around the pet;
- an activity recognizer configured to recognize an activity of the pet using a support vector machine (SVM) based on the pet-related data acquired from the plurality of sensors, wherein the recognizing comprises recognizing continuous primary activities of the pet based on data that is continuously acquired from the at least one sensor;
- a separation anxiety determiner configured to determine whether the pet is in a separation anxiety state by analyzing input data that is the activity of the pet recognized by the activity recognizer using a complex event processing (CEP), wherein the determining comprises:
- determining whether the pet is in the separation anxiety state by analyzing the continuous primary activities using a hierarchical model, wherein the primary activities comprise at least one of a posture activity of the pet comprising walking, standing, sitting, or lying;
- a head motion activity of the pet comprising head up, head down, or horizontal head movement a voice activity of the pet comprising howling, barking, or whining;
- and a motion activity of the pet comprising digging or jumping
- setting each of the primary activities as an event of a first level;
- determining an event of a second level of the pet comprising at least one of sniffing and head shake based on the event of the first level;
- determining an event of a third level of the pet comprising at least one of an exploratory behavior, a play behavior, a destructive behavior, an escape behavior, and a voice behavior based on at least one of the event of the first level and the event of the second level;
- and
- determining whether the pet is in the separation anxiety state based on at least one of sequence in which a behavior corresponding to the event of the third level occurs and frequency thereof;
- a user interface device configured to receive data indicating that the pet is in the separation anxiety state, output from the separation anxiety determiner based on a determination of the separation anxiety determiner, and to output a command for a measure for the separation anxiety state;
- a separation anxiety handler configured to receive the command output from the user interface device, to generate a measure command for taking a measure for relieving the separation anxiety of the pet, and to transmit the measure command to at least one actuator;
- and
- the at least one actuator configured to take the measure for relieving the separation anxiety of the pet in response to the measure command.
Dependencies are inferred from the claim text.