Animal attack prevention system and operation method thereof

Abstract

The invention discloses an animal attack prevention system and an operation method thereof. The animal attack prevention system comprises a microphone, a processor and a response subsystem. The microphone is configured to collect sound data from an animal. The processor is configured to cause the response subsystem to execute one of M pre-specified barring responses (i) and pre-specified neutral responses based on the sound data, where M is a positive integer, i = 1,..., M. For each value of i, the processor is configured to cause the response subsystem to execute a pre-specified barring response (i) that makes the animal feel uncomfortable in response to the sound data indicating that the attack by the animal is an attack category (i). The processor is configured to cause the response subsystem to perform a pre-specified neutral response that does not make the animal uncomfortable in response to the sound data indicating that the animal is not attacked.

Description

[Technical field]

[0001] The present invention relates to an animal attack prevention system and an operating method thereof. [Background technology]

[0002] There is a need for a system to stop pet aggression to ensure the safety of pet owners, nearby residents, and the pets themselves. Aggressive behavior from pets can lead to accidents and harm people around them. This system can help train and modify pet behavior to create a safe environment for pets and people. [Summary of the invention]

[0003] The present invention discloses an animal attack prevention system, which includes: a microphone, a processor, and a response subsystem. The microphone is configured to collect sound data from an animal. The processor is configured to enable the response subsystem to execute one of M pre-specified prevention responses (i) and pre-specified neutral responses based on the sound data, wherein M is a positive integer, i=1, ..., M. For each value of i, the processor is configured to enable the response subsystem to respond to the sound data indicating that the attack of the animal is an attack category (i) and execute the pre-specified prevention response (i) that makes the animal feel uncomfortable. The processor is configured to enable the response subsystem to respond to the sound data indicating that the animal is not attacking and execute the pre-specified neutral response that does not make the animal feel uncomfortable.

[0004] In one aspect, M=1 or M>1.

[0005] In one aspect, the animal attack deterrent system further comprises an attachment device configured to strap the microphone to the animal such that the microphone can pick up sounds made by the animal.

[0006] In one aspect, the attachment device is a collar configured to strap the microphone to the neck of the animal.

[0007] In one aspect, the response subsystem is connected to the microphone.

[0008] In one aspect, the response subsystem is not connected to the microphone.

[0009] In one aspect, the processor is connected to the microphone.

[0010] In one aspect, the processor is not connected to the microphone.

[0011] In one aspect, the processor implements a categorized artificial neural network having M+1 pre-specified output results corresponding to the M pre-specified stop responses (i) and the pre-specified neutral responses, respectively, so that the response subsystem executes the one of the M pre-specified stop responses (i) and the pre-specified neutral response based on the sound data.

[0012] In one aspect, the processor implements the categorized artificial neural network in the following manner: extracting sound samples from the sound data, each of which is the sound intensity of a certain sound frequency at a certain point in time; feeding the sound samples to the categorized artificial neural network to generate one of the M+1 pre-specified output results; and causing the response subsystem to execute the one of the M pre-specified suppression responses (i) and the pre-specified neutral responses that corresponds to the one of the M+1 pre-specified output results.

[0013] In one aspect, the animal is a cat, dog, or bird.

[0014] In one aspect, each of the M pre-specified restraining responses (i) includes the response subsystem generating a sound, voice, or ultrasound that causes discomfort to the animal, or the response subsystem releasing an electric shock that causes discomfort to the animal, or the response subsystem generating a sound, voice, or ultrasound that causes discomfort to the animal and releasing an electric shock that causes discomfort to the animal.

[0015] In one aspect, the pre-specified neutral response comprises the response subsystem taking no action with respect to the animal.

[0016] The present invention discloses an operating method of an animal attack prevention system, the operating method comprising: using a microphone to collect sound data from an animal; and the processor causing a response subsystem to execute one of M pre-specified prevention responses (i) and a pre-specified neutral response based on the sound data.

[0017] In one aspect, the operating method further comprises recording the one of the M pre-designated suppress responses (i) and the pre-designated neutral responses to keep track of the frequency with which the one of the M pre-designated suppress responses (i) and the pre-designated neutral responses are executed.

Brief Description of the Drawings

[0018] Figure 1 An animal attack deterrent system according to an embodiment is schematically illustrated.

[0019] Figure 2A flow chart outlining the operation of an animal attack deterrence system according to an embodiment is shown. [Specific implementation method]

[0020] Animal Attack Deterrence System

[0021] Figure 1 Schematically, an animal attack deterrent system 100 is shown according to an embodiment. In one embodiment, the animal attack deterrent system 100 may include a microphone 110, a processor 120, and a response subsystem 130.

[0022] In one embodiment, microphone 110 may collect sound data (e.g., sounds from the animal's mouth) from animal 190 and send the collected sound data to processor 120 for processing. In one embodiment, microphone 110 may be close to animal 190 to be able to pick up sounds made by animal 190. In one embodiment, animal 190 may be a pet, such as a cat, dog, or bird.

[0023] In one embodiment, processor 120 may receive and process sound data collected from animal 190 (collected by microphone 110) and then cause response subsystem 130 to perform one of M+1 pre-specified responses based on the collected sound data, where M is a positive integer.

[0024] Specifically, in one embodiment, the M+1 pre-specified responses may include (A) M pre-specified restraining responses (pre-specified restraining responses (i), i=1, ..., M) that make the animal 190 feel uncomfortable and (B) a pre-specified neutral response that does not make the animal 190 feel uncomfortable.

[0025] M pre-specified suppression responses

[0026] In one embodiment, for i=1, ..., M (i.e., for each value of i), the processor 120 may cause the response subsystem 130 to execute a pre-specified stop response (i) in response to the processor 120 finding that the collected sound data indicates that the attack by the animal 190 is of the attack category (i). In other words, there are M attack categories (i.e., attack categories (i), i=1, ..., M) corresponding to M pre-specified stop responses (i.e., pre-specified stop responses (i), i=1, ..., M), respectively.

[0027] For illustration, assume that the animal 190 is a dog and M=2. When the dog charges at a person, the dog makes a certain sound (hereinafter referred to as a charge sound). The microphone 110 collects this charge sound from the dog and sends the charge sound to the processor 120 for processing. Based on the charge sound, the processor 120 finds that the charge sound indicates that the dog's attack is an attack category (1) among the M attack categories, and accordingly causes the response subsystem 130 to execute a pre-specified stop response (1) among the M pre-specified stop responses (for example, the response subsystem 130 generates an ultrasonic wave that makes the dog feel uncomfortable), thereby preventing the dog from doing this (i.e., charging at a person).

[0028] In addition, when the dog bites a person, the dog will make a certain sound (hereinafter referred to as a biting sound). The microphone 110 collects the biting sound from the dog and sends the biting sound to the processor 120 for processing. Based on the biting sound, the processor 120 finds that the biting sound indicates that the dog's attack is an attack category (2) among the M attack categories, and accordingly causes the response subsystem 130 to execute a pre-specified stop response (2) among the M pre-specified stop responses (for example, the response subsystem 130 generates an electric shock that makes the dog feel uncomfortable), thereby preventing the dog from doing this behavior (i.e., biting a person).

[0029] In one embodiment, each of the M pre-specified restraining responses may include the response subsystem generating a sound, voice, or ultrasound that causes discomfort to the animal 190, or the response subsystem releasing an electric shock that causes discomfort to the animal 190, or the response subsystem generating a sound, voice, or ultrasound that causes discomfort to the animal 190 and releasing an electric shock that causes discomfort to the animal 190.

[0030] Pre-specified neutral response

[0031] In one embodiment, processor 120 may cause response subsystem 130 to execute a pre-specified neutral response in response to processor 120 finding that the collected acoustic data indicates that animal 190 is not attacking.

[0032] For example, when a dog sniffs a person, the dog will make a certain sound (hereinafter referred to as a sniffing sound). The microphone 110 collects the sniffing sound from the dog and sends the sniffing sound to the processor 120 for processing. Based on the sniffing sound, the processor 120 finds that the sniffing sound indicates that the dog is not attacking, and accordingly causes the response subsystem 130 to perform a pre-specified neutral response (for example, the response subsystem 130 does not take any action on the dog).

[0033] Method of operating the animal attack prevention system 100

[0034] In one embodiment, the operation of the animal attack deterrence system 100 may be as follows. The microphone 110 may collect sound data from the animal 190 and then send the collected sound data to the processor 120 for processing. Then, in one embodiment, the processor 120 may cause the response subsystem 130 to execute one of the M+1 pre-specified responses (i.e., the M pre-specified deterrence responses and the pre-specified neutral response) based on the collected sound data.

[0035] Flowchart outlining the operation of the Animal Attack Deterrence System

[0036] Figure 2 shows a summary according to an embodiment Figure 1 Flowchart 200 of the operation of the animal attack deterrence system 100.

[0037] In step 210, the operation may include using a microphone to collect sound data from the animal. For example, in the above embodiment, reference Figure 1 , microphone 110 collects sound data from animal 190.

[0038] In step 220, the operation may include the processor causing the response subsystem to execute one of M pre-specified suppression responses and pre-specified neutral responses based on the sound data. Figure 1 , the processor 120 causes the response subsystem 130 to execute one of M pre-specified suppression responses and pre-specified neutral responses based on the sound data.

[0039] Other embodiments

[0040] Attachment device

[0041] In one embodiment, reference Figure 1 , the animal attack deterrent system 100 may include a collar 150 configured to tie the microphone 110 around the neck of the animal 190. For example, the microphone 110 may be located on a printed circuit board (PCB) 140, and the collar 150 may be connected to the printed circuit board 140 (eg, Figure 1 Thus, the collar 150 can tie the microphone 110 around the neck of the animal 190.

[0042] Generally speaking, the animal attack deterrent system 100 may include an attachment device (not shown) configured to strap the microphone 110 to the animal 190 such that the microphone 110 can pick up sounds made by the animal 190 .

[0043] Responding to subsystem connected or not connected to microphone

[0044] In one embodiment, reference Figure 1 , the response subsystem 130 can be connected to the microphone 110. For example, the microphone 110 and the response subsystem 130 can be located on a printed circuit board 140 (eg, Figure 1 ). Therefore, the response subsystem 130 is connected to the microphone 110 via the printed circuit board 140. Alternatively, the response subsystem 130 may not be connected to the microphone 110.

[0045] Processor connected or not connected to microphone

[0046] In one embodiment, reference Figure 1 , the processor 120 may be connected to the microphone 110. For example, the microphone 110 and the processor 120 may be located on a printed circuit board 140 (eg, Figure 1 As shown in FIG. 1 , the processor 120 is connected to the microphone 110 via the printed circuit board 140. Alternatively, the processor 120 may not be connected to the microphone 110.

[0047] Artificial Neural Networks

[0048] In one embodiment, reference Figure 1 , the processor 120 can enable the response subsystem 130 to execute one of the M+1 pre-specified responses (i.e., M pre-specified stop responses and a pre-specified neutral response) based on the sound data collected from the animal 190 by implementing a categorized artificial neural network having M+1 pre-specified output results corresponding to the M+1 pre-specified responses, respectively.

[0049] For example, assume that the animal 190 is a dog. To train a dog using the animal attack prevention system 100, two attack categories (i.e., M=2) may be assigned to the dog's behavior: (A) the first attack category may include less aggressive behaviors, such as barking and approaching a person, and (B) the second attack category may include more aggressive behaviors, such as biting and scratching a person. In addition to the behaviors in the first and second attack categories, the dog's other behaviors may be considered non-aggressive (i.e., belonging to the non-aggression category).

[0050] In addition, the response subsystem 130 may be assigned (A) a first pre-specified restraining response (e.g., the response subsystem 130 generates ultrasonic waves that cause discomfort to the dog), (B) a second pre-specified restraining response (e.g., the response subsystem 130 releases an electric shock that causes discomfort to the dog), and (C) a pre-specified neutral response (e.g., the response subsystem 130 does nothing with respect to the dog) in response to the dog's behavior being considered (i.e., determined by the processor 120 based on the collected sound data) to belong to (A) the first attack category, (B) the second attack category, and (C) the non-aggression category, respectively.

[0051] Therefore, a classification artificial neural network can be created, and the three output results 1, 2 and 3 correspond to the first attack category, the second attack category and the non-attack category respectively. Therefore, the three output results 1, 2 and 3 correspond to the first pre-specified suppression response, the second pre-specified suppression response and the pre-specified neutral response respectively.

[0052] Next, in one embodiment, during the training phase, the classification artificial neural network can be trained using sound samples emitted by animals of the same species as the animal 190 (for example, if the animal 190 is a dog, the classification artificial neural network can be trained using sound samples emitted by the dog).

[0053] In one embodiment, after the classification artificial neural network is trained, in the prediction stage, the sound samples from the sound data collected from the dog by the microphone 110 can be fed into the trained classification artificial neural network to generate output results 1, 2 or 3.

[0054] If the output result is 1, the processor 120 causes the response subsystem 130 to execute a first pre-specified restraining response (ie, the response subsystem 130 generates ultrasonic waves that cause discomfort to the dog).

[0055] If the output result is 2, the processor 120 causes the response subsystem 130 to execute a second pre-specified restraining response (ie, the response subsystem 130 delivers an electric shock that causes discomfort to the dog).

[0056] If the output result is 3, the processor 120 causes the response subsystem 130 to execute a pre-specified neutral response (ie, the response subsystem 130 does nothing with respect to the dog).

[0057] In one embodiment, each sound sample used in the training phase or the prediction phase may be the sound intensity of a certain sound frequency at a certain time point. Alternatively, each sound sample used in the training phase or the prediction phase may include multiple sound intensities of certain frequencies at certain time points.

[0058] Record keeping

[0059] In one embodiment, reference Figure 1 , the animal attack deterrence system 100 may record the frequency (ie, the number of times within a period of time) at which the response subsystem 130 executes each of the M+1 pre-specified responses (ie, the M pre-specified deterrence responses and the pre-specified neutral response).

[0060] Although various aspects and embodiments have been disclosed herein, other aspects and embodiments will be apparent to those skilled in the art. The various aspects and embodiments disclosed herein are for illustrative purposes and are not intended to be limiting, with the true scope and spirit being indicated by the appended claims.

Claims

1. An animal attack prevention system, characterized in that: include: microphone; processor; as well as Response subsystem, wherein the microphone is configured to collect sound data from the animal, wherein the processor is configured to cause the response subsystem to execute one of M pre-specified suppression responses (i) and pre-specified neutral responses based on the sound data, wherein M is a positive integer, i=1, ..., M, wherein, for each value of i, the processor is configured to cause the response subsystem to execute the pre-specified restraining response (i) that causes discomfort to the animal in response to the sound data indicating that the aggression by the animal is of aggression category (i), and Wherein, the processor is configured to cause the response subsystem to execute the pre-specified neutral response that does not cause discomfort to the animal in response to the sound data indicating that the animal is not attacking.

2. The animal attack prevention system according to claim 1, characterized in that: M＝1。 3. The animal attack prevention system according to claim 1, characterized in that: M>1。 4. The animal attack prevention system according to claim 1, characterized in that: Also included is an attachment device configured to strap the microphone to the animal so that the microphone can pick up sounds made by the animal.

5. The animal attack prevention system according to claim 4, characterized in that: The attachment device is a collar configured to strap the microphone to the neck of the animal.

6. The animal attack prevention system according to claim 1, characterized in that: The response subsystem is connected to the microphone.

7. The animal attack prevention system according to claim 1, characterized in that: The response subsystem is not connected to the microphone.

8. The animal attack prevention system according to claim 1, characterized in that: The processor is connected to the microphone.

9. The animal attack prevention system according to claim 1, characterized in that: The processor is not connected to the microphone.

10. The animal attack prevention system according to claim 1, characterized in that: The processor implements a classified artificial neural network having M+1 pre-specified output results corresponding to the M pre-specified stop responses (i) and the pre-specified neutral responses, respectively, so that the response subsystem executes the one of the M pre-specified stop responses (i) and the pre-specified neutral response based on the sound data.

11. The animal attack prevention system according to claim 10, characterized in that: The processor implements the classification artificial neural network in the following manner: Extracting sound samples from the sound data, each of the sound samples being the sound intensity of a certain sound frequency at a certain time point; Feeding the sound sample to the categorized artificial neural network to generate one of the M+1 pre-specified output results; as well as The response subsystem is caused to execute the one of the M pre-specified suppression responses (i) and the pre-specified neutral response that corresponds to the one of the M+1 pre-specified output results.

12. The animal attack prevention system according to claim 1, characterized in that: The animal is a cat, dog or bird.

13. The animal attack prevention system according to claim 1, characterized in that: Each of the M pre-specified restraining responses (i) includes the response subsystem generating a sound, voice or ultrasound that causes discomfort to the animal, or the response subsystem releasing an electric shock that causes discomfort to the animal, or the response subsystem generating a sound, voice or ultrasound that causes discomfort to the animal and releasing an electric shock that causes discomfort to the animal.

14. The animal attack prevention system according to claim 1, characterized in that: The pre-specified neutral response comprises the response subsystem taking no action with respect to the animal.

15. A method for operating an animal attack prevention system, characterized in that: The operation method comprises: Using microphones, collect sound data from animals; and Using a processor, the response subsystem is caused to execute one of M pre-specified suppression responses (i) and pre-specified neutral responses based on the sound data, where M is a positive integer.

16. The operating method according to claim 15, characterized in that: Also included is recording the one of the M pre-designated suppress responses (i) and the pre-designated neutral responses to keep track of how often the one of the M pre-designated suppress responses (i) and the pre-designated neutral responses are executed.

Images