Electronic and software-based autonomous Anti-attacking and Anti-mating collar for stray dogs or owned / unowned dogs

Abstract

The present invention relates to a smart dog collar that is worn on dogs to prevent aggressive stray dogs or owned / unowned dogs from attacking people individually or in packs, has autonomous control capability, detects the dog's aggression or mating activity using the sensors and artificial intelligence software thereon, removes the aggressive dog from people without harming the dog, and prevents dogs from reproducing by mating uncontrollably.

Description

[0001] DESCRIPTION

[0002] ELECTRONIC AND SOFTWARE-BASED AUTONOMOUS ANTI-ATTACKING AND ANTI-MATING COLLAR FOR STRAY DOGS OR OWNED / UNOWNED DOGS

[0003] Field of the Invention

[0004] The present invention relates to a smart dog collar that is worn on dogs to prevent aggressive stray dogs or owned / unowned dogs from attacking people individually or in packs, has autonomous control capability, detects the dog's aggression or mating activity using the sensors and artificial intelligence software thereon, removes the aggressive dog from people without harming the dog, and prevents dogs from reproducing by mating uncontrollably. In the following part of the description, the word stray dogs will be used for stray dogs or owned / unowned dogs.

[0005] State of the Art

[0006] Although stray dogs are beneficial for cities, due to their hunting nature, they can cause undesirable situations in their interactions with people. Stray dogs can sometimes display an aggressive attitude due to their nature or sometimes because they are provoked by people, and they can harm the people they encounter. People use various dog repellent devices and tools to protect themselves from these aggressive dogs.

[0007] Today, there are dog repellent devices / apparatus that prevent aggressive dogs from attacking humans, and human-controlled devices or apparatus that do not have autonomous features. People who encounter stray dogs, without knowing for sure whether the dog will attack them or not, hold dog repellent devices or apparatus towards the dogs and drive them away before they attack them. However, all these existing dog repellent devices / apparatuses are designed to hurt the dog without harming the dog.

[0008] People carry existing dog repellent tools / devices on themselves, and when necessary, they take them out of their bags and point them at the dog. Since dog attacks often occur suddenly, it may be too late to intervene with such dog repellent devices or apparatus. In the state of the art, when dogs mate, physical intervention is required to prevent this mating. Efforts are made to prevent dogs of the opposite sex from coming together so that dogs do not mate. When a mating dog is seen, the dog is either physically intervened directly or an attempt is made to prevent them from mating with dog repellent tools / devices. With the devices / apparatus available in the state of the art, uncontrolled mating of dogs cannot be sufficiently prevented, and it is too late to prevent mating. Such apparatus / devices are insufficient to prevent stray dogs from attacking humans or to prevent uncontrolled mating of these dogs.

[0009] Object of the Invention

[0010] The object of the present invention is to develop a smart collar with an artificial intelligence algorithm that prevents stray dogs from attacking humans and prevents dogs from mating uncontrollably by listening to the barking of the stray dog, listening to the sound values around the dog and following the dog's pulse rate.

[0011] The smart collar of our invention achieves this purpose by using the autonomous artificial intelligence system within it. Our smart collar first understands that the dog will attack, or the dog will mate uncontrollably, then it emits ultrasonic sound waves towards the dogs to remove the attacking dog or the dog that will mate, and then applies electroshock to remove the attacking or mating dog. Thus, it prevents unnecessary intervention to dogs by distinguishing between aggressive dogs and non-aggressive dogs, or by distinguishing between mating and non-mating dogs.

[0012] The most important feature of the collar of our invention, is that it decides to intervene on its own with the artificial intelligence program thereon, removes the aggressive dog from the environment without harming the dog spontaneously without attacking the human, or intervenes in the uncontrolled mating dog and removes it from the environment without mating.

[0013] In order to fulfill the above-mentioned purposes with our collar of the present invention; the collar includes

[0014] • at least one sound-detecting microphone that allows collection of sounds from the dog's environment,

[0015] • a processor to perform the identification operations of the audio data collected by the said audio sensor, • software algorithm that converts the sound data collected from the aforementioned processor and sound-sensing microphone into a 2-dimensional mathematical data format after first filtering it and then classifies it using convolutional neural networks,

[0016] • Piezo, which is controlled by artificial intelligence according to the environmental status information mentioned and emits ultrasonic sound waves in the 20-25 KHz sound range that scares dogs away,

[0017] • at least one high voltage transmission unit that provides the transmission of non- lethal electrical current to calm an overly aggressive dog.

[0018] In addition, in order to prevent uncontrolled mating with our inventive collar, with the sensor and artificial intelligence algorithm on the collar, the mating activities of the dogs are detected and electroshock and ultrasonic sound are given to the mating dogs for prevention. Our collar of the present invention includes;

[0019] • tilt and pulse sensors on the smart collar worn around the dog's neck

[0020] • artificial intelligence algorithm that detects the body movement of male dogs during mating after processing the data collected from this tilt and pulse sensor.

[0021] If the collar subject to our invention is found on both a male and a female dog, the artificial intelligence algorithm calculates the distance between the two collars and compares this calculated data with the dog's body behavior data to decide whether the dog has mated or not. When artificial intelligence detects dogs' mating activity, dogs are prevented from mating by activating the electroshock transmitter and ultrasonic sound generator on the collar.

[0022] The structural and characteristic features of the present invention will be understood clearly by the following drawings and the detailed description made with reference to these drawings and therefore the evaluation shall be made by taking these figures and the detailed description into consideration.

[0023] Description of the Figures:

[0024] Figure 1: Side exterior detail view of our smart collar of the present invention

[0025] Figure 2: Front exterior detail view of our smart collar of the present invention

[0026] Figure 3: Side exterior detail view of our smart collar of the present invention Figure 4: Inside exterior detail view of the parts forming the invention attached to the bite-protected collar of our smart collar of the present invention

[0027] Figure 5: Outside exterior detail view of the parts forming the invention attached to the bite-protected collar of our smart collar of the present invention

[0028] Figure 6: Inside interior detail view of the parts forming the invention attached to the bite- protected collar of our smart collar of the present invention

[0029] Figure 7: Artificial intelligence flow chart that enables the detection and prevention of attacking by our smart collar of the present invention

[0030] Figure 8: Artificial intelligence flow chart that enables the detection and prevention of mating by our smart collar of the present invention

[0031] Figure 9: Flowchart of the invention showing matrix formation in attacking mode

[0032] Figure 10: Flowchart of the invention showing matrix formation in mating mode

[0033] Description of Parts Indicated by Reference Numbers in the Figures

[0034] 1. Sound Sensor (Microphone)

[0035] 1.1 Barking sound and human voice detecting microphone

[0036] 1.2 Pulse detector microphone

[0037] 2. Processor Card

[0038] 3. Ultrasonic Sound Generator (Speaker (Piezo))

[0039] 4. Artificial Intelligence

[0040] 4.1 Acoustic Categorizer Software (HNR)

[0041] 4.2 Convolutional Neural Network (CNN)

[0042] 5. Electric Shock Unit

[0043] 6. Solar Panel

[0044] 7. Bite Protection Collar

[0045] 8. Tilt sensor

[0046] 9. Acceleration sensor

[0047] 10. Proximity sensor

[0048] 11. Speaker

[0049] 12. Shock Pins

[0050] 13. Battery

[0051] Description of the Invention The present invention is an artificial intelligence-supported smart dog control collar developed to prevent attacking or uncontrolled mating of stray dogs.

[0052] Our invention is an autonomous electromechanical artificial intelligence-supported smart dog control collar that is worn on dogs, automatically manages itself with the artificial intelligence (4) thereon, and removes dogs at 2 different levels by emitting ultrasonic sound waves and applying electroshock in cases of aggression and mating.

[0053] With the smart collar of our present invention, it is determined whether the dog will attack the human or not by analyzing the barking sound of the dog that will attack and the human voice around the barking dog, together with the pulse rate of the dog at the moment of barking, with an artificial intelligence (Figure- 1). Among the sounds around the dog; “Help”, “Help”, “Shoo”, “Go away”, “Shhh”, “Don't”, “Stop”, “Mom”, “Dad”, “Ah”, “Ay”, when human sounds are detected, KET Ultrasonic Repeller and electroshock switches to active mode. When it is understood that the dog will attack a human, in attacker mode, a command is sent to the ultrasonic sound generator (3) and the electric shock generator (5) to apply ultrasonic sound and electric shock together to the attacking dog. In our invention, in addition to electroshock application, a more effective deterrent is achieved by applying magneto-based electroshock to stray dogs with a high pain threshold. With this application, the dog moves away from its environment without attacking humans. When our smart collar understands that the dog is not going to attack, it remains in listening mode and does not intervene. Thus, with the collar of our invention, it is possible to both intervene instantly with the dog that will attack and prevent the dog from harming people by removing the dog from the person it attacks and prevent people from intervening by being afraid that the dog will attack the dog that will not attack.

[0054] In the collar of the present invention; there is a pulse sound sensor microphone (1.2) placed on the dog to detect the heartbeat sound of the dog when encountering a human, and a bark and human sound sensor microphone (1.1) that detects the sound of barking during the dog's barking and the sound of people around.

[0055] The barking sound, human voice and pulse sound detected by the microphones (1.1, 1.2) on our collar are analyzed by an artificial intelligence software (4) and the probability of the dog attacking is calculated. In the artificial intelligence software (4) found in the collar of the present invention, trained models and deep neural networks are used to detect dog barking. By working with 1000 barks recorded from 100 dogs of different breeds, ages and sexes, reference values were obtained against which the artificial intelligence can compare whether the dog will attack or not, or whether the dog will mate uncontrollably.

[0056] The collar of the present invention has Acoustic Categorizer Software (HNR) (4.1) and Convolutional Neural Network (CNN) (4.2) used in artificial intelligence (4), which determines the attacking dog from the sound of barking, human voices in the environment and pulse rate at the time of barking and intervenes accordingly.

[0057] Another feature of the collar of the present invention is that; it is possible to determine whether a dog will mate with another dog or not, and if mating is the case, intervention can be made to prevent this mating (Figure-2). The collar of the present invention has a tilt sensor (8) that measures the angle of tilt of the dog relative to the horizontal and vertical reference at the time of mating, an acceleration sensor (9) that measures the acceleration caused by the movement of the dog at the time of mating, and a proximity sensor (10) that measures the proximity of the dog to each other due to the desire to mate. After detecting the tilt, acceleration and proximity values of the dog at the time of mating with the collar of the present invention, they are directly analyzed by the Convolutional Neural Network (CNN) (4.2) in the artificial intelligence (4). Convolutional Neural Network (CNN) decides whether the dog has mated or not by comparing it with the tilt, acceleration and proximity values previously determined in artificial intelligence, when it decides that mating will occur, it prevents uncontrolled mating by allowing the dogs to separate from each other by applying the ultrasonic sound provided by the ultrasonic sound generator (3) or the electroshock obtained from the electroshock generator (5) in the mating mode. If the dogs that come close to each other are not in mating mode, then it remains in listening mode and no intervention is made to the dogs.

[0058] When a barking dog is encountered, the barking sound and human sounds in the environment are detected and collected with the microphone on the collar. The sounds collected by this microphone are categorized by the Acoustic Categorizer Software (HNR) and digitized as frequency, amplitude, tone and rhythm.

[0059] When a dog with our collar of the present invention encounters a human, it categorizes the dog's condition into three classes: calm, aggressive and attacking, after categorizing the barking, the human voice around it and the pulse sound. When the dog is in a calm or aggressive state, no intervention is made, but when the dog is in an attacking state, it activates the ultra-shock sound generator (3) and the electroshock generator (5) to prevent the dog's aggression, removing the dog from its environment and preventing it from harming people.

[0060] In preventing aggression with our smart collar of the present invention, the sound data collection required for artificial intelligence is performed with the Acoustic Categorizer Software (HNR) as follows:

[0061] Dog barking sounds from different breeds, ages, sizes and environments were created from shelters and using open data sets. These collected sounds were labeled as aggressive or non-aggressive. This labeled data was converted into an artificial intelligence model and loaded onto the processor card on the smart collar as "Sound Matrix".

[0062] When sound is transmitted to the system from the speaker on the smart collar, data is pre- processed and the relevant data is made suitable for deep learning analysis.

[0063] • Sampling Rate: All sound files are brought to a standard sampling rate,

[0064] • Duration: Audio clips are trimmed to a consistent length,

[0065] • Normalization: Sound levels are equalized.

[0066] • Noise Reduction: background noise is filtered out

[0067] Spectrograms: Audio signals are converted into spectrogram images that show the distribution of frequencies over time.

[0068] Using a Convolutional Neural Network (CNN) deep learning method, which is ideal for processing data such as spectrograms, the preprocessed audio data is compared to the audio model containing dog barks to determine in real time whether the barking dog is Aggressive or Non-Aggressive.

[0069] In the collar of the present invention, a CNN matrix is created with the digitalized barking sound and the data in the environment (Figure-3). By comparing the CNN matrix with the barking matrices previously taught to the artificial intelligence on the collar, the probability that the barking is in an aggressive mood is calculated. The artificial intelligence in the collar of the present invention decides whether the dog will be aggressive or not based on this probability. When the collar of the present invention is active on the dog, it automatically prevents dogs from attacking humans by emitting ultrasonic sound waves between 21 and 25 KHz at 360 degrees, allowing dogs to move away from the environment.

[0070] In our invention, there is an aggression detection and prevention management algorithm that calculates the probability of a dog attacking and makes a decision accordingly by using an artificial intelligence-supported dog barking sounds from different breeds, ages, sizes and environments by recording dog barking sounds from shelters and open audio data sets. This algorithm is characterized by; o receiving real-time ambient sound of dog barking, human voice and dog pulse beats through a microphone, o digitization of these sounds through Acoustic Categorizer Software (HNR) by normalization and reduction of background noise and categorization (determining which sound belongs to whom and to what), and as a result, determining frequency, amplitude, tone and rhythm values and creating a sound matrix (SMx) based on these four data, o determining which of these sound matrices (SMx) correspond to the dog's resting, aggressive and aggression state, o determining whether the dog is in a resting, aggressive or non-aggressive state by comparing a sound matrix (SMx) obtained from sounds taken from the environment with the values in the sound data matrix taught through the Convolutional Neural Network (CNN) (4.2), o not giving any warning as a result of determining that the dog is resting or aggressive, o giving ultrasonic sound to the dog through a loudspeaker and at the same time, removing dog from the environment and eliminating its aggression by giving an electric shock as a result of determining that the dog is aggressive.

[0071] The present invention comprises an uncontrolled mating detection and management algorithm that calculates the probability of mating of stray dogs and makes a decision accordingly by using the sound data matrix of aggressive dogs by recording dog barking sounds from shelters and open sound data sets from different breeds, ages, sizes and environments in advance using the taught vibration data matrix consisting of the proximity distance data of male and female dogs at the time of mating, the motion acceleration values of the collars attached to the dogs at the time of mating, and the tilt angle data according to the horizontal and vertical position of the dogs at the time of mating. This algorithm is characterized by; o measuring the tilt value by measuring the angle of tilt with respect to the horizontal and vertical position of the dog at the time of mating with an tilt sensor (8), an acceleration value with an acceleration sensor (9) due to the movement of the dog at the time of mating, and a proximity value with a proximity sensor (10) due to the dog's desire to mate, o creating a vibration matrix (TMx) from real-time measured tilt, acceleration and proximity values at the time of dog mating, o creating a taught vibration data matrix from vibration matrices (TMx) created with Convolutional Neural Network (CNN) and determining which values in the matrix will correspond to the mating mode, o determining the mating status of male and female dogs by comparing the vibration matrix (TMx) obtained from the dog with the values in the taught vibration data matrix using Convolutional Neural Network (CNN), o not giving any warning as a result of determining that the dog is at rest, o preventing the dog from mating by giving an electric shock to the female and male dog through a loudspeaker by giving Ultrasonic Sound as a result of determining that the dog is in a mating situation.

[0072] When the collar of the present invention detects that the dog will attack or mate, the ultrasonic sound generator (3) on the collar emits a sound tone of 21-25 Khz (inaudible to humans) at 360 degrees through 2 speakers (11) on the collar until the dog stops barking. This sound frequency is a tone that dogs do not like and causes them to move away from their environment. However, when ultrasonic sound is constantly applied to the dog at the same frequency, dogs may develop immunity to this sound frequency and may not be affected by this ultrasonic sound. Ultrasonic sound varying between 21-25 KHz frequency is applied to prevent the dog's immune mechanism to this determined fixed frequency.

[0073] When the collar of the present invention detects that the dog is about to attack or mate, it applies an electroshock to the dog with the Electro Shock generator (5) thereon, along with an ultrasonic sound broadcast. When the electroshock generator is active (ESA), it applies an electroshock to the dog on which the collar is attached, with a voltage range of 80 mA and 70000-100000 V from the shocker nipples on the collar. Artificial intelligence decides how much and when to apply electroshock to the attacking dog. Electroshocks continue to be given as long as the dog's aggression continues. Electroshock is applied in an amount and duration that is painful and non-lethal to the dog.

[0074] The collar of the present invention is designed such that dogs cannot bite or tear it off. Dogs may bite and damage the collar while fighting among themselves. To prevent damage to the collar, hardened plastic, iron fibers and barbs that prevent biting during biting are used.

[0075] The collar of the present invention is designed to be self-charged by means of the solar panels (6) on the collar. There is a solar collector (6) on the collar so that it can charge itself. The energy obtained from the solar collector is continuously transferred to the battery (6) charge control unit, and the required energy is provided by the solar collectors. Once the collar is energized, it charges itself autonomously without human supervision.

[0076] The sensors and artificial intelligence algorithm thereon detect the mating activities of dogs as follows and intervene:

[0077] • It uses an artificial intelligence algorithm to detect the body movement of male dogs during mating using data from the tilt and pulse sensors on the smart collar worn around the dog's neck.

[0078] • If both male and female dogs have a collar, it also calculates the distance between the two collars and compares this data with dog body behavior data.

[0079] • When artificial intelligence detects the mating activity of dogs, it prevents dogs from mating by activating the electroshock transmitter and ultrasonic dog repellent units on the collar.

[0080] The collar of the present invention, is a smart collar that generates its own energy through the solar panels thereon without the need for maintenance and charging in the form of attach / forget after it is attached to stray dogs as a collar, and can automatically detect the aggressive behavior of dogs through artificial intelligence algorithms and sensors on the processor and remove single and pack dogs.

[0081] In the collar of the present invention; outdoor sounds collected through the microphones on the smart collar are analyzed, dog barking analysis, human voice analysis and data merging process are subjected to data merging process through the processor and software algorithms on the collar, as a result of the data analysis, the dog's aggression status from the barking sound and the call for help status from the human voice are analyzed. The system also includes electro-shock systems that will put an end to the dog's aggressive behavior without permanent damage to the dog based on the sound data collected and analyzed by the sound sensor (1) microphone, and ultrasonic dog repellent sound wave emitting systems that will remove stray dogs in the vicinity.

[0082] In the invention, people are protected from dog attacks in all weather conditions, by automatically analyzing the aggression of single or herding dogs from the sounds of single or herding dogs and giving dog repellent ultrasonic sound and electroshock through the collar attached to the dogs.

[0083] The most important point where our invention differs from the techniques in the art and brings innovation is the performance of all these functions in a collar that can be carried on the dog by combining voice recognition technology, which is the technology of our age, and convolutional neural networks in a system software algorithm (4). The smart collar thus automatically prevents the stray dogs to which it is attached from attacking people.

Claims

CLAIMS1. An autonomous electromechanical artificial intelligence-supported smart dog control collar that is worn on dogs, automatically manages itself with the artificial intelligence (4) thereon, and moves dogs away in case of attacking by emitting ultrasonic sound waves and applying electroshock, characterized by comprising; microphone (1), which detects the barking sound of the dog and the frequency, amplitude, tone and rhythm of the human voice around the dog and transfers the same to the artificial intelligence, microprocessor (2), which creates the electronic infrastructure that operates and controls the mechanism of our smart collar and runs the artificial intelligence software, rechargeable battery (13), which provides the energy of the smart collar, ultrasonic sound generator (3), which creates the ultrasonic sound to be applied to the dog in case the dog's aggression is detected, electric shock unit (5), which transmits the power received from the battery on the collar by the voltage amplifier through the electrodes contacting the dog's neck and shocks the dog in the event that the dog's aggression is detected, microphone (1.2) on the smart collar that detects environmental sounds such as barking and human voices, pulse detector microphone (1.2) on the smart collar, which detects the pulse sound, artificial intelligence algorithm (4), which detects whether the dog will attack or not from the real-time barking sound signal, real-time pulse sound signal and the real-time sound signal of the people around the dog, and in case of attack, emits ultrasonic sound to the dog and applies electric shock to the dog to make it move away.

2. Aggression detection and prevention management algorithm, which calculates the probability of dog attack and makes a decision accordingly by using the sound data matrix of the dog barking sounds from different breeds, ages, sizes and environments, which are used in the artificial intelligence supported smart dog control collar of the present invention, by recording dog barking sounds from shelters and open sound data sets andusing the sound data matrix of the aggressive dogs, according to any one of the preceding claims, characterized by; o receiving real-time ambient sound of dog barking, human voice and dog pulse beats through a microphone, o digitizing these sounds through Acoustic Categorizer Software (HNR) by normalization and reduction of background noise, categorization and digitization, and as a result, determining frequency, amplitude, tone and rhythm values and creation of a sound matrix (SMx) based on these four data, o determining which of these sound matrices (SMx) correspond to the dog's resting, aggressive and attacking state, o determining whether the dog is in a resting, aggressive or non-aggressive state by comparing a sound matrix (SMx) obtained from sounds taken from the environment with the values in the sound data matrix taught through the Convolutional Neural Network (CNN) (4.2), o not giving any warning as a result of determining that the dog is resting or aggressive, o giving ultrasonic sound to the dog through a speaker and at the same time, moving the dog away from the environment and eliminating its aggression by giving an electric shock as a result of determining that the dog is attacking.

3. Autonomous electromechanical artificial intelligence-supported smart dog control collar that is worn on dogs, automatically manages itself with the artificial intelligence (4) thereon, and moves the dogs away in case of mating by emitting ultrasonic sound waves and applying electroshock, characterized by comprising; tilt sensor (8), which measures the dog's tilt angle relative to the horizontal and vertical reference where the dog is at the time of mating, acceleration sensor (9), which measures the acceleration caused by the movement at the moment of mating, proximity sensor (10), which measures the dogs' proximity to each other due to their desire to mate, microprocessor (2), which creates the electronic infrastructure that operates and controls the mechanism of our smart collar and runs the artificial intelligence software,rechargeable battery (13), which provides the energy of the smart collar, Ultrasonic Sound Generator (3), which creates the ultrasonic sound to be applied to the dog if the dog's mating activity is detected,Electric Shock Unit (5) that transmits the power received from the battery on the collar by the voltage amplifier through the electrodes contacting the dog's neck and shocks the dog in the event that the dog's aggression is detected,Artificial intelligence algorithm that autonomously detects whether the dog is mating with the opposite sex in the mating position of the dog during the mating of the dog with the opposite sex, from the dog's tilt, acceleration due to movement and proximity to the opposite sex, and prevents the dog from mating by emitting ultrasonic sound and applying electric shock to the dog in case of mating (4).

4. Uncontrolled mating detection and management algorithm according to claim 3, which calculates the probability of mating of stray dogs and makes a decision accordingly by using the sound data matrix of aggressive dogs by recording dog barking sounds from shelters and open sound data sets from different breeds, ages, sizes and environments in advance using the taught vibration data matrix consisting of the proximity distance data of male and female dogs at the time of mating, the motion acceleration values of the collars attached to the dogs at the time of mating, and the tilt angle data according to the horizontal and vertical position of the dogs at the time of mating, characterized by comprising; o measuring the tilt value by measuring the angle of tilt with respect to the horizontal and vertical position of the dog at the time of mating with an tilt sensor (8), an acceleration value with an acceleration sensor (9) due to the movement of the dog at the time of mating, and a proximity value with a proximity sensor (10) due to the dog's desire to mate, o creating a vibration matrix (TMx) from real-time measured tilt, acceleration and proximity values at the time of dog mating, o creating a taught vibration data matrix from vibration matrices (TMx) created with Convolutional Neural Network (CNN) and determining which values in the matrix will correspond to the mating mode, o determining the mating status of male and female dogs by comparing the vibration matrix (TMx) obtained from the dog with the values in the taught vibration data matrix using Convolutional Neural Network (CNN),o not giving any warning as a result of determining that the dog is at rest, o preventing the dog from mating by giving an electric shock to the female and male dog through a loudspeaker by giving Ultrasonic Sound as a result of determining that the dog is in a mating situation.

5. Autonomous electromechanical artificial intelligence-supported smart dog control collar that is worn on the neck of the dogs, automatically manages itself with the artificial intelligence (4) thereon, and moves the dogs away in case of attacking and mating situation by emitting ultrasonic sound waves and applying electroshock, characterized by comprising; microphone (1), which detects the barking sound of the dog and the frequency, amplitude, tone and rhythm of the human voice around the dog and transfers the same to the artificial intelligence, microprocessor (2), which creates the electronic infrastructure that operates and controls the mechanism of our smart collar and runs the artificial intelligence software, tilt sensor (8), which measures the dog's tilt angle relative to the horizontal and vertical reference where the dog is at the time of mating, acceleration sensor (9), which measures the acceleration caused by the movement at the moment of mating, proximity sensor (10), which measures the dogs' proximity to each other due to their desire to mate, rechargeable battery (13) that provides the energy of the smart collar, ultrasonic sound generator (3), which creates the ultrasonic sound to be applied to the dog in case the dog's aggression is detected,Electric Shock Unit (5) that transmits the power received from the battery on the collar by the voltage amplifier through the electrodes contacting the dog's neck and shocks the dog in the event that the dog's aggression is detected and shock transfer pins (12), which enable this shock to be transferred to the dog, microphone (1.2) on the smart collar that detects environmental sounds such as barking and human voices, pulse detector microphone (1.2) on the smart collar, which detects the pulse sound, artificial intelligence algorithm (4) that autonomously detects whether the dog is attacking by detecting whether the dog is attacking or not from the real-timebarking sound signal, real-time pulse sound signal and real-time sound signal of people around the dog, or during the mating of the dog, from the dog's tilt in the mating position of the dog with the opposite sex, from the acceleration caused by the movement of the dog and the proximity to the opposite sex, emits ultrasonic sound to the dog in one of the attacking or mating situations and applies an electric shock to the dog, causing the dog to move away from the situation when attacking or mating.

6. Artificial intelligence supported smart dog control collar according to any one of the claims 1, 3 or 5, characterized by comprising a magneto-based electro-shock unit that applies an electro-shock.

7. Artificial intelligence supported smart dog control collar according to any one of the claims 1, 3 or 5, characterized by comprising Bite Protection Collar (7) with hardened plastic, iron fibers and bite-preventing barbs thereon so that the dog will not be harmed by the dog's bite while the dog is struggling.

Images