KR20230090422A - Risk prevention system using artificial intelligence - Google Patents

Abstract

According to the risk prevention system using artificial intelligence of the present invention, an information collection unit for acquiring motion information of a user in which movement occurs, recognizing motion from the motion information obtained from the information collection unit, and learning during a preset learning period It includes a motion recognition unit for recognizing motion and a current motion after the learning period, and an artificial intelligence control unit for learning the learning motion through an artificial intelligence algorithm and comparing the learning motion with the current motion to determine whether a dangerous situation exists.
The effect of the present invention is to detect and learn the user's motion, and if an abnormality is found in the normal motion according to the learned result, an alarm is generated in the user's terminal so that the user recognizes the abnormality, and the alarm is not turned off. Otherwise, it is possible to provide a risk prevention system using artificial intelligence that automatically contacts a pre-set third party.

Description

Risk prevention system using artificial intelligence

The present invention relates to a risk prevention system using artificial intelligence.

In the current era when the number of one-person households in Korea is over 660,000, the proportion of sudden and lonely deaths is increasing, emerging as a social problem. The National Statistical Office predicted that by 2047, the number of single-person households would exceed 4 million, accounting for 20% of all households.

When shock or abnormalities occur due to sudden health problems or external shocks of single-person households or people with underlying diseases, there are cases where the golden time is missed due to failure to respond quickly. This is very likely to result in immediate death.

In order to prevent this problem, a manager regularly visits and checks, or a photographing device for photographing the user is installed in the user's household, and the manager monitors the photographed image of the photographing device at the central center.

In this way, although users are managed by monitoring visits or captured images, many difficulties and difficulties have occurred as one manager manages a plurality of users.

In particular, when monitoring is performed through a recording device, as the recording device (IP camera) is connected to the Internet, there is a risk that privacy may be exposed when an IP address is hacked.

Korean Registered Patent No. 10-2134771 Korean Registered Patent No. 10-2303633

An object of the present invention, which has been devised to solve the above problems, is to detect a user's motion, learn, and generate an alarm in the user's terminal when an abnormality is found in the normal motion according to the learned result. An object of the present invention is to provide a risk prevention system using artificial intelligence that allows a user to recognize an abnormality and to automatically contact a pre-set third party if the alarm is not turned off.

In addition, an object of the present invention is to learn and analyze the motion information captured by the information collection unit and prevent it from being transmitted to the outside, thereby providing a risk prevention system using artificial intelligence that is stable from invading a user's personal privacy.

According to the risk prevention system using artificial intelligence of the present invention for achieving the above object, an information collection unit for obtaining motion information of a user in which movement occurs, and recognizing motion from the motion information obtained from the information collection unit However, a motion recognition unit for recognizing the learning motion during a preset learning period and the current motion after the learning period and learning the learning motion through an artificial intelligence algorithm, and comparing the learning motion and the current motion to determine whether a dangerous situation exists A risk prevention system using artificial intelligence including an artificial intelligence control unit that judges.

In addition, the artificial intelligence controller may include a time measuring unit for measuring a first holding time for maintaining the learning motion and a second holding time for maintaining the current motion, and learning the learning motion and the first holding time for the learning motion. When the current motion is recognized, an analysis unit that compares and analyzes the second holding time of the current motion with the first holding time of the learning motion corresponding to the current motion, and the second holding time of the current motion. When the first holding time of the learning motion is exceeded, a situation determining unit for determining a dangerous situation is included.

In addition, a first communication unit that is worn on the user and measures the heart rate of the user, an alarm unit that generates an alarm, and receives a control signal from the artificial intelligence controller and transmits the heart rate to the artificial intelligence controller. The terminal further includes a; wherein the learning unit learns the first heart rate of the user according to the learning motion, and the analysis unit, when the second heart rate of the current motion is received, determines the second heart rate of the current motion is compared with the first heart rate of the learning motion corresponding to the current motion, and the situation determination unit determines that, when the second heart rate of the current motion exceeds the first heart rate of the learning motion, it is determined as a dangerous situation, and the artificial The intelligent control unit includes a second communication unit that transmits a control signal to the terminal to generate an alarm in the terminal when it is determined as a dangerous situation by the situation determination unit.

In addition, the terminal further includes a status information request unit that periodically requests status information including at least one of an emotion item, an action item, and a media item, and the first communication unit includes one or more statuses selected by the user. Information is transmitted to the artificial intelligence control unit, the learning unit learns a first holding time and a first heart rate of each learning motion for each state information, and the analysis unit recognizes the current motion and analyzes the state information, Compare and analyze the holding time and heart rate of the current motion and the learning motion corresponding to the analyzed state information, and the situation determination unit determines that the second holding time exceeds the first holding time, the second heart rate is the first heart rate Exceeding heart rate is set, but set differently in the state information to determine a dangerous situation, and among the emotion items, the excess time and excess heart rate are set to decrease as you go toward joy, sadness, worry, dislike, displeasure, and anger , Among the action items, the overtime and overtime are set to increase as you go to exercise, cleaning, eating, rest, and sleep, and among the media items, the overtime is set to increase as you go toward comedy, romance, action, crime, and fear. and the excess heart rate is set to decrease, and when the situation determination unit sets the excess time and excess heart rate with priority, the emotion item is set prior to the media item, but when viewing media, the first setting is made from the viewing start time of the media. Up to time, emotion items are set first, media items are set first from the first set time until the second set time after the end of media viewing, and emotion items are set first after the second set time, exceeding the excess heart rate of the action item. It is characterized by setting the time first.

As described above, the effect of the present invention is to detect and learn the user's motion, and if an abnormality is found in the normal motion according to the learned result, an alarm is generated in the user's terminal so that the user recognizes the abnormality, In this case, if the alarm is not turned off, a risk prevention system using artificial intelligence can be provided so that a pre-set third party is automatically contacted.

In addition, the effect of the present invention is to learn and analyze the motion information acquired by the information collection unit and prevent it from being transmitted to the outside, thereby providing a risk prevention system using artificial intelligence that is stable from invading the user's personal privacy.

1 and 2 are configuration diagrams showing a risk prevention system using artificial intelligence according to a preferred embodiment of the present invention.

Advantages and features of the present invention, and methods of achieving them, will become clear with reference to the detailed description of the following embodiments taken in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only these embodiments make the disclosure of the present invention complete, and common knowledge in the art to which the present invention belongs. It is provided to completely inform the person who has the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numbers designate like elements throughout the specification.

Hereinafter, the present invention will be described with reference to drawings for explaining a risk prevention system using artificial intelligence according to embodiments of the present invention.

1 and 2, the risk prevention system using artificial intelligence according to the present invention includes an information collection unit 10, a motion recognition unit 20, and an artificial intelligence control unit 30.

First, the information collection unit 10 acquires motion information of a user whose movement occurs. In this case, the motion information may be video in units of 1 second or image information in intervals of 1 second.

Here, the information collection unit 10 may include motion detection means such as a camera, lidar sensor, radar sensor, and motion sensor.

The motion recognition unit 20 recognizes motion from the motion information obtained from the information collection unit 10, but recognizes a learning motion during a preset learning period and a current motion after the learning period.

That is, the motion recognition unit 20 recognizes the learning motion and the current motion based on the learning period.

At this time, the preset learning period may be 1 month, but since this may vary according to settings, it is not limited thereto.

Here, the motion recognition unit 20 separates the background and the user area from the motion information, extracts a skeleton using joint information based on the user's body part recognition in the separated user area, and uses the extracted skeleton to extract a posture, that is, motion. can recognize

The artificial intelligence control unit 30 learns the learning motion through an artificial intelligence algorithm, and compares the learning motion with the current motion to determine whether or not a dangerous situation exists.

Here, the artificial intelligence control unit 30 includes a time measurement unit 31, a learning unit 32, an analysis unit 33, and a situation determination unit 34.

The time measurement unit 31 measures a first holding time during which the learning motion is maintained and a second holding time during which the current motion is maintained. That is, how much time one motion lasts is measured.

The learning unit 32 learns the learning motion and the first holding time of the learning motion.

Here, the learning motion can be learned through artificial intelligence algorithms, and examples of artificial intelligence algorithms include machine learning, neural networks, genes, deep learning, and classification algorithms. The learning motion can be learned using at least one of these algorithms, and the current motion can be accurately recognized using the learned result value.

When the current motion is recognized, the analyzer 33 compares and analyzes the second holding time of the current motion with the first holding time of the learning motion corresponding to the current motion. That is, the analysis unit 33 not only compares and analyzes the learning motion corresponding to the current motion with the current motion in the result of learning the learning motion, but also analyzes the corresponding current motion and the time for which the learning motion is maintained. Accordingly, it is possible to find a learning motion corresponding to the current motion, and compare and analyze the first retention time of the found learning motion and the second retention time of the current motion.

When the second holding time of the current motion exceeds the first holding time of the learning motion, the situation determination unit 34 determines it to be a dangerous situation.

Here, the first holding time of the learning motion may be an average time or a maximum time during which the learning motion is maintained during the learning period. That is, when the second holding time of the current motion exceeds the average time or maximum time of the learning motion, it can be determined as a dangerous situation.

For example, as a result of learning the motion during sleep, that is, the bedtime learning motion, in the case of the bedtime learning motion, if the average time when tossing and turning occurs once is 30 minutes, even if the holding time of the current motion is 33 minutes, tossing and turning does not occur. can be judged as a risk situation.

Alternatively, if the maximum time is 35 minutes when one turn of the bedtime learning motion occurs, it can be determined as a dangerous situation when toss and turns do not occur even after the duration of the current motion for 35 minutes.

At this time, if the average time is exceeded, it is divided into a first dangerous situation, and if the maximum time is exceeded, it is divided into a second dangerous situation, and in the first dangerous situation, the terminal 40 to be described later vibrates, and in the second dangerous situation, an alarm to be described later sounds. can do.

As another example, the learning motion of staying in the bathroom can be learned by dividing it into a business learning motion for urinating or defecation and a shower learning motion for taking a shower. At this time, the artificial intelligence control unit 30 is connected to recognize the operation of the pre-installed boiler, and if the hot water operation of the boiler is recognized for a long time, it can be learned as a shower learning motion, and if it is recognized or not recognized for a short time, it can be learned as a business learning motion.

As described above, the dangerous situation may be determined as a dangerous situation when the second holding time of the current motion exceeds the first holding time (average time or maximum time) of the learning motion.

Here, the situation judgment unit 34 determines a dangerous situation immediately when the first maintenance time is exceeded when determining the dangerous situation, or the dangerous situation based on the excess time in which the second maintenance time exceeds the first maintenance time. can also judge. In this case, the overtime may be set by recognizing the health condition of the user.

On the other hand, the risk prevention system using artificial intelligence according to the present invention further includes a terminal 40.

The terminal 40 has the function of a watch and is worn on a user's wrist or the like. In such a terminal 40, a heart rate measurement unit 41 for measuring the user's heart rate, an alarm unit 42 for generating an alarm, and a control signal received from the artificial intelligence control unit 30 are received and the heart rate is measured by the artificial intelligence control unit 30. ) and a first communication unit 43 for transmission.

Here, the learning unit 32 learns the user's first heart rate according to the learning motion.

When the second heart rate of the current motion is received, the analyzer 33 compares and analyzes the second heart rate of the current motion with the first heart rate of the learning motion corresponding to the current motion.

At this time, when the second heart rate of the current motion exceeds the first heart rate of the learning motion, the situation determination unit 34 determines that the situation is dangerous.

Here, the first heart rate of the learning motion may be a minimum heart rate, an average heart rate, or a maximum heart rate. In addition, the situation determination unit 34 may determine that the second heart rate of the current motion is less than the first heart rate of the learning motion as a dangerous situation.

In addition, the artificial intelligence control unit 30 uses a second communication unit 35 for transmitting a control signal for generating an alarm in the terminal 40 to the terminal 40 when it is determined as a dangerous situation by the situation determination unit 34. include

Then, the alarm unit 42 sounds an alarm when the situation determination unit 34 determines that the situation is dangerous. At this time, the user can turn off the alarm.

Here, after the first alarm starts ringing, if the user fails to turn off the alarm before a preset waiting time, the second communication unit 35 may transmit a preset message to the preset third party 50 . At this time, the third person may be an acquaintance, 119, or 112, but is not limited thereto.

Meanwhile, the terminal 40 further includes a status information request unit that periodically requests status information including at least one of an emotion item, an action item, and a media item.

At this time, the emotional items include detailed items such as joy, sadness, worry, dislike, displeasure, and anger, and the action items include detailed items such as exercise, cleaning, eating, rest, and sleeping, and the media items include comedy, comedy, and the like. , Romance, Action, Crime, Horror, etc. That is, the terminal 40 displays detailed items for each item and requests input of the detailed items from the user.

For example, appraisal items can be requested at wake-up, noon, and 6 p.m. In addition, action items and media items can be requested when the retention time of the user's learning motion or current motion is drastically shortened or lengthened.

In addition, the user can directly input status information when a change in emotion occurs, when there is an action to be performed, or when a user wants to watch media. For example, when exercising, a detailed exercise item of an action item may be selected, and when a media to be watched is a comedy, a detailed comedy item may be selected.

By selecting detailed items according to emotion, behavior, and media, the learning unit 32 can classify and learn the learning motion, holding time, and heart rate during the learning period for each state information. That is, learning motion, holding time, and heart rate according to detailed items of the user's emotion items may be learned. Similarly, learning motion, maintenance time, and heart rate according to detailed items of action items and media items can be learned.

At this time, the first communication unit 43 may transmit one or more status information selected by the user to the artificial intelligence control unit 30 .

On the other hand, motion appears differently for each emotion, action, and media viewing. Also, it appears differently for each detailed item. When the emotion is happy, a motion of clapping may appear, and when the emotion is sad, a motion of wiping tears may appear. Likewise, the behavior is different when cleaning or exercising, and whether the media is coordinating or fearful. By a user inputting each of these detailed items or by periodically requesting an input from the terminal 40, the learning unit 32 can learn the learning motion, holding time, and heart rate according to the input detailed items, and the recognized When comparing the holding time and heart rate of the current motion and the learning motion, it is possible to determine the dangerous situation in detail and quickly for each detailed item.

Accordingly, the learning unit 32 learns the first holding time and the first heart rate of each learning motion for each state information.

The analysis unit 33 recognizes the current motion, analyzes state information, and compares and analyzes the current motion corresponding to the analyzed state information and the duration and heart rate of the learning motion.

Then, the situation determination unit 34 sets an excess time in which the second holding time exceeds the first holding time and an excess heart rate in which the second heart rate exceeds the first heart rate, but sets the excess time and excess heart rate differently from the state information. It is set and judged as a dangerous situation. In addition, in the case of the heart rate, the second heart rate is less than the first heart rate may be set differently for each state information.

Accordingly, among the emotion items, the excess time and the excess heart rate may be set to decrease in proportion to joy, sadness, worry, dislike, displeasure, and anger.

Among the action items, the excess time and the excess heart rate may be set to increase as you move toward exercise, cleaning, eating, resting, and sleeping.

Among the media items, the excess time may be set to increase, and the excess heart rate may be set to decrease in proportion to comedy, romance, action, crime, and horror.

In addition, when the condition determination unit 34 preferentially sets the excess time and the excess heart rate, it may set the emotion item prior to the media item. Here, when watching media, from the start time of media viewing to the first set time, the emotion item is set first, from the first set time to the second set time after media viewing is over, the media item is set first, and from the second set time onward, the emotion item is set first. Items can be prioritized.

In addition, the excess time may be set prior to the excess heart rate of the action item, but is not limited thereto.

In this way, by setting the excess time and the excess heart rate differently for each detailed item, a dangerous situation is determined at a different timing for each detailed item, and an alarm is quickly sounded on the user's terminal 40 so that the user recognizes it and Emotions, behaviors, media viewing, etc. can be controlled, and shock can be prevented by preventing dangerous situations from continuing.

Those skilled in the art to which the present invention pertains will understand that the present invention can be embodied in other specific forms without changing its technical spirit or essential features. Therefore, the embodiments described above should be understood as illustrative in all respects and not limiting. The scope of the present invention is indicated by the claims to be described later rather than the detailed description above, and all changes or modifications derived from the meaning and scope of the claims and equivalent concepts thereof are included in the scope of the present invention. should be interpreted In addition, the operation sequence of the components described in the above-described process does not necessarily have to be performed in a time-series order, and even if the execution sequence of each component and step is changed, if the gist of the present invention is satisfied, these processes will fall within the scope of the present invention. Of course you can.

10: information collection unit 20: motion recognition unit
30: artificial intelligence control unit 31: time measurement unit
32: learning unit 33: analysis unit
34: situation judgment unit 35: second communication unit
40: terminal 41: heart rate measuring unit
42: alarm unit 43: first communication unit
50: third party

Claims (4)

An information collection unit for obtaining motion information of a user whose movement occurs;
a motion recognition unit recognizing a motion from the motion information acquired from the information collection unit and recognizing a learning motion during a preset learning period and a current motion after the learning period; and
an artificial intelligence control unit that learns the learning motion through an artificial intelligence algorithm and compares the learning motion with the current motion to determine whether or not a dangerous situation exists;
Risk prevention system using artificial intelligence that includes.
According to claim 1,
The artificial intelligence control unit,
a time measurement unit for measuring a first holding time during which the learning motion is maintained and a second holding time during which the current motion is maintained;
A learning unit for learning the learning motion and a first holding time of the learning motion;
When the current motion is recognized, an analyzer for comparing and analyzing a second holding time of the current motion with a first holding time of a learning motion corresponding to the current motion;
A risk prevention system using artificial intelligence including a situation judgment unit that determines a dangerous situation when the second holding time of the current motion exceeds the first holding time of the learning motion.
According to claim 2,
A heart rate measurement unit worn on the user to measure the heart rate of the user, an alarm unit to generate an alarm, and a first communication unit to receive a control signal from the artificial intelligence controller and transmit the heart rate to the artificial intelligence controller It further includes a terminal;
The learning unit learns a first heart rate of the user according to the learning motion;
When the second heart rate of the current motion is received, the analyzer compares and analyzes the second heart rate of the current motion with the first heart rate of a learning motion corresponding to the current motion;
The situation determination unit determines that a dangerous situation occurs when the second heart rate of the current motion exceeds the first heart rate of the learning motion;
The artificial intelligence control unit includes a second communication unit for transmitting a control signal to the terminal to generate an alarm in the terminal when it is determined as a dangerous situation from the situation determination unit.
According to claim 3,
The terminal further includes a status information requesting unit that periodically requests status information including at least one of an emotion item, an action item, and a media item;
The first communication unit transmits one or more status information selected by the user to the artificial intelligence controller,
The learning unit learns a first holding time and a first heart rate of each learning motion for each state information;
The analysis unit recognizes the current motion, analyzes state information, compares and analyzes the duration and heart rate of the current motion and the learning motion corresponding to the analyzed state information,
The situation determination unit sets an excess time in which the second holding time exceeds the first holding time and an excess heart rate in which the second heart rate exceeds the first heart rate, but sets them differently in the state information to determine a dangerous situation;
Among the emotion items, the excess time and excess heart rate are set to decrease as you go toward joy, sadness, worry, dislike, displeasure, and anger.
Among the action items, the excess time and excess heart rate are set to increase as you move toward exercise, cleaning, eating, resting, and sleeping.
Among the media items, the excess time is set to increase and the excess heart rate is set to decrease in proportion to comedy, romance, action, crime, and horror;
When the situation judgment unit preferentially sets the overtime and the overtime heart rate,
Priority is given to the emotion item over the media item, but when viewing media, the emotion item is set first from the start time of viewing the media to the first set time, and the media item from the first set time to the second set time after the media is viewed. Priority is set, and after the second set time, the emotion item is set first,
A risk prevention system using artificial intelligence, characterized in that the excess time is prioritized over the excess heart rate of the action item.

Images