KR102285632B1 - Health abnormality detection system and method using gait pattern - Google Patents

Abstract

Embodiments of the present invention are a non-wearable sensor for generating image information by photographing a user or measuring a distance to the user, a wearable sensor worn by the user and sensing an impact with the ground or a user's movement, the image information Or a first analysis unit for extracting a first gait parameter based on the measured distance, a second analysis unit for extracting a second gait parameter based on the sensed impact or movement, and the first gait parameter and the second gait parameter Provides a health abnormality detection system and method using a gait pattern comprising a fusion unit for calculating the user's gait pattern information by fusion.
Accordingly, embodiments of the present invention can easily and accurately analyze a gait pattern in a daily environment at a low cost.

Description

Health abnormality detection system and method using gait pattern {HEALTH ABNORMALITY DETECTION SYSTEM AND METHOD USING GAIT PATTERN}

The present invention relates to a system and method for detecting health abnormalities using a gait pattern, and more particularly, to a system and method for detecting health abnormalities using a gait pattern for accurately and efficiently detecting health abnormalities using a gait pattern. .

Human gait is a repetitive motion that alternately moves from one point to another using both feet, and is the result of a complex mechanism of balance and strength.

Therefore, by analyzing the gait pattern, it is possible to diagnose physical and neurological abnormalities, and in particular, early diagnosis of the health of the elderly in an aging society.

For example, a change in the gait pattern such as a decrease in gait speed, decrease in stride length, decrease in arm sway, and a posture in which the body bends forward while walking may appear as mild antecedent symptoms of the elderly.

In addition, since an incorrect gait motion can cause abnormal symptoms in the brain, human body structure, joints, etc., abnormal symptoms can be prevented by analyzing the gait pattern to correct the gait motion.

As such, the gait pattern analysis can be used for various purposes such as diagnosis of disease, maintenance of health, and rehabilitation treatment.

However, in order to analyze a gait pattern by the conventional method, the user has to attach expensive sensors to various parts of the body, use expensive equipment such as a pressure mat, or wear shoes that measure pressure. Pattern analysis may be inconvenient (ii) it may take a long time to set up the sensor, (iii) it may increase the cost, (iv) the place for analyzing the gait pattern may be limited, and (v) walking in a different environment than usual. The gait pattern may be analyzed inaccurately. One of the health abnormality detection systems using a conventional gait pattern is as follows.

International Patent Publication No. 2017/131347 relates to a gait analysis system, wherein the gait analysis system includes an inertial sensor attached to the lumbar spine, hip joint, knee and ankle of a pedestrian, at least two cameras for measuring the position of the inertial sensor, and the A arithmetic unit for calculating the distance between the inertial sensors, a receiver for receiving the joint angle and acceleration values from the inertial sensor, and a change in distance between the inertial sensor, the joint angle and the acceleration value to analyze the walking state of the pedestrian It is characterized in that it includes an analysis unit.

In order to solve the above problem, embodiments of the present invention are a non-wearable sensor that generates image information by photographing a user or measures a distance to the user, is worn by the user and detects an impact with the ground or a user's movement a wearable sensor, a first analysis unit for extracting a first gait parameter based on the image information or the measured distance, a second analysis unit for extracting a second gait parameter based on the sensed shock or movement, and the second analysis unit An object of the present invention is to provide a system and method for detecting health abnormalities using a gait pattern comprising a fusion unit for calculating gait pattern information of the user by fusing the first gait parameter and the second gait parameter.

In addition, embodiments of the present invention calculate the user's foot position, arm position or skeletal information based on the image information or the measured distance, and calculate the first gait parameter based on the calculated foot position, arm position or skeletal information. An object of the present invention is to provide a health abnormality detection system using a gait pattern, characterized in that extraction.

In addition, embodiments of the present invention provide a system for detecting health abnormalities using a walking pattern, characterized in that the gait cycle is extracted using an inertial sensor that detects an impact with the ground when a user steps on the ground. .

In addition, embodiments of the present invention calculate the position of the user's arm using an inertial sensor that detects the movement of the user's arm, and extract the swing cycle of the user's arm based on the calculated arm position. An object of the present invention is to provide a health abnormality detection system using a gait pattern.

In addition, embodiments of the present invention correct or correspond to the first gait parameter based on the gait cycle or arm swing cycle of the second gait parameter to calculate the gait pattern information of the user. An object of the present invention is to provide a detection system.

In addition, embodiments of the present invention select one similar to the first gait parameter from among a plurality of second gait parameters extracted for each user and correct the first gait parameter based on the gait cycle or arm swing cycle of the selected second gait parameter An object of the present invention is to provide a health abnormality detection system using a gait pattern, characterized in that the gait pattern information for the user corresponding to the selected second gait parameter is calculated.

In addition, embodiments of the present invention have an object to provide a health abnormality detection system using a gait pattern, characterized in that it comprises a determination unit for determining whether or not a normal gait state based on the calculated gait pattern information.

In addition, the embodiments of the present invention distinguish and store the gait pattern information calculated for each user or store the gait pattern information calculated according to the walking time or place for each user, and determine whether the stored gait pattern information is in a normal state. An object of the present invention is to provide a health abnormality detection system using a gait pattern, characterized in that determining whether or not converges to an abnormal gait state.

In addition, embodiments of the present invention detect health abnormalities using a gait pattern, characterized in that it includes a notification unit that transmits health abnormality information when it is determined that the calculated gait pattern information corresponds to an abnormal state or converges to an abnormal state. The purpose is to provide a system.

An embodiment of the present invention for achieving the above object is a non-wearable sensor that generates image information by photographing a user or measures a distance to the user, is worn by the user and detects an impact with the ground or a user's movement a wearable sensor, a first analysis unit for extracting a first gait parameter based on the image information or the measured distance, a second analysis unit for extracting a second gait parameter based on the sensed impact or movement; And it provides a health abnormality detection system using a gait pattern comprising a fusion unit for calculating the gait pattern information of the user by fusing the first gait parameter and the second gait parameter.

In one embodiment, the first gait parameter includes at least one of a stride length, a gait cycle, a joint angle, an upper body tilt, and a swing cycle of the arm, and the first analysis unit is the image information or the measured distance based on the The user's foot position, arm position, or skeletal information may be calculated, and the first gait parameter may be extracted based on the calculated foot position, arm position or skeletal information.

In one embodiment, the wearable sensor includes an inertial sensor that detects an impact with the ground when the user steps on the ground, the second gait parameter includes a gait cycle, and the second analyzer includes the impact The gait cycle may be extracted based on the occurrence time.

In an embodiment, the wearable sensor includes an inertial sensor for detecting the movement of the user's arm, the second gait parameter includes a swing cycle of the arm, and the second analysis unit includes the detected arm movement The position of the user's arm may be calculated based on , and the swing cycle of the user's arm may be extracted based on the calculated position of the arm.

In one embodiment, the second gait parameter includes a gait cycle or arm swing cycle, and the fusion unit corrects or corresponds to the first gait parameter based on the gait cycle or arm swing cycle of the second gait parameter. It is possible to calculate the user's walking pattern information.

In one embodiment, the second analysis unit extracts a second gait parameter for each user based on the user's movement or the impact with the ground sensed by the plurality of wearable sensors worn by the plurality of users, and the fusion unit Selecting one similar to the first gait parameter from among a plurality of the second gait parameters and correcting or matching the first gait parameter based on the gait cycle or arm swing cycle of the selected second gait parameter to determine the selected second gait parameter It is possible to calculate gait pattern information for the user corresponding to .

In an embodiment, the health abnormality detection system using the gait pattern may include a determination unit that determines whether the gait pattern is in a normal gait state based on the calculated gait pattern information.

In one embodiment, the health abnormality detection system using the gait pattern is a management unit that distinguishes and stores the calculated gait pattern information for each user or distinguishes and stores the calculated gait pattern information according to a walking time or a walking place for each user. Including, the determination unit may determine whether the stored gait pattern information is in a normal state or whether it converges to an abnormal gait state.

In one embodiment, the health abnormality detection system using the gait pattern may include a notification unit that transmits health abnormality information when it is determined that the calculated gait pattern information corresponds to an abnormal state or converges to an abnormal state.

In addition, another embodiment of the present invention for achieving the above object receives image information or distance information from a non-wearable sensor that captures a user or measures a distance to the user, and detects an impact with the ground or a user's movement An input step of receiving information on impact or movement from a wearable sensor, a first analysis step of extracting a first gait parameter based on the image information or distance information, a second analysis step on the basis of the information on the impact or movement A second analysis step of extracting a gait parameter, a convergence step of calculating the gait pattern information of the user by fusing the first gait parameter and the second gait parameter Provides a health abnormality detection method using a gait pattern .

As described above, embodiments of the present invention include a non-wearable sensor for generating image information by photographing a user or measuring a distance from a user, a wearable sensor worn by the user and sensing an impact with the ground or a user's movement , a first analysis unit for extracting a first gait parameter based on the image information or the measured distance, a second analysis unit for extracting a second gait parameter based on the sensed impact or movement, and the first gait parameter and By providing a health abnormality detection system and method using a gait pattern comprising a convergence unit for calculating the gait pattern information of the user by fusion of the second gait parameter, it is possible to easily and accurately analyze a gait pattern in a daily environment at low cost. there is.

In addition, embodiments of the present invention calculate the user's foot position, arm position or skeletal information based on the image information or the measured distance, and calculate the first gait parameter based on the calculated foot position, arm position or skeletal information. By providing a health abnormality detection system using a gait pattern, characterized in that it is extracted, it is possible to easily obtain various gait parameters and analyze the gait pattern from various angles at low cost.

In addition, embodiments of the present invention provide a health abnormality detection system using a gait pattern characterized in that the user extracts the gait cycle using an inertial sensor that detects an impact with the ground when the user steps on the ground, making it easy at low cost In this way, the accuracy of the walking parameters can be improved.

In addition, embodiments of the present invention calculate the position of the user's arm using an inertial sensor that detects the movement of the user's arm, and extract the swing cycle of the user's arm based on the calculated arm position. By providing a health abnormality detection system using a gait pattern, it is possible to easily improve the accuracy of a gait parameter at a low cost.

In addition, embodiments of the present invention correct or correspond to the first gait parameter based on the gait cycle or arm swing cycle of the second gait parameter to calculate the gait pattern information of the user. By providing the sensing system, the accuracy of various gait parameters (first gait parameter) extracted at low cost can be easily improved at low cost.

In addition, embodiments of the present invention select one similar to the first gait parameter from among a plurality of second gait parameters extracted for each user and correct the first gait parameter based on the gait cycle or arm swing cycle of the selected second gait parameter By providing a health abnormality detection system using a gait pattern, characterized in that the gait pattern information for the user corresponding to the selected second gait parameter is calculated in correspondence with the second gait parameter, the user corresponding to the first gait parameter among a plurality of users is facilitated. It is possible to analyze the gait pattern for each user by estimating it.

In addition, embodiments of the present invention provide a gait pattern analysis by providing a health abnormality detection system using a gait pattern, characterized in that it comprises a determination unit for determining whether or not a normal gait state based on the calculated gait pattern information. It is possible to provide a health abnormality detection service used.

In addition, the embodiments of the present invention distinguish and store the gait pattern information calculated for each user or store the gait pattern information calculated according to the walking time or place for each user, and determine whether the stored gait pattern information is in a normal state. By providing a health abnormality detection system using a gait pattern, which is characterized by determining whether judging or converges to an abnormal gait state, it is possible to provide a customized health abnormality detection service, improve the accuracy of health abnormality detection, and Health abnormalities can be predicted before the transition from a normal gait state to an abnormal gait state.

It is not limited to the above-mentioned effects, and other effects not mentioned can be clearly understood by those of ordinary skill in the art to which the present invention belongs from the following description.

1 is a view showing a health abnormality detection system using a gait pattern according to an embodiment of the present invention.
2 is a view showing a gait pattern analysis apparatus according to an embodiment of the present invention.
3 is a diagram exemplarily showing a data flow in the gait pattern analysis apparatus according to an embodiment of the present invention.
4 is a graph exemplarily showing an example in which the value of the stride converges to the abnormal gait state.
5 is a flowchart illustrating a method for detecting health abnormalities using a walking pattern according to an embodiment of the present invention.

Since the description of the present invention is merely embodiments for structural or functional description, the scope of the present invention should not be construed as being limited by the embodiments described in the text. That is, since the present embodiments are capable of various modifications and may have various forms, it should be understood that the scope of the present invention includes equivalents capable of realizing the technical idea.

In addition, since the object or effect presented in the present invention does not mean that a specific embodiment should include all of them or only such effects, it should not be understood that the scope of the present invention is limited thereby.

In addition, the accompanying drawings are provided to help the understanding of the present invention, and provide embodiments together with the detailed description. However, the technical features of the present invention are not limited to specific drawings, and features disclosed in each drawing may be combined with each other to form a new embodiment.

A health abnormality detection system using a gait pattern disclosed in the following embodiments will be described in more detail with reference to each drawing.

1 is a view showing a health abnormality detection system using a gait pattern according to an embodiment of the present invention.

Referring to FIG. 1 , the health abnormality detection system 10 using a gait pattern according to an embodiment includes a non-wearable sensor 100 , a wearable sensor 200 and a gait pattern analysis device 300 . .

The non-wearable sensor 100 may be fixed to a specific place such as a wall or ceiling, or may be mounted on a robot as shown in FIG. 1 and move together with the robot. In addition, the non-wearable sensor 100 may be connected to the gait pattern analysis apparatus 300 through a communication network. Here, the communication network may refer to a network of a broad concept including a wired or wireless communication network.

The non-wearable sensor 100 may generate image information by photographing a user or may measure a distance from the user and transmit it to the gait pattern analysis apparatus 300 . The non-wearable sensor 100 may (i) include an RGB camera to generate image information by photographing the user, and (ii) include a laser scanner to measure the distance to the user. and (iii) it may be configured to include an RGB-D camera, etc. to measure the distance to the user while generating image information by photographing the user.

At this time, the non-wearable sensor 100 may set a photographing or measurement time and place, and transmit the set photographing or measurement time or place to the walking pattern analysis apparatus 300 together.

Here, the photographing or measurement location may mean the positional coordinates of the non-wearable sensor 100 in which photographing or measurement is performed, or may mean the positional coordinates of a user who is the subject of photographing or measurement, and may refer to the non-wearable sensor ( 100) or the width of the space where the user is located.

Also, here, the non-wearable sensor 100 may set a photographing or measurement location in different ways depending on whether it is fixed or stayed in a specific location.

When the non-wearable sensor 100 is fixed or stays in a specific place, the photographing or measuring location may be preset in the non-wearable sensor 100 .

On the other hand, when the non-wearable sensor 100 is mounted on the robot and moves to various areas, the non-wearable sensor 100 receives and receives the photographing or measurement location calculated using the map stored in the robot from the robot. You can set the shooting or measurement location as the location.

Hereinafter, image information or distance information measured at a distance from a user is determined to include the time or place of shooting or measurement, unless otherwise specified.

Here, the distance to the user may refer to distance information calculated by 3D scanning the user with, for example, infrared rays.

At this time, the non-wearable sensor 100 may calculate a point cloud for the user based on the measured distance after measuring the distance to the user, and transmit the calculated point cloud to the walking pattern analysis device 300 . . Here, the point cloud may mean a set of coordinates of vertices constituting the 3D model (user).

On the other hand, the non-wearable sensor 100 may be configured to include a plurality of cameras or binocular cameras.

The wearable sensor 200 may correspond to an inertial sensor provided in a wearable or portable device such as a smart watch or a smart phone. In addition, the wearable sensor 200 may be connected to the gait pattern analysis apparatus 300 through a communication network.

Here, the inertial sensor converts a dynamic force such as motion or impact into an electrical signal and may correspond to an acceleration sensor or an angular velocity sensor.

The wearable sensor 200 may be worn by the user and may detect an impact with the ground or the user's movement and transmit it to the gait pattern analysis device 300 . Here, the impact with the ground may mean an impact transmitted to the wearable sensor 200 when the user steps on the ground, and the user's movement is the arm or wrist wearing the smart watch equipped with the wearable sensor 200 . may mean the movement of

At this time, the wearable sensor 200 may transmit the user's identifier and the time at which the shock or movement is sensed to the gait pattern analysis apparatus 300 together. Here, the user's identifier may mean an identifier of a user using a smart watch equipped with the wearable sensor 200 or an identifier assigned to the electronic device itself, such as a smart watch. For example, it may correspond to a user's resident registration number, a serial number such as a smart watch, or a phone number.

Hereinafter, information on the impact or movement is determined to include the time at which the impact or movement is sensed, unless otherwise specified.

On the other hand, the wearable sensor 200 may transmit to the walking pattern analysis apparatus 300 a place where an impact or movement is sensed using a GPS function built into a device such as a smart watch. In this case, the non-wearable sensor 100 does not need to transmit the photographed or measured location to the walking pattern analysis device 300 . However, since the GPS function may not work indoors, it may be desirable to transmit the location where the non-wearable sensor 100 is photographed or measured.

The gait pattern analysis apparatus 300 may be connected to the non-wearable sensor 100 and the wearable sensor 200 through a communication network. A detailed configuration of the gait pattern analysis apparatus 300 will be described with reference to FIGS. 2 and 3 .

As such, the health abnormality detection system 10 using a gait pattern includes a non-wearable sensor 100 such as an RGB-D camera, a wearable sensor 200 such as a smart watch, and a gait pattern analysis device 300 . By being configured, it is possible to analyze a gait pattern easily and accurately in a daily environment at a low cost. Here, the ability to accurately analyze a gait pattern will be described later in the fusion unit 340 .

2 is a view showing a gait pattern analysis apparatus according to an embodiment of the present invention.

Referring to FIG. 2 , the gait pattern analysis apparatus 300 according to an embodiment includes an input unit 310 , a first analysis unit 320 , a second analysis unit 330 , a fusion unit 340 , and a management unit 350 . , a determination unit 360 , a notification unit 370 , a transmission/reception unit 380 , a database 390 and a control unit 395 .

3 is a diagram exemplarily showing a data flow in the gait pattern analysis apparatus according to an embodiment of the present invention.

Referring to FIG. 3 , among the data received by the gait pattern analysis apparatus 300 according to an embodiment, (i) the data received from the non-wearable sensor 100 is an input unit 310 and a first analysis unit 320 . is transmitted to the fusion unit 340 through (ii) and (ii) data received from the wearable sensor 200 is transmitted to the fusion unit 340 through the input unit 310 and the second analysis unit 330 . The data fused in the fusion unit 340 may be transmitted to the determination unit 360 through the management unit 350 .

Let's look at each configuration of the gait pattern analysis device (300).

The input unit 310 transmits image information or measured distance information generated by the non-wearable sensor 100 , information on an impact or movement detected by the wearable sensor 200 and an identifier of the user through the transceiver 380 . The input may be received or the input may be received through the database 390 .

Here, the distance information may refer to information obtained by measuring the distance to the user or may refer to a point cloud calculated based on the distance measured after measuring the distance to the user.

Also, here, the image information or the distance information may include a time or place taken or measured.

In addition, the input unit 310 may transmit the received image information or distance information to the first analysis unit 320 , and may transmit information on the received impact or movement and the user's identifier to the second analysis unit 330 . .

The first analyzer 320 may receive image information or distance information from the input unit 310 , extract a first gait parameter based on the image information or distance information, and transmit it to the fusion unit 340 .

Here, the first gait parameter may include at least one of a step length or stride length, a gait cycle, a joint angle, an upper body tilt, and a swing cycle of the arm. . In addition, the first gait parameter may include a time or place at which the image information or distance information was photographed or measured.

Here, the step length means the distance between the heel strike of one foot and the heel strike of the opposite foot, and the stride length means the distance between the heel strike of one foot and the heel strike of the same foot.

Also, here, the gait cycle may refer to a process constituting a gait cycle, a gait speed, or a stride. The process of constructing a step is the stance phase, in which the foot is touching the ground, from the heel strike of one foot until the toe of the same foot leaves the ground, and one foot is in the air away from the ground. A swing phase and a double stance phase in which both feet are touching the ground can be distinguished.

In addition, here, the arm swing cycle refers to the cycle, speed, and movement width of the wrist or hand during one arm swing, the movement angle, or the process during one swing of the arm during walking. can do.

Specifically, the first analysis unit 320 is based on the input image information or distance information, the user's foot position (toe or heel position), arm position (wrist, wrist or hand position) or skeletal information ( skeleton) and based on the calculated foot position, arm position, or skeletal information, step length or stride length, gait cycle, joint angle, upper body tilt, and arm swing cycle ( swing cycle) can be extracted.

At this time, the first analysis unit 320 may detect the user's silhouette or outline from the input image information, and calculate the position of the toe or heel by using the detected silhouette or outline. Here, the silhouette may mean an image in which the inside of the outline is filled with a single color.

As such, the health abnormality detection system 10 using a gait pattern can easily obtain various gait parameters by extracting the gait parameter (first gait parameter) of the user based on image information or distance information, thereby reducing the gait pattern at low cost. can be analyzed from various angles.

The second analysis unit 330 receives information on the impact or movement and the user's identifier from the input unit 310, and extracts the second gait parameter for each user based on the information on the impact or movement and the user's identifier to extract the fusion unit It can be passed to (340).

Here, as described above, the impact may mean an impact transmitted to the wearable sensor 200 when the user steps on the ground, and the movement is the arm wearing the smart watch provided with the wearable sensor 200 . Alternatively, it may mean the movement of the wrist.

Also, here, the second gait parameter may include at least one of a gait cycle and a swing cycle of the arm. In addition, the second gait parameter may include a time at which an impact or movement is sensed.

Specifically, the second analysis unit 330 obtains the time at which the impact transmitted to the wearable sensor 200 reaches the maximum value (peak) when the heel touches the ground from the information on the impact received from the input unit 310 . It is possible to extract the gait cycle such as the time interval of heel strike by considering the time when the impact peaks as the time when the heel of the foot touches the ground (heel strike time).

In addition, the second analysis unit 330 is the value of the acceleration or angular velocity sensed by the inertial force when the arm or wrist of the user wearing the wearable sensor 200 moves from the information about the movement received from the input unit 310 . can be used to calculate the arm or wrist position, and the user's arm or wrist position can be used to determine the frequency, speed, and width of wrist or hand movement during one arm swing, movement angle, or arm movement speed during walking. It is possible to extract the swing cycle of the user's arm, such as during one swing.

In this way, the health abnormality detection system 10 using a gait pattern extracts the user's gait parameter (second gait parameter) based on the information on the shock or movement sensed from the wearable sensor 200, so that it is easy at low cost In this way, the accuracy of the walking parameters can be improved.

In other words, improving the accuracy of the walking parameters may mean that by using the wearable sensor 200 , an impact or movement can be more accurately detected than when the non-wearable sensor 100 is used.

The fusion unit 340 may receive the first gait parameter from the first analysis unit 320 and receive the second gait parameter and the user's identifier from the second analysis unit 330, and the received first gait parameter and the second gait parameter By fusing the two gait parameters, gait pattern information for the user identified by the user's identifier may be calculated, and the calculated gait pattern information may be transmitted to the management unit 350 or the determiner 360 together with the user's identifier.

Here, the gait pattern information may be composed of a type of information similar to that of the first gait parameter. In addition, the gait pattern information may include a gait time or gait location calculated based on the time or place taken or measured included in the first gait parameter or the time when an impact or movement included in the second gait parameter is sensed. .

Here, since the walking place may correspond to a photographed or measured place included in the first gait parameter, the walking place may mean the positional coordinates of the non-wearable sensor 100 in which photographing or measurement is performed, or may be photographed or measured. It may mean the coordinates of the user's location, which is the target of , or it may mean the width of the non-wearable sensor 100 or the space in which the user is located.

Specifically, the fusion unit 340 corrects or responds to the gait cycle or arm swing cycle included in the first gait parameter based on the gait cycle or arm swing cycle included in the second gait parameter to correct or correspond to the user's gait pattern. information can be calculated.

For example, the fusion unit 340 can compare the first gait parameter and the second gait parameter, and when the heel strike time interval is different from each other, the first gait parameter is the heel strike time interval of the second gait parameter. Correct the heel strike time interval of the first gait parameter based on image information or distance information corresponding to the corrected heel strike time interval, step length or stride length, gait cycle, joint angle angle), upper body inclination, etc. can be corrected.

In this way, the fusion unit 340 calculates gait pattern information by correcting or matching the first gait parameter extracted based on image information or distance information based on the second gait parameter extracted based on the impact or motion information. , it is possible to easily improve the accuracy of various gait parameters (first gait parameter) extracted at low cost at low cost.

In addition, the fusion unit 340 selects one similar to the first gait parameter from among the plurality of second gait parameters extracted for each user, and corrects the first gait parameter based on the gait cycle or arm swing cycle of the selected second gait parameter. Alternatively, the gait pattern information for the user corresponding to the selected second gait parameter may be calculated.

For example, the fusion unit 340 selects the second gait parameter including the value most similar to the value of the stride length of the first gait parameter, and selects the first gait parameter based on the gait cycle or arm swing cycle of the selected second gait parameter. By correcting or corresponding to the gait parameter, gait pattern information for the user corresponding to the selected second gait parameter may be calculated.

In this way, the fusion unit 340 selects a thing similar to the first gait parameter from among a plurality of second gait parameters extracted for each user and calculates gait pattern information, thereby making it easy for a user corresponding to the first gait parameter from among a plurality of users. It is possible to analyze the gait pattern for each user by estimating it.

Meanwhile, the fusion unit 340 may simply merge the first gait parameter and the second gait parameter. For example, when the swing cycle of the arm is not included in the first gait parameter and only the swing cycle of the arm is included in the second gait parameter, the fusion unit 340 merges the first gait parameter and the second gait parameter. The arm swing cycle can be included in the gait pattern information.

In this way, the fusion unit 340 may include various gait parameters in the gait pattern information by merging the first gait parameter and the second gait parameter.

The management unit 350 may receive the gait pattern information and the identifier of the user corresponding to the gait pattern information from the convergence unit 340 , and may store the gait pattern information for each user by distinguishing it in the database 390 , and the stored gait pattern for each user. The information may be transmitted to the determination unit 360 .

In addition, the management unit 350 may classify the walking pattern information according to the walking time or walking place included in the walking pattern information for each user and store it in the database 390, and store the walking pattern information for each user according to the walking time or the walking place. It may be transmitted to the determination unit 360 .

The determination unit 360 may receive the gait pattern information and the user's identifier corresponding to the gait pattern information from the convergence unit 340 or the management unit 350, and may determine whether the gait pattern information corresponds to a normal gait state. and when the gait pattern information corresponds to an abnormal gait state, the gait pattern information and the user's identifier may be transmitted to the notification unit 370 .

For example, the determination unit 360 may receive the value of the stride length and the identifier of the user from the management unit 350 , and if the transmitted value of the stride length is smaller than a specific threshold value, the determination unit 360 determines that it corresponds to an abnormal state, and thus the notification unit 370 . ) can be passed the value of the stride length and the identifier of the user.

Here, since the threshold value may vary depending on the user's age, gender, height, etc., the determination unit 360 may obtain the user's age, gender, height, etc. based on the received identifier of the user and set the correct threshold value.

In addition, the determination unit 360 may receive the user's identifier corresponding to the walking pattern information and the walking pattern information differentiated according to the walking time or the walking place from the management unit 350, and for each user, each walking time, or each walking place. It may be determined whether the gait pattern information corresponds to an abnormal gait state.

For example, the determination unit 360 may receive the value of the stride length at a specific time or at a specific place and the identifier of the user from the management unit 350, and if the received stride value is less than a specific threshold value, it corresponds to an abnormal state. It is determined that the user is able to transmit the value of the stride length and the identifier of the user to the notification unit 370 .

At this time, since the threshold value may vary according to time or place, the determination unit 360 receives information on the walking time or the walking place from the management unit 350 to set the threshold value corresponding to the walking time or the walking place correctly. can

For example, the threshold value of stride length at dawn or evening time may be smaller than the threshold value for stride length at day time.

Also, the threshold value of the stride length in a narrow place such as a room may be smaller than the threshold value of the stride length in a wide place such as a living room. For example, the threshold value of the stride length may be set to be proportional to a log value of the width of the place where the walking is performed.

At this time, the determination unit 360 may require a map to calculate the area of the place where walking is performed from the walking place.

As such, the determination unit 360 can provide a customized health abnormality detection service by determining whether the gait pattern information for each user, each walking time, or each walking place corresponds to an abnormal gait state, and improve the accuracy of health abnormality detection. can be improved

In addition, the determination unit 360 may determine whether the gait pattern information received from the management unit 350 converges to the abnormal gait state, and when the gait pattern information converges to the abnormal gait state, the gait to the notification unit 370 Pattern information and user identifiers can be transmitted.

4 is a graph exemplarily showing an example in which the value of the stride converges to the abnormal gait state.

Referring to FIG. 4 , the determination unit 360 may receive the value of the stride length and the user's identifier for each day from the management unit 350 , and as a result of analyzing the change in the value of the stride length received during a specific period, the value of the stride length is When it is determined that the value corresponding to the normal gait state converges to a threshold value that distinguishes the normal gait state and the abnormal gait state, that is, if it is determined that the value of the stride converges to the abnormal gait state, the notification unit 370 displays the value of the stride and the user You can pass the identifier of

As such, by determining whether the gait pattern information converges to the abnormal gait state, the determination unit 360 may predict health abnormalities before the user transitions from the normal gait state to the abnormal gait state.

The notification unit 370 may receive the gait pattern information and the user's identifier from the determination unit 360 , and may transmit health abnormality information informing of the health abnormality to the terminal of the user or a registered guardian.

Here, the registered guardian may mean a guardian registered in advance in association with the user's identifier, and the terminal is a server, laptop, mobile phone, smart phone (2G/3G/4G/LET, smart phone), PMP (Portable Media Player) ), PDA (Personal Digital Assistant), and tablet PC (Tablet PC) may correspond to any one.

In this case, the notification unit 370 may include walking pattern information in the health abnormality information.

Transceiver 380 receives or notifies image information, distance information, impact or movement information, user's identifier, etc. from the non-wearable sensor 100 or the wearable sensor 200 connected to the input unit 310 through a communication network. The unit 370 may transmit health abnormality information to the terminal of the user or registered guardian.

The database 390 includes image information, distance information, information on impact or movement, first gait parameters, second gait parameters, gait pattern information, age, height, gender, walking place, threshold of gait parameters according to walking time, etc. It is possible to store a value, information about the user's or guardian's terminal, and the like.

The database 390 is a concept including a computer-readable recording medium, and refers not only to a narrow database, but also a database in a broad sense including a data record based on a file system, and even a simple set of logs is searched for. If data can be extracted by doing so, it is included in the scope of the database in the present invention.

The control unit 395 includes an input unit 310 , a first analysis unit 320 , a second analysis unit 330 , a fusion unit 340 , a management unit 350 , a determination unit 360 , a notification unit 370 , and transmit/receive Control flow or data flow between the unit 380 and the database 390 may be controlled.

5 is a flowchart illustrating a method for detecting health abnormalities using a walking pattern according to an embodiment of the present invention.

Referring to FIG. 5 , the method 400 for detecting health abnormalities using a gait pattern according to an embodiment includes an input step (S410), a first analysis step (S420), a second analysis step (S410), and a fusion step (S440). It may be composed of

In the input step (S410), the gait pattern analysis device 300 is the image information or measured distance information generated by the non-wearable sensor 100, information on the impact or movement detected by the wearable sensor 200, and the user may be inputted through the transceiver 380 or may be inputted through the database 390 .

In the first analysis step (S420), the gait pattern analysis apparatus 300 may extract the first gait parameter based on the image information or distance information.

Specifically, the gait pattern analysis device 300 is based on the input image information or distance information, the position of the user's foot (position of the toes or heel), the position of the arm (position of the wrist, wrist or hand) or skeletal information ( skeleton) and based on the calculated foot position, arm position, or skeletal information, step length or stride length, gait cycle, joint angle, upper body tilt, and arm swing cycle ( swing cycle) can be extracted.

In the second analysis step (S430), the gait pattern analysis apparatus 300 may extract the second gait parameter based on the information on the impact or movement.

Specifically, the gait pattern analysis device 300 can obtain the time at which the shock transmitted to the wearable sensor 200 reaches the maximum value (peak) when the heel touches the ground from the information about the received shock, and the shock is the maximum Considering the peak time as the time when the heel of the foot touches the ground (heel strike time), the gait cycle such as the heel strike time interval can be extracted.

In addition, the gait pattern analysis device 300 uses the value of the acceleration or angular velocity sensed by the inertial force when the arm or wrist of the user wearing the wearable sensor 200 moves from the information on the received movement to the arm or wrist. can calculate the position of the user's arm or wrist and use the user's arm or wrist position to move the arm back and forth during walking, the speed, the width of the wrist or hand movement during one arm swing, the movement angle or the duration of the arm swing It is possible to extract the swing cycle of the user's arm, such as the process.

In the fusion step (S440), the gait pattern analysis apparatus 300 may calculate the user's gait pattern information by fusing the first gait parameter and the second gait parameter.

Specifically, the gait pattern analysis device 300 corrects or responds to the gait cycle or arm swing cycle included in the first gait parameter based on the gait cycle or arm swing cycle included in the second gait parameter, thereby providing the user's Walking pattern information can be calculated.

For example, the gait pattern analysis apparatus 300 can compare the first gait parameter and the second gait parameter, and when the heel strike time interval is different from each other, the first gait parameter is the heel strike time interval of the second gait parameter. The heel strike time interval of the gait parameter is corrected, and the step length (or stride length), the gait cycle, and the joint angle of the first gait parameter based on the image information or distance information corresponding to the corrected heel strike time interval. (joint angle) and upper body inclination can be corrected.

In addition, the gait pattern analysis apparatus 300 selects one similar to the first gait parameter from among the plurality of second gait parameters extracted for each user, and selects the first gait parameter based on the gait cycle or arm swing cycle of the selected second gait parameter. By correcting or corresponding to the gait pattern information for the user corresponding to the selected second gait parameter may be calculated.

In addition, the gait pattern analysis apparatus 300 may simply merge the first gait parameter and the second gait parameter. For example, when the swing cycle of the arm is not included in the first gait parameter and only the swing cycle of the arm is included in the second gait parameter, the gait pattern analysis apparatus 300 calculates the first gait parameter and the second gait parameter. By merging, the swing cycle of the arm can be included in the gait pattern information.

Unlike FIG. 5 , the health abnormality detection method 400 using a walking pattern may further include a management step S450 , a determination step S460 , and a notification step S470 .

In the management step S450 , the gait pattern analysis apparatus 300 may classify gait pattern information for each user by using the identifier of the user corresponding to the gait pattern information and store it in the database 390 .

In addition, the gait pattern analysis apparatus 300 may classify gait pattern information according to the gait time or gait location for each user by using the gait time or gait place included in the gait pattern information and store it in the database 390 .

In the determination step (S460), the gait pattern analysis apparatus 300 may determine whether the gait pattern information corresponds to a normal gait state. In addition, the gait pattern analysis apparatus 300 may determine whether the gait pattern information corresponds to an abnormal gait state for each user, each gait time, or each gait location.

In addition, the gait pattern analysis apparatus 300 may determine whether the gait pattern information converges to the abnormal gait state.

Specifically, the gait pattern analysis apparatus 300 may receive gait pattern information periodically, and as a result of analyzing the change trend of the gait pattern information for a specific period, the gait pattern information corresponds to the normal gait state, but then the gait pattern information corresponds to the normal gait state and the abnormal gait state. It can be determined whether or not the convergence to the critical state for discriminating the state, that is, whether the gait pattern information converges to the abnormal gait state.

In the notification step (S470), when the gait pattern information corresponds to an abnormal state or converges to an abnormal gait state, the gait pattern analysis device 300 transmits health abnormality information informing of a health abnormality to the terminal of the user or registered guardian. there is.

As described above, although described with reference to preferred embodiments of the present application, those skilled in the art can variously modify the present application within the scope without departing from the spirit and scope of the present invention described in the claims below. and may be changed.

10: Health abnormality detection system using gait pattern
100: non-wearable sensor 200: wearable sensor
300: walking pattern analysis device
310: input unit 320: first analysis unit
330: second analysis unit 340: fusion unit
350: management unit 360: judgment unit
370: notice unit 380: transceiver unit
390: database 395: control unit
400: Health abnormality detection method using gait pattern

Claims (10)

a non-wearable sensor that generates image information by photographing a user or measures a distance to the user and transmits a measurement time;
a wearable sensor that is worn by the user and senses an impact with the ground or a movement of the user;
a first analysis unit for extracting a first gait parameter based on the image information or the measured distance;
a second analysis unit for extracting a second gait parameter based on the sensed impact or movement; and
It is configured to include a fusion unit for calculating the user's gait pattern information by fusing the first gait parameter and the second gait parameter,
The first gait parameter includes at least one of stride length, gait cycle, joint angle, upper body tilt, and arm swing cycle,
The second gait parameter includes a gait cycle or a swing cycle of the arm and a time when an impact or movement is sensed,
The fusion unit calculates the user's gait pattern information by correcting or corresponding to the first gait parameter based on the gait cycle or arm swing cycle of the second gait parameter,
The gait cycle of the first gait parameter is corrected based on the gait cycle of the second gait parameter by comparing the gait cycle of the first gait parameter and the second gait parameter, and image information corresponding to the time interval of the corrected gait cycle Alternatively, a health abnormality detection system using a gait pattern that calculates gait pattern information for a user by correcting at least one of stride length, gait cycle, joint angle, and upper body inclination of the first gait parameter based on distance information.
According to claim 1,
The first analysis unit calculates the user's foot position, arm position or skeletal information based on the image information or the measured distance, and the first gait based on the calculated foot position, arm position or skeletal information A health abnormality detection system using a gait pattern, characterized in that extracting parameters.
According to claim 1,
The wearable sensor includes an inertial sensor for detecting an impact with the ground when the user steps on the ground,
The second gait parameter includes a gait cycle,
The second analysis unit health abnormality detection system using a gait pattern, characterized in that extracting the gait cycle based on the time when the impact occurs.
According to claim 1,
The wearable sensor includes an inertial sensor for detecting the movement of the user's arm,
The second gait parameter includes a swing cycle of the arm,
The second analysis unit calculates the position of the user's arm based on the detected arm movement and extracts the swing cycle of the user's arm based on the calculated arm position. Health abnormality detection system.
According to claim 1,
The fusion unit corrects or corresponds to the first gait parameter based on the gait cycle or arm swing cycle of the second gait parameter to calculate the gait pattern information of the user.
6. The method of claim 5,
The second analysis unit extracts a second gait parameter for each user based on the user's movement or the impact with the ground sensed by the plurality of wearable sensors worn by the plurality of users,
The fusion unit selects one similar to the first gait parameter from among a plurality of the second gait parameters and corrects or corresponds to the first gait parameter based on the gait cycle or arm swing cycle of the selected second gait parameter A health abnormality detection system using a gait pattern, characterized in that the gait pattern information for the user corresponding to the two gait parameters is calculated.
According to claim 1,
The health abnormality detection system using the gait pattern is characterized in that it comprises a determination unit that determines whether the gait is in a normal gait state based on the calculated gait pattern information.
8. The method of claim 7,
The health abnormality detection system using the gait pattern includes a management unit that distinguishes and stores the calculated gait pattern information for each user or distinguishes and stores the calculated gait pattern information according to a walking time or a walking place for each user,
The determination unit determines whether the stored gait pattern information is in a normal state or converges to an abnormal gait state.
9. The method of claim 7 or 8,
The health abnormality detection system using the gait pattern is configured to include a notification unit that transmits health abnormality information when it is determined that the calculated gait pattern information corresponds to an abnormal state or converges to an abnormal state. Health abnormality detection system.
The input unit receives image information or distance information from a non-wearable sensor that captures the user or measures the distance to the user, and receives information on impact or movement from a wearable sensor that detects an impact with the ground or a user's movement. input step;
a first analysis step of extracting a first gait parameter based on the image information or distance information by a first analysis unit;
a second analysis step of extracting a second gait parameter based on the information on the impact or movement by the second analysis unit;
Performing a fusion step of calculating the user's gait pattern information by fusing the first gait parameter and the second gait parameter by the fusion unit,
The first gait parameter includes at least one of stride length, gait cycle, joint angle, upper body tilt, and arm swing cycle,
The second gait parameter includes a gait cycle or a swing cycle of the arm and a time when an impact or movement is sensed,
The fusion unit calculates the user's gait pattern information by correcting or corresponding to the first gait parameter based on the gait cycle or arm swing cycle of the second gait parameter,
The gait cycle of the first gait parameter is corrected based on the gait cycle of the second gait parameter by comparing the gait cycle of the first gait parameter and the second gait parameter, and image information corresponding to the time interval of the corrected gait cycle Alternatively, a method for detecting health abnormalities using a gait pattern that calculates gait pattern information for a user by correcting at least one of stride length, gait cycle, joint angle, and upper body inclination of the first gait parameter based on distance information.

Images