# US 12,597,292 B2 · Gait analysis apparatus and method

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 12,597,292 B2
- Document ID: US12597292B2
- Status: Granted
- Kind: B2
- Application number: 18/374,170
- Filed: 2023-09-28
- Published: 2024-07-04
- Granted: 2026-04-07
- Earliest claimed: 2021-03-29
- Assignees: Seoul National University R&Db Foundation
- Applicants: Seoul National University R&Db Foundation
- Inventors: Je Hee Lee; Beom Seok Jeon; Jung Hwan Shin; Ri Yu
- Web page: https://patentgrep.com/patent/US12597292B2

## CPC

- G06V 40/25: {Recognition of walking or running movements, e.g. gait recognition}
- A61B 5/11: Measuring movement of the entire body or parts thereof, e.g. head or hand tremor or mobility of a limb {(A61B5/1038 takes precedence; motion detection to correct for motion artifacts in physiological signals A61B5/721)}
- G06T 7/20: Analysis of motion (motion estimation for coding, decoding, compressing or decompressing digital video signals H04N19/43, H04N19/51)
- G06T 7/73: using feature-based methods
- G06T 2207/10016: Video; Image sequence
- G06T 2207/30196: Human being; Person
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

A gait analysis method includes collecting a two-dimensional gait video of a subject's gait situation during a preset time, which is captured using a camera; generating a three-dimensional coordinate system based on at least one frame of the two-dimensional gait video by receiving three-dimensional information of a capturing space; calculating three-dimensional feature point coordinates by receiving a plurality of feature points on the three-dimensional coordinate system, and calculating two-dimensional feature point coordinates for the plurality of feature points using resolution information of the camera; estimating mapping information between the three-dimensional feature point coordinates and the two-dimensional feature point coordinates; calculating a plurality of two-dimensional joint position information for the subject from the two-dimensional gait video; calculating a plurality of three-dimensional joint position information corresponding to the plurality of two-dimensional joint position information; and calculating gait parameters for the subject using the three-dimensional joint position information.

## Claims

### Claim 1

A gait analysis method for analyzing gait of a subject, the method comprising:
collecting a two-dimensional gait video of a subject's gait situation during a preset time, which is captured using a camera;
generating a three-dimensional coordinate system based on at least one frame of the two-dimensional gait video by receiving three-dimensional information of a capturing space;
calculating three-dimensional feature point coordinates by receiving a plurality of feature points on the three-dimensional coordinate system, and calculating two-dimensional feature point coordinates for the plurality of feature points using resolution information of the camera;
estimating mapping information between the three-dimensional feature point coordinates and the two-dimensional feature point coordinates by using the three-dimensional feature point coordinates, the two-dimensional feature point coordinates, and an internal index of the camera
calculating a plurality of two-dimensional joint position information for the subject from the two-dimensional gait video;
calculating a plurality of three-dimensional joint position information corresponding to the plurality of two-dimensional joint position information using the plurality of two-dimensional joint position information, the internal index of the camera, and the mapping information; and
calculating gait parameters for the subject using the three-dimensional joint position information.

### Claim 2 (depends on 1)

The gait analysis method of claim 1, wherein the collecting of a two-dimensional gait video includes:
collecting two-dimensional gait videos obtained by capturing a first gait section in which the subject walks forward in a straight line a preset distance from a starting point toward the camera, a second gait section in which the subject walks to turn a direction while passing through a direction turning point after the first gait section, and a third gait section in which the subject walks backward to return to the starting point after the second gait section.

### Claim 3 (depends on 2)

The gait analysis method of claim 2, wherein the generating of a three-dimensional coordinate system includes:
receiving three-dimensional information of a capturing space including a first parameter that represents a width of a floor corresponding to a starting point within the capturing space, a second parameter that represents a height of a wall in contact with the width of the floor, and a third parameter that represents a distance between the starting point and the direction turning point;
receiving a starting point of the three-dimensional coordinate system in the capturing space; and
generating the three-dimensional coordinate system using three-dimensional information of the capturing space and the starting point.

### Claim 4 (depends on 2)

The gait analysis method of claim 2, wherein the calculating of a plurality of two-dimensional joint position information includes:
extracting position information of both feet including the big toe, little toe, and ankle, and position information of the both shoulders from the plurality of two-dimensional joint position information.

### Claim 5 (depends on 4)

The gait analysis method of claim 4, wherein the calculating of gait parameters includes:
generating a velocity graph for the both feet using the three-dimensional position information of the both feet, and the distance and time corresponding to those between the start of the first gait section to the completion of the third gait section;
setting a first period in which one foot moves at the fastest velocity as a swing phase on the velocity graph, and a second period in which the other foot comes into contact with the ground corresponding to the first period as a stance phase; and
calculating gait parameters including one or more of step length, gait velocity, number of strides per hour, variability of step length and gait velocity, and difference between left and right step lengths using the swing phase and the stance phase for each of both feet.

### Claim 6 (depends on 5)

The gait analysis method of claim 5, wherein, in the calculating of gait parameters,
the length is calculated as a distance between two stance phases for one foot,
the gait velocity is calculated as a sum of the step lengths per unit time,
the number of strides is calculated as the number of swing phases per unit time,
the variability of step length and gait velocity is calculated by dividing a standard deviation of the step length and the gait velocity by an average step length or an average gait velocity, and
the difference between the left and right step lengths is calculated as a difference between the step length calculated for the left foot and the step length calculated for the right foot of the both feet.

### Claim 7 (depends on 6)

The gait analysis method of claim 6, wherein the calculating of gait parameters includes:
calculating direction turning information including the direction turning time from the second gait section and the number of strides at the direction turning time using the three-dimensional position information of both shoulders; and
providing the direction turning information to the gait parameters.

### Claim 8 (depends on 7)

The gait analysis method of claim 7, wherein the calculating of direction turning information includes:
generating a change graph of the distance between both shoulder joints using the three-dimensional position information of the both shoulders and the distance between both shoulder joints according to time using the time corresponding between the start of the first gait section to the completion of the third gait section;
extracting a first point at which the distance between the both shoulder joints gradually increases and vertically descends from a first value to a second value, and a second point at which the distance between the both shoulder joints vertically rises from the second value to a third value on the change graph;
calculating a difference between a first time corresponding to the first point and a second time corresponding to the second point as the direction turning time; and
calculating the number of swing phases corresponding between the first time and the second time as the number of strides during the turning of the direction on the velocity graph for the both feet.

### Claim 9 (depends on 1)

The gait analysis method of claim 1, wherein the collecting of a two-dimensional gait video further includes:
allowing collection of two-dimensional video data that satisfies the preset condition by determining whether the two-dimensional gait video satisfies a preset condition, and
the preset condition includes a first condition in which the camera has to capture an image at a fixed position, a second condition in which the subject's feet has to be included in the two-dimensional gait video, and a third condition in which a zoom-in or zoom-out operation is not performed while capturing the two-dimensional gait video.

### Claim 10 (depends on 1)

The gait analysis method of claim 1, wherein in the estimating of mapping information,
the internal index of the camera includes focal length information given based on resolution of the camera and center information of a two-dimensional gait video frame included in the two-dimensional gait video, and
the mapping information includes rotation information of the camera and horizontal movement information of the camera.

### Claim 11

A gait analysis apparatus for analyzing gait of a subject, the apparatus comprising:
a processor; and
a memory that is operatively connected to the processor and stores one more instructions executed by the processor,
wherein when the memory is executed through the processor, the instruction cause the processor to
collect a two-dimensional gait video of a subject's gait situation during a preset time, which is captured using a camera,
generate a three-dimensional coordinate system based on at least one frame of the two-dimensional gait video by receiving three-dimensional information of a capturing space,
calculate three-dimensional feature point coordinates by receiving a plurality of feature points on the three-dimensional coordinate system, and calculate two-dimensional feature point coordinates for the plurality of feature points using resolution information of the camera,
estimate mapping information between the three-dimensional feature point coordinates and the two-dimensional feature point coordinates by using the three-dimensional feature point coordinates, the two-dimensional feature point coordinates, and an internal index of the camera
calculate a plurality of two-dimensional joint position information for the subject from the two-dimensional gait video,
calculate a plurality of three-dimensional joint position information corresponding to the plurality of two-dimensional joint position information using the plurality of two-dimensional joint position information, the internal index of the camera, and the mapping information, and
calculate gait parameters for the subject using the three-dimensional joint position information.

### Claim 12 (depends on 11)

The gait analysis apparatus of claim 11, wherein
when collecting the two-dimensional gait video, the instruction causes the processor to collect two-dimensional gait videos obtained by capturing a first gait section in which the subject walks forward in a straight line a preset distance from a starting point toward the camera, a second gait section in which the subject walks to turn a direction while passing through a direction turning point after the first gait section, and a third gait section in which the subject walks backward to return to the starting point after the second gait section.

### Claim 13 (depends on 12)

The gait analysis apparatus of claim 12, wherein,
when generating the three-dimensional coordinate system, the instruction causes the processor to
receive three-dimensional information of a capturing space including a first parameter that represents a width of a floor corresponding to a starting point within the capturing space, a second parameter that represents a height of a wall in contact with the width of the floor, and a third parameter that represents a distance between the starting point and the direction turning point,
receive a starting point of the three-dimensional coordinate system in the capturing space, and
generate the three-dimensional coordinate system using three-dimensional information of the capturing space and the starting point.

### Claim 14 (depends on 12)

The gait analysis apparatus of claim 12, wherein
when calculating the plurality of two-dimensional joint position information, the instruction causes the processor to extract position information of both feet including the big toe, little toe, and ankle, and position information of the both shoulders from the plurality of two-dimensional joint position information.

### Claim 15 (depends on 14)

The gait analysis apparatus of claim 14, wherein,
when calculating the gait parameters, the instruction causes the processor to
generate a velocity graph for the both feet using the three-dimensional position information of the both feet, and the distance and time corresponding to those between the start of the first gait section to the completion of the third gait section,
set a first period in which one foot moves at the fastest velocity as a swing phase on the velocity graph, and a second period in which the other foot comes into contact with the ground corresponding to the first period as a stance phase, and
calculate gait parameters including one or more of step length, gait velocity, number of strides per hour, variability of step length and gait velocity, and difference between left and right step lengths using the swing phase and the stance phase for each of both feet.

### Claim 16 (depends on 15)

The gait analysis apparatus of claim 15, wherein
when calculating the gait parameters, the instruction causes the processor to make the step length be calculated as a distance between two stance phases for one foot, the gait velocity be calculated as a sum of the step lengths per unit time, the number of strides be calculated as the number of swing phases per unit time, the variability of step length and gait velocity be calculated by dividing a standard deviation of the step length and the gait velocity by an average step length or an average gait velocity, and the difference between the left and right step lengths be calculated as a difference between the step length calculated for the left foot and the step length calculated for the right foot of the both feet.

### Claim 17 (depends on 16)

The gait analysis apparatus of claim 16, wherein,
when calculating the gait parameters, the instruction causes the processor to
calculate direction turning information including the direction turning time from the second gait section and the number of strides at the direction turning time using the three-dimensional position information of both shoulders, and
provide the direction turning information to the gait parameters.

### Claim 18 (depends on 17)

The gait analysis apparatus of claim 17, wherein,
when calculating the direction turning information, the instruction causes the processor to
generate a change graph of the distance between both shoulder joints using the three-dimensional position information of the both shoulders and the distance between both shoulder joints according to time using the time corresponding between the start of the first gait section to the completion of the third gait section,
extract a first point at which the distance between the both shoulder joints gradually increases and vertically descends from a first value to a second value, and a second point at which the distance between the both shoulder joints vertically rises from the second value to a third value on the change graph,
calculate a difference between a first time corresponding to the first point and a second time corresponding to the second point as the direction turning time, and
calculate the number of swing phases corresponding between the first time and the second time as the number of strides during the turning of the direction on the velocity graph for the both feet.

### Claim 19 (depends on 11)

The gait analysis apparatus of claim 11, wherein
when collecting the two-dimensional gait video, the instruction causes the processor to determine whether the two-dimensional gait video satisfies a preset condition to allow collection of two-dimensional video data that satisfies the preset condition, and
the preset condition includes a first condition in which the camera has to capture an image at a fixed position, a second condition in which the subject's feet has to be included in the two-dimensional gait video, and a third condition in which a zoom-in or zoom-out operation is not performed while capturing the two-dimensional gait video.

### Claim 20 (depends on 11)

The gait analysis apparatus of claim 11, wherein,
when estimating the mapping information, the instruction causes the processor to
provide the internal index of the camera with focal length information given based on resolution of the camera and center information of a two-dimensional gait video frame included in the two-dimensional gait video, and
provide the mapping information with rotation information of the camera and horizontal movement information of the camera.