CN111582081A - Multi-Kinect serial gait data space-time combination method and measuring device - Google Patents

Multi-Kinect serial gait data space-time combination method and measuring device Download PDF

Info

Publication number
CN111582081A
CN111582081A CN202010335064.2A CN202010335064A CN111582081A CN 111582081 A CN111582081 A CN 111582081A CN 202010335064 A CN202010335064 A CN 202010335064A CN 111582081 A CN111582081 A CN 111582081A
Authority
CN
China
Prior art keywords
kinect
gait data
gait
kinects
data
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202010335064.2A
Other languages
Chinese (zh)
Inventor
刘跃虎
何晓娟
马霜逊
陈成成
张翰桢
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xian Jiaotong University
Original Assignee
Xian Jiaotong University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Xian Jiaotong University filed Critical Xian Jiaotong University
Priority to CN202010335064.2A priority Critical patent/CN111582081A/en
Publication of CN111582081A publication Critical patent/CN111582081A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/103—Static body considered as a whole, e.g. static pedestrian or occupant recognition
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
    • A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
    • A61B5/1116—Determining posture transitions
    • A61B5/1117—Fall detection
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
    • A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
    • A61B5/112—Gait analysis

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Medical Informatics (AREA)
  • Biomedical Technology (AREA)
  • Veterinary Medicine (AREA)
  • Physiology (AREA)
  • Dentistry (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Public Health (AREA)
  • Biophysics (AREA)
  • Pathology (AREA)
  • General Health & Medical Sciences (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • Molecular Biology (AREA)
  • Surgery (AREA)
  • Human Computer Interaction (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)

Abstract

The invention discloses a multi-Kinect series connection gait data space-time merging method and a measuring device, which comprise at least three Kinects, wherein one Kinect is arranged right opposite to a footpath, the other Kinects are arranged on the same side of the footpath, the acquisition range between two adjacent Kinects arranged on the right side of the footpath has an intersection, each Kinect is connected with a computer, and the Kinects are used for acquiring three-dimensional coordinate values of human body joint points; and the computer is used for processing the three-dimensional coordinate value acquired by the Kinect to obtain the combined gait data. The method can not only enlarge the data acquisition range and increase the length of the gait data sequence, but also avoid the potential safety risk that patients with walking obstacles or old people may fall down on the treadmill due to a non-invasive and non-contact data acquisition method.

Description

Multi-Kinect serial gait data space-time combination method and measuring device
Technical Field
The invention belongs to the technical field of gait analysis, and particularly relates to a multi-Kinect serial gait data space-time combination method and a measuring device.
Background
Walking is one of the key features for distinguishing human beings from other animals, normal walking does not need thinking, however, the control of walking is very complicated, and involves the coordinated movement of a plurality of links, and the gait can be influenced by the disorder of any link. Therefore, correct and effective gait analysis is particularly important.
The gait analysis is to perform objective qualitative analysis or quantitative analysis on the state of the human walking function by utilizing a mechanical concept and a kinematics means, is beneficial to evaluating the rehabilitation function state of a patient with neurological diseases, and is helpful to making a rehabilitation treatment scheme and evaluating the rehabilitation curative effect. Due to the remote non-contact recognizable characteristic of gait, gait analysis has become a research hotspot in the field of computer vision as an emerging biometric analysis recognition technology.
At this stage, physicians clinically assess human gait by using standardized scales and tests, and make diagnostic results based on patient results compared to results from matching healthy samples. The traditional mode is easy to operate and short in time consumption, but the traditional mode has poor accuracy and is influenced by subjective factors of doctors, and a unified standard cannot be established. In view of these problems, in recent years, researchers have proposed a method for accurately estimating spatiotemporal parameters through ground reaction force during walking, which can completely and relatively simply change gait and time phase, and the corresponding instruments are a footpath pad and a pressure plate, however, estimating kinematic parameters requires tracking specific positions of a plurality of joint points during a test process, and the footpath pad cannot meet the requirement, because in the process of movement, a swing phase is performed in suspension and cannot be in contact with the ground at this time, so the gait pad and the pressure plate cannot capture a complete gait cycle. And the footpath pad and the pressure plate have single capture information and cannot completely record the gait information. The development of some sensors attached to specific body parts is indirectly facilitated. Among them, optical motion tracking has become the most common method for accurately studying motion gait parameters. Although optical instruments can perform accurate spatio-temporal and kinematic analysis, they are usually used in large laboratories or large clinical centers due to their high cost, which is not conducive to the popularization and promotion of gait analysis techniques.
Therefore, the following problems are mainly to be solved: 1. reducing the price of measuring equipment, optical motion tracking has become the most common method for accurately studying motion gait parameters at present. Although optical instruments can perform accurate spatio-temporal and kinematic analysis, they are often used in large laboratories or large clinical centers due to their high cost, and therefore, it is necessary to reduce the price of the measuring equipment to further facilitate widespread use of gait analysis; 2. the measurement precision is improved, so that the accuracy and the reliability of gait analysis can be improved; 3. the experimental steps are simplified, the operability is improved, and the repeatability of the experiment is improved; 4. the natural flat ground measurement is selected, so that the safety risk of the treadmill on the falling of the old, children, pregnant women and other people is eliminated; 5. the method and the device realize the measurement of long-distance natural gait sequences and reduce the randomness caused by too short measurement distance.
Disclosure of Invention
The invention provides a multi-Kinect serial gait data space-time merging method and a measuring device, which can expand the data acquisition range and increase the length of a gait data sequence, and avoid the safety risk of possible falling when walking on a treadmill due to a non-invasive and non-contact data acquisition method. The reliability and the accuracy of gait analysis and identification are greatly improved, and no special limitation is imposed on a testee.
In order to achieve the above object, the gait measuring device with multiple Kinect connected in series according to the present invention comprises at least two Kinect, wherein: one Kinect is arranged right opposite to the footpath, and the other Kinects are arranged on the same side of the footpath; the method comprises the following steps that an intersection exists in the acquisition range between two adjacent Kinects placed on the right side of a footpath, each Kinect is connected with a computer, and the Kinects are used for acquiring three-dimensional coordinate values of human body joint points; and the computer is used for processing the three-dimensional coordinate value acquired by the Kinect to obtain the combined gait data.
Further, the lens of the Kinect placed on one side of the footpath makes an angle of 45 degrees with the footpath.
Further, the Kinect is mounted on a support bracket with adjustable height and angle.
Furthermore, computers connected with different Kinects communicate through a wireless local area network.
A gait data space-time combination method of multiple Kinects connected in series based on the gait measuring device is characterized by comprising the following steps:
step 1, initializing Kinect parameters: setting a Kinect to acquire three-dimensional coordinate values of human body joint points at a rate of 30 frames/second;
step 2, carrying out data acquisition on the same testee by using a plurality of Kinects to acquire gait data of the testee, wherein the gait data is a three-dimensional coordinate value of a human body joint point;
step 3, performing coordinate conversion on the gait data acquired by the Kinects, and converting the gait data acquired by the Kinects into a target coordinate system to obtain the gait data after coordinate conversion;
and 4, dividing the gait data obtained in the step 3 after coordinate conversion into two parts, wherein one part is the gait data only acquired by the Kinect, the other part is the gait data acquired simultaneously with the adjacent Kinect, and then combining all the gait data acquired simultaneously with the adjacent Kinect to obtain long-distance natural gait data.
Further, in step 3, gait data in an optimal measuring range of the Kinect is selected for coordinate conversion, wherein the optimal measuring range refers to that the distance from the testee to the Kinect is 1.8m-3.5 m.
Further, in step 3, the target coordinate system is the coordinate system of the Kinect facing the footpath.
Further, in step 3, the coordinate transformation formula is:
Figure BDA0002466273330000031
Ra,bis the rotation matrix between Kinect a and Kinect b, τa,bAre the translation vectors of Kinect a and Kinect b,
Figure BDA0002466273330000032
is a three-dimensional coordinate value of a human body joint point acquired by Kinect a,
Figure BDA0002466273330000033
the value is obtained after coordinate conversion of three-dimensional coordinate values of human body joint points acquired by Kinect a, and j represents the frame number, wherein the Kinect a is adjacent to the Kinect b.
Further, in step 4, the process of merging the gait data of each frame of two adjacent Kinect is as follows:
Figure BDA0002466273330000034
Figure BDA0002466273330000035
in the formula, the first step is that,
Figure BDA0002466273330000036
a frame of gait data in an overlapped area after the gait data are merged, Z is the gait data after the gait data are merged, vn-1Is the proportion of Kinect # (n-1) in the overlap region, vnIs the proportion of Kinect # (n) in the overlap region; in order to balance the gait data collected by the two Kinects; when v isn＝vn-1When the content is 0.5, good effect can be obtained;
Figure BDA0002466273330000037
is gait data that only Kinect # (n-1) can collect,
Figure BDA0002466273330000038
is gait data collected by Kinect # (n-1) in the overlap region,
Figure BDA0002466273330000041
is gait data that only Kinect # n can collect,
Figure BDA0002466273330000042
merging the gait data to obtain the gait data of the overlapped area;
Figure BDA0002466273330000043
is gait data collected by Kinect # n in the overlap region,
Figure BDA0002466273330000044
is gait data collected by Kinect # (n-1) in the overlap region.
Compared with the prior art, the invention has at least the following beneficial technical effects:
a gait measuring device with multiple Kinects connected in series can enlarge the data acquisition range, increase the length of a gait data sequence and finally obtain long-distance, stable and natural gait data, so that the reliability and the accuracy of gait analysis are greatly improved.
The expansibility is very strong. If the field and the equipment are not limited, the number of Kinects can be superposed infinitely, namely the measuring distance can be lengthened infinitely. And the used hardware equipment has low price, the main equipment used comprises a Kinect, a computer and a Kinect bracket, and the total cost is low, so that the gait analysis is very favorable for popularization and generalization.
The Kinect can directly capture the object to be measured, so that a marker does not need to be pasted during data measurement, relatively accurate human body joint point coordinate values can still be obtained, and errors caused by pasting the marker are eliminated;
a multi-Kinect serial gait data space-time merging method is characterized in that a collection site selects indoor flat ground to collect gait, the method is a non-invasive and non-contact data collection method, meanwhile, in order to ensure that a tested person is always in the optimal measurement range of Kinect, a treadmill is used for replacing the flat ground, however, the requirement of the treadmill on the tested person is high, the method is very limited to be suitable for the tested person, the potential safety risk that a patient with walking disorder or an old person can fall down on the treadmill can be avoided, and the method is suitable for various types of tested persons, for example: the old, pregnant women, patients and the like, and has wide applicability. The invention only uses the depth value of the human body joint point in the coordinate conversion process, does not need other auxiliary data, simplifies the operation process and is easy to realize.
Furthermore, only gait data within the Kinect optimal test range is selected when gait data are collected, so that the reliability and the accuracy of gait analysis are improved;
drawings
FIG. 1 illustrates a layout relationship of hardware devices according to the present invention;
FIG. 2 is an overlapping relationship of adjacent Kinect measurement ranges;
FIG. 3 is the process of Kinect measuring the body joint points;
FIG. 4 is a data flow state of the neighboring Kinect before merging of gait data;
FIG. 5 shows the flow state of the neighboring Kinect after merging of gait data;
FIG. 6 shows Z-axis coordinate values of three Kinects before merging of gait data in the invention;
FIG. 7 shows Z-axis coordinates of three Kinects in the present invention after merging gait data.
Detailed Description
In order to make the objects and technical solutions of the present invention clearer and easier to understand. The present invention will be described in further detail with reference to the following drawings and examples, wherein the specific examples are provided for illustrative purposes only and are not intended to limit the present invention.
Referring to fig. 1, the gait measuring device with multiple Kinects connected in series comprises at least two Kinects, wherein one Kinect is arranged right opposite to a footpath, the other Kinects are arranged on the same side of the footpath, and a lens of the Kinect arranged on one side of the footpath forms an angle of 45 degrees with the footpath. The method comprises the following steps that an intersection exists in the acquisition range between two adjacent Kinects placed on the right side of a footpath, each Kinect is connected with a computer, and the Kinects are used for acquiring three-dimensional coordinate values of human body joint points; and the computer is used for processing the three-dimensional coordinate value acquired by the Kinect to obtain the combined gait data. The Kinect is installed on a supporting bracket with adjustable height and angle. And the computers connected with different Kinects communicate through the wireless local area network.
Referring to fig. 2 to 5, a method for spatio-temporal merging of gait data of multiple Kinect tandem mainly includes the following steps (in the present invention, the number of Kinect may be infinitely superposed, and since the field is limited, only four Kinect are taken as an example here):
step 1, establishing a hardware environment for gait data acquisition:
the hardware equipment mainly comprises four Kinects, a supporting bracket capable of adjusting height and angle and four computers. The hardware of this embodiment is placed in a manner as shown in fig. 1, each Kinect needs to be connected to a computer for measuring gait data, one Kinect is placed on the front side of the footpath, the rest of the kinects are placed on the right side of the footpath, and the lens of the Kinect forms an angle of 45 degrees with the footpath. Because at 45 degrees, the Kinect is better able to track the walking subject and this position is the best position to capture sagittal gait data. Meanwhile, the distance between the Kinect and the ground is 0.8m, the vertical distance from the center of the footpath is 1.5m, the acquisition range between two adjacent Kinects placed on the right side of the footpath has intersection, and the horizontal distance between two adjacent Kinects placed on the right side of the footpath is 2 m.
Meanwhile, in order to improve the accuracy of gait data measurement, the optimal measuring range of the Kinect is introduced into the gait acquisition environment design for the first time, according to research, when an interaction environment is designed, the optimal acquiring range of the Kinect is within the range of 3.5 meters, the height of a tested person is considered, the embodiment selects the distance between the tested person and the Kinect, which is 1.8-3.5 meters, as the optimal measuring range of a gait sequence, and the specific acquiring range is shown in figure 2. In fig. 2, L is a depth recognition range of the Kinect, S1 is an overlapping region of the Kinect #1 and the Kinect #2, S2 is an overlapping region of the Kinect #2 and the Kinect #3, S3 is an overlapping region of the Kinect #3 and the Kinect #4, and θ is a horizontal view angle of the Kinect.
Step 2, initializing Kinect parameters: a measured speed of 30 frames/second is set, and during the gait data measurement, the Kinect collects data of the human joint points at a rate of 30 frames/second. In order not to influence the walking state of the testee, different testees only need to walk on the footpath at different speeds and at uniform speeds in the experiment.
And 3, obtaining three-dimensional coordinate values of human body joint points of the testee by using a plurality of Kinects:
kinect is originally applied to Xbox360 peripheral equipment, is developed mainly for motion sensing games, but because they can collect real-time color data and depth information, so it is often used in three-dimensional reconstruction, human skeleton extraction and gait recognition, through Kinect can get the 25 human body joint point three-dimensional coordinates, at the same time, Kinect reads the human body joint point three-dimensional coordinates at 30 frames/second rate, to complete the continuous measurement of gait data;
however, one Kinect can only be connected with one computer, and in order to realize the simultaneous start and stop of three computers, local area network communication is established among different computers. The method is to establish a TCP/IP protocol in the same wireless local area network and utilize a socket to carry out communication. Therefore, a plurality of Kinects can simultaneously acquire continuous gait data, and the tracking state of the Kinects on human body joint points is shown in fig. 3.
The invention measures 30 testees, and then selects gait data in the optimal measuring range of Kinect to construct a human body gait database, which is also natural gait data of the testees to the maximum extent. And then, the gait data of each testee is processed in the steps 4 to 5, and the combined gait data of each testee can be obtained.
And 4, performing coordinate conversion (completing time consistency) on the gait data acquired by the Kinects:
after the gait data are collected, in order to maintain the spatial consistency of the gait data, a very important step is to solve the problem of coordinate conversion among a plurality of Kinects. Data are acquired on the same target object by adopting a plurality of Kinects, each data set corresponds to a coordinate system under the camera of the data set, so that the data cannot be directly operated, and the data need to be converted into a certain target coordinate system. As shown in fig. 1, in the present embodiment, the Kinect facing the walkway is selected as the target coordinate system, and the rest of the Kinect are all source coordinate systems. That is, the data in all source coordinate systems is converted to be under the target coordinate system. The reasons for this selection are mainly the following: firstly, the Kinect has strong recognition capability on the human body, and the self-shielding limit can be reduced as much as possible when the human body is over against the Kinect; secondly, gait data acquired during Kinect driving is better reflected to symmetry parameters in gait properties; thirdly, when a plurality of Kinect gait data are spliced later, the depth value extending along the driving direction is used, so that the depth values of all the gait data are converted to be parallel to the driving direction.
The method used by the invention directly utilizes the depth information, namely the three-dimensional coordinates of the human body joint points under two Kinect coordinate systems, and directly calculates the space transformation of the human body joint points according to the collected three-dimensional coordinate information. The calculation process is as follows:
Figure BDA0002466273330000071
in the formula
Figure BDA0002466273330000072
The gait data collected for the target coordinate system,
Figure BDA0002466273330000073
is gait data collected by a source coordinate system, R is a rotation matrix between two Kinects, tau is a translation vector,
Figure BDA0002466273330000074
is the three-dimensional coordinates of the human body joint points acquired by a target coordinate system, wherein i is the number of Kinects. By the above formula, the problem of coordinate transformation can be converted into solving the least square transformation between two three-dimensional point sets, the solution can be rapidly carried out by using singular value decomposition, and the detailed calculation process is not repeated as the least square transformation is a very general method.
Therefore, under the condition that the three-dimensional coordinates of the human body joint points measured by the two Kinects are known, the rotation matrix and the translation vector between the two Kinect coordinate systems can be obtained. The solving method is not influenced by the self-calibration precision of the camera depth lens, and the calculating speed is relatively high. After the rotation matrix and the translation vector between the adjacent Kinects are obtained, the converted gait data can be obtained through the following formula, and the conversion relationship between every two adjacent Kinects is the same, so that only one group of adjacent Kinects is taken as an example in the invention, and the specific conversion relationship is as follows:
Figure BDA0002466273330000081
in the formula, Rn,n-1Is a rotation matrix, tau, between the nth Kinect and the (n-1) th Kinect calculated by a least square methodn,n-1Is the calculated translation vector and the translation vector,
Figure BDA0002466273330000082
is the three-dimensional coordinate value of the collected human body joint point,
Figure BDA0002466273330000083
is the gait data after coordinate transformation, j represents the frame number.
And 5, merging gait data (completing space consistency):
after coordinate conversion, as the acquisition ranges of adjacent Kinects have overlapped areas, the acquired gait data have repeated stages in time and the z axis, and therefore, in order to ensure the time continuity of the data, the gait data acquired by the adjacent Kinects must be merged in the overlapped areas.
As shown in fig. 4, the depth value can be directly obtained by the Kinect, and the spine base (spine center point) in the collected joint points is a joint point which is relatively stable in the walking process, so that the data is divided by using the depth value of the joint point, and the gait data collected by the Kinect is divided into two parts according to whether the depth values of the adjacent Kinect coincide or not: one part is the gait data collected by only the Kinect, and the other part is the gait data collected simultaneously with the adjacent Kinect (namely the data with the same depth value).
After the gait data division is finished, since the collection ranges of adjacent Kinect have an overlapping region, as shown in fig. 5, in order to ensure the time continuity of the data, it is necessary to obtain the long-distance natural gait data by merging the gait sequences in the overlapping region. The process of merging each frame is as follows:
Figure BDA0002466273330000084
Figure BDA0002466273330000085
in the formula, the first step is that,
Figure BDA0002466273330000086
the gait data of the j-th frame in the overlapped area after the gait data are merged, Z is the gait data after the gait data are merged, and the corresponding data flow after the gait data are merged is shown in fig. 7. v. ofn-1Is the proportion of Kinect # (n-1) in the overlap region, vnIs the proportion of Kinect # (n) in the overlap region. In order to balance the gait data collected by two adjacent Kinects. Good results were obtained at that time.
Figure BDA0002466273330000091
Is gait data that only Kinect # (n-1) can collect,
Figure BDA0002466273330000092
is gait data collected by Kinect # (n-1) in the overlap region,
Figure BDA0002466273330000093
is gait data that only Kinect # n can collect,
Figure BDA0002466273330000094
merging the gait data to obtain the gait data of the overlapped area;
Figure BDA0002466273330000095
is gait data collected by Kinect # n in the overlap region,
Figure BDA0002466273330000096
is gait data collected by Kinect # (n-1) in the overlap region. Since the merging method is the same between all adjacent Kinect, only one group of adjacent Kinect is taken as an example for illustration.
FIG. 6 shows Z-axis coordinate values of hip joints corresponding to three Kinects before spatio-temporal merging of gait data, which are shown to be discontinuous; the gait data collected by the Kinect can be combined by the method of the invention, as shown in fig. 7, so that the invention achieves the aim of collecting long-distance natural gait sequences.
The above-mentioned contents are only for illustrating the technical idea of the present invention, and the protection scope of the present invention is not limited thereby, and any modification made on the basis of the technical idea of the present invention falls within the protection scope of the claims of the present invention.

Claims (9)

1. A gait measurement device with multiple Kinects connected in series is characterized by comprising at least two Kinects, wherein: one Kinect is arranged right opposite to the footpath, and the other Kinects are arranged on the same side of the footpath; the method comprises the following steps that an intersection exists in the acquisition range between two adjacent Kinects placed on the right side of a footpath, each Kinect is connected with a computer, and the Kinects are used for acquiring three-dimensional coordinate values of human body joint points; and the computer is used for processing the three-dimensional coordinate value acquired by the Kinect to obtain the combined gait data.
2. A gait measurement device with multiple Kinect in series as claimed in claim 1, wherein the lens of the Kinect placed on one side of the footpath is at an angle of 45 degrees to the footpath.
3. A gait measurement device with multiple Kinect in series connection as claimed in claim 1, wherein said Kinect is mounted on a height and angle adjustable support bracket.
4. A gait measurement device with multiple Kinect in series as claimed in claim 1, wherein said computers connected to different Kinect communicate via wireless lan.
5. A gait data space-time combination method of multiple Kinect tandem based on the gait measuring device of claim 1, characterized by comprising the following steps:
step 1, initializing Kinect parameters: setting a Kinect to acquire three-dimensional coordinate values of human body joint points at a rate of 30 frames/second;
step 2, carrying out data acquisition on the same testee by using a plurality of Kinects to acquire gait data of the testee, wherein the gait data is a three-dimensional coordinate value of a human body joint point;
step 3, performing coordinate conversion on the gait data acquired by the Kinects, and converting the gait data acquired by the Kinects into a target coordinate system to obtain the gait data after coordinate conversion;
and 4, dividing the gait data obtained in the step 3 after coordinate conversion into two parts, wherein one part is the gait data only acquired by the Kinect, the other part is the gait data acquired simultaneously with the adjacent Kinect, and then combining all the gait data acquired simultaneously with the adjacent Kinect to obtain long-distance natural gait data.
6. The method as claimed in claim 5, wherein in step 3, the gait data in the optimal measuring range of Kinect is selected for coordinate transformation, and the optimal measuring range is 1.8m-3.5m away from Kinect.
7. The method as claimed in claim 5, wherein in step 3, the target coordinate system is the coordinate system of Kinect directly facing the footpath.
8. The method as claimed in claim 5, wherein in step 3, the coordinate transformation formula is:
Figure FDA0002466273320000021
Ra,bis the rotation matrix between Kinect a and Kinect b, τa,bAre the translation vectors of Kinect a and Kinect b,
Figure FDA0002466273320000022
is a three-dimensional coordinate value of a human body joint point acquired by Kinect a,
Figure FDA0002466273320000023
the value is obtained after coordinate conversion of three-dimensional coordinate values of human body joint points acquired by Kinect a, and j represents the frame number, wherein Kinect a is adjacent to Kinect b.
9. The method as claimed in claim 5, wherein in step 4, the gait data of each frame of two adjacent Kinects are merged as follows:
Figure FDA0002466273320000024
Figure FDA0002466273320000025
in the formula, the first step is that,
Figure FDA0002466273320000026
a frame of gait data in an overlapped area after the gait data are merged, Z is the gait data after the gait data are merged, vn-1Is the proportion of Kinect # (n-1) in the overlap region, vnIs the proportion of Kinect # (n) in the overlap region; in order to balance the gait data collected by the two Kinects; when v isn＝vn-1When the content is 0.5, good effect can be obtained;
Figure FDA0002466273320000027
only Kinect # (n-1) can be collectedThe data of the gait of the arrival time,
Figure FDA0002466273320000028
is gait data collected by Kinect # (n-1) in the overlap region,
Figure FDA0002466273320000029
is gait data that only Kinect # n can collect,
Figure FDA00024662733200000210
merging the gait data to obtain the gait data of the overlapped area;
Figure FDA00024662733200000211
is gait data collected by Kinect # n in the overlap region,
Figure FDA00024662733200000212
is gait data collected by Kinect # (n-1) in the overlap region.
CN202010335064.2A 2020-04-24 2020-04-24 Multi-Kinect serial gait data space-time combination method and measuring device Pending CN111582081A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010335064.2A CN111582081A (en) 2020-04-24 2020-04-24 Multi-Kinect serial gait data space-time combination method and measuring device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010335064.2A CN111582081A (en) 2020-04-24 2020-04-24 Multi-Kinect serial gait data space-time combination method and measuring device

Publications (1)

Publication Number Publication Date
CN111582081A true CN111582081A (en) 2020-08-25

Family

ID=72116765

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010335064.2A Pending CN111582081A (en) 2020-04-24 2020-04-24 Multi-Kinect serial gait data space-time combination method and measuring device

Country Status (1)

Country Link
CN (1) CN111582081A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113223056A (en) * 2021-06-01 2021-08-06 北京理工大学 Unmarked motion tracking system based on multiple Kinect V2
CN114722939A (en) * 2022-04-07 2022-07-08 山东大学 Data expansion method and system for gait analysis

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140303937A1 (en) * 2013-04-05 2014-10-09 The Boeing Company Creating Ergonomic Manikin Postures and Controlling Computer-Aided Design Environments Using Natural User Interfaces
CN104933734A (en) * 2015-06-26 2015-09-23 西安理工大学 Multi-Kinect-based human body gesture data fusion method
CN107341844A (en) * 2017-06-21 2017-11-10 上海大学 A kind of real-time three-dimensional people's object plotting method based on more Kinect
CN107577451A (en) * 2017-08-03 2018-01-12 中国科学院自动化研究所 More Kinect human skeletons coordinate transformation methods and processing equipment, readable storage medium storing program for executing
CN107993249A (en) * 2017-08-23 2018-05-04 北京航空航天大学 A kind of body gait data fusion method based on more Kinect
CN108564653A (en) * 2018-03-16 2018-09-21 中国传媒大学 Human skeleton tracing system and method based on more Kinect
CN108572731A (en) * 2018-03-16 2018-09-25 中国科学院自动化研究所 Dynamic based on more Kinect and UE4 catches Data Representation method and device
CN109166134A (en) * 2018-08-17 2019-01-08 安徽信息工程学院 Human body dynamic three-dimensional reconstruction method based on more Kinect
CN111035393A (en) * 2019-12-13 2020-04-21 中国科学院深圳先进技术研究院 Three-dimensional gait data processing method, system, server and storage medium

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140303937A1 (en) * 2013-04-05 2014-10-09 The Boeing Company Creating Ergonomic Manikin Postures and Controlling Computer-Aided Design Environments Using Natural User Interfaces
CN104933734A (en) * 2015-06-26 2015-09-23 西安理工大学 Multi-Kinect-based human body gesture data fusion method
CN107341844A (en) * 2017-06-21 2017-11-10 上海大学 A kind of real-time three-dimensional people's object plotting method based on more Kinect
CN107577451A (en) * 2017-08-03 2018-01-12 中国科学院自动化研究所 More Kinect human skeletons coordinate transformation methods and processing equipment, readable storage medium storing program for executing
CN107993249A (en) * 2017-08-23 2018-05-04 北京航空航天大学 A kind of body gait data fusion method based on more Kinect
CN108564653A (en) * 2018-03-16 2018-09-21 中国传媒大学 Human skeleton tracing system and method based on more Kinect
CN108572731A (en) * 2018-03-16 2018-09-25 中国科学院自动化研究所 Dynamic based on more Kinect and UE4 catches Data Representation method and device
CN109166134A (en) * 2018-08-17 2019-01-08 安徽信息工程学院 Human body dynamic three-dimensional reconstruction method based on more Kinect
CN111035393A (en) * 2019-12-13 2020-04-21 中国科学院深圳先进技术研究院 Three-dimensional gait data processing method, system, server and storage medium

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
赵春晖等著: "《视频图像运动目标分析》", 30 June 2011, 北京：国防工业出版社 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113223056A (en) * 2021-06-01 2021-08-06 北京理工大学 Unmarked motion tracking system based on multiple Kinect V2
CN114722939A (en) * 2022-04-07 2022-07-08 山东大学 Data expansion method and system for gait analysis
CN114722939B (en) * 2022-04-07 2024-05-31 山东大学 Data expansion method and system for gait analysis

Similar Documents

Publication Publication Date Title
CN106580294A (en) Physiological signal remote monitoring system based on multimodal imaging technique and application thereof
CN107705848B (en) Method and system for recommending conditioning scheme according to health condition of user
CN107115102A (en) A kind of osteoarticular function appraisal procedure and device
CN111582081A (en) Multi-Kinect serial gait data space-time combination method and measuring device
Ohberg et al. Chronic whiplash associated disorders and neck movement measurements: an instantaneous helical axis approach
CN109745046A (en) A kind of electrical impedance imaging electrode and system suitable under motion state
KR100349338B1 (en) Clinical diagnosis system for orthopedic pathological disease using three-dimensional human motion measurement
CN115515480A (en) Disease detection system
CN116999057A (en) Hemiplegia gait recognition and hemiplegia gait evaluation method based on wearable sensor
CN210277171U (en) Health detection device with multiple sensor information fused in small biological force line station
Boldo et al. On the reliability of single-camera markerless systems for overground gait monitoring
Iosa et al. Assessment of waveform similarity in electromyographical clinical gait data: the linear fit method
CN113367688B (en) Hemiplegia rating system based on human body electrostatic gait and TOF camera
Rana et al. 3-D gait abnormality detection employing contactless IR-UWB sensing phenomenon
Matthew et al. Reachable workspace and proximal function measures for quantifying upper limb motion
CN113768471B (en) Parkinson disease auxiliary diagnosis system based on gait analysis
CN115105062A (en) Hip and knee joint coordination evaluation method, device and system and storage medium
CN114743664A (en) Gait analysis and determination learning-based spinal cervical spondylosis auxiliary diagnosis system
Zhu et al. Extraction and selection of gait recognition features using skeleton point detection and improved fuzzy decision
Gao Kinect-Based Gait Assessment Method for Hemiplegic Patients
Jamsrandorj et al. Deep Learning Networks for View-independent Knee and Elbow Joint Angle Estimation
CN209269681U (en) Quantitative blood flow Fraction analysis device
Ono et al. Assessment of Joint Range of Motion Measured by a Stereo Camera
Lin et al. Quantitative measurement of Parkinsonian gait from walking in monocular image sequences using a centroid tracking algorithm
CN204839527U (en) Gait analytical equipment

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20200825