CN111460908B - Human body fall recognition method and system based on OpenPose - Google Patents
Human body fall recognition method and system based on OpenPose Download PDFInfo
- Publication number
- CN111460908B CN111460908B CN202010148742.4A CN202010148742A CN111460908B CN 111460908 B CN111460908 B CN 111460908B CN 202010148742 A CN202010148742 A CN 202010148742A CN 111460908 B CN111460908 B CN 111460908B
- Authority
- CN
- China
- Prior art keywords
- human body
- human
- joint
- inclination angle
- openpose
- 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.)
- Active
Links
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
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/02—Alarms for ensuring the safety of persons
- G08B21/04—Alarms for ensuring the safety of persons responsive to non-activity, e.g. of elderly persons
Abstract
The invention discloses a human body tumbling identification method and system based on OpenPose, which comprises the steps of firstly obtaining a monitoring video shot by a monitoring camera, and processing the monitoring video by using the OpenPose to obtain human body skeleton node data in each frame of image in the video; then, judging the following three conditions based on human skeleton joint points, and if all the three conditions are met, considering that the falling occurs: (1) The descending speed of the central point of the hip joint of the human body is greater than the preset critical speed; (2) The inclination angle between the longitudinal center line of the human body and the ground is smaller than a preset critical inclination angle; (3) The aspect ratio of the rectangle circumscribed by the human body is larger than a preset critical ratio. The invention can timely and accurately find the falling event without disturbing the normal activities of people, and can accelerate the response rescue speed after the falling event occurs.
Description
Technical Field
The invention relates to the field of human behavior recognition, in particular to a human body fall recognition method and system based on OpenPose.
Background
People often fall down under walking, running and other conditions. For normal people, people can give an alarm or inform the families immediately or by making a call under the condition that the fall is not particularly serious, however, for some special people, such as the empty nest old people, the people can not stand up when falling, and the people can not give an alarm or inform the families to help in time, so that a method capable of accurately and timely identifying the fall and timely carrying out coping is required to be established.
Disclosure of Invention
In order to solve the technical problems, the invention provides a human body fall identification method and system based on OpenPose, and aims to establish a method capable of accurately and timely identifying falls and timely performing coping treatments.
According to a first aspect of the present invention, in order to solve the technical problem, the method for identifying a human body falling based on openPose provided by the present invention comprises the following steps:
s1, acquiring a monitoring video shot by a monitoring camera, and processing the monitoring video by using OpenPose to obtain human skeleton joint point data in each frame of image in the video;
s2, judging the following three conditions based on the human skeleton joint point data, and if all the three conditions are met, considering that the falling occurs:
(1) The descending speed of the central point of the hip joint of the human body is greater than the preset critical speed;
(2) The inclination angle between the longitudinal center line of the human body and the ground is smaller than a preset critical inclination angle;
(3) The aspect ratio of the rectangle circumscribed by the human body is larger than a preset critical ratio.
Further, in the openPose-based human body fall identification method of the invention, the method further comprises the steps of:
s3, judging whether the fallen person stands again in a period of time, if not, carrying out emergency treatment according to a preset rule; wherein, the rule for judging the re-standing is as follows: after a human body falls, if the inclination angle of the longitudinal center line of the human body and the ground is not smaller than a preset angle and the aspect ratio of the external rectangle is not smaller than a preset ratio in a period of time, the human body is re-standing by the falling person as long as the two conditions are met, and otherwise, the human body is not re-standing.
Further, in the openPose-based human body fall recognition method of the present invention, the falling speed of the central point of the hip joint of the human body is obtained according to the following method:
determining the positions of two hip joints in human skeleton joint point data as s 8 (t)＝(x t8 ,y t8 ) S 11 (t)＝(x t11 ,y t11 ) Let t be 1 Ordinate value of moment hip joint midpointt 2 Ordinate value of the moment hip joint midpoint +.>The time interval is Δt=t 2 -t 1 The midpoint descending speed of the hip joint of the human body is +.>Where t represents time.
Further, in the openPose-based human body fall identification method of the present invention, the inclination angle between the longitudinal center line of the human body and the ground is obtained by the following method:
determining data s of head articulation point, knee articulation point and foot articulation point which are positioned on the same leg with the knee articulation point in human skeleton articulation point data 0 (t)＝(x t0 ,y t0 )、s 12 (t)＝(x t12 ,y t12 )、s 13 (t)＝(x t13 ,y t13 ) ThenI.e. < ->At t, the inclination angle formed by the longitudinal center line of the human body and the ground isWhere t represents time.
Further, in the openPose-based human body fall recognition method of the present invention, the critical inclination angle=45 degrees, and the critical ratio=1.
According to another aspect of the present invention, in order to solve the technical problem, the openPose-based human body fall recognition system provided by the present invention includes the following modules:
the OpenPose processing module is used for acquiring the monitoring video shot by the monitoring camera, and processing the monitoring video by using the OpenPose to obtain human skeleton joint point data in each frame of image in the video;
the fall judgment module is used for judging the following three conditions based on the human skeleton joint point data, and if all the three conditions are met, the fall is considered to occur:
(1) The descending speed of the central point of the hip joint of the human body is greater than the preset critical speed;
(2) The inclination angle between the longitudinal center line of the human body and the ground is smaller than a preset critical inclination angle;
(3) The aspect ratio of the rectangle circumscribed by the human body is larger than a preset critical ratio.
Further, the openPose-based human body fall recognition system is characterized by further comprising:
the emergency processing module is used for judging whether the fallen person stands again in a period of time, if not, carrying out emergency processing according to a preset rule; wherein, the rule for judging the re-standing is as follows: after a human body falls, if the inclination angle of the longitudinal center line of the human body and the ground is not smaller than a preset angle and the aspect ratio of the external rectangle is not larger than a preset ratio in a period of time, the human body is re-standing by the falling person as long as the two conditions are met, and otherwise, the human body is not re-standing.
Further, in the openPose-based human body fall recognition system, the falling speed of the central point of the hip joint of the human body is obtained according to the following method:
determining the positions of two hip joints in human skeleton joint point data as s 8 (t)＝(x t8 ,y t8 ) S 11 (t)＝(x t11 ,y t11 ) Let t be 1 Ordinate value of moment hip joint midpointt 2 Ordinate value of the moment hip joint midpoint +.>The time interval is Δt=t 2 -t 1 The midpoint descending speed of the hip joint of the human body is +.>Where t represents time.
Further, in the openPose-based human body fall recognition system of the present invention, the inclination angle between the longitudinal center line of the human body and the ground is obtained by the following method:
determining data s of head articulation point, knee articulation point and foot articulation point which are positioned on the same leg with the knee articulation point in human skeleton articulation point data 0 (t)＝(x t0 ,y t0 )、s 12 (t)＝(x t12 ,y t12 )、s 13 (t)＝(x t13 ,y t13 ) ThenI.e. < ->At t, the inclination angle formed by the longitudinal center line of the human body and the ground isWhere t represents time.
Further, in the openPose-based human body fall recognition system of the present invention, the critical inclination angle=45 degrees, and the critical ratio value=1.
By implementing the human body falling identification method and system based on OpenPose, falling events can be timely and accurately found under the condition that normal activities of people are not interfered, and response rescue speed after the falling events occur can be accelerated.
Drawings
The invention will be further described with reference to the accompanying drawings and examples, in which:
FIG. 1 is a schematic diagram of an OpenPose processing picture of the present invention;
FIG. 2 is a schematic illustration of various joints and numbering thereof according to the present invention;
FIG. 3 is a flowchart of one embodiment of an OpenPose-based human fall recognition method of the present invention;
FIG. 4 is a schematic illustration of a human fall process of the present invention;
FIG. 5 is a schematic view of the inclination angle of the longitudinal centerline of the human body with the ground;
FIG. 6 is a schematic view of a human circumscribed rectangular frame;
fig. 7 is a schematic view of a person standing by himself after falling.
Detailed Description
For a clearer understanding of technical features, objects and effects of the present invention, a detailed description of embodiments of the present invention will be made with reference to the accompanying drawings.
The openPose human gesture recognition project is a convolutional neural network and supervised learning open source library developed by the university of Carcinyl Meuron (CMU). In 2017, researchers at university of Carcinyl Mercury have issued source codes of the OpenPose human skeleton recognition system, thus realizing real-time tracking of targets under video monitoring. It can capture human skeletal information in color video and provide joint information in the scene. As shown in fig. 1, a frame of photo in a video is subjected to openelse processing to obtain a human skeleton diagram.
The video processed by OpenPose can not only obtain a human skeleton diagram, but also obtain the coordinate position of a corresponding joint point, and the coordinate position data corresponding to the human joint point of a certain frame of picture is given in a table 1. For a specific joint corresponding to the joint number in the table, as shown in fig. 2, each joint number and corresponding joint name in fig. 2 are shown in table 2.
Human body joint point data obtained in Table 1
For convenience of representation, we define joint point coordinates Joint Coordinates (JC), s= { S 0 ,s 1 ,…,s 13 And represents the set of joint points. Defining the representation mode of the node j at the t moment as s j (t)＝(x tj ,y tj ),j∈{0,1,…,13}。
Table 2 each joint number and corresponding joint name
Referring to fig. 3, a flowchart of an embodiment of a human body fall recognition method based on openPose according to the present invention includes the following steps:
s1, acquiring a monitoring video shot by a monitoring camera, and processing the monitoring video by using OpenPose to obtain human skeleton joint point data in each frame of image in the video;
s2, judging the following three conditions based on the human skeleton joint point data, and if all the three conditions are met, considering that the falling occurs:
(1) The descending speed of the central point of the hip joint of the human body is greater than the preset critical speed;
(2) The inclination angle between the longitudinal center line of the human body and the ground is smaller than a preset critical inclination angle;
(3) The aspect ratio of the rectangle circumscribed by the human body is larger than a preset critical ratio.
The three determination conditions in step S2 will be specifically described below.
Human hip joint midpoint descending speed (first judgment condition)
As shown in fig. 4, the center of gravity of the human body is changed sharply in the vertical direction during a sudden fall, and the hip joint center point of the human body can be approximately representedThe center of gravity of the human body reflects this feature. Since the person is a very short process from standing to falling, the time taken is also very short, so every 5 adjacent frames are detected at intervals of 0.25 seconds. Human body skeleton joint point data acquired through OpenPose, wherein two hip joint points are s respectively 8 (t)＝(x t8 ,y t8 ) S 11 (t)＝(x t11 ,y t11 ) Assume that the ordinate value of the hip joint midpoint at time t1Ordinate value of hip joint midpoint at time t2 +.>The time interval is Δt=t 2 -t 1 The mid-point of the hip joint decreases at a rate +.>
When (when)(/>Critical speed), M 1 =1, it is considered that the first determination condition for occurrence of the fall time is satisfied.
Inclination angle of human body center line (determination condition II)
The most obvious feature is that the human body is inclined during a fall, and the inclination angle of the human body with the ground is continuously reduced. To reflect this feature, a human body center line is set (set up node s 12 And the articulation point s 13 Is the midpoint of (1)And joint withPoint s 0 The connection is the human body central line L).
As shown in fig. 5, θ is the inclination angle between the longitudinal center line of the human body and the ground, and is calculated as follows: data s of the joints 0, 12, 13 obtained by OpenPose 0 (t)＝(x t0 ,y t0 )、s 12 (t)＝(x t12 ,y t12 )、s 13 (t)＝(x t13 ,y t13 ). ThenI.e. < ->At t, the included angle formed by the human body central line and the ground is +.>
When theta is less than theta 0 (critical angle 45 °), M 2 When=1, that is, the inclination angle of the human longitudinal center line with respect to the ground is smaller than the critical angle, it is considered that the fall judgment condition two is satisfied.
Human body external rectangle aspect ratio (determination condition three)
When a human body falls down, the body contour also changes correspondingly, and simply comparing the height and width of the body contour is not reasonable, because the height and width of the body contour both change due to the distance from the camera, but the aspect ratio can solve the problem. Thus, whether the human body falls down or not is judged by detecting the human body height-width ratio.
The aspect ratio of the rectangle circumscribed by the human body is set to be p=width/Height, as shown in fig. 6, when the rectangle circumscribed by the human body is obviously changed in the standing walking and falling processes, the P value is also different.
T represents the critical value of the aspect ratio (T=1), and according to the investigation of the actual situation, the aspect ratio P is smaller than 1 when the human body walks normally, and the ratio is larger than 1 when the human body falls down, namely M 3 When=1, a fall occurs, and the determination condition three is satisfied.
If a person can stand up by himself in a short time after a fall occurs, emergency treatment, such as an alarm, is not necessary. Most of the fall identification is focused on analysis of the fall process, and the process of self-standing after the fall of the human body is rarely considered. As shown in fig. 7, the process of standing after a fall can be regarded as an inverse of the fall, the only difference being that the process occurs more slowly than a fall. According to the analysis of the first three distinguishing conditions, if the inclination angle of the longitudinal center line of the human body and the ground is not smaller than a preset angle and the aspect ratio of the external rectangle is not larger than a preset ratio in a period of time after the human body falls, as long as the two conditions are met, the user can be considered to stand again, the preset angle can be the same as the critical angle, namely 45 degrees, or different, and the preset ratio can be the same as the critical ratio, namely 1, or different. The key to judging whether a person can stand up after falling is to reduce unnecessary alarms, because sometimes the falling does not cause serious injury to the human body.
The embodiments of the present invention have been described above with reference to the accompanying drawings, but the present invention is not limited to the above-described embodiments, which are merely illustrative and not restrictive, and many forms may be made by those having ordinary skill in the art without departing from the spirit of the present invention and the scope of the claims, which are to be protected by the present invention.
Claims (4)
1. The human body fall recognition method based on OpenPose is characterized by comprising the following steps of:
s1, acquiring a monitoring video shot by a monitoring camera, and processing the monitoring video by using OpenPose to obtain human skeleton joint point data in each frame of image in the video;
s2, judging the following three conditions based on the human skeleton joint point data, and if all the three conditions are met, considering that the falling occurs:
(1) The descending speed of the central point of the hip joint of the human body is greater than the preset critical speed;
(2) The inclination angle between the longitudinal center line of the human body and the ground is smaller than a preset critical inclination angle;
(3) The aspect ratio of the rectangle externally connected with the human body is larger than a preset critical ratio;
s3, judging whether the fallen person stands again in a period of time, if not, carrying out emergency treatment according to a preset rule; wherein, the rule for judging the re-standing is as follows: after a human body falls, if the inclination angle of the longitudinal center line of the human body and the ground is not smaller than a preset angle and the aspect ratio of the external rectangle is not larger than a preset ratio in a period of time, if both conditions are met, the human body is re-standing, otherwise, the human body is not re-standing;
the descending speed of the central point of the hip joint of the human body is obtained according to the following method:
determining that two hip joints in human skeleton joint point data are s respectively 8 (t)＝(x t8 ,y t8 ) S 11 (t)＝(x t11 ,y t11 ) Let t be 1 Ordinate value of moment hip joint midpointt 2 Ordinate value of moment hip joint midpointThe time interval is Δt=t 2 -t 1 The mid-point of the hip joint decreases at a rate +.>Wherein t represents time;
the inclination angle between the longitudinal center line of the human body and the ground is obtained by the following method:
determining data of head joint points, knee joint points and foot joint points which are positioned on the same leg with the knee joint points in human skeleton joint point data, and s 0 (t)＝(x t0 ,y t0 )，s 12 (t)＝(x t12 ,y t12 )，s 13 (t)＝(x t13 ,y t13 ) The method comprises the steps of carrying out a first treatment on the surface of the ThenI.e. < ->At t, the included angle formed by the longitudinal center line of the human body and the ground isWhere t represents time.
2. The openPose-based human fall recognition method according to claim 1, wherein the critical inclination angle=45 degrees, and the critical ratio value=1.
3. Human body fall recognition system based on OpenPose, which is characterized by comprising the following modules:
the OpenPose processing module is used for acquiring the monitoring video shot by the monitoring camera, and processing the monitoring video by using the OpenPose to obtain human skeleton joint point data in each frame of image in the video;
the fall judgment module is used for judging the following three conditions based on the human skeleton joint point data, and if all the three conditions are met, the fall is considered to occur:
(1) The descending speed of the central point of the hip joint of the human body is greater than the preset critical speed;
(2) The inclination angle between the longitudinal center line of the human body and the ground is smaller than a preset critical inclination angle;
(3) The aspect ratio of the rectangle externally connected with the human body is larger than a preset critical ratio;
the emergency processing module is used for judging whether the fallen person stands again in a period of time, if not, carrying out emergency processing according to a preset rule; wherein, the rule for judging the re-standing is as follows: after a human body falls, if the inclination angle of the longitudinal center line of the human body and the ground is not smaller than a preset angle and the aspect ratio of the external rectangle is not smaller than a preset ratio in a period of time, if both conditions are met, the human body is re-standing, otherwise, the human body is not re-standing;
the descending speed of the central point of the hip joint of the human body is obtained according to the following method: determining that two hip joints in human skeleton joint point data are s respectively 8 (t)＝(x t8 ,y t8 ) S 11 (t)＝(x t11 ,y t11 ) Let t be 1 Ordinate value of moment hip joint midpointt 2 Ordinate value of the moment hip joint midpoint +.>The time interval is Δt=t 2 -t 1 The mid-point of the hip joint decreases at a rate +.>Wherein t represents time;
the inclination angle between the longitudinal center line of the human body and the ground is obtained by the following method:
determining data of head joint points, knee joint points and foot joint points which are positioned on the same leg with the knee joint points in human skeleton joint point data, and s 0 (t)＝(x t0 ,y t0 )，s 12 (t)＝(x t12 ,y t12 )，s 13 (t)＝(x t13 ,y t13 ) The method comprises the steps of carrying out a first treatment on the surface of the ThenI.e. < ->At t, the included angle formed by the longitudinal center line of the human body and the ground isWhere t represents time.
4. The openpase-based human fall recognition system of claim 3, wherein the critical inclination angle = 45 degrees and the critical ratio = 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010148742.4A CN111460908B (en) | 2020-03-05 | 2020-03-05 | Human body fall recognition method and system based on OpenPose |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010148742.4A CN111460908B (en) | 2020-03-05 | 2020-03-05 | Human body fall recognition method and system based on OpenPose |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111460908A CN111460908A (en) | 2020-07-28 |
| CN111460908B true CN111460908B (en) | 2023-09-01 |
Family
ID=71685585
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010148742.4A Active CN111460908B (en) | 2020-03-05 | 2020-03-05 | Human body fall recognition method and system based on OpenPose |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111460908B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111931733B (en) * | 2020-09-25 | 2021-02-26 | 西南交通大学 | Human body posture detection method based on depth camera |
| CN113033437A (en) * | 2021-03-29 | 2021-06-25 | 合肥工业大学 | Fall monitoring method and system based on improved pattern recognition |
| CN113256938A (en) * | 2021-05-31 | 2021-08-13 | 苏州优函信息科技有限公司 | Fall monitoring alarm scheme based on spectrum camera |
| US20230004740A1 (en) * | 2021-06-30 | 2023-01-05 | Ubtech North America Research And Development Center Corp | Human lying posture detection method and mobile machine using the same |
| CN115100734A (en) * | 2022-05-12 | 2022-09-23 | 燕山大学 | Openpos-based ski field dangerous action identification method and system |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN87211648U (en) * | 1987-08-15 | 1988-04-13 | 北京市东城实用技术研究所开发部 | Analogical automatic reset humanbody for judo and wrestling training |
| CN103186902A (en) * | 2011-12-29 | 2013-07-03 | 爱思开电讯投资（中国）有限公司 | Trip detecting method and device based on video |
| CN103955699A (en) * | 2014-03-31 | 2014-07-30 | 北京邮电大学 | Method for detecting tumble event in real time based on surveillance videos |
| EP2763116A1 (en) * | 2013-02-01 | 2014-08-06 | FamilyEye BVBA | Fall detection system and method for detecting a fall of a monitored person |
| CN105448041A (en) * | 2016-01-22 | 2016-03-30 | 苏州望湖房地产开发有限公司 | A human body falling intelligent control system and method |
| CN106205322A (en) * | 2016-08-22 | 2016-12-07 | 叶强 | Human synovial measurement for Evaluation system |
| CN106503643A (en) * | 2016-10-18 | 2017-03-15 | 上海电力学院 | Tumble detection method for human body |
| CN108629946A (en) * | 2018-06-14 | 2018-10-09 | 清华大学深圳研究生院 | A kind of tumble detection method for human body based on RGBD sensors |
| CN108846996A (en) * | 2018-08-06 | 2018-11-20 | 浙江理工大学 | One kind falling down detecting system and method |
| CN109241841A (en) * | 2018-08-01 | 2019-01-18 | 甘肃未来云数据科技有限公司 | The acquisition methods and device of video human movement |
| CN109670396A (en) * | 2018-11-06 | 2019-04-23 | 华南理工大学 | A kind of interior Falls Among Old People detection method |
| CN109919132A (en) * | 2019-03-22 | 2019-06-21 | 广东省智能制造研究所 | A kind of pedestrian's tumble recognition methods based on skeleton detection |
| CN109920208A (en) * | 2019-01-31 | 2019-06-21 | 深圳绿米联创科技有限公司 | Tumble prediction technique, device, electronic equipment and system |
| JP2019121045A (en) * | 2017-12-28 | 2019-07-22 | コニカミノルタ株式会社 | Posture estimation system, behavior estimation system, and posture estimation program |
| CN110598536A (en) * | 2019-07-31 | 2019-12-20 | 杭州连帆科技有限公司 | Falling detection method and system based on human skeleton motion model |
| CN110706255A (en) * | 2019-09-25 | 2020-01-17 | 马可 | Fall detection method based on self-adaptive following |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103735268B (en) * | 2013-09-29 | 2015-11-25 | 沈阳东软医疗系统有限公司 | A kind of position detection method and system |
-
2020
- 2020-03-05 CN CN202010148742.4A patent/CN111460908B/en active Active
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN87211648U (en) * | 1987-08-15 | 1988-04-13 | 北京市东城实用技术研究所开发部 | Analogical automatic reset humanbody for judo and wrestling training |
| CN103186902A (en) * | 2011-12-29 | 2013-07-03 | 爱思开电讯投资（中国）有限公司 | Trip detecting method and device based on video |
| EP2763116A1 (en) * | 2013-02-01 | 2014-08-06 | FamilyEye BVBA | Fall detection system and method for detecting a fall of a monitored person |
| CN103955699A (en) * | 2014-03-31 | 2014-07-30 | 北京邮电大学 | Method for detecting tumble event in real time based on surveillance videos |
| CN105448041A (en) * | 2016-01-22 | 2016-03-30 | 苏州望湖房地产开发有限公司 | A human body falling intelligent control system and method |
| CN106205322A (en) * | 2016-08-22 | 2016-12-07 | 叶强 | Human synovial measurement for Evaluation system |
| CN106503643A (en) * | 2016-10-18 | 2017-03-15 | 上海电力学院 | Tumble detection method for human body |
| JP2019121045A (en) * | 2017-12-28 | 2019-07-22 | コニカミノルタ株式会社 | Posture estimation system, behavior estimation system, and posture estimation program |
| CN108629946A (en) * | 2018-06-14 | 2018-10-09 | 清华大学深圳研究生院 | A kind of tumble detection method for human body based on RGBD sensors |
| CN109241841A (en) * | 2018-08-01 | 2019-01-18 | 甘肃未来云数据科技有限公司 | The acquisition methods and device of video human movement |
| CN108846996A (en) * | 2018-08-06 | 2018-11-20 | 浙江理工大学 | One kind falling down detecting system and method |
| CN109670396A (en) * | 2018-11-06 | 2019-04-23 | 华南理工大学 | A kind of interior Falls Among Old People detection method |
| CN109920208A (en) * | 2019-01-31 | 2019-06-21 | 深圳绿米联创科技有限公司 | Tumble prediction technique, device, electronic equipment and system |
| CN109919132A (en) * | 2019-03-22 | 2019-06-21 | 广东省智能制造研究所 | A kind of pedestrian's tumble recognition methods based on skeleton detection |
| CN110598536A (en) * | 2019-07-31 | 2019-12-20 | 杭州连帆科技有限公司 | Falling detection method and system based on human skeleton motion model |
| CN110706255A (en) * | 2019-09-25 | 2020-01-17 | 马可 | Fall detection method based on self-adaptive following |
Non-Patent Citations (1)
| Title |
|---|
| 基于人体模型的上下楼梯动作的生成与评价;邱俊佳;《中国优秀硕士学位论文全文数据库 (信息科技辑)》(第2期);第I138-4269页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111460908A (en) | 2020-07-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111460908B (en) | Human body fall recognition method and system based on OpenPose | |
| CN109919132B (en) | Pedestrian falling identification method based on skeleton detection | |
| Hazelhoff et al. | Video-based fall detection in the home using principal component analysis | |
| Tao et al. | Fall incidents detection for intelligent video surveillance | |
| US20080181499A1 (en) | System and method for feature level foreground segmentation | |
| JP2006522959A (en) | Method and apparatus for fall prevention and detection | |
| CN112784987B (en) | Target nursing method and device based on multistage neural network cascade | |
| Eng et al. | A bayesian framework for robust human detection and occlusion handling human shape model | |
| Abdo et al. | Fall detection based on RetinaNet and MobileNet convolutional neural networks | |
| Poonsri et al. | Improvement of fall detection using consecutive-frame voting | |
| Stone et al. | Silhouette classification using pixel and voxel features for improved elder monitoring in dynamic environments | |
| Cardile et al. | A vision-based system for elderly patients monitoring | |
| JP5771039B2 (en) | Abandoned person detection device | |
| WO2017193679A1 (en) | Method for automatically detecting whether bicycle has fallen onto the ground | |
| TWI493510B (en) | Falling down detection method | |
| Abd et al. | Human fall down recognition using coordinates key points skeleton | |
| CN112766183A (en) | Alarm system and method for people entering forbidden zone based on AI analysis | |
| CN112633179A (en) | Farmer market aisle object occupying channel detection method based on video analysis | |
| Miaou et al. | A smart vision-based human fall detection system for telehealth applications | |
| JP2009187348A (en) | Automatic suspicious object detecting device | |
| Lee et al. | Automated abnormal behavior detection for ubiquitous healthcare application in daytime and nighttime | |
| CN111814590B (en) | Personnel safety state monitoring method, equipment and computer readable storage medium | |
| CN114463838A (en) | Human behavior recognition method, system, electronic device and storage medium | |
| JP6543546B2 (en) | Specific motion detection device and specific motion detection method | |
| JP5669648B2 (en) | Anomaly detection device |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |