CN108629300A - A kind of fall detection method - Google Patents
A kind of fall detection method Download PDFInfo
- Publication number
- CN108629300A CN108629300A CN201810373224.5A CN201810373224A CN108629300A CN 108629300 A CN108629300 A CN 108629300A CN 201810373224 A CN201810373224 A CN 201810373224A CN 108629300 A CN108629300 A CN 108629300A
- Authority
- CN
- China
- Prior art keywords
- target person
- tumble
- node
- point
- fall detection
- 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.)
- Granted
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
- G06V40/23—Recognition of whole body movements, e.g. for sport training
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- 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
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
- G06V20/42—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items of sport video content
-
- 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
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
-
- 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
Abstract
The present invention provides a kind of fall detection method, can improve the accuracy rate of fall detection.The method includes：Obtain the video for including target person；According to the video for including target person of acquisition, the characteristic point of target person is extracted；For each frame image, according to the characteristic point of the target person extracted, tumble judgement is carried out from the front of target person, side, three, back side direction；If the judging result of the either direction in front, side or the back side, which is target person, is in tumble state, judge whether target person is always maintained at tumble state in continuous certain frame number, if so, confirming that target person is fallen.The present invention is operated suitable for human body fall detection.
Description
Technical field
The present invention relates to image processing fields, particularly relate to a kind of fall detection method.
Background technology
With advances in technology and the development of medical level, the world enter aging society comprehensively, and China is used as population
Big country, aging trend are also very severe.However the endowment service industry present situation in China is not matched that with aging level, is supplied
It is very big to give demand difference.On the one hand there is not small gap in terms of bed of supporting parents, on the other hand in terms of the attendant that supports parents, I
The current situation of state is also personnel's wretched insufficiency.In this case, the monitoring of video auxiliary is particularly important.
World Health Organization points out, fall down have become lethality in unexpected and unintentional property injury it is second largest because
Element has substantially 424000 adults to die of the accident of falling down every year in the world.It falls down and easily causes phobic anxiety mood, bring brain
Damage, hip joint is impaired, and the complications such as heart arrest bring huge harm and the pain of injury to the crowd of suffering a calamity or disaster.Tumble is old
Common one of the injury of people, and Falls in Old People is to now result in the elderly's disability and dead one of the major reasons.
Since monitoring technology occurs from the 1960s, quickly grows, experienced first generation analog video monitoring technology,
Second generation automated video monitoring technology, has marched toward the intelligent Video Surveillance Technology of the third generation now.Intelligent video monitoring
Technology and previous monitoring technology have the difference of essence, are mainly characterized by the method using computer vision, are almost being not required to
In the case of wanting human intervention, automatically analyzed come to the target in dynamic scene by the image sequence to video camera recording
It positioned, tracked and is identified, and analyze and judge the behavior of target on this basis, to accomplish that daily management can be completed
It can in time make a response when abnormal conditions occur again.Big portion of the intelligent video monitoring system instead of people in monitor task
The division of labor is made, and is the monitoring technology with height intelligence of a new generation.Intelligent Video Surveillance Technology has shown that huge market
Value.If intelligent Video Surveillance Technology is applied to the elderly to fall down in detection, undoubtedly can also have huge application value
And market value.
Elderly population is the disadvantaged group of society, and with the increase at age, every function all can slowly decline, and compare year
Light people is easier that accident occurs and comes to harm, once accident occurs cannot often find and cause much draw at the first time
Return injury or cannot timely be treated, injury of gently then fracturing, heavy then entail dangers to life.Therefore, when the elderly occurs
When dangerous, timely relief be very it is necessary to.
The study on monitoring based on environment information acquisition used in the prior art is mainly by installation in a room each
Kind sensor is acquired the daily routines data of old man, analyzes, and the interference of various information is larger (for example, working as in by environment
When personage wears close with the colors such as carpet clothes, light it is dark when, monitoring result is disturbed larger), easily judge by accident.
Invention content
The technical problem to be solved in the present invention is to provide a kind of fall detection methods, to solve the base present in the prior art
In environment information acquisition monitoring method easily in by environment various information interfered, lead to the problem that False Rate is high.
In order to solve the above technical problems, the embodiment of the present invention provides a kind of fall detection method, including：
Obtain the video for including target person；
According to the video for including target person of acquisition, the characteristic point of target person is extracted；
For each frame image, according to the characteristic point of the target person extracted, from the front, side, the back of the body of target person
Three, face direction carries out tumble judgement；
If the judging result of the either direction in front, side or the back side, which is target person, is in tumble state, judge to connect
Whether target person is always maintained at tumble state in continuous certain frame number, if so, confirming that target person is fallen.
Further, the video for including target person according to acquisition, the characteristic point for extracting target person include：
According to the video for including target person of acquisition, image procossing is carried out using Attitude estimation algorithm, extracts target person
The characteristic point of object.
Further, the characteristic point of extraction includes：Nose node, neck node, pereonite point, right-hand man's node, left and right toggle point,
One or more of stern node, left and right knee node, left and right foot node.
Further, described to be directed to each frame image, according to the characteristic point of the target person extracted, just from target person
Face carries out tumble judgement：
Judge whether target person is in handstand state, if so, judgement target person is in tumble state；Otherwise, from mesh
The side of mark personage carries out tumble judgement.
Further, described to judge whether target person is in handstand state, if so, judgement target person is in tumble shape
State includes：
If detecting the foot node of target person, judge whether the height of the neck node of target person is less than pereonite point
Whether height and the height of pereonite point are less than the height of foot node；
If the two is both less than, judge that target person is in tumble state.
Further, described to judge whether target person is in handstand state, if so, judgement target person is in tumble shape
State further includes：
If detecting the foot node of fall short personage, judge whether the height of the neck node of target person is less than pereonite
The height of point；
If being less than, judge that target person is in tumble state.
Further, described to be directed to each frame image, according to the characteristic point of the target person extracted, from target person side
Face carries out tumble judgement：
Determine the first slope and the second slope for indicating target person body direction；
If first slope meets preset first threshold range and the second slope meets preset second threshold range, sentence
The personage that sets the goal is in tumble state, otherwise, tumble judgement is carried out from the back side of target person；
Wherein, first slope is expressed as：
σ 1=ychest-yneck/xchest-xneck；
Second slope is expressed as：
σ 2=ychest-yfoot/xchest-xfoot
Wherein, σ 1, σ 2 indicate first slope, the second slope respectively, ychest, yneck, yfoot indicate respectively pereonite point,
Neck node, the coordinate value of foot node in the vertical direction, xchest, xneck, xfoot indicate pereonite point, neck node, foot respectively
The coordinate value of node in the horizontal direction.
Further, described to be directed to each frame image, according to the characteristic point of the target person extracted, carried on the back from target person
Face carries out tumble judgement：
It is preset to judge whether the target person difference in height between preset two frame is more than using the method for relative height differential
Third threshold value；
Whether the difference in height if more than preset third threshold value, then the hand node for judging target person and foot node is less than in advance
If the 4th threshold value；
If being less than preset 4th threshold value, judge that target person is in tumble state.
Further, after confirming that target person is fallen, the method further includes：
Send out alarm.
The above-mentioned technical proposal of the present invention has the beneficial effect that：
In said program, by the video for including target person of acquisition, the characteristic point of target person is extracted；For each
Frame image is fallen according to the characteristic point of the target person extracted from the front of target person, side, three, back side direction
Judge；If the judging result of the either direction in front, side or the back side, which is target person, is in tumble state, judge continuous
Whether target person is always maintained at tumble state in certain frame number, if so, confirming that target person is fallen.Know in this way, falling
Influencing for the various factors such as other process is not easy the clothing of personage in by environment, light is influenced, can improve the accuracy rate of fall detection.
Description of the drawings
Fig. 1 is the flow diagram one of fall detection method provided in an embodiment of the present invention；
Fig. 2 is the flow diagram two of fall detection method provided in an embodiment of the present invention；
Fig. 3 is positive tumble algorithm flow schematic diagram provided in an embodiment of the present invention；
Fig. 4 is tumble algorithm flow schematic diagram in side provided in an embodiment of the present invention；
Fig. 5 is tumble algorithm flow schematic diagram in the back side provided in an embodiment of the present invention.
Specific implementation mode
To keep the technical problem to be solved in the present invention, technical solution and advantage clearer, below in conjunction with attached drawing and tool
Body embodiment is described in detail.
The present invention for the existing monitoring method based on environment information acquisition easily in by environment various information interfered, lead
The problem for causing False Rate high, provides a kind of fall detection method.
As shown in Figure 1, fall detection method provided in an embodiment of the present invention, including：
S101 obtains the video for including target person；
S102 extracts the characteristic point of target person according to the video for including target person of acquisition；
S103, for each frame image, according to the characteristic point of the target person extracted, from the front of target person, side
Face, three, back side direction carry out tumble judgement；
S104 sentences if the judging result of positive, side or the back side either direction, which is target person, is in tumble state
Disconnection continues whether target person in certain frame number is always maintained at tumble state, if so, confirming that target person is fallen.
Fall detection method described in the embodiment of the present invention extracts target by the video for including target person of acquisition
The characteristic point of personage；For each frame image, according to the characteristic point of the target person extracted, from the front of target person, side
Face, three, back side direction carry out tumble judgement；If the judging result of the either direction in front, side or the back side is at target person
In tumble state, then judge whether target person is always maintained at tumble state in continuous certain frame number, if so, confirming target
Personage falls.In this way, influencing for the various factors such as tumble identification process is not easy the clothing of personage in by environment, light is influenced, it can
Improve the accuracy rate of fall detection.
It is further, described that target person is included according to acquisition in the specific implementation mode of aforementioned fall detection method
The video of object, the characteristic point for extracting target person include：
According to the video for including target person of acquisition, image procossing is carried out using Attitude estimation algorithm, extracts target person
The characteristic point of object.
In the present embodiment, the video comprising target person can be obtained by camera or video monitoring, then, according to
The video for including target person obtained carries out image procossing using Attitude estimation algorithm, extracts the characteristic point of target person, has
Body process is：
For example, the libraries c++ that open source projects OpenPose is provided can be utilized, to the video comprising target person of acquisition into
Row initialization, then matches the image in video using the learning model in library, identifies the characteristic point of target person.
In the present embodiment, the characteristic point of the target person of mark may include：Nose node, neck node, pereonite point, right-hand man
One or more of node, left and right toggle point, stern node, left and right knee node, left and right foot node.In practical applications, may be used
To combine practical application scene, required characteristic point is determined.
In the present embodiment, the difference of tumble posture determines that single fall detection algorithm can not be realized accurately to each side
To tumble judge, the fall detection method described in the present embodiment, using the characteristic point of acquisition, mainly from front, side, the back side
Three directions carry out tumble judgement, realize the fusion of multi-direction fall detection algorithm, and flow is as shown in Figure 2.
It is further, described to be directed to each frame image in the specific implementation mode of aforementioned fall detection method, according to carrying
The characteristic point for the target person got, carrying out tumble judgement from target person front includes：
Judge whether target person is in handstand state, if so, judgement target person is in tumble state；Otherwise, from mesh
The side of mark personage carries out tumble judgement.
It is to carry out tumble judgement from target person front first, if personage front is fallen, in personage's appearance in the present embodiment
Significant change can occur in state, be mainly manifested in personage and handstand state is presented, it therefore, can when target person is in handstand state
It is in tumble state with preliminary judgement target person.
In the present embodiment, target person can be judged using neck node, pereonite point and these three main nodes of foot node
Whether it is in handstand state；Specifically：
If detecting the foot node of target person, judge whether the height (Hneck) of the neck node of target person is less than
Whether the height (Hchest) of pereonite point and the height (Hchest) of pereonite point are less than the height (Hfoot) of foot node；If two
Person is both less than (Hneck<Hchest and Hchest<Hfoot), then judge that target person is in tumble state.
If detecting the foot node of fall short personage, judge the neck node of target person height (Hneck) whether
Less than the height (Hchest) of pereonite point；If being less than (Hneck<Hchest), then judge that target person is in tumble state, such as scheme
Shown in 3；
It is further, described to be directed to each frame image in the specific implementation mode of aforementioned fall detection method, according to carrying
The characteristic point for the target person got, carrying out tumble judgement from target person side includes：
Determine the first slope and the second slope for indicating target person body direction；
If first slope meets preset first threshold range and the second slope meets preset second threshold range, sentence
The personage that sets the goal is in tumble state, otherwise, tumble judgement is carried out from the back side of target person；
Wherein, first slope is expressed as：
σ 1=ychest-yneck/xchest-xneck；
Second slope is expressed as：
σ 2=ychest-yfoot/xchest-xfoot
Wherein, σ 1, σ 2 indicate first slope, the second slope respectively, ychest, yneck, yfoot indicate respectively pereonite point,
Neck node, the coordinate value of foot node in the vertical direction, xchest, xneck, xfoot indicate pereonite point, neck node, foot respectively
The coordinate value of node in the horizontal direction.
In the present embodiment, if not can determine that target person is in tumble state from front, it can continue from target person
Object side carries out tumble judgement, and the center of gravity that side tumble is mainly reflected in people can change, therefore, be used in the present embodiment
A kind of dual slope method carries out tumble judgement；Specifically：Determine the first slope σ 1 and second for indicating target person body direction
Slope σ 2；If first slope σ 1 meets preset first threshold range and the second slope σ 2 meets preset second threshold range,
Then judge that target person is in tumble state, otherwise, tumble judgement is carried out from the back side of target person.Such as：
Second slope σ 2<1.4 and first slope σ 1=1 or first slope σ 1<The 1 and second slope σ 2=1 or second
Slope σ 2<1.4 and first slope σ 1<1 thrin is then judged to falling, and flow chart is as shown in Figure 4.
It is further, described to be directed to each frame image in the specific implementation mode of aforementioned fall detection method, according to carrying
The characteristic point for the target person got, carrying out tumble judgement from the target person back side includes：
It is preset to judge whether the target person difference in height between preset two frame is more than using the method for relative height differential
Third threshold value；
Whether the difference in height if more than preset third threshold value, then the hand node for judging target person and foot node is less than in advance
If the 4th threshold value；
If being less than preset 4th threshold value, judge that target person is in tumble state.
In the present embodiment, if not can determine that target person is in tumble state from side, it can continue from target person
The object back side carries out tumble judgement, and the back side is fallen, and variation is less but fast with the time of falling in posture, the big spy of height change
Point, therefore the detection algorithm that relative height differential may be used carries out tumble judgement, specifically：If personage is high in certain frame number
Degree difference meets a certain distance (alternatively, personage's height meets certain proportion condition), then carries out second step and confirm other rows of exclusion
For (for example, judging whether the hand node of target person is less than preset 4th threshold value with the difference in height of foot node, if less than default
The 4th threshold value, then judge that target person is in tumble state).
In the present embodiment, for example, the personage's height Hnow (Hn) for recording current frame number and personage's height before 50 frames
Hbefore (Hb), if Hn<0.6*Hb, that carries out second step confirmation, is not otherwise to fall, when second step confirms, if mesh
Mark hand node and the foot node of personage difference in height (Hh) be less than height five/be first tumble state, flow chart such as Fig. 5
It is shown.
In the present embodiment, the result that S103 is obtained is the judging result of a certain frame, cannot represent whole process, therefore also
It need to judge whether target person is always maintained at tumble state in continuous certain frame number, if the target in continuous certain frame number
Personage is always maintained at tumble state and then confirms that target person is fallen, and sends out alarm.
In the present embodiment, if within the time during continuous 50 frame is about 2s, there is continuous 40 frame to be judged target person
In tumble state, then it can be confirmed that target person is fallen, and send out alarm.
Fall detection method provided in an embodiment of the present invention can make full use of the common camera of universalness in the family,
Situations such as public endowment environment or old solitary people, is monitored, personal or more people can be effectively distinguished and squat down, falls
Behavior can also judge the behavior of more people's close contacts.
To sum up, fall detection method provided in an embodiment of the present invention can identify the tumble row of personage by fall detection
To notify caregiver or surrounding people in time by alarm, target person being made timely to be helped to give first aid to, avoid relief not
There is life problem in time, is suitable for the indoor nurse place such as home for destitute, family, hospital.
It should be noted that herein, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any actual relationship or order or sequence.
The above is the preferred embodiment of the present invention, it is noted that for those skilled in the art
For, without departing from the principles of the present invention, several improvements and modifications can also be made, these improvements and modifications
It should be regarded as protection scope of the present invention.
Claims (9)
1. a kind of fall detection method, which is characterized in that including：
Obtain the video for including target person；
According to the video for including target person of acquisition, the characteristic point of target person is extracted；
For each frame image, according to the characteristic point of the target person extracted, from the front of target person, side, the back side three
A direction carries out tumble judgement；
If the judging result of the either direction in front, side or the back side, which is target person, is in tumble state, continuous one is judged
Whether target person is always maintained at tumble state in fixed frame number, if so, confirming that target person is fallen.
2. fall detection method according to claim 1, which is characterized in that it is described according to acquisition comprising target person
Video, the characteristic point for extracting target person include：
According to the video for including target person of acquisition, image procossing is carried out using Attitude estimation algorithm, extracts target person
Characteristic point.
3. fall detection method according to claim 2, which is characterized in that the characteristic point of extraction includes：Nose node, neck section
One or more of point, pereonite point, right-hand man's node, left and right toggle point, stern node, left and right knee node, left and right foot node.
4. fall detection method according to claim 1, which is characterized in that it is described to be directed to each frame image, according to extraction
The characteristic point of the target person arrived, carrying out tumble judgement from target person front includes：
Judge whether target person is in handstand state, if so, judgement target person is in tumble state；Otherwise, from target person
The side of object carries out tumble judgement.
5. fall detection method according to claim 4, which is characterized in that described to judge whether target person is in handstand shape
State, if so, judging that target person is in tumble state and includes：
If detecting the foot node of target person, judge whether the height of the neck node of target person is less than the height of pereonite point
And whether the height of pereonite point is less than the height of foot node；
If the two is both less than, judge that target person is in tumble state.
6. fall detection method according to claim 5, which is characterized in that described to judge whether target person is in handstand shape
State, if so, judgement target person be in tumble state, further include：
If detecting the foot node of fall short personage, judge whether the height of the neck node of target person is less than pereonite point
Highly；
If being less than, judge that target person is in tumble state.
7. fall detection method according to claim 1, which is characterized in that it is described to be directed to each frame image, according to extraction
The characteristic point of the target person arrived, carrying out tumble judgement from target person side includes：
Determine the first slope and the second slope for indicating target person body direction；
If first slope meets preset first threshold range and the second slope meets preset second threshold range, mesh is judged
Mark personage is in tumble state, otherwise, tumble judgement is carried out from the back side of target person；
Wherein, first slope is expressed as：
σ 1=ychest-yneck/xchest-xneck；
Second slope is expressed as：
σ 2=ychest-yfoot/xchest-xfoot
Wherein, σ 1, σ 2 indicate that first slope, the second slope, ychest, yneck, yfoot indicate pereonite point, neck section respectively respectively
Point, the coordinate value of foot node in the vertical direction, xchest, xneck, xfoot indicate pereonite point, neck node, foot node respectively
Coordinate value in the horizontal direction.
8. fall detection method according to claim 1, which is characterized in that it is described to be directed to each frame image, according to extraction
The characteristic point of the target person arrived, carrying out tumble judgement from the target person back side includes：
Judge whether the target person difference in height between preset two frame is more than preset third using the method for relative height differential
Threshold value；
If more than preset third threshold value, then whether the difference in height of the hand node and foot node that judge target person is less than preset
4th threshold value；
If being less than preset 4th threshold value, judge that target person is in tumble state.
9. fall detection method according to claim 1, which is characterized in that described after confirming that target person is fallen
Method further includes：
Send out alarm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810373224.5A CN108629300B (en) | 2018-04-24 | 2018-04-24 | Fall detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810373224.5A CN108629300B (en) | 2018-04-24 | 2018-04-24 | Fall detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108629300A true CN108629300A (en) | 2018-10-09 |
| CN108629300B CN108629300B (en) | 2022-01-28 |
Family
ID=63694302
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810373224.5A Active CN108629300B (en) | 2018-04-24 | 2018-04-24 | Fall detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108629300B (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109522793A (en) * | 2018-10-10 | 2019-03-26 | 华南理工大学 | More people's unusual checkings and recognition methods based on machine vision |
| CN109993063A (en) * | 2019-03-05 | 2019-07-09 | 福建天晴数码有限公司 | A kind of method and terminal identified to rescue personnel |
| CN110490070A (en) * | 2019-07-12 | 2019-11-22 | 浙江省北大信息技术高等研究院 | A kind of fall detection method based on human body attitude estimation |
| CN111241913A (en) * | 2019-12-19 | 2020-06-05 | 北京文安智能技术股份有限公司 | Method, device and system for detecting falling of personnel |
| CN111401296A (en) * | 2020-04-02 | 2020-07-10 | 浙江大华技术股份有限公司 | Behavior analysis method, equipment and device |
| CN111767888A (en) * | 2020-07-08 | 2020-10-13 | 北京澎思科技有限公司 | Object state detection method, computer device, storage medium, and electronic device |
| CN111783702A (en) * | 2020-07-20 | 2020-10-16 | 杭州叙简科技股份有限公司 | Efficient pedestrian tumble detection method based on image enhancement algorithm and human body key point positioning |
| CN111881898A (en) * | 2020-09-27 | 2020-11-03 | 西南交通大学 | Human body posture detection method based on monocular RGB image |
| CN111931733A (en) * | 2020-09-25 | 2020-11-13 | 西南交通大学 | Human body posture detection method based on depth camera |
| CN111986690A (en) * | 2020-08-27 | 2020-11-24 | 三星电子（中国）研发中心 | Voice noise reduction method and device for video |
| CN112287759A (en) * | 2020-09-26 | 2021-01-29 | 浙江汉德瑞智能科技有限公司 | Tumble detection method based on key points |
| WO2021164654A1 (en) * | 2020-02-20 | 2021-08-26 | 艾科科技股份有限公司 | Time-continuity-based detection determination system and method |
| US11182923B2 (en) | 2019-03-22 | 2021-11-23 | Wistron Corporation | Method, computing device, and system for monitoring postures |
| CN114694252A (en) * | 2022-03-30 | 2022-07-01 | 电子科技大学 | Old people falling risk prediction method |
| CN115273243A (en) * | 2022-09-27 | 2022-11-01 | 深圳比特微电子科技有限公司 | Fall detection method and device, electronic equipment and computer readable storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060145874A1 (en) * | 2002-11-21 | 2006-07-06 | Secumanagement B.V. | Method and device for fall prevention and detection |
| CN105279483A (en) * | 2015-09-28 | 2016-01-27 | 华中科技大学 | Fall-down behavior real-time detection method based on depth image |
| CN106503643A (en) * | 2016-10-18 | 2017-03-15 | 上海电力学院 | Tumble detection method for human body |
| CN107103733A (en) * | 2017-07-06 | 2017-08-29 | 司马大大（北京）智能系统有限公司 | One kind falls down alarm method, device and equipment |
| CN107194967A (en) * | 2017-06-09 | 2017-09-22 | 南昌大学 | Human fall detection method and device based on Kinect depth image |
-
2018
- 2018-04-24 CN CN201810373224.5A patent/CN108629300B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060145874A1 (en) * | 2002-11-21 | 2006-07-06 | Secumanagement B.V. | Method and device for fall prevention and detection |
| CN105279483A (en) * | 2015-09-28 | 2016-01-27 | 华中科技大学 | Fall-down behavior real-time detection method based on depth image |
| CN106503643A (en) * | 2016-10-18 | 2017-03-15 | 上海电力学院 | Tumble detection method for human body |
| CN107194967A (en) * | 2017-06-09 | 2017-09-22 | 南昌大学 | Human fall detection method and device based on Kinect depth image |
| CN107103733A (en) * | 2017-07-06 | 2017-08-29 | 司马大大（北京）智能系统有限公司 | One kind falls down alarm method, device and equipment |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109522793B (en) * | 2018-10-10 | 2021-07-23 | 华南理工大学 | Method for detecting and identifying abnormal behaviors of multiple persons based on machine vision |
| CN109522793A (en) * | 2018-10-10 | 2019-03-26 | 华南理工大学 | More people's unusual checkings and recognition methods based on machine vision |
| CN109993063A (en) * | 2019-03-05 | 2019-07-09 | 福建天晴数码有限公司 | A kind of method and terminal identified to rescue personnel |
| US11182923B2 (en) | 2019-03-22 | 2021-11-23 | Wistron Corporation | Method, computing device, and system for monitoring postures |
| CN110490070A (en) * | 2019-07-12 | 2019-11-22 | 浙江省北大信息技术高等研究院 | A kind of fall detection method based on human body attitude estimation |
| CN111241913A (en) * | 2019-12-19 | 2020-06-05 | 北京文安智能技术股份有限公司 | Method, device and system for detecting falling of personnel |
| WO2021164654A1 (en) * | 2020-02-20 | 2021-08-26 | 艾科科技股份有限公司 | Time-continuity-based detection determination system and method |
| CN111401296A (en) * | 2020-04-02 | 2020-07-10 | 浙江大华技术股份有限公司 | Behavior analysis method, equipment and device |
| CN111401296B (en) * | 2020-04-02 | 2023-09-29 | 浙江大华技术股份有限公司 | Behavior analysis method, device and apparatus |
| CN111767888A (en) * | 2020-07-08 | 2020-10-13 | 北京澎思科技有限公司 | Object state detection method, computer device, storage medium, and electronic device |
| CN111783702A (en) * | 2020-07-20 | 2020-10-16 | 杭州叙简科技股份有限公司 | Efficient pedestrian tumble detection method based on image enhancement algorithm and human body key point positioning |
| CN111986690A (en) * | 2020-08-27 | 2020-11-24 | 三星电子（中国）研发中心 | Voice noise reduction method and device for video |
| CN111931733A (en) * | 2020-09-25 | 2020-11-13 | 西南交通大学 | Human body posture detection method based on depth camera |
| CN112287759A (en) * | 2020-09-26 | 2021-01-29 | 浙江汉德瑞智能科技有限公司 | Tumble detection method based on key points |
| CN111881898A (en) * | 2020-09-27 | 2020-11-03 | 西南交通大学 | Human body posture detection method based on monocular RGB image |
| CN114694252A (en) * | 2022-03-30 | 2022-07-01 | 电子科技大学 | Old people falling risk prediction method |
| CN114694252B (en) * | 2022-03-30 | 2023-04-28 | 电子科技大学 | Old people falling risk prediction method |
| CN115273243A (en) * | 2022-09-27 | 2022-11-01 | 深圳比特微电子科技有限公司 | Fall detection method and device, electronic equipment and computer readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108629300B (en) | 2022-01-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108629300A (en) | A kind of fall detection method | |
| US11819344B2 (en) | Systems for automatic assessment of fall risk | |
| JP4587067B2 (en) | Method and apparatus for monitoring the fall of an object | |
| Jansen et al. | Context aware inactivity recognition for visual fall detection | |
| CN109558865A (en) | A kind of abnormal state detection method to the special caregiver of need based on human body key point | |
| CN112784662A (en) | Video-based fall risk evaluation system | |
| CN110477925B (en) | Fall detection and early warning method and system for elderly people in nursing home | |
| Debard et al. | Camera-based fall detection on real world data | |
| CN109963539B (en) | Nursing support system, control method thereof, and computer-readable recording medium | |
| CN107072548B (en) | Device, system and method for automatic detection of orientation and/or position of a person | |
| Jansen et al. | 3D human pose recognition for home monitoring of elderly | |
| WO2019013257A1 (en) | Monitoring assistance system and method for controlling same, and program | |
| JP6692987B2 (en) | Devices, Systems, and Methods for Patient Monitoring to Predict and Prevent Bed Fall | |
| CN112489368A (en) | Intelligent falling identification and detection alarm method and system | |
| CN115116133A (en) | Abnormal behavior detection system and method for monitoring solitary old people | |
| Liciotti et al. | Human activity analysis for in-home fall risk assessment | |
| Li et al. | Detection of patient's bed statuses in 3D using a Microsoft Kinect | |
| Xue et al. | Vision-based gait analysis for senior care | |
| CN109325474A (en) | A kind of abnormal state detection method of couple of special caregiver of need | |
| CN112071022A (en) | Fall monitoring method based on visual sensing and voice feedback | |
| WO2019003859A1 (en) | Monitoring system, control method therefor, and program | |
| Kittipanya-Ngam et al. | Computer vision applications for patients monitoring system | |
| CN112949417A (en) | Tumble behavior identification method, equipment and system | |
| CN108010579A (en) | Health monitoring system | |
| US20210007676A1 (en) | Systems and methods for detecting attempted bed exit |
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 |