CN110472614A - A kind of recognition methods for behavior of falling in a swoon - Google Patents
A kind of recognition methods for behavior of falling in a swoon Download PDFInfo
- Publication number
- CN110472614A CN110472614A CN201910778281.6A CN201910778281A CN110472614A CN 110472614 A CN110472614 A CN 110472614A CN 201910778281 A CN201910778281 A CN 201910778281A CN 110472614 A CN110472614 A CN 110472614A
- Authority
- CN
- China
- Prior art keywords
- data
- cluster
- swoon
- falling
- threshold value
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of recognition methods of behavior of falling in a swoon.The invention belongs to monitoring technology field more particularly to a kind of recognition methods for behavior of falling in a swoon.It solves and needs to be arranged multiple cameras in the prior art to detect to the behavior of falling in a swoon, and detect inaccurate problem.Technical solution of the present invention: the data moved in video data are obtained using frame difference method, character contour similarity and area similarity are matched by clustering algorithm, it is tracked, calculate boundary rectangle, if there is abrupt change in target speed and the length-width ratio of boundary rectangle is more than threshold value or is inverted, it is judged as and falls down, is judged as if falling down the time more than threshold value and falls in a swoon.The present invention can solve finds that monitoring objective occurred does the case where falling in a swoon and in time with corresponding processing in time in the management of gymnasium, and target cluster is analyzed and calculated, and can judge to fall down, posture of falling in a swoon, to accurately judge that the behavior of falling in a swoon.
Description
Technical field
The invention belongs to monitoring technology field more particularly to a kind of recognition methods for behavior of falling in a swoon.
Background technique
Syncope is that of short duration sexual dysfunction occurs for the contraction and diastolic function due to blood vessel, caused by causing brain transient ischemic,
If people is stimulated suddenly, extremely frightened and severe pain, anaemia patient sitting or couchant start up are come, and have heart disease or spondylodynia
Deng all may cause and fall in a swoon suddenly.
Currently, monitoring screen is monitored come long-time by the way that multiple monitoring systems are arranged in safety monitoring system,
When discovery picture exception, alarmed when falling in a swoon by warning device to be monitored to this.
In the prior art, the connection type being connected directly using camera and monitoring system host, wherein monitoring system is wanted
The problem of several hundred, even thousands of a cameras are set, complicated construction, higher cost will be will cause using this kind of connection type.
Summary of the invention
The behavior of falling in a swoon is detected for needing to be arranged multiple cameras in the prior art, and detects inaccurate ask
Topic, the present invention provide a kind of recognition methods of behavior of falling in a swoon, its object is to: it solves to find prison in time in the management of gymnasium
What control target occurred does the case where falling in a swoon and in time with corresponding processing, and target cluster is analyzed and calculated, and can be sentenced
It is disconnected fall down, posture of falling in a swoon, to accurately judge that the behavior of falling in a swoon.
The technical solution adopted by the invention is as follows:
A kind of recognition methods for behavior of falling in a swoon, comprising the following steps:
Step 1: the mobile data in video data being obtained using frame difference method, is subtracted with the gradation data of mobile data present frame
The gradation data for removing mobile data previous frame, obtained difference data, the difference data are delta data;
Step 2: by line line filter, the Gaussian noise of delta data is eliminated, by Gaussian filter to delta data
It is weighted and averaged, to other pixel values in itself and contiguous range by obtaining each pixel put after weighted average
Value;
Step 3: median filtering is carried out to pixel value, by thresholding, constitutes two dimension pattern plate, it will be all in two dimension pattern plate
Pixel value by arrive greatly it is small be ranked up, generate the 2-D data sequence of monotone increasing, export are as follows:
G (x, y)=med { f (x-k, y-l), (k, l ∈ W) },
Wherein, W is two dimension pattern plate, and the initialization of this template is one 5 × 5 line-like area, and f (x, y), G (x, y) are respectively
Image after original image and thresholding;
Step 4: 2-D data sequence obtains cluster data by the clustering algorithm of weighting coefficient, and clustering algorithm is based on density
Cluster clusters (dbscan), and dbscan is a kind of Classic Clustering Algorithms based on density, the image ash handled through the above steps
Entire gray level image is defined as density space in by degree, and defining first can be by as cluster in neighbor point around a point
Adjacent domain radius epsilon, re-define this region include at least point number minPts;Its midpoint is above-mentioned steps
Pixel value is the pixel of (255,255,255) after processing;Epsilon is 5;MinPts is data width 5;Wherein epsilon
It is all empirical data with minPts;The length-width ratio of cluster data boundary rectangle and rectangle is added, by cluster data and human body
Outline data compares, and by SIFT algorithm, when similarity is greater than 55%, addition judges queue;
(1) according to the above parameter, the point in image data can be divided into three classes:
A: epipole (core), arbitrary point p meet in the proximity point p in (radius is less than or equal to epsilon) and this region
The quantity of point is more than or equal to minPts, then the point is epipole;
B: marginal point (border) meets in the proximity p (radius is less than or equal to epsilon) quantity and is less than minPts, still
Fall in the point in core neighborhood of a point;
C: outlier (outlier): neither core point is also not marginal point, just belong to outlier
(2) select any point p, according to pesilon and minPts judge this point whether be core point, marginal point or
It is outlier, and the point deletion that will peel off;
(3) if the distance between core point is less than minPts, just two core points are linked together, result in formation of
Several groups cluster;
(4) according to minPts, marginal point is assigned in closest to him core point range;
(5) above step is repeated, until all the points meet in cluster or are outlier;
Step 5: target following is carried out to this cluster data, the attribute change of cluster data is recorded by target following,
Data set M is created, if the data of M are sky, this cluster data are tracked through the above way, M is added in data;
Step 6: calculating the perhaps movement velocity of inversion and cluster data and be more than threshold value or transient change occur, be arranged
The cluster is target cluster;Define the mass center (x, y) of target person trunk;The length-width ratio and cluster numbers of cluster data boundary rectangle
According to movement speed, if cluster data boundary rectangle length-width ratio be more than threshold value;
Wherein, n: the skeleton point of torso portion;
By the pixel difference between present frame and the centroid position of previous frame target person, carried according to frame per second and program operation
The speed of service of body server calculates the movement velocity of target person,
V=pixel/s × (FPS × Vrun)
Wherein, s: the dimension of object ratio of pixel size and logic judgment, Vnm: every frame method every frame threshold value, pixel: preceding
Euclidean distance between two frame mass centers afterwards;
Step 7: in lateral distance, target cluster boundary rectangle is made comparisons with other cluster boundary rectangles, if other poly-
Class boundary rectangle width is less than the width of target cluster boundary rectangle, and the length-width ratio of this target cluster is greater than threshold value, then judges
To fall down, the length-width ratio of boundary rectangle is inverted in addition to falling down under normal circumstances, and there are also the arm actions of target person suddenly
Variation, at this time, it may be necessary to carry out carrying out ratio judgement with the personage of lateral coordinates in space, it is ensured that target person is in image space
Height change, can just be judged as and fall down;
Step 8: calculating target and cluster the time fallen down, if the time is more than threshold time, target cluster is not sent out again
Raw speed transient change, movement velocity is not above threshold value, and length-width ratio is not above threshold value, and conditions above is full up, and foot is judged as
It falls in a swoon.
Wherein, clustering algorithm is to be clustered based on density cluster in the step 4.
Wherein, include 2 kinds of situations in the step 6:
Situation 1: if the length-width ratio of cluster data boundary rectangle is not above threshold value, judgement is exited;
Situation 2: if cluster data boundary rectangle length-width ratio is more than threshold value or inversion, and the movement velocity clustered is more than threshold
It is worth or occurs transient change, then enters and judge in next step.
Wherein, the threshold time in the step 8 is -40 seconds 20 seconds.
In conclusion by adopting the above-described technical solution, the beneficial effects of the present invention are:
1. falling in a swoon of finding that monitoring objective occurs in time in the management of gymnasium and is done in time with corresponding place situation
Reason, moves personage and movement is analyzed and calculated, and preferably can judge to fall down, to accurately judge that row of falling in a swoon
For.
Detailed description of the invention
Examples of the present invention will be described by way of reference to the accompanying drawings, in which:
Fig. 1 is flow diagram of the present invention.
Fig. 2 is target cluster boundary rectangle schematic diagram of the present invention.
Specific embodiment
All features disclosed in this specification or disclosed all methods or in the process the step of, in addition to mutually exclusive
Feature and/or step other than, can combine in any way.
It elaborates below with reference to Fig. 1, Fig. 2 to the present invention.
A kind of recognition methods for behavior of falling in a swoon, comprising the following steps:
Step 1: the mobile data in video data being obtained using frame difference method, is subtracted with the gradation data of mobile data present frame
The gradation data for removing mobile data previous frame, obtained difference data, the difference data are delta data;
Step 2: by line line filter, the Gaussian noise of delta data is eliminated, by Gaussian filter to delta data
It is weighted and averaged, to other pixel values in itself and contiguous range by obtaining each pixel put after weighted average
Value;
Step 3: median filtering is carried out to pixel value, by thresholding, constitutes two dimension pattern plate, it will be all in two dimension pattern plate
Pixel value by arrive greatly it is small be ranked up, generate the 2-D data sequence of monotone increasing, export are as follows:
G (x, y)=med { f (x-k, y-l), (k, l ∈ W) },
Wherein, W is two dimension pattern plate, and the initialization of this template is the line-like area of a 5*5, and f (x, y), G (x, y) are respectively
Image after original image and thresholding;
Step 4: 2-D data sequence obtains cluster data by the clustering algorithm of weighting coefficient, and clustering algorithm is based on density
Cluster clusters (dbscan), and dbscan is a kind of Classic Clustering Algorithms based on density, the image ash handled through the above steps
Entire gray level image is defined as density space in by degree, and defining first can be by as cluster in neighbor point around a point
Adjacent domain radius epsilon, re-define this region include at least point number minPts;Its midpoint is above-mentioned steps
Pixel value is the pixel of (255,255,255) after processing;Epsilon is 5;MinPts is data width 5;Wherein epsilon
It is all empirical data with minPts;The length-width ratio of cluster data boundary rectangle and rectangle is added, by cluster data and human body
Outline data compares, and by SIFT algorithm, when similarity is greater than 55%, addition judges queue;
(1) according to the above parameter, the point in image data can be divided into three classes:
A: epipole (core), arbitrary point p meet in the proximity point p in (radius is less than or equal to epsilon) and this region
The quantity of point is more than or equal to minPts, then the point is epipole;
B: marginal point (border) meets in the proximity p (radius is less than or equal to epsilon) quantity and is less than minPts, still
Fall in the point in core neighborhood of a point;
C: outlier (outlier): neither core point is also not marginal point, just belong to outlier
(2) select any point p, according to pesilon and minPts judge this point whether be core point, marginal point or
It is outlier, and the point deletion that will peel off;
(3) if the distance between core point is less than minPts, just two core points are linked together, result in formation of
Several groups cluster;
(4) according to minPts, marginal point is assigned in closest to him core point range;
(5) above step is repeated, until all the points meet in cluster or are outlier;
Step 5: target following is carried out to this cluster data, the attribute change of cluster data is recorded by target following,
Data set M is created, if the data of M are sky, this cluster data are tracked through the above way, M is added in data;
Step 6: the length-width ratio of cluster data boundary rectangle and the movement speed of cluster data are calculated, if outside cluster data
Connecing rectangular aspect ratio, perhaps the movement velocity of inversion and cluster data is more than threshold value or transient change occurs more than threshold value, if
The cluster is set as target cluster;Define the mass center (x, y) of target person trunk;
Wherein, n: the skeleton point of torso portion;
By the pixel difference between present frame and the centroid position of previous frame target person, carried according to frame per second and program operation
The speed of service of body server calculates the movement velocity of target person,
V=pixel/s × (FPS × Vrun)
Wherein, s: the dimension of object ratio of pixel size and logic judgment, Vnm: every frame method every frame threshold value, pixel: preceding
Euclidean distance between two frame mass centers afterwards;
Wherein, the threshold value value process of length-width ratio is to calculate the existing rectangle of target long (y-axis length) wide (x-axis length) ratio
Value, spevalue=x:y record the value if current ratio is greater than 1 (ratio greater than 1 generally falls into abnormal ranges), and
Calculate inverse ratio.If changed in a certain frame rectangular profile, ratio tends to 1 or is greater than record value, record current value and record
Variable quantity between value, variation are more than that the ratio of 60% and the length-width ratio at this time and record value length-width ratio before of record value is greater than
Equal to 1.5, then it is considered and falls down;Otherwise record process is skipped, if ratio variation is more than or equal to 1.5 or is inverted, directly
It connects and regards as falling down process.
Wherein, the threshold value value process of movement velocity are as follows: this value is a preset value, obtains the movement velocity of target person
Set (movement velocity of same personage only takes once) is added, when set sizes are greater than 10, calculates average movement velocity.If
In a certain frame, the movement speed of target person is more than the 30% of average movement velocity, then belongs to more than threshold value.This threshold value needs reality
Shi Gengxin, and need to exclude following data addition and recalculate:
(1), it has been judged as the angular movement speed data more than threshold value
(2), it is considered to be (speed is in the case where 40 pixels/s and there is no realities for the angular movement speed data of noise data
Personage's speed of border displacement).
Step 7: in lateral distance, target cluster boundary rectangle is made comparisons with other cluster boundary rectangles, if other poly-
Class boundary rectangle width is less than the width of target cluster boundary rectangle, and the length-width ratio of this target cluster is greater than threshold value, then judges
To fall down, the length-width ratio of boundary rectangle is inverted in addition to falling down under normal circumstances, and there are also the arm actions of target person suddenly
Variation, at this time, it may be necessary to carry out carrying out ratio judgement with the personage of lateral coordinates in space, it is ensured that target person is in image space
Height change, can just be judged as and fall down;
Step 8:
(1) it calculates target and clusters the time fallen down, the time is more than threshold time;
(2) target cluster is not above threshold value that speed transient change, movement velocity do not occur again;
(3) length-width ratio is not above threshold value;
Meet (1), (2) and (3) simultaneously and be then judged as and falls in a swoon;
Wherein, clustering algorithm is to be clustered based on density cluster in the step 4.
Wherein, include 2 kinds of situations in the step 6:
Situation 1: if the length-width ratio of cluster data boundary rectangle is not above threshold value, judgement is exited;
Situation 2: if cluster data boundary rectangle length-width ratio is more than threshold value or inversion, and the movement velocity clustered is more than threshold
It is worth or occurs transient change, then enters and judge in next step.
Wherein, the threshold time in the step 8 is -40 seconds 20 seconds.
The specific embodiment of the application above described embodiment only expresses, the description thereof is more specific and detailed, but simultaneously
The limitation to the application protection scope therefore cannot be interpreted as.It should be pointed out that for those of ordinary skill in the art
For, under the premise of not departing from technical scheme design, various modifications and improvements can be made, these belong to this
The protection scope of application.
Claims (4)
1. a kind of recognition methods for behavior of falling in a swoon, which comprises the following steps:
Step 1: the mobile data in video data being obtained using frame difference method, subtracts shifting with the gradation data of mobile data present frame
The gradation data of dynamic data previous frame, obtained difference data, the difference data are delta data;
Step 2: delta data being weighted and averaged by Gaussian filter, by the picture for obtaining each point after weighted average
Element value;
Step 3: median filtering being carried out to pixel value, by thresholding, two dimension pattern plate is constituted, by all pixels in two dimension pattern plate
Value by arrive greatly it is small be ranked up, generate the 2-D data sequence of monotone increasing, the 2-D data sequence output are as follows:
G (x, y)=med { f (x-k, y-l), (k, l ∈ W) },
Wherein, W is two dimension pattern plate, and the initialization of this template is one 5 × 5 line-like area, and f (x, y) is original image, G (x, y)
For the image after thresholding;
Step 4: 2-D data sequence obtains cluster data by the clustering algorithm of weighting coefficient, be added cluster boundary rectangle and
The length-width ratio of rectangle compares the outline data of cluster data and human body;
Step 5: target following being carried out to the cluster data, the attribute change of cluster data is recorded by target following;
Step 6: the attribute change includes the length-width ratio and movement speed of boundary rectangle, calculates the length of cluster data boundary rectangle
Width than and cluster data movement speed, if cluster data boundary rectangle length-width ratio is more than threshold value or inversion, and cluster numbers
According to movement velocity be more than threshold value or transient change occur, it is that target clusters that the cluster, which is arranged,;
Step 7: in lateral distance, target cluster boundary rectangle is made comparisons with other cluster boundary rectangles, if outside other clusters
Width of the rectangle width less than target cluster boundary rectangle is connect, and the length-width ratio of this target cluster is greater than threshold value, then is judged as and falls
;
Step 8: calculating target and cluster the time fallen down, if the time is more than threshold time, target cluster without occurring speed again
Transient change is spent, movement velocity is not above threshold value, and length-width ratio is not above threshold value, and conditions above is full up, and foot is then judged as dizzy
.
2. a kind of recognition methods of behavior of falling in a swoon according to claim 1, which is characterized in that cluster and calculate in the step 4
Method is to be clustered based on density cluster.
3. a kind of recognition methods of behavior of falling in a swoon according to claim 1, which is characterized in that include 2 kinds in the step 6
Situation:
Situation 1: if the length-width ratio of cluster data boundary rectangle is not above threshold value, judgement is exited;
Situation 2: if cluster data boundary rectangle length-width ratio be more than threshold value or inversion, and cluster movement velocity be more than threshold value or
There is transient change in person, then enters and judge in next step.
4. a kind of recognition methods of behavior of falling in a swoon according to claim 1, which is characterized in that the threshold value in the step 8
Time is -40 seconds 20 seconds.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910778281.6A CN110472614B (en) | 2019-08-22 | 2019-08-22 | Identification method for motion sickness |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910778281.6A CN110472614B (en) | 2019-08-22 | 2019-08-22 | Identification method for motion sickness |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110472614A true CN110472614A (en) | 2019-11-19 |
CN110472614B CN110472614B (en) | 2023-06-30 |
Family
ID=68513455
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910778281.6A Active CN110472614B (en) | 2019-08-22 | 2019-08-22 | Identification method for motion sickness |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110472614B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112560723A (en) * | 2020-12-22 | 2021-03-26 | 中电海康集团有限公司 | Fall detection method and system based on form recognition and speed estimation |
CN113158783A (en) * | 2021-03-10 | 2021-07-23 | 重庆特斯联智慧科技股份有限公司 | Community resident health monitoring method and system based on human body recognition |
CN113239874A (en) * | 2021-06-01 | 2021-08-10 | 平安科技(深圳)有限公司 | Behavior posture detection method, device, equipment and medium based on video image |
CN113628413A (en) * | 2021-08-30 | 2021-11-09 | 中山大学附属第三医院(中山大学肝脏病医院) | Automatic alarm and help-seeking technology for accidents of wearing and taking off protective clothing |
Citations (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2009266052A (en) * | 2008-04-28 | 2009-11-12 | Hitachi Ltd | Abnormal behavior detector |
US20100124358A1 (en) * | 2008-11-17 | 2010-05-20 | Industrial Technology Research Institute | Method for tracking moving object |
CN102044131A (en) * | 2010-07-20 | 2011-05-04 | 北京紫峰华纳科技有限公司 | Alarm signal processing method and human body position information-based alarm method |
WO2011143711A1 (en) * | 2010-05-19 | 2011-11-24 | Australasian Pork Research Institute Ltd | Image analysis for making animal measurements |
JP2012036005A (en) * | 2010-07-12 | 2012-02-23 | Mitsubishi Electric Corp | Fall detecting device and passenger conveyor |
CN102722715A (en) * | 2012-05-21 | 2012-10-10 | 华南理工大学 | Tumble detection method based on human body posture state judgment |
CN103093197A (en) * | 2013-01-15 | 2013-05-08 | 信帧电子技术(北京)有限公司 | Monitoring method and system for recognizing hanging behavior |
CN103118249A (en) * | 2013-03-13 | 2013-05-22 | 胡茂林 | Housebound elder intelligence video surveillance system |
CN203328700U (en) * | 2013-06-06 | 2013-12-11 | 刘翔 | Motion-state tracking and real-time transmission system |
CN103927743A (en) * | 2014-03-27 | 2014-07-16 | 中国科学院长春光学精密机械与物理研究所 | Method for detecting man-made targets in remotely-sensed images |
CN104574441A (en) * | 2014-12-31 | 2015-04-29 | 浙江工业大学 | Fall real-time detection method based on GMM and time sequence model |
FR3028343A1 (en) * | 2014-11-10 | 2016-05-13 | Centre Nat Rech Scient | METHOD FOR DETECTING THE FALL OF A HUMAN SUBJECT AND CORRESPONDING ACTIMETRIC DEVICE |
CN106204640A (en) * | 2016-06-29 | 2016-12-07 | 长沙慧联智能科技有限公司 | A kind of moving object detection system and method |
CN106557812A (en) * | 2016-11-21 | 2017-04-05 | 北京大学 | The compression of depth convolutional neural networks and speeding scheme based on dct transform |
CN106708084A (en) * | 2016-11-24 | 2017-05-24 | 中国科学院自动化研究所 | Method for automatically detecting and avoiding obstacles for unmanned aerial vehicle under complicated environments |
CN109635721A (en) * | 2018-12-10 | 2019-04-16 | 山东大学 | Video human fall detection method and system based on track weighting depth convolution sequence poolization description |
CN109670396A (en) * | 2018-11-06 | 2019-04-23 | 华南理工大学 | A kind of interior Falls Among Old People detection method |
-
2019
- 2019-08-22 CN CN201910778281.6A patent/CN110472614B/en active Active
Patent Citations (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2009266052A (en) * | 2008-04-28 | 2009-11-12 | Hitachi Ltd | Abnormal behavior detector |
US20100124358A1 (en) * | 2008-11-17 | 2010-05-20 | Industrial Technology Research Institute | Method for tracking moving object |
WO2011143711A1 (en) * | 2010-05-19 | 2011-11-24 | Australasian Pork Research Institute Ltd | Image analysis for making animal measurements |
JP2012036005A (en) * | 2010-07-12 | 2012-02-23 | Mitsubishi Electric Corp | Fall detecting device and passenger conveyor |
CN102044131A (en) * | 2010-07-20 | 2011-05-04 | 北京紫峰华纳科技有限公司 | Alarm signal processing method and human body position information-based alarm method |
CN102722715A (en) * | 2012-05-21 | 2012-10-10 | 华南理工大学 | Tumble detection method based on human body posture state judgment |
CN103093197A (en) * | 2013-01-15 | 2013-05-08 | 信帧电子技术(北京)有限公司 | Monitoring method and system for recognizing hanging behavior |
CN103118249A (en) * | 2013-03-13 | 2013-05-22 | 胡茂林 | Housebound elder intelligence video surveillance system |
CN203328700U (en) * | 2013-06-06 | 2013-12-11 | 刘翔 | Motion-state tracking and real-time transmission system |
CN103927743A (en) * | 2014-03-27 | 2014-07-16 | 中国科学院长春光学精密机械与物理研究所 | Method for detecting man-made targets in remotely-sensed images |
FR3028343A1 (en) * | 2014-11-10 | 2016-05-13 | Centre Nat Rech Scient | METHOD FOR DETECTING THE FALL OF A HUMAN SUBJECT AND CORRESPONDING ACTIMETRIC DEVICE |
CN104574441A (en) * | 2014-12-31 | 2015-04-29 | 浙江工业大学 | Fall real-time detection method based on GMM and time sequence model |
CN106204640A (en) * | 2016-06-29 | 2016-12-07 | 长沙慧联智能科技有限公司 | A kind of moving object detection system and method |
CN106557812A (en) * | 2016-11-21 | 2017-04-05 | 北京大学 | The compression of depth convolutional neural networks and speeding scheme based on dct transform |
CN106708084A (en) * | 2016-11-24 | 2017-05-24 | 中国科学院自动化研究所 | Method for automatically detecting and avoiding obstacles for unmanned aerial vehicle under complicated environments |
CN109670396A (en) * | 2018-11-06 | 2019-04-23 | 华南理工大学 | A kind of interior Falls Among Old People detection method |
CN109635721A (en) * | 2018-12-10 | 2019-04-16 | 山东大学 | Video human fall detection method and system based on track weighting depth convolution sequence poolization description |
Non-Patent Citations (4)
Title |
---|
刘楠楠: "基于多特征融合的跌倒行为识别与研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
刘皓: "基于视频的多特征融合摔倒检测算法研究" * |
王鹏等: "基于FPGA的视频监控摔倒检测报警***", 《电机与控制学报》 * |
蔡耀萱: "基于智能手机的人体跌倒检测方法研究与实现" * |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112560723A (en) * | 2020-12-22 | 2021-03-26 | 中电海康集团有限公司 | Fall detection method and system based on form recognition and speed estimation |
CN112560723B (en) * | 2020-12-22 | 2023-10-17 | 中电海康集团有限公司 | Fall detection method and system based on morphological recognition and speed estimation |
CN113158783A (en) * | 2021-03-10 | 2021-07-23 | 重庆特斯联智慧科技股份有限公司 | Community resident health monitoring method and system based on human body recognition |
CN113239874A (en) * | 2021-06-01 | 2021-08-10 | 平安科技(深圳)有限公司 | Behavior posture detection method, device, equipment and medium based on video image |
WO2022252642A1 (en) * | 2021-06-01 | 2022-12-08 | 平安科技(深圳)有限公司 | Behavior posture detection method and apparatus based on video image, and device and medium |
CN113239874B (en) * | 2021-06-01 | 2024-05-03 | 平安科技(深圳)有限公司 | Behavior gesture detection method, device, equipment and medium based on video image |
CN113628413A (en) * | 2021-08-30 | 2021-11-09 | 中山大学附属第三医院(中山大学肝脏病医院) | Automatic alarm and help-seeking technology for accidents of wearing and taking off protective clothing |
Also Published As
Publication number | Publication date |
---|---|
CN110472614B (en) | 2023-06-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110472614A (en) | A kind of recognition methods for behavior of falling in a swoon | |
St-Charles et al. | Flexible background subtraction with self-balanced local sensitivity | |
US9208675B2 (en) | Loitering detection in a video surveillance system | |
CN109614922A (en) | A kind of dynamic static gesture identification method and system | |
US8494267B2 (en) | Image processing device, image processing method, and program for normalizing a histogram | |
CN104134077B (en) | A kind of gait recognition method unrelated with visual angle based on the determination theories of learning | |
CN108198206A (en) | The multi-object tracking method combined based on multiple features combining and Camshift algorithms | |
US20030165193A1 (en) | Method for abstracting multiple moving objects | |
US20220254171A1 (en) | Event detector and method of generating textural image based on event count decay factor and net polarity | |
CN113168520A (en) | Method of tracking objects in a scene | |
KR101646000B1 (en) | Surveillance System And Method To Object And Area | |
JP2019032588A (en) | Image analysis apparatus | |
Vasuhi et al. | Real time multiple human tracking using kalman filter | |
JP2017068705A (en) | Detection program, detection method and detection device | |
CN109299702B (en) | Human behavior recognition method and system based on depth space-time diagram | |
CN109426797A (en) | Doze detection device and doze detection method thereof | |
CN117132949B (en) | All-weather fall detection method based on deep learning | |
KR101542206B1 (en) | Method and system for tracking with extraction object using coarse to fine techniques | |
CN113536849A (en) | Crowd gathering identification method and device based on image identification | |
Khashman | Automatic detection, extraction and recognition of moving objects | |
Zweng et al. | Introducing a statistical behavior model into camera-based fall detection | |
CN112347830A (en) | Factory epidemic prevention management method and system | |
Juang et al. | Vision-based human body posture recognition using support vector machines | |
Fang et al. | A vision-based infant monitoring system using PT IP camera | |
Kavya et al. | Human Fall Detection Using Video Surveillance |
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 |