CN109325474A - A kind of abnormal state detection method of couple of special caregiver of need - Google Patents
A kind of abnormal state detection method of couple of special caregiver of need Download PDFInfo
- Publication number
- CN109325474A CN109325474A CN201811371951.4A CN201811371951A CN109325474A CN 109325474 A CN109325474 A CN 109325474A CN 201811371951 A CN201811371951 A CN 201811371951A CN 109325474 A CN109325474 A CN 109325474A
- Authority
- CN
- China
- Prior art keywords
- posture
- human
- human body
- caregiver
- image
- 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.)
- Withdrawn
Links
- 230000002159 abnormal effect Effects 0.000 title claims abstract description 20
- 238000001514 detection method Methods 0.000 title claims abstract description 11
- 238000000034 method Methods 0.000 claims abstract description 22
- 230000004438 eyesight Effects 0.000 claims abstract description 12
- 238000012423 maintenance Methods 0.000 claims abstract description 11
- 238000012360 testing method Methods 0.000 claims abstract description 11
- 238000012544 monitoring process Methods 0.000 claims abstract description 8
- 238000010801 machine learning Methods 0.000 claims abstract description 5
- 238000012549 training Methods 0.000 claims description 6
- 230000000007 visual effect Effects 0.000 claims description 5
- 230000014759 maintenance of location Effects 0.000 claims description 4
- 230000002040 relaxant effect Effects 0.000 claims description 3
- 238000013528 artificial neural network Methods 0.000 claims description 2
- 238000012706 support-vector machine Methods 0.000 claims description 2
- 238000013135 deep learning Methods 0.000 claims 1
- 238000000638 solvent extraction Methods 0.000 claims 1
- 238000011410 subtraction method Methods 0.000 claims 1
- 230000005856 abnormality Effects 0.000 abstract 1
- 230000007812 deficiency Effects 0.000 description 2
- 238000009434 installation Methods 0.000 description 2
- 206010000117 Abnormal behaviour Diseases 0.000 description 1
- 208000028752 abnormal posture Diseases 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000036772 blood pressure Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 210000003414 extremity Anatomy 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000035479 physiological effects, processes and functions Effects 0.000 description 1
- 210000000707 wrist Anatomy 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Human Computer Interaction (AREA)
- Alarm Systems (AREA)
Abstract
The present invention relates to a kind of to the abnormal state detection method for needing special caregiver.The method includes following features: the maintenance of background scene image intelligent keeps the background scene image of last state at any time;Scene areas, boundary and specific objective identify in advance;Human testing and tracking;Human body attitude judgement is done to the human body image detected;Comprehensive preceding feature, by the time parameter of caregiver's posture different in monitoring scene, different zones, whether comprehensive descision is abnormal for analysis.This method is based on computer vision field, is detected using machine learning method to human body abnormality, possesses the advantages of equipment cost is low, discovery situation is timely, accuracy in detection is high, strong robustness.
Description
Technical field
The present invention relates to a kind of based on computer vision to the abnormal state detection method for needing special caregiver.
Background technique
With the development of the times, society increasingly pays close attention to the disadvantaged group such as old man, patient, physical disabilities.It is many more crisp
Weak personnel have to live alone because of various reasons, such as the patient in old solitary people, blind person and the ward in family.With
The increase of human cost, very more weak tendency personnel effectively cannot be nursed and be ensured.They once encounter emergency case and lead
Cause falls down, falls in a swoon, and loses consciousness or self-saving ability, and cannot be found in time by other people, it is more likely that threat to life.Cut-off 2018
Year, the existing the elderly in China about 2.4 hundred million, disabled person about 85,000,000 have become the important of today's society to the nurse of weak tendency personnel
Project.
The equipment for needing special caregiver to be monitored and assist nurse can be divided into three categories at present, the first kind is intelligence
Energy wearable device refers to the physiology such as human motion acceleration, heart rate, blood pressure by being worn on the positions such as wrist, arm, chest
Target perception, situation and alarms to determine whether facing a danger.Such equipment precision is high, but worn for long periods is uncomfortable, inconvenient,
It charges relatively complicated.Second class is to monitor people needing caregiver zone of action such as floor, bed and sofa to be laid with sensor
It alarms when body abnormal behavior.Such equipment installation cost is high, and difficult in maintenance, environmental suitability is poor.Third class is by taking the photograph
As head acquisition video information, the human body target in video is extracted with microprocessor, analyzes current human's shape in conjunction with related algorithm
State.Such equipment installation maintenance is simple, will not influence the adventure in daily life for needing caregiver, but portioned product detection accuracy is not
It is high.
With the increase of hardware computing capability and the rise of artificial intelligence, computer vision application field is more and more wider.Mesh
The preceding method analyzed based on video human body attitude is had: 1, it determines the body part of people, the connection of each limbs is identified, this method
Operand is big, to hardware requirement height.2, human object is tracked, is estimated with manikin algorithm and needs caregiver current
Posture, according to testing result and whether human body is in nonstatic region to judge to need whether caregiver falls down, for difference
The abnormal conditions of grade are alarmed.This method determines the depth-width ratio for being identified by human body tracking frame of human body attitude, right
In seat, squat and kneel etc. similar in posture be easy erroneous judgement.3, by identifying that Face datection finds human body, to frame human body appearance every in video
State is judged.It is blocked for face, fails not judging the case where identifying face.
There is also following deficiencies for current gesture recognition method: 1, posture is generally based on single frames judgement, will not be more
Frame joint judgement, easily occurs erroneous judgement situation.2, recognition methods is more single, lacks the maintenance to background, shadow variation, furniture and
The mobile of object can disturbance ecology.3, the multiple parameters such as human body attitude and time, position are not combined analysis, it cannot be fine
Differentiation current human's state.4, special training is not carried out to application scenarios, training pictures are downloaded from the Internet, with product
Camera parameter, actual use scene mismatch, and training classifying quality is bad.
Summary of the invention
It is an object of the present invention to provide a kind of based on computer vision to the monitoring abnormal state side for needing special caregiver
Method.By needing special caregiver's indoor location video capture device, video and instant analysis in real-time collection room are found different
It normal posture and abnormal posture retention time, alerts in time.For up to this purpose of realization and for aforementioned the deficiencies in the prior art, this hair
It is bright to adopt the technical scheme that:
M1 background scene image intelligent maintenance: to the continuous intelligent maintenance of background scene image, at any time according to background scene
Shadow variation and the movement of non-human target are updated, and the moment safeguards the background scene image of last state;
The mark and intelligent maintenance of M2 scene areas, boundary and specific objective:
1) mark such as zone of action, relaxing area, so that whether subsequent step intelligent decision difference posture is abnormal in corresponding region;
2) toilet doorway is identified, so that whether the time of subsequent step judgement into toilet is abnormal;
3) balcony, monitoring visual angle blind zone etc. are identified, so that subsequent step judges whether entry time overlength;
4) access door mouth is identified, subsequent step is judged as normal exit.
M3 human testing and tracking: by doing gaussian filtering to instant background image and instant video image, figure is reused
As frame difference method subduction, by result binaryzation and corroded and expanded, obtains foreground target.Human testing is done to foreground image again
And it tracks;
The judgement of M4 posture:
1) machine learning method is used, in advance using the image that simulated scenario is shot as training library, trains human body attitude feature
Library;
2) human body attitude judgement is done to the human body image that M3 is detected, judges current human's posture;
M5 combination judgement: comprehensive abovementioned steps, the posture that combinatory analysis needs caregiver different in monitoring scene is not
With region and retention time and human body in different zones disengaging and residence time, whether extremely according to preparatory comprehensive descision.
The present invention carries out auxiliary nurse to specific people using computer vision technique, have it is at low cost, discovery situation and
When etc. advantages, while the judgement of abnormal conditions is not limited to individually to judge certain abnormal operation, further includes time factor, to personnel
It falls down at a slow speed, in some regions, for example the toilet residence time is too long or bed rest time is too long etc. is difficult to judge by the prior art
Abnormal conditions in time, accurately can find and alarm.
Detailed description of the invention
Fig. 1 is that functional module of the invention forms;
Specific embodiment
It is connected with 1080P, the Freescale mainboard device of USB camera at 100 degree of visual angle is installed on and needs caregiver room
Between one jiao, about 2 meters away from ground, camera is against room, and the vision signal of acquisition is by the journey for having following functions pre-installed in equipment
Sequence module is handled:
The maintenance of background scene image intelligent: the background scene figure of each frame is analyzed, continuous intelligent maintenance determines background at any time
The shadow variation of scene and non-human movement avoid the background scene figure of maintenance of non-human target area moment last state
The misjudgment for leading to subsequent step is influenced with object variation in scene by shadow variation using simple Background;
Scene areas, boundary and specific objective identify in advance:
1) mark such as zone of action, relaxing area, so that whether subsequent step intelligent decision difference posture is abnormal in corresponding region;
2) toilet doorway is identified, so that whether the time of subsequent step judgement into toilet is abnormal;
3) balcony, monitoring visual angle blind zone etc. are identified, so that subsequent step judges whether entry time overlength;
4) access door mouth is identified, subsequent step is judged as normal exit.
Human testing and tracking: frames differencing subduction is done by instant background image and instant video image, obtains prospect
Image, then human body outline identification is done to foreground image by DNN neural network and is tracked;
Posture judgement:
1) support vector machine method is used, in advance using the image that simulated scenario is shot as training library, extracts Surf feature, with
BOW bag of words method trains human body attitude feature database;
2) human body attitude judgement is done to the human body image that M3 is detected, judges current human's posture;
Combination judgement: comprehensive abovementioned steps, the posture that comprehensive analysis needs caregiver different in monitoring scene is in difference
Region and retention time and human body judge whether exception according to set comprehensive in different zones disengaging and residence time, such as:
1) sitting posture on the ground is more than 1 hour or lying posture is more than 10 minutes, is considered as exception;
2) lying posture in bed or on sofa is more than 12 hours, is considered as exception;
3) it is sitting on chair more than 6 hours, is considered as exception;
4) entering toilet is more than 1 hour, is considered as exception;
Etc..
Claims (5)
1. a kind of based on computer vision to the abnormal state detection method for needing special caregiver.It is characterized by:
M1) background scene image intelligent is safeguarded: to the continuous intelligent maintenance of background scene image, determining the shadow of background scene at any time
Variation and non-human movement keep the background scene image of last state at any time;
M2) scene areas, boundary and specific objective identify in advance;
M3) human testing and tracking;
M4) posture judges: doing human body attitude judgement to the human body image that M3 is detected, judges current human's posture;
M5) combination judgement: comprehensive preceding feature, the posture that comprehensive analysis needs special caregiver different in monitoring scene is not
With region and retention time and human body is passed in and out in different zones and whether residence time, comprehensive descision are abnormal.
2. one kind according to claim 1 is based on computer vision to the abnormal state detection for needing special caregiver
Method.It is characterized in that, the scene areas, boundary and specific objective identify in advance, including the following contents:
1) zone of action, relaxing area are identified;
2) toilet doorway is identified;
3) balcony, monitoring visual angle blind zone are identified;
4) access door mouth is identified.
3. one kind according to claim 1 is based on computer vision to the abnormal state detection for needing special caregiver
Method.It is characterized in that, one of the human testing and tracking, including following methods:
1) by by the background image safeguarded immediately and instant received video image do frames differencing method or background subtraction method into
Row human testing and tracking;
2) human testing and tracking are carried out by DNN neural network;
3) human testing and tracking are carried out using GrabCut partitioning algorithm;
4) human testing and tracking are carried out using other computer visions and machine learning method.
4. one kind according to claim 1 is based on computer vision to the abnormal state detection for needing special caregiver
Method.It is characterized in that, the posture judges, training picture library is instructed using the image of simulated scenario shooting as machine learning
Practice library, judgment method includes one of following methods:
1) posture judgement is carried out by Haar feature combination Adaboost classifier;
2) SVM support vector machines combination Hog feature or surf feature or sift feature carry out posture judgement;
3) posture judgement is carried out using OpenPose;
4) posture judgement is carried out to human body attitude feature database using the method for other machine learning (including deep learning).
5. one kind according to claim 1 is based on computer vision to the abnormal state detection for needing special caregiver
Method.It is characterized in that, the combination judges, including the following contents:
1) sitting posture on the ground and lying posture time anomaly;
2) the lying posture time anomaly in bed or on sofa;
3) it is sitting in time anomaly on chair;
4) enter toilet time anomaly;
5) enter the positions such as balcony and visual angle blind zone time anomaly.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811371951.4A CN109325474A (en) | 2018-11-14 | 2018-11-14 | A kind of abnormal state detection method of couple of special caregiver of need |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811371951.4A CN109325474A (en) | 2018-11-14 | 2018-11-14 | A kind of abnormal state detection method of couple of special caregiver of need |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109325474A true CN109325474A (en) | 2019-02-12 |
Family
ID=65258547
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811371951.4A Withdrawn CN109325474A (en) | 2018-11-14 | 2018-11-14 | A kind of abnormal state detection method of couple of special caregiver of need |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109325474A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110110710A (en) * | 2019-06-03 | 2019-08-09 | 北京启瞳智能科技有限公司 | A kind of scene abnormality recognition methods, system and intelligent terminal |
CN110633652A (en) * | 2019-08-27 | 2019-12-31 | 宇龙计算机通信科技(深圳)有限公司 | Unexpected situation determination method and device and electronic equipment |
CN110992629A (en) * | 2019-12-06 | 2020-04-10 | 江西洪都航空工业集团有限责任公司 | Method for detecting static human body based on video monitoring |
CN111753648A (en) * | 2020-05-12 | 2020-10-09 | 高新兴科技集团股份有限公司 | Human body posture monitoring method, device, equipment and storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080119961A1 (en) * | 2006-11-16 | 2008-05-22 | Samsung Electronics Co., Ltd. | Methods, apparatus, and medium for estimating pose of mobile robot using particle filter |
CN101739550A (en) * | 2009-02-11 | 2010-06-16 | 北京智安邦科技有限公司 | Method and system for detecting moving objects |
CN105718857A (en) * | 2016-01-13 | 2016-06-29 | 兴唐通信科技有限公司 | Human body abnormal behavior detection method and system |
CN106022284A (en) * | 2016-05-30 | 2016-10-12 | 重庆大学 | Living-alone old person abnormal behavior detecting method based on panoramic infrared camera |
-
2018
- 2018-11-14 CN CN201811371951.4A patent/CN109325474A/en not_active Withdrawn
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080119961A1 (en) * | 2006-11-16 | 2008-05-22 | Samsung Electronics Co., Ltd. | Methods, apparatus, and medium for estimating pose of mobile robot using particle filter |
CN101739550A (en) * | 2009-02-11 | 2010-06-16 | 北京智安邦科技有限公司 | Method and system for detecting moving objects |
CN105718857A (en) * | 2016-01-13 | 2016-06-29 | 兴唐通信科技有限公司 | Human body abnormal behavior detection method and system |
CN106022284A (en) * | 2016-05-30 | 2016-10-12 | 重庆大学 | Living-alone old person abnormal behavior detecting method based on panoramic infrared camera |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110110710A (en) * | 2019-06-03 | 2019-08-09 | 北京启瞳智能科技有限公司 | A kind of scene abnormality recognition methods, system and intelligent terminal |
CN110633652A (en) * | 2019-08-27 | 2019-12-31 | 宇龙计算机通信科技(深圳)有限公司 | Unexpected situation determination method and device and electronic equipment |
CN110992629A (en) * | 2019-12-06 | 2020-04-10 | 江西洪都航空工业集团有限责任公司 | Method for detecting static human body based on video monitoring |
CN111753648A (en) * | 2020-05-12 | 2020-10-09 | 高新兴科技集团股份有限公司 | Human body posture monitoring method, device, equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109558865A (en) | A kind of abnormal state detection method to the special caregiver of need based on human body key point | |
CN109325474A (en) | A kind of abnormal state detection method of couple of special caregiver of need | |
CN112784662A (en) | Video-based fall risk evaluation system | |
Stone et al. | Fall detection in homes of older adults using the Microsoft Kinect | |
Tzeng et al. | Design of fall detection system with floor pressure and infrared image | |
Liu et al. | A fall detection system using k-nearest neighbor classifier | |
Dubois et al. | Human activities recognition with RGB-Depth camera using HMM | |
Zhang et al. | A viewpoint-independent statistical method for fall detection | |
Shoaib et al. | View-invariant fall detection for elderly in real home environment | |
CN110321767A (en) | Image acquiring apparatus and method, behavior analysis system and storage medium | |
CN108629300A (en) | A kind of fall detection method | |
US20220189264A1 (en) | Monitoring device, suspicious object detecting method, and recording medium | |
CN113257440A (en) | ICU intelligent nursing system based on patient video identification | |
CN112489368A (en) | Intelligent falling identification and detection alarm method and system | |
WO2017098265A1 (en) | Method and apparatus for monitoring | |
Debard et al. | Camera based fall detection using multiple features validated with real life video | |
SG188111A1 (en) | Condition detection methods and condition detection devices | |
Ezzahout et al. | Conception and development of a video surveillance system for detecting, tracking and profile analysis of a person | |
WO2019003859A1 (en) | Monitoring system, control method therefor, and program | |
Albawendi et al. | Video based fall detection using features of motion, shape and histogram | |
CN107958572A (en) | A kind of baby monitoring systems | |
Stone et al. | Silhouette classification using pixel and voxel features for improved elder monitoring in dynamic environments | |
CN115116133A (en) | Abnormal behavior detection system and method for monitoring solitary old people | |
Khan et al. | Video analytic for fall detection from shape features and motion gradients | |
Tasoulis et al. | Statistical data mining of streaming motion data for fall detection in assistive environments |
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 | ||
WW01 | Invention patent application withdrawn after publication | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190212 |