CN106295524A - A kind of human motion recognition method of view-based access control model word bag - Google Patents
A kind of human motion recognition method of view-based access control model word bag Download PDFInfo
- Publication number
- CN106295524A CN106295524A CN201610615930.7A CN201610615930A CN106295524A CN 106295524 A CN106295524 A CN 106295524A CN 201610615930 A CN201610615930 A CN 201610615930A CN 106295524 A CN106295524 A CN 106295524A
- Authority
- CN
- China
- Prior art keywords
- word bag
- image
- model
- access control
- based access
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- 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
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Social Psychology (AREA)
- Psychiatry (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention belongs to computer vision field, particularly relate to human action Computer Vision therein and identification.The invention discloses the human motion recognition method of a kind of view-based access control model word bag, it is characterised in that: (1) carries out pretreatment to raw video image, detects moving target.(2) moving target is carried out image characteristics extraction and classification model construction.(3) the standard operation image library utilizing view-based access control model word bag to set up carries out Classification and Identification to human action, and is identified Deviant Behavior according to normal behaviour model and Deviant Behavior model and judges.The human motion recognition method of the view-based access control model word bag of the present invention utilizes visual dictionary to classify in advance, for different samples, there is preferable discrimination, and it is fast to calculate speed, can quickly carry out classification and the identification of substantial amounts of human action, can be used for the human action identification system of complexity.
Description
1. technical field
The invention belongs to computer vision field, particularly relate to human action Computer Vision therein and identification.
2. background technology
Human action identification refers to the data that computer gathers based on video image acquisition equipment, analyzes and identifies human body attitude
Process with motor pattern.It relates to the subjects such as machine learning, artificial intelligence, image procossing, is computer vision neck
The research direction that territory is important, the most also has extensive and important application.Bank safety monitoring system, old man nurses
System, the new demand new opplication such as Smart Home constantly produces, and human action identification technology is gradually socially reintegrated each angle of life
Fall, become the study hotspot in action recognition field.
Monitoring system is widely used in each unit and public place in recent years, but most monitoring system is only
The real time imaging collected by image capture device is transferred to backstage by network, could be found by naked eyes when event occurs,
The most also the function that video is looked into it is provided solely for.This traditional method cause its in real time monitoring difficulty increasing, and
Workload increases day by day.How to solve this problem, introduce artificial intelligence technology and Computer Vision and identification technology, replace
Manually completing above-mentioned work is following development trend.The present invention proposes the Research Thinking of oneself, a kind of new to providing
Method.Especially propose the human motion recognition method of view-based access control model word bag, extract human action according to real-time monitor video
Behavior characteristics, in analysis monitoring region, the feature in the attitude of human body and motor pattern, with model database is compared, and identifies
Deviant Behavior also sends warning.This method can reduce monitoring system to artificial dependency so that monitoring system becomes more
Intelligence is with efficient.
3. summary of the invention
The present invention is the human action automatic identifying method of a kind of view-based access control model word bag designed according to above-mentioned thinking.
The technical scheme is that the human motion recognition method that a kind of view-based access control model word bag is provided, it is characterised in that:
It includes following step:
(1) raw video image is carried out pretreatment, detect moving target.
(2) moving target is carried out image characteristics extraction and classification model construction.
(3) utilize standard operation image library that human action carries out Classification and Identification, and according to normal behaviour model and exception
Deviant Behavior is identified and judges by behavior model.
The most further comprising the steps of in above-mentioned steps (1):
A () gathers human body standard operation video, including normal behaviour and Deviant Behavior, normal behaviour includes standing, walks,
Jog, sit down, Deviant Behavior includes quickly running, embrace beat, clustering, fall down.
B () extracts human action feature, build visual word bag model, Criterion motion images storehouse.
The present invention proposes the human action feature extracting method of view-based access control model word bag model, including following components:
(1) human action feature analysis
The image table that video extraction goes out reveals great uncertainty, either illumination variation, and angle changes, or human body
Movement range changes, and is all difficult to capture stable feature.Extraction based on global image feature and calculating are the most inadvisable, we
Method carries out partial analysis, research human action feature and key point change to moving target region in image, it is ensured that extract
The effectiveness of feature.
(2) image characteristics extraction is carried out according to visual word bag model
Image can analogize to document, the i.e. set of several " visual vocabularies ", and the vocabulary in image can be defined as one
The characteristic vector of individual image block, visual vocabulary not order each other, visual word bag model is i.e. " all image blocks in image
The rectangular histogram that obtains of characteristic vector ".The present invention utilizes SIFT conversion to extract the eigenmatrix of visual word bag.Calculate key point
The gradient of each pixel in the window of 16*16 around, and use Gauss decreasing function to reduce deep weight.This
Sample just can form description of a 4*4*8=128 dimension to each feature.
(3) generate characteristics of image pyramid and carry out standard operation image classification
The eigenmatrix obtained previous step with K-means algorithm clusters, and is generated by image word bag feature clustering
Visual dictionary, uses visual dictionary to generate feature pyramid.Use the middle lexical representation image of dictionary sheet, generate rectangular histogram.
The human motion recognition method of the view-based access control model word bag of the present invention utilizes visual dictionary to classify in advance, for not
Same sample has preferable discrimination, and it is fast to calculate speed, can quickly carry out classification and the knowledge of substantial amounts of human action
, not can be used for the human action identification system of complexity.
4. accompanying drawing explanation
Fig. 1 is the structured flowchart of the system of the application present invention.
5. detailed description of the invention
Below the detailed description of the invention of the present invention is described in further detail.
As it is shown in figure 1, system corresponding to the human motion recognition method of a kind of view-based access control model word bag of the present invention includes: figure
As pretreatment, image characteristics extraction, visual word bag model builds, the foundation of standard operation image library, the classification of motion and identification.
Image semantic classification is that video file is carried out sub-frame processing, retains the image containing key operations, extracts grader
Training and the final important indicator identifying judgement.
Image characteristics extraction uses SIFT conversion to convert pretreated image, calculates the feature on each point
Vector, in conjunction with the multiple key points in image, obtains the local invariant feature of image.
It is to utilize the rectangular histogram of the characteristic vector of all image blocks in image to obtain that visual word bag model builds.
It is according to the Euclidean distance of eigenmatrix between different images that standard operation image library is set up, the most close for distance
Image is classified as a class, and the action of identification has a few class to be generated as several cluster centre, obtains several standard operation model.Each action
In model, comprise the angled moving image of this classification of motion.
The classification of motion and identification include detecting moving target, model construction of SVM classification and action recognition.Utilize and extract
The method of HOG feature and calculus of finite differences obtain in video whether presence of people determine human body position, it is thus achieved that comprise target person
The consecutive image of body.Support vector machine is by the human body attitude Feature Mapping that extracts to higher dimensional space, and support vector machine belongs to
Linear classifier, feature is can to minimize experience error simultaneously and maximize Geometry edge district.Final recognition result be with
On detect key operations sequence and standard operation image library on the basis of, based on KNN sorting algorithm realize.
Above example is only the present invention a kind of embodiment therein, and it describes more concrete, but can not therefore and
It is interpreted as the restriction to the scope of the claims of the present invention.For those skilled in the art, without departing from the inventive concept of the premise,
Can also make some deformation and improvement, these broadly fall into protection scope of the present invention.Therefore, the protection domain of patent of the present invention
Should be as the criterion with claims.
Claims (3)
1. the technical scheme is that and the human motion recognition method of a kind of view-based access control model word bag be provided, it is characterised in that: its
Including following step:
(1) raw video image is carried out pretreatment, detect moving target;
(2) moving target is carried out image characteristics extraction and classification model construction;
(3) utilize standard operation image library that human action carries out Classification and Identification, and according to normal behaviour model and Deviant Behavior
Deviant Behavior is identified and judges by model.
The human motion recognition method of view-based access control model word bag the most according to claim 1, it is characterised in that: in above-mentioned steps
(1) the most further comprising the steps of:
A () gathers human body standard operation video, including normal behaviour and Deviant Behavior, normal behaviour includes standing, walking, slowly
Run, sit down, Deviant Behavior includes quickly running, embrace beat, clustering, fall down;
B () extracts human action feature, build visual word bag model, Criterion motion images storehouse.
The human motion recognition method of view-based access control model word bag the most according to claim 1, the present invention proposes view-based access control model
The human action feature extracting method of word bag model, including following components:
(1) human action feature analysis: moving target region in image carries out partial analysis, studies human action feature
Change with key point, it is ensured that extract the effectiveness of feature;
(2) image characteristics extraction is carried out according to visual word bag model: utilize SIFT conversion to extract the eigenmatrix of visual word bag;
(3) generate characteristics of image pyramid and carry out standard operation image classification: use the spy that previous step is obtained by K-means algorithm
Levy matrix to cluster, image word bag feature clustering is generated visual dictionary, use visual dictionary to generate characteristics of image pyramid.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610615930.7A CN106295524A (en) | 2016-08-01 | 2016-08-01 | A kind of human motion recognition method of view-based access control model word bag |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610615930.7A CN106295524A (en) | 2016-08-01 | 2016-08-01 | A kind of human motion recognition method of view-based access control model word bag |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106295524A true CN106295524A (en) | 2017-01-04 |
Family
ID=57663429
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610615930.7A Pending CN106295524A (en) | 2016-08-01 | 2016-08-01 | A kind of human motion recognition method of view-based access control model word bag |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106295524A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107203745A (en) * | 2017-05-11 | 2017-09-26 | 天津大学 | A kind of across visual angle action identification method based on cross-domain study |
CN107272468A (en) * | 2017-08-01 | 2017-10-20 | 刘太龙 | Electronic security(ELSEC) based on communication ensures platform |
CN107688790A (en) * | 2017-09-01 | 2018-02-13 | 东软集团股份有限公司 | Human bodys' response method, apparatus, storage medium and electronic equipment |
CN107786848A (en) * | 2017-10-30 | 2018-03-09 | 周燕红 | The method, apparatus of moving object detection and action recognition, terminal and storage medium |
CN109583307A (en) * | 2018-10-31 | 2019-04-05 | 东华大学 | A kind of Cashmere and Woolens fiber recognition method based on local feature Yu word packet model |
CN109871771A (en) * | 2019-01-21 | 2019-06-11 | 北京轻舟科技有限公司 | A kind of method and system of the automatic detection human body based on single-view videos |
CN111860231A (en) * | 2020-07-03 | 2020-10-30 | 厦门欧准卫浴有限公司 | Universal water module based on household occasions |
CN111985413A (en) * | 2020-08-22 | 2020-11-24 | 深圳市信诺兴技术有限公司 | Intelligent building monitoring terminal, monitoring system and monitoring method |
CN114120770A (en) * | 2021-03-24 | 2022-03-01 | 张银合 | Barrier-free communication method for hearing-impaired people |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110295688A1 (en) * | 2010-05-28 | 2011-12-01 | Microsoft Corporation | Defining user intent |
CN104268520A (en) * | 2014-09-22 | 2015-01-07 | 天津理工大学 | Human motion recognition method based on depth movement trail |
CN104616006A (en) * | 2015-03-11 | 2015-05-13 | 湖南智慧平安科技有限公司 | Surveillance video oriented bearded face detection method |
CN104616316A (en) * | 2014-05-23 | 2015-05-13 | 苏州大学 | Method for recognizing human behavior based on threshold matrix and characteristics-fused visual word |
CN104866841A (en) * | 2015-06-05 | 2015-08-26 | 中国人民解放军国防科学技术大学 | Human body object running behavior detection method |
CN104978561A (en) * | 2015-03-25 | 2015-10-14 | 浙江理工大学 | Gradient and light stream characteristics-fused video motion behavior identification method |
-
2016
- 2016-08-01 CN CN201610615930.7A patent/CN106295524A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110295688A1 (en) * | 2010-05-28 | 2011-12-01 | Microsoft Corporation | Defining user intent |
CN104616316A (en) * | 2014-05-23 | 2015-05-13 | 苏州大学 | Method for recognizing human behavior based on threshold matrix and characteristics-fused visual word |
CN104268520A (en) * | 2014-09-22 | 2015-01-07 | 天津理工大学 | Human motion recognition method based on depth movement trail |
CN104616006A (en) * | 2015-03-11 | 2015-05-13 | 湖南智慧平安科技有限公司 | Surveillance video oriented bearded face detection method |
CN104978561A (en) * | 2015-03-25 | 2015-10-14 | 浙江理工大学 | Gradient and light stream characteristics-fused video motion behavior identification method |
CN104866841A (en) * | 2015-06-05 | 2015-08-26 | 中国人民解放军国防科学技术大学 | Human body object running behavior detection method |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107203745A (en) * | 2017-05-11 | 2017-09-26 | 天津大学 | A kind of across visual angle action identification method based on cross-domain study |
CN107272468A (en) * | 2017-08-01 | 2017-10-20 | 刘太龙 | Electronic security(ELSEC) based on communication ensures platform |
CN107688790A (en) * | 2017-09-01 | 2018-02-13 | 东软集团股份有限公司 | Human bodys' response method, apparatus, storage medium and electronic equipment |
CN107688790B (en) * | 2017-09-01 | 2020-09-04 | 东软集团股份有限公司 | Human behavior recognition method and device, storage medium and electronic equipment |
CN107786848A (en) * | 2017-10-30 | 2018-03-09 | 周燕红 | The method, apparatus of moving object detection and action recognition, terminal and storage medium |
CN109583307A (en) * | 2018-10-31 | 2019-04-05 | 东华大学 | A kind of Cashmere and Woolens fiber recognition method based on local feature Yu word packet model |
CN109871771A (en) * | 2019-01-21 | 2019-06-11 | 北京轻舟科技有限公司 | A kind of method and system of the automatic detection human body based on single-view videos |
CN111860231A (en) * | 2020-07-03 | 2020-10-30 | 厦门欧准卫浴有限公司 | Universal water module based on household occasions |
CN111985413A (en) * | 2020-08-22 | 2020-11-24 | 深圳市信诺兴技术有限公司 | Intelligent building monitoring terminal, monitoring system and monitoring method |
CN114120770A (en) * | 2021-03-24 | 2022-03-01 | 张银合 | Barrier-free communication method for hearing-impaired people |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106295524A (en) | A kind of human motion recognition method of view-based access control model word bag | |
Singha et al. | Recognition of Indian sign language in live video | |
CN104978550A (en) | Face recognition method and system based on large-scale face database | |
Alawwad et al. | Arabic sign language recognition using Faster R-CNN | |
JP2012053756A (en) | Image processor and image processing method | |
CN105046195A (en) | Human behavior identification method based on asymmetric generalized Gaussian distribution model (AGGD) | |
CN111046732A (en) | Pedestrian re-identification method based on multi-granularity semantic analysis and storage medium | |
Wang et al. | Automatic recognition of facial movement for paralyzed face | |
CN106778501A (en) | Video human face ONLINE RECOGNITION method based on compression tracking with IHDR incremental learnings | |
Mohane et al. | Object recognition for blind people using portable camera | |
CN103955671A (en) | Human behavior recognition method based on rapid discriminant common vector algorithm | |
CN105320948A (en) | Image based gender identification method, apparatus and system | |
Sheng et al. | Siamese networks for weakly supervised human activity recognition | |
CN110458022A (en) | It is a kind of based on domain adapt to can autonomous learning object detection method | |
Vo et al. | Deep learning for Vietnamese Sign Language recognition in video sequence | |
Travieso et al. | Bimodal biometric verification based on face and lips | |
CN108596057B (en) | Information security management system based on face recognition | |
Firdaus et al. | Recognizing fingerspelling in sibi (sistem isyarat bahasa indonesia) using openpose and elliptical fourier descriptor | |
Batool et al. | Fundamental recognition of ADL assessments using machine learning engineering | |
Hou et al. | Visual detection of Parkinson’s disease via facial features recognition | |
Viswanathan et al. | Recognition of hand gestures of English alphabets using HOG method | |
Nabi et al. | Abnormal event recognition in crowd environments | |
Sani et al. | Evaluation of face recognition system using Support Vector Machine | |
Vo et al. | Automatic hand gesture segmentation for recognition of Vietnamese sign language | |
Gilorkar et al. | A review on feature extraction for Indian and American sign language |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170104 |