CN101286238A - Bionic intelligent monitoring system - Google Patents
Bionic intelligent monitoring system Download PDFInfo
- Publication number
- CN101286238A CN101286238A CNA2008100993771A CN200810099377A CN101286238A CN 101286238 A CN101286238 A CN 101286238A CN A2008100993771 A CNA2008100993771 A CN A2008100993771A CN 200810099377 A CN200810099377 A CN 200810099377A CN 101286238 A CN101286238 A CN 101286238A
- Authority
- CN
- China
- Prior art keywords
- bionic
- monitoring system
- scene
- information
- intelligent monitoring
- 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
Images
Landscapes
- Alarm Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention starts from the bionic idea and comprehensively applies digital image processing, computer vision, biometric identification and other methods to realize the intellectualization of a monitoring system. The realization mode is characterized in that: video data with variable scale is obtained by constructing a bionic video monitoring network with the combination of a panoramic camera and a detailed camera, then the extraction, detection and tracking of character features of a panoramic monitoring field, as well as the human face identification of the detailed data are carried out on the basis of the high-dimensional bionic information concept, thus achieving the purpose of understanding the monitoring field and finally turning into the corresponding response mechanism by the autonomous decision-making link of the monitoring system. The monitoring system of the invention has part of the human brain decision-making functions, the security function is introduced into the monitoring system to distinguish the specific people groups, the system can autonomously understand the contents of the field and take the corresponding measure in the absence of human assistance, improve the safety level and the response speed of the system and have broad application prospect under the background of growing importance of the public safety.
Description
Technical field
The present invention relates to bionical and area of pattern recognition, relate in particular to the bionic intelligent monitoring system that a kind of supervisory system has part human brain decision making function.
Background technology
At present, supervisory system has been widely used in all trades and professions, no matter be colliery, oil field, the hydrology, or government, bank, public security, traffic, monitoring all becomes the important information management supplementary means of these industries.In these are used, the information data that front monitoring front-end is gathered can be multi-form with video, voice, picture, literal etc. by wired or wireless network, send to the monitor terminal that is positioned at the control and management center, wherein, especially with video monitoring system, comprise that the application mode of simulation monitoring and digital supervision is more popular.
But there are some intrinsic shortcomings in traditional video monitoring system, such as the situation of user for surveillance monitor point, must have staff on duty before monitor.In the zone of bank cashier machine, safe deposit box storehouse, the warning of airport security place equal altitudes, still need under manpower is auxiliary, carry out policy-making operation to the information result of supervisory system collection, processing gained especially at the control and management center.In addition, in family and company's sight, for working clan who has child or patient to look after in the leader of enterprise who goes on business or go to work, the family etc., they need understand the situation in unit and the family anywhere or anytime, and make suitable response rapidly, this also be traditional supervisory system can't realize.Moreover in busy day by day society, monitoring scene becomes increasingly complex, and the single manpower that leans on can not be dealt with all situations, and traditional manpower judgement has careless omission unavoidably, and monitoring efficiency is lower.
Therefore, for the supervisory system research staff that security level required day by day improves, how to make supervisory system have the function of part human brain decision-making, specific crowd is distinguished, system can independently understand scene content and take appropriate measures under the situation that does not have manpower to assist, improving security of system grade and response speed problem is the important directions of researching and developing at present.
Summary of the invention
Defective at above-mentioned prior art existence, the objective of the invention is to: a kind of bionic intelligent monitoring system is provided, to solve the problem of supervisory system intellectuality, high efficiency, need not under the auxiliary situation of manpower, the security of system grade promotes relative slow problem with response speed.
For reaching above-mentioned purpose, the present invention proposes following technical scheme:
Bionic intelligent monitoring system, it is characterized by: based on the starting point of higher-dimension bio-information, integrated use means such as computer vision, Digital Image Processing realize the bionic intelligenceization of supervisory system, by object of observation in higher dimensional space abstract mapping and different mode under the self-adaptation of geometrical body cover the functional bionic of having realized people's imaginal thinking capability; Its specific implementation method is at first by the multi-scale information in the bionic video monitor network tap/dip deep into scene, and then scene information classified, analyzes and understand, the final Decision Fusion method of learning based on the higher-dimension bio-information that adopts is discerned abnormal conditions, and changes corresponding response mechanism over to.
Further, implementation step is:
(1) changes the mechanism with the visual field based on the vision fusion of biology, take a plurality of panoramas, details video camera to make up the bionic video monitor network, by the coordinated scheduling between the video camera with have complementary functions, realize that the seamless covering of scene and the multi-scale information of object of observation obtain;
(2) learn " homology continuity " principle based on the higher-dimension bio-information and carry out data " roughing " processing, by making up reasonable mapping space, extract the target signature vector of low redundance, and the intrinsic contact in this mapping space when guaranteeing same target different mode, thereby be implemented in robust ground fast detecting in the large scale scene picture, cut apart and tracking target;
(3) learn the data that the details video camera is obtained based on the higher-dimension bio-information and carry out " deep processing " processing, by understanding to the large scale scene, scheduling details video camera emphasis track up people face information, the self-adaptation of utilization geometrical body of face characteristic vector in higher dimensional space covers carries out recognition of face, to distinguish specific crowd;
(4) based on understandability to scene, and the result of face characteristic identification, the decision-making link of establishment supervisory system by the processing to scene information, is analyzed specific abnormal conditions, and is changed corresponding responder over to, improves monitoring efficiency;
(5) video data acquiring, transmission course are carried out " singularity " and encrypt, the ability of the anti-invasion of raising system.
Bionic intelligent monitoring system provided by the invention, its beneficial effect that has is:
This bionic intelligent monitoring system has the function of part human brain decision-making, the safety check function is incorporated into supervisory system, specific crowd is distinguished, system can independently understand scene content and take appropriate measures under the situation that does not have manpower to assist, improve security of system grade, response speed and monitoring efficiency, under the background that public safety becomes more and more important, have wide practical use.
Description of drawings
Fig. 1 is the principle schematic of bionic intelligent monitoring system construction of the present invention;
Fig. 2 is the structure of bionic intelligent monitoring system one embodiment of the present invention and the schematic flow sheet of function.
Fig. 3 is data storage and the index synoptic diagram of bionic intelligent monitoring system one embodiment of the present invention.
Embodiment
For making the above-mentioned purpose of the present invention, feature and advantage can more obvious easy understanding, below the spy in conjunction with the embodiments and relevant drawings, be described in detail below:
As shown in Figure 1, be the principle schematic of bionic intelligent monitoring system of the present invention.Along with supervisory system is applied to more and more widely field, response accuracy and the intelligent requirements to supervisory system also day by day improves simultaneously.For this reason, the scientific research personnel of the industry has the exploitation direction seeking breakthrough of part human brain decision making function with great concentration towards supervisory system, and has obtained certain progress.As shown in Figure 1: based on the starting point of higher-dimension bio-information, integrated use means such as Digital Image Processing, computer vision realize the bionic intelligenceization of supervisory system, by object of observation in higher dimensional space abstract mapping and different mode under the self-adaptation of geometrical body cover the functional bionic of having realized people's imaginal thinking capability; Its specific implementation principle is at first carrying out data acquisition by the bionic video monitor network, multi-scale information in the tap/dip deep into scene, and then the scene information data are classified, analyze and understand, the final Decision Fusion method of learning based on the higher-dimension bio-information that adopts is discerned abnormal conditions, and changes corresponding response mechanism over to.
Wherein, data analysis (scene understanding) is included four big major function steps, be respectively: cut apart interested part in the image data; The tracking target motion; Face characteristic extracts; Recognition of face.
For further specifying the exploitativeness of the principle of the invention,, elaborate the function of its job step and realization below in conjunction with example structure shown in the accompanying drawing 2 and functional sequence synoptic diagram.
Step 1: the vision fusion based on biology is changed the mechanism with the visual field, takes the autonomous variable panorama of a plurality of yardsticks, the bionic video monitoring data collection network that the details video camera is formed.By the coordinated scheduling between the video camera with have complementary functions, realize that the seamless covering of scene and the multi-scale information of object of observation obtain.
For the single yardstick of supervisory system is observed, adopt high-resolution data mode collection can obtain a large amount of garbages, increase the burden that late time data is handled, be difficult to guarantee the requirement of real-time; And adopt the mode of low resolution to gather contextual data, and can not obtain the detailed information of enough interesting targets, influence the system works reliability.The present invention is based on the bionical angle of principle considers, observe the mode of things from the imitation people, realize the fusion of multi-scale information: after wide scope visual field information is carried out preparatory processing, dispatch details video camera primary part observation area-of-interest in the mode that vision merges.This monitoring scene data acquisition modes has been taken into account the requirement of supervisory system to scene range and precision, can solve the resource consumption and the contradiction of observing sharpness of field of information processing, realizes monitoring range and the unification of monitoring precision.
Step 2: the characteristic that does not have sudden change based on the scene content of monitor video in time, utilize the higher-dimension bio-information to learn " homology continuity " principle and carry out data " roughing " processing, by making up reasonable mapping space, extract the target signature vector of low redundance, and the intrinsic contact in this mapping space when guaranteeing same target different mode, thereby be implemented in robust ground fast detecting in the large scale scene picture, cut apart and tracking target.
Wherein so-called " homology continuity " is meant: come from the target in the same monitoring scene, its sample should be continuous in higher dimensional space.Be to undergo mutation between the sample, all have a continually varying curve between any two sample points, make one of them sample point carry out the transition to the another one sample point smoothly.
Step 3: learn the data that the details video camera is obtained based on the higher-dimension bio-information and carry out " deep processing " processing.By the understanding to the large scale scene, scheduling details video camera emphasis track up people face information utilizes the self-adaptation covering of the geometrical body of face characteristic vector in higher dimensional space to carry out recognition of face, to distinguish specific crowd.
Face characteristic has singularity, and can carry out identity authentication after the non-contacting vision sensor collection, has high using value in monitoring scene.People's face sample is treated observation station through shining upon of becoming in the higher dimensional space after the feature extraction.According to the continuity of similar sample point in higher dimensional space, training sample set should be covered by limited, the continuous zone of a slice, treat that by analysis the relation of observation station and training sample set overlay area can obtain the face feature sample of this portrait and the matching degree between training sample set, can realize the identity authentication to improper crowd in the monitoring scene.
Through obtaining the low redundance eigenvector of target after the suitable mapping, the sample of same target has cluster character in higher dimensional space in target signature.By adaptive learning to known sample, select one or more suitable occluding surfaces, the continuous complicated geometirc physique that forms a higher dimensional space rationally covers training sample.And this complicated geometirc physique need carry out self-adaptation according to the distribution of training sample and chooses, and covers training sample set with reasonable manner.The final face characteristic self-adaptation of learning based on the higher-dimension bio-information that realizes is identified.This function is incorporated into supervisory system with the safety check function, and specific crowd is distinguished, and system has taken into account and guarded against improper personnel and the privacy of protecting normal personnel not having can to improve the warning coefficient to it under the human-aided situation.
Step 4: based on understandability to scene, and the result of face characteristic identification, the decision-making link of establishment supervisory system.By processing to scene information, analyze specific abnormal conditions, and change corresponding responder over to, improve monitoring efficiency.
Data processing by the higher-dimension bio-information is learned can detect the personage in the scene, follow the tracks of and discern.Need not manpower auxiliary time, whether the personage that supervisory system can identify rapidly in the scene is internal staff or some population at risk (for example category-A wanted criminal), thereby makes a strategic decision corresponding response mechanism.To the internal staff, protect its privacy, supervisory system does not process; For non-internal staff, improve the security of system grade, video tracking, video recording are carried out in its behavior; For the population at risk, system changes special alert mechanism over to, and selective analysis is carried out in its behavior.
In addition; for further protection collection, transmission safety of data and accuracy; by making up " the only contact " between rig camera and the supervisory system, guarantee that system only accepts to formulate the data of video capture device, resist data obtain with transmission course in invasion.
Moreover, because intelligent monitor system of the present invention has scene and understands function, therefore realized the case index of abnormal conditions in the video recording evidence obtaining, replace traditional time index (shown in Fig. 3,4).
Generally, the video recording of supervisory system evidence obtaining function all is the index condition with time, consults the video recording of corresponding period with the form of manpower work and collect evidence after the time period of determining incident.With time is the condition index, needs the work of recording function all the period of time, takies ample resources, and need call great amount of manpower during evidence obtaining and consult.Bionic intelligent monitoring system has the function that scene content is understood, and can analyze the abnormal conditions in the scene, the face characteristic that identification occurs.Therefore, the appearance that can utilize personage in the scene is as the switch of control video recording system work, and with the personage as the index condition of consulting evidence obtaining.Saved a large amount of storage resources, and realized that video consults the facilitation of evidence obtaining.
In sum, bionic intelligent monitoring system of the present invention has the function of part human brain decision-making, specific crowd is distinguished, system can independently understand scene content and take appropriate measures under the situation that does not have manpower to assist, improve security of system grade and response speed, under the background that public safety becomes more and more important, have wide practical use.
Below only be to the detailed description of the present invention's design, this case protection domain do not constituted any limitation that all employing equivalents or equivalence are replaced and the technical scheme of formation, all drop within the rights protection scope of the present invention.
Claims (8)
1. bionic intelligent monitoring system, starting point based on higher-dimension bio-information, the integrated use Digital Image Processing, means such as computer vision realize the bionic intelligenceization of supervisory system, by object of observation in higher dimensional space abstract mapping and different mode under the self-adaptation of geometrical body cover, realization is to the functional bionic of people's imaginal thinking capability, it is characterized in that: at first by the multi-scale information in the bionic video monitor network tap/dip deep into scene, and then scene information classified, analyze and understanding, the final Decision Fusion method of learning based on the higher-dimension bio-information that adopts is discerned abnormal conditions, and changes corresponding response mechanism over to.
2. bionic intelligent monitoring system according to claim 1 is characterized in that: described intelligent monitor system, the step that its function realizes be---
(1) changes the mechanism with the visual field based on the vision fusion of biology, take a plurality of panoramas, details video camera to make up the bionic video monitor network, by the coordinated scheduling between the video camera with have complementary functions, realize that the seamless covering of scene and the multi-scale information of object of observation obtain;
(2) learn " homology continuity " principle based on the higher-dimension bio-information and carry out data " roughing " processing, by making up reasonable mapping space, extract the target signature vector of low redundance, and the intrinsic contact in this mapping space when guaranteeing same target different mode, thereby be implemented in robust ground fast detecting in the large scale scene picture, cut apart and tracking target;
(3) learn the data that the details video camera is obtained based on the higher-dimension bio-information and carry out " deep processing " processing, by understanding to the large scale scene, scheduling details video camera emphasis track up people face information, the self-adaptation of utilization geometrical body of face characteristic vector in higher dimensional space covers carries out recognition of face, to distinguish specific crowd;
(4) based on understandability to scene, and the result of face characteristic identification, the decision-making link of establishment supervisory system by the processing to scene information, is analyzed specific abnormal conditions, and is changed corresponding responder over to, improves monitoring efficiency;
(5) video data acquiring, transmission course are carried out " singularity " and encrypt, the ability of the anti-invasion of raising system.
3. bionic intelligent monitoring system according to claim 1 is characterized in that: make up and form the panorama that is independently changed by yardstick, the bionic video monitoring data collection network that the details video camera is formed.
4. bionic intelligent monitoring system according to claim 1 is characterized in that: supervisory system has certain understandability to scene, learns angle from the higher-dimension bio-information, realizes the functional bionic of thinking in images.
5. bionic intelligent monitoring system according to claim 1 is characterized in that: learn angle from the higher-dimension bio-information, utilize and based on the face identification method of higher dimensional space geometrical body self-adaptation covering theory personage's identity in the monitoring scene is differentiated.
6. bionic intelligent monitoring system according to claim 1 is characterized in that: described bionic intelligent monitoring system has the function of making decisions on one's own, and can change corresponding response link under the auxiliary situation of no manpower.
7. bionic intelligent monitoring system according to claim 1 is characterized in that: described bionic intelligent monitoring system has scene and understands function, and the case index of realizing abnormal conditions in the video recording evidence obtaining is to the replacement of time index in the past.
8. bionic intelligent monitoring system according to claim 1, it is characterized in that: by making up " the only contact " between rig camera and the supervisory system, guarantee that system only accepts to formulate the data of video capture device, resist data obtain with transmission course in invasion.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNA2008100993771A CN101286238A (en) | 2008-05-06 | 2008-05-06 | Bionic intelligent monitoring system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNA2008100993771A CN101286238A (en) | 2008-05-06 | 2008-05-06 | Bionic intelligent monitoring system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN101286238A true CN101286238A (en) | 2008-10-15 |
Family
ID=40058428
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNA2008100993771A Pending CN101286238A (en) | 2008-05-06 | 2008-05-06 | Bionic intelligent monitoring system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101286238A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102223520A (en) * | 2011-04-15 | 2011-10-19 | 北京易子微科技有限公司 | Intelligent face recognition video monitoring system and implementation method thereof |
CN103714250A (en) * | 2013-12-25 | 2014-04-09 | 广东电网公司电力科学研究院 | Fast collision detecting algorithm method for high voltage warning region of transformer substation |
CN104092972A (en) * | 2014-07-15 | 2014-10-08 | 北京小鱼儿科技有限公司 | Communication terminal and tool installed on mobile terminal |
CN105007395A (en) * | 2015-07-22 | 2015-10-28 | 深圳市万姓宗祠网络科技股份有限公司 | Privacy processing method for continuously recording video |
CN109978358A (en) * | 2019-03-18 | 2019-07-05 | 中国科学院自动化研究所 | Marketing risk point detecting method, system, device based on semi-supervised learning |
-
2008
- 2008-05-06 CN CNA2008100993771A patent/CN101286238A/en active Pending
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102223520A (en) * | 2011-04-15 | 2011-10-19 | 北京易子微科技有限公司 | Intelligent face recognition video monitoring system and implementation method thereof |
CN103714250A (en) * | 2013-12-25 | 2014-04-09 | 广东电网公司电力科学研究院 | Fast collision detecting algorithm method for high voltage warning region of transformer substation |
CN103714250B (en) * | 2013-12-25 | 2017-04-12 | 广东电网公司电力科学研究院 | Fast collision detecting method for high voltage warning region of transformer substation |
CN104092972A (en) * | 2014-07-15 | 2014-10-08 | 北京小鱼儿科技有限公司 | Communication terminal and tool installed on mobile terminal |
CN105007395A (en) * | 2015-07-22 | 2015-10-28 | 深圳市万姓宗祠网络科技股份有限公司 | Privacy processing method for continuously recording video |
CN105007395B (en) * | 2015-07-22 | 2018-02-16 | 深圳市万姓宗祠网络科技股份有限公司 | A kind of continuous record video, the privacy processing method of image |
CN109978358A (en) * | 2019-03-18 | 2019-07-05 | 中国科学院自动化研究所 | Marketing risk point detecting method, system, device based on semi-supervised learning |
CN109978358B (en) * | 2019-03-18 | 2021-08-13 | 中国科学院自动化研究所 | Sale risk point detection system and device based on semi-supervised learning |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111898514B (en) | Multi-target visual supervision method based on target detection and action recognition | |
CN111967393B (en) | Safety helmet wearing detection method based on improved YOLOv4 | |
CN105809679B (en) | Mountain railway side slope rockfall detection method based on visual analysis | |
CN102521565B (en) | Garment identification method and system for low-resolution video | |
CN102629320B (en) | Ordinal measurement statistical description face recognition method based on feature level | |
CN105426820A (en) | Multi-person abnormal behavior detection method based on security monitoring video data | |
CN104378582A (en) | Intelligent video analysis system and method based on PTZ video camera cruising | |
CN102254394A (en) | Antitheft monitoring method for poles and towers in power transmission line based on video difference analysis | |
CN104239851A (en) | Intelligent cell inspection system based on behavior analysis and control method thereof | |
CN103824070A (en) | Rapid pedestrian detection method based on computer vision | |
CN105590099A (en) | Multi-user behavior identification method based on improved convolutional neural network | |
CN106951838B (en) | A kind of overboard alarm system and its method based on image water wave | |
CN104504377A (en) | Bus passenger crowding degree identification system and method | |
CN104732236B (en) | A kind of crowd's abnormal behaviour intelligent detecting method based on layered shaping | |
CN108647619A (en) | The detection method and device that safety cap is worn in a kind of video based on deep learning | |
CN101286238A (en) | Bionic intelligent monitoring system | |
CN109948455A (en) | One kind leaving object detecting method and device | |
Ozcelik et al. | A vision based traffic light detection and recognition approach for intelligent vehicles | |
CN114998830A (en) | Wearing detection method and system for safety helmet of transformer substation personnel | |
Wang et al. | A safety helmet and protective clothing detection method based on improved-yolo v 3 | |
Miao et al. | Abnormal Behavior Learning Based on Edge Computing toward a Crowd Monitoring System | |
Sonkar et al. | Crowd Abnormal behaviour detection using deep learning | |
Xu et al. | [Retracted] Multiview Fusion 3D Target Information Perception Model in Nighttime Unmanned Intelligent Vehicles | |
CN116030405A (en) | Pedestrian target safety protection tool wearing monitoring method based on artificial intelligence | |
Xu et al. | Crowd density estimation based on improved Harris & OPTICS Algorithm |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C12 | Rejection of a patent application after its publication | ||
RJ01 | Rejection of invention patent application after publication |
Open date: 20081015 |