CN101635834A - Automatic tracing identification system for artificial neural control - Google Patents
Automatic tracing identification system for artificial neural control Download PDFInfo
- Publication number
- CN101635834A CN101635834A CN200810133598A CN200810133598A CN101635834A CN 101635834 A CN101635834 A CN 101635834A CN 200810133598 A CN200810133598 A CN 200810133598A CN 200810133598 A CN200810133598 A CN 200810133598A CN 101635834 A CN101635834 A CN 101635834A
- Authority
- CN
- China
- Prior art keywords
- module
- visual field
- identification
- image
- field video
- 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
- Image Analysis (AREA)
- Closed-Circuit Television Systems (AREA)
- Alarm Systems (AREA)
Abstract
The invention provides an automatic tracing identification system for artificial neural control. The system comprises a fixed viewing field video acquisition module, a full-functional variable viewing field video acquisition module, a video image identification and judgment algorithm module, an artificial neural control module, a moving object track tracing module, a database comparison and alarm judgment module, a monitoring characteristic entering and rule setting module, a light monitoring and controlling module, a background light source module, an alarm output, display and storage module and a safety monitoring sensor. The system fully solves the problems of unclear doubtful target characteristics, single function, low intelligence degree and the like in post processing and images of the prior image monitoring system, uses the artificial neural control, biological identification technology and intelligent image identification algorithm, and has the characteristics of high intelligence degree, precaution, clear doubtful object images, high cost performance and the like.
Description
Technical field
The present invention relates to a kind of automatic tracing identification system, relate in particular to the neural automatic tracing identification system of a kind.
Background technology
Along with the development of human civilization, people are more and more higher to the demand of safety.From investigating afterwards, become " strick precaution in advance, timely prevention in the thing and fast processing afterwards ", become the direction of wisdom security and guard technology development.
There is following problem in traditional safety monitoring system: 1. monitoring image is fixed, and is difficult to capture suspect article feature clearly under a lot of situations, especially has the dead angle of image, makes investigation afterwards have big difficulty; 2. do not take precautions against ability in advance, there are not wisdom identification algorithm ability and system's self-learning capability, can't detect the motion thing and whether violate rule of conduct (as cross the border, direction of motion mistake, pilferage etc.), more can not discern suspicious figure's identity, thereby not reach the purpose of taking precautions against in advance by biological identification technology; 3. can't determine the definite movement locus and the minutia of suspicious object; In addition, for legacy system, also need a large amount of cameras, a large amount of storage device and great amount of manpower to come supervisory control system.
Some system developments have been arranged to discerning, judge whether the motion thing breaks the rules at present according to set rule of conduct, or pilferage is not arranged, leave over behaviors such as article, but these systems are owing to the employed fixing video camera matching computer of visual field of mostly being is discerned warning, therefore, still it is many to keep away unavoidable image dead angle, can't determine the suspicious object movement locus, problems such as the suspicious object minutia is unintelligible, what is more important many times because image is too small, has difficulties and screen the suspicious figure.
Summary of the invention
Main purpose of the present invention is to provide a kind neural automatic tracing identification system, fixes, do not have the definite movement locus taking precautions against ability in advance, can't determine suspicious object and minutia, shortcoming that the wisdom degree is low in order to overcome the prior art monitoring image.
For achieving the above object, the invention provides a kind of automatic tracing identification system for artificial neural control, it comprises: fixing visual field video acquisition module, it utilizes fixedly the rifle formula or the ball-shaped camera of visual field, obtains the interior image of fixed area scope; Full function variable visual field video acquisition module, utilize automated variable Jiao's high speed monopod video camera or quick ball-shaped camera, in 360 degree scopes, catch the suspicious motion object detail image that occurs in the fixedly visual field video camera coverage in the fixing visual field video acquisition module rapidly; Video image identification is judged the algorithm module, from the video image that fixedly visual field video acquisition module and full function variable visual field video acquisition module collect, extract object, classify, determine the characteristic portion of object, license plate of people's face, automobile etc. for example, the recognition objective feature is as the number-plate number, face characteristic etc.; The neural control module of class, it controls the full function variable visual field video camera and the fixing angle coupling of visual field video camera in this full function variable visual field video acquisition module, according to the fixing orientation angles of visual field motion thing, control full function variable visual field video camera rotational angle, change lens focus and aperture etc., all the time follow the trail of suspicious motion thing with assurance full function variable visual field video camera, and make full function variable visual field video camera can capture detailed object features; Motion thing trajectory track module, judge the formation feature of the classification selection track of the target object that the algorithm module is determined according to video image identification, people's face or object center of gravity in this way, and judge the outcome record of algorithm identification in the algorithm module and form the movement locus of suspicious object according to video image identification; Data bank comparison and alarm decision module, judge the result of algorithm identification in the algorithm module according to this video image identification, extracting the information of feature data bank compares, as extract people's face data or suspicious object feature data in the data bank, and cross the border, enter the alarm decision of problems such as forbidden zone, direction of motion mistake according to the rule of setting; Monitoring features typing and regular setting module are input object features, as facial image, form data bank, and set the rule and the sensitivity level of reporting to the police according to the requirement of guarded region; Light monitoring and control module, according to the situation of the graphical analysis ambient light of video acquisition module, when ambient light was unfavorable for surveying, control started background light source; The background light source module according to the difference of site of deployment, selects different light source such as infrared light or artificial light so that background light source to be provided; Warning output, demonstration and memory module, it carries out the demonstration of warning message according to this data bank comparison and the alarm decision of alarm decision module output, and show the motion thing track that this motion thing trajectory track module is exported, monitoring video and warning message are put in order, stored; The security monitoring sensor, it is used for merging with the safety monitoring system of other types, when the security monitoring sensor moves, also is used to transfer the information that full function variable visual field video camera is caught needs.
The present invention also provides a kind of track identification method based on above-mentioned automatic tracing identification system for artificial neural control, and it comprises step 1, typing target signatures data base and set the track identification rule that use for network analysis, it is further comprising the steps of:
The setting of step 2, video camera and tracking are proofreaied and correct;
IMAQ is carried out in the neural control module control of step 3, class fixedly visual field video acquisition module and full function variable visual field video acquisition module, and the suspicious object minutia image that collects stores in hard disk or the image recorder with the fixing image of visual field video camera;
Step 6, the data bank comparative analysis and the output of reporting to the police.
The present invention has used light monitoring and control module and background light source module, makes system can work in night or the unfavorable environment of light condition effectively.Wherein when using near-infrared as a setting during light source, institute's use video camera has night vision function.
System of the present invention comprises a series of behavior identification algorithms, the neural control of class algorithm, track following algorithm and people's face or special characteristic identification algorithm.Different with wisdom frequency image monitoring system in the past, system of the present invention no longer only is confined to simple function, and has possessed multi-functional monitoring, recognition function.
System of the present invention has solved the problem that the traditional images supervisory control system exists, connect a plurality of fixed visual field video camera and a cover full function variable visual field video camera by a cover tracking and identifying system main frame, both can finish the monitoring function of traditional images supervisory control system, can further realize the behavior detecting function of people or motion thing again, and the suspicious object of violating rule of conduct carried out center of gravity or head movement track following (image range movement locus and architectural plane movement locus) in the zone, the system that the more important thing is obtains the minutia of suspicious object fast by the neural control of class full function variable visual field video camera, catch and comprise people's face, the frame that car plate etc. are important, and carry out bio-identification and wisdom graphical analysis, information in characteristic parameter and the data bank compares the most at last, provides dissimilar warning messages.
Beneficial effect of the present invention is, ((1) system has full function variable visual field monitoring function, by the neural control of class video camera, can at utmost extract the minutia image of suspicious object, thereby realize the bio-identification and the image recognition of suspicious object people's face and privileged site, extract characteristic parameter, and then compare, report to the police timely with the data bank data.(2) system has the behavior recognition function, can analyze at the behavior of people or other motion things, with the behavior of determining to break the rules, and then sends warning message.(3) system has suspicious object motion tracking recognition function, can represent the movement locus of suspicious object on video image and architectural plane, determines the direction of motion.Above advantage makes system of the present invention have early detection, strick precaution in advance, mid-event control and the advantage of fast processing afterwards, and this is that in the past frequency image monitoring system hardly matches.
Description of drawings
Fig. 1 is the functional module frame diagram of automatic tracing identification system for artificial neural control of the present invention;
Fig. 2 is the operation operating procedure of system of the present invention;
Fig. 3 is a kind of area tracking recognition system pie graph based on computer or DSP treatment system of the present invention;
Fig. 4 A, 4B for system of the present invention fixedly the coarse adjustment of mapping angle one by one of visual field video camera and full function variable visual field video camera proofread and correct schematic diagram;
Fig. 5 A, 5B set the rule of conduct schematic diagram for the present invention;
Fig. 6 A, 6B, 6C, 6D catch the contrast schematic diagram of suspicious object image and traditional images supervisory control system image for system of the present invention.
Description of reference numerals:
1-is the visual field video acquisition module fixedly; 10-background light source module; 11-security monitoring sensor; 2-full function variable visual field video acquisition module; The 3-video image identification is judged the algorithm module; Comparison of 4-data bank and alarm decision module; 5-warning output, demonstration and memory module; 6-motion thing trajectory track module; Typing of 7-monitoring features and regular setting module; Monitoring of 8-light and control module; The neural control module of 9-class.
Embodiment
Below in conjunction with accompanying drawing above-mentioned and other feature and advantage of the present invention are elaborated.
Fig. 1 is the functional module frame diagram of automatic tracing identification system for artificial neural control of the present invention.Total system is by fixedly visual field video acquisition module 1, full function variable visual field video acquisition module 2, video image identification are judged algorithm module 3, the neural control module 9 of class, motion thing trajectory track module 6, data bank is compared and alarm decision module 4, monitoring features typing and regular setting module 7, light monitoring are formed with control module 8, warning output, demonstration and memory module 5, background light source module 10, security monitoring sensor 11.Fixedly visual field video acquisition module 1 is mainly used the rifle formula or the ball-shaped camera of plurality of fixed visual field, according to the requirement of institute's environment for use, and the video camera that can select common camera for use or have night vision function.A common cover district system can be used three fixedly visual field video cameras, to obtain the image in the extensive region.Full function variable visual field video acquisition module 2 is main uses the high speed monopod video cameras of the burnt control of automated variables or ball-shaped cameras fast, catches the suspicious motion object detail image that occurs in the fixing visual field video camera coverage in 360 degree scopes rapidly.According to the requirement of institute's environment for use, the video camera that can select common camera for use or have night vision function.This module is accepted the control signal of the neural control module of class, carries out the adjustment of angle adjustment and focal length.The method of control is to allow the characteristic portion of suspicious object, as people's face, is in all the time on the middle part and certain scale of image.Promptly this system uses the neural control method of similar human eye captured target thing, earlier by being somebody's turn to do fixedly visual field video camera locking suspected target, system controls this full function variable visual field video camera rotation, amplification, large scale lock onto target thing afterwards, and emphasis is caught key character.
Fixedly visual field and full function variable visual field video camera can be charge coupled device (Charge CoupledDevice:CCD), CMOS (Complementary Metal Oxide Semiconductor) (Comple-mentaryMetal-Oxicle-Semiconductor:CMOS) video camera or thermal imaging camera.
Video image identification judges that algorithm module 3 mainly comprises the feature identification algorithm of humanoid detecting identification algorithm, number of people location algorithm, motion thing identification algorithm, behavior identification algorithm, system failure algorithm and other types etc., it extracts object from this video image that fixedly visual field video acquisition module and full function variable visual field video acquisition module collect, classify, determine the characteristic portion of object, and discerned.The object type that motion thing identification algorithm can be discerned is as required determined different match parameter, and for example human body identification can be by depth-width ratio, limbs are discerned than the characteristic parameters such as ratio of, head and health up and down.Behavior identification algorithm, then generally comprise cross the border, enter the security area, the direction of motion is unusual, movement velocity is unusual, steal, leave over different behavioural characteristic identification such as article, for the target that breaks the rules, all can be regarded as suspicious object, and then start the automatic orbit tracking function.Feature identification calculation rule is after extracting the characteristic image of details, by the parameter at a series of algorithm calculated characteristics position, for example colour of skin of body ratio, people's face, people's face skeleton feature, interpupillary distance, skeletal structure, face feature, the number-plate number, vehicle color etc.
The neural control module 9 of class constitutes closed-loop control system by the global function camera review as feedback.The neural control module of class can according to after the artificial adjustment fixedly the coarse adjustment corresponding angle quick control global function video camera of visual field video camera and variable field of view video camera turn to the coarse adjustment angle, afterwards by the fine setting angle with focus automatically and catch suspicious object thing characteristic portion, be in the central of image and account for image area up to characteristic portion and be not less than 15% position.Along with moving of suspicious object, the neural control module of class also will be controlled the global function video camera and follow the tracks of rotation and adjust focal length, continuous capturing suspicious object characteristic portion.
The suspicious object that motion thing trajectory track module 6 is primarily aimed in the fixing visual field camera review is carried out track following, the track that both can finish the video image scope calculates, follows the tracks of and sign, can also finish that the track coordinate is mapped in the architectural plane coordinate, and the track of suspicious object is identified on the architectural plane.Can select the formation feature of track according to the classification of target object, people's face or object center of gravity in this way, and form the movement locus of suspicious object according to the outcome record of algorithm identification.Movement locus can be converted to the coordinate position of architectural plane as required, and system has corresponding coordinate transformation algorithms.
Data bank comparison and alarm decision module 4, it judges the result of algorithm identification in the algorithm module according to this video image identification, extracting the information of feature data bank compares, as extract people's face data or suspicious object feature data in the data bank, and cross the border, enter the alarm decision of problems such as forbidden zone, direction of motion mistake according to the rule of setting.The rule that the suspicious object minutia is reported to the police may be prescribed as: meet the warning of data bank feature, or do not meet the warning of data bank feature.
Monitoring features typing and regular setting module 7 mainly are input object features, as facial image, form data bank, and set the rule and the sensitivity level of reporting to the police according to the requirement of guarded region.The foundation of target signatures data base both can using system typing of monitoring objective feature and the instant typing of regular setting module, for example the personnel that want typing are carried out data immediately and obtain, also can use the image library or the direct import system of data bank that have recorded to use as data bank.In order to guarantee the reliability of system, this module can be upgraded fixed data feature or store in the data bank in addition.
Light monitoring and control module 8, it is according to the situation of the graphical analysis ambient light of video acquisition module, and when ambient light was unfavorable for surveying, control started background light source.
Background light source module 10 can select to use different light source such as near-infrared light source or artificial light so that background light source to be provided according to the difference of site of deployment, and the control command that it accepts light monitoring and control module starts light source.When using this near-infrared light source, above-mentioned video camera possesses night vision function, to obtain the near-infrared image of motion thing.
Warning output, demonstration and memory module 5, it carries out demonstration, the relay output of warning message according to this data bank comparison and the alarm decision of alarm decision module output, and show the motion thing track that this motion thing trajectory track module is exported, monitoring video and warning message are put in order, stored.
Fig. 2 is a system operation operating procedure of the present invention.This figure relates to a kind of track identification method based on above-mentioned automatic tracing identification system for artificial neural control, mainly may further comprise the steps:
12: step 1, typing target signatures data base and setting track identification rule are used for network analysis.Mode that on the one hand can be by fixing or the instant typing one by one of variable field of view video camera is photographic subjects one by one, and by algorithm the clarification of objective positional parameter is extracted, and finally deposits in the data bank; On the other hand, also image or the target signature parameter library that has recorded directly can be imported in the data bank of system of the present invention.Determining for suspicious object in addition; need preestablish rule of conduct by the client; usually can comprise cross the border, behaviors such as the direction of motion is unusual, movement velocity is unusual, pilferage, in case the rule of conduct that moving object violate to be set promptly can be confirmed as " suspicious object " immediately.
13: step 2, the setting of video camera and tracking are proofreaied and correct.Main purpose be with full function variable visual field video camera and fixedly the monitor area of visual field video camera carry out relationship maps.As shown in Figure 4, angle and focal length by manual adjustment global function video camera, some selected points that the global function video camera can be obtained image carry out corresponding one by one with the corresponding points of fixed cameras image, main purpose is to provide a parameter for the control of global function video camera initial angle in the neural control of class, the assurance system can quick control global function video camera turns to fixed cameras to be had on the coarse adjustment angle of certain matching degree, again by the further minutia of fine setting and meticulous control seizure suspicious object.
14: step 3, IMAQ is carried out in the neural control module control of class fixedly visual field video acquisition module and full function variable visual field video acquisition module, and the suspicious object minutia image that collects stores in hard disk or the image recorder with the fixing image of visual field video camera.The fixedly image of visual field video camera is constantly gathered by system, and offer main frame analysis, when finding suspicious object, system is by the neural control of automatic class global function video camera quick rotation and adjust focal length, make the characteristic portion of suspicious object be positioned at the central part of image, size is not less than 15% of general image, and then this image is gathered.In addition when the system discovery ambient light is not enough to satisfy requiring of groundwork, will control background light source starts, and the while is according to the difference of light source, control the mode of operation of each video camera, when using near-infrared light source (Near Infraredray), system need switch to black and white and night vision function pattern, to obtain the near-infrared image of motion thing.
15: step 4, suspicious object trajectory track.System carries out the wisdom graphical analysis according to the fixedly visual field camera review that collects, determine whether to have cross the border, direction of motion mistake, movement velocity are unusual, rule of conduct is violated in pilferage etc. problem exists, in case determine, promptly is denoted as suspicious object.And then system follows the trail of the movement locus of suspicious object automatically.Movement locus can use center of gravity, people's head etc. according to the difference of monitoring object, and movement locus can be divided into two kinds of track in the image visual field and the tracks in the building plane, needs to be changed by linear transformation between two kinds of tracks.
16: step 5, the suspicious object feature identification is extracted.System is at the full function variable visual field camera review that collects, and use biological identification technology etc. carries out the wisdom graphical analysis, extracts the characteristic portion and the relevant parameter of suspicious object.Relevant parameter comprises face complexion, interocular distance, skeleton feature, face feature etc., or license plate, color etc.Described suspicious object feature identification is extracted and is specifically comprised: step 51, each serial image of fixedly visual field video acquisition module acquisition is analyzed, extract the motion thing, calculate ratio characteristic parameter, distribution of color feature, the sign that breaks the rules, motion thing center of gravity, movement locus, the direction prediction parameter of motion thing; Step 52, at the suspicious object that breaks the rules, by the neural control module control of class full function variable visual field video acquisition module, obtain amplification, the detail pictures of suspicious object, extract the characteristic portion of suspicious object; The characteristic portion parameter of step 53, calculating suspicious object comprises body ratio, people's face skeleton feature, interpupillary distance, face feature, the colour of skin, the number-plate number, vehicle color; Step 54, at a plurality of suspicious object that occur in the guarded region, system uses the mode of poll to screen one by one, when the motion thing is a colony, the identification algorithm will be carried out cluster analysis, the emphasis chosen distance is near, feature the most clearly suspicious object survey, tracking and signature analysis; When suspicious object was dispersion, system promptly can screen one by one.
17: step 6, the data bank comparative analysis and the output of reporting to the police.At the above suspicious object characteristic parameter that calculates, system will compare analysis with the characteristic parameter of data bank, for not meeting in time reporting to the police of imposing a condition.Usually can be divided into and meet the data bank feature and report to the police and do not meet rule and condition such as data bank feature warning.For the feature of confirming, system will carry out self study, store the characteristic parameter of up-to-date acquisition.
Fig. 3 is a kind of area tracking recognition system pie graph based on computer or DSP treatment system of the present invention, be the embodiment of a domain type system of the present invention, system is by fixedly visual field rifle formula video camera, full function variable visual field video camera, track identification main frame, input-output equipment, and the network equipment is formed.The track identification main frame has been finished IMAQ, identification, judgement, the neural control of class, movement locus tracking, light monitoring and control, monitoring features typing and the regular a series of functions such as output, storage of setting, report to the police, and is the core of system.The global function video camera is then accepted the neural control command of class of system host, and quick rotation and adjustment focal length are to catch the minutia image of suspicious object.System can export warning message by monitor etc.Every suit system can cover a bigger zone, as a construction level, has solved security fields early detections, strick precaution in advance, mid-event control and the problem of fast processing afterwards.System also has networking function, carries out the prompting of the data passes and the direction of motion between the system of networking, makes total system can effectively monitor bigger zone of protection, to be used to tackle security monitoring problem large-scale, important place.
System of the present invention can follow the tracks of at a plurality of suspicious object in the overlay area, and catches the minutia of suspicious object.General method is: the first, to some fixedly the zone that covers of visual field video camera carry out poll, lock the suspicious object that all break the rules; The second, carry out cluster analysis, for people or the motion thing that a group is associated, system often emphasis catches recently and characteristic portion suspicious object the most clearly; The 3rd, for the suspicious object of disperseing, system uses polling mode, catches the suspicious object characteristic portion one by one, discerns respectively and calculates.System can follow the tracks of the target more than 6 at most simultaneously.
System of the present invention can also cover bigger building protection area by formation networked systems such as Intranet.When suspicious object exceeded the monitoring range of a cover system, the adjacent system of networking entered tracking mode immediately, and the suspicious object characteristic parameter of having caught transmits between networked hosts, and system can realize the trajectory track of suspicious object on a large scale like this.
Fig. 5 sets the rule of conduct schematic diagram for the present invention, i.e. the setting example of behavior rule enters regional setting, allows the setting of the direction of motion etc. comprising forbidding, can reach the rule of conduct requirement of user expectation by systems soft ware.
Fig. 6 catches the contrast of suspicious object image and traditional frequency image monitoring system image for system of the present invention.One of them is the surveillance on the bank doorway, and conventional camera system can only be recorded to ambiguous image, does not see messenger's the face and the license plate of motorcycle, as shown in Figure 6A; System of the present invention then can get access to messenger's facial detail feature and license plate, and for meeting criminal's feature in the data bank, system is in time reported to the police, shown in Fig. 6 B.Another then is some no-parking zones, conventional camera system can't be determined the minutia of the vehicle of parking ticket rule, shown in Fig. 6 C, system of the present invention then can obtain the image of license plate, and then the number-plate number that identification obtains placed on record, so that handle, shown in Fig. 6 D afterwards.
The above only is the preferred embodiments of the present invention; more than to the invention description only be illustrative and nonrestrictive; those skilled in the art is understood; within spirit that following appended claim limits and scope, can make many modifications to it; change or equivalence, but they all will fall within the scope of protection of the present invention.
Claims (10)
1, a kind of automatic tracing identification system for artificial neural control comprises the fixedly visual field video acquisition module that is used to obtain image in the fixed area scope, it is characterized in that it also comprises:
Full function variable visual field video acquisition module, it catches rapidly this suspicious motion object detail image that fixedly occurs in the fixedly visual field video camera coverage in the video acquisition module of visual field in 360 degree scopes;
Video image identification is judged the algorithm module, and it extracts object from this video image that fixedly visual field video acquisition module and full function variable visual field video acquisition module collect, classify, and determines the characteristic portion of object, the recognition objective feature;
The neural control module of class, it is controlled the full function variable visual field video camera in this full function variable visual field video acquisition module and is somebody's turn to do the fixedly angle coupling of visual field video camera, according to the fixing orientation angles of visual field motion thing, control this full function variable visual field video camera rotational angle, change lens focus and aperture etc., guaranteeing that this full function variable visual field video camera follows the trail of suspicious motion thing all the time, and make this full function variable visual field video camera capture detailed object features;
Motion thing trajectory track module, it judges that according to this video image identification the classification of the target object that algorithm module is determined selects the formation feature of track, and judges the outcome record of algorithm identification in algorithm module and form the movement locus of suspicious object according to this video image identification;
Data bank comparison and alarm decision module, it judges the result that algorithm is discerned in the algorithm module according to this video image identification, extracts the information of feature data bank and compares, and carry out alarm decision according to the rule of setting;
Monitoring features typing and regular setting module, it is an input object feature, forms data bank, and sets the rule and the sensitivity level of reporting to the police according to the requirement of guarded region;
Light monitoring and control module, it is according to the situation of the graphical analysis ambient light of this video acquisition module, and when ambient light was unfavorable for surveying, control started background light source;
The background light source module, it is according to the difference of site of deployment, and the light source of selecting infrared light or artificial light is to provide background light source;
Warning output, demonstration and memory module, it carries out the demonstration of warning message according to this data bank comparison and the alarm decision of alarm decision module output, and show the motion thing track that this motion thing trajectory track module is exported, monitoring video and warning message are put in order, stored;
The security monitoring sensor, it is used for merging with the safety monitoring system of other types, when this security monitoring sensor action, also is used to transfer the information that this full function variable visual field video camera is caught needs.
2, automatic tracing identification system for artificial neural control according to claim 1 is characterized in that, this is the rifle formula or the ball-shaped camera of visual field video acquisition module use plurality of fixed visual field fixedly; This full function variable visual field video acquisition module is used the high speed monopod video camera or the quick ball-shaped camera of autozoom; This video camera uses a kind of in charge coupled device camera, CMOS (Complementary Metal Oxide Semiconductor) video camera or thermal imaging camera.
3, automatic tracing identification system for artificial neural control according to claim 1, it is characterized in that this video image identification judges that algorithm comprises humanoid detecting identification algorithm, number of people location algorithm, motion thing identification algorithm vehicle license identification algorithm, behavior identification algorithm, system failure algorithm.
4, automatic tracing identification system for artificial neural control according to claim 1, it is characterized in that, it uses the neural control method of similar human eye captured target thing, earlier by being somebody's turn to do fixedly visual field video camera locking suspected target, system controls this full function variable visual field video camera rotation, amplification, large scale lock onto target thing afterwards, and emphasis is caught key character.
5, automatic tracing identification system for artificial neural control according to claim 1, it is characterized in that it is set rule of conduct and determines suspicious object, for the target that breaks the rules, all can be regarded as suspicious object, and then start this motion thing trajectory track module automatic orbit tracking function.
6, automatic tracing identification system for artificial neural control according to claim 1 is characterized in that, this background light source module is used near-infrared light source or visible light source; When using this near-infrared light source, this video camera possesses night vision function, to obtain the near-infrared image of motion thing.
7, a kind of track identification method based on above-mentioned automatic tracing identification system for artificial neural control, it comprises step 1, typing target signatures data base and sets the track identification rule, uses for network analysis, it is characterized in that it is further comprising the steps of:
The setting of step 2, video camera and tracking are proofreaied and correct;
IMAQ is carried out in the neural control module control of step 3, class fixedly visual field video acquisition module and full function variable visual field video acquisition module, and the suspicious object minutia image that collects stores in hard disk or the image recorder with the fixing image of visual field video camera;
Step 4, suspicious object trajectory track;
Step 5, suspicious object feature identification are extracted;
Step 6, the data bank comparative analysis and the output of reporting to the police.
8, the track identification method based on above-mentioned automatic tracing identification system for artificial neural control according to claim 7, it is characterized in that, the monitoring features typing of setting up using system of target signatures data base and the instant typing of regular setting module in the step 1, or use the image library or the direct import system of data bank that have recorded to use as data bank.
9, the track identification method based on above-mentioned automatic tracing identification system for artificial neural control according to claim 7, it is characterized in that, the track that the trajectory track of suspicious object described in the step 4 had both been finished the video image scope calculates, follows the tracks of and sign, finish that also the track coordinate is mapped in the architectural plane coordinate, and the track of suspicious object is identified on the architectural plane.
10, the track identification method based on above-mentioned automatic tracing identification system for artificial neural control according to claim 7 is characterized in that, the feature identification of suspicious object described in the step 5 is extracted and specifically comprised:
Step 51, each serial image that fixing visual field video acquisition module is obtained are analyzed, extract the motion thing, calculate ratio characteristic parameter, distribution of color feature, the sign that breaks the rules, motion thing center of gravity, movement locus, the direction prediction parameter of motion thing;
Step 52, at the suspicious object that breaks the rules, by the neural control module control of class full function variable visual field video acquisition module, obtain amplification, the detail pictures of suspicious object, extract the characteristic portion of suspicious object;
The characteristic portion parameter of step 53, calculating suspicious object comprises body ratio, people's face skeleton feature, interpupillary distance, face feature, the colour of skin, the number-plate number, vehicle color;
Step 54, at a plurality of suspicious object that occur in the guarded region, system uses the mode of poll to screen one by one, when the motion thing is a colony, the identification algorithm will be carried out cluster analysis, the emphasis chosen distance is near, feature the most clearly suspicious object survey, tracking and signature analysis; When suspicious object was dispersion, system promptly can screen one by one.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200810133598A CN101635834A (en) | 2008-07-21 | 2008-07-21 | Automatic tracing identification system for artificial neural control |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200810133598A CN101635834A (en) | 2008-07-21 | 2008-07-21 | Automatic tracing identification system for artificial neural control |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101635834A true CN101635834A (en) | 2010-01-27 |
Family
ID=41594868
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200810133598A Pending CN101635834A (en) | 2008-07-21 | 2008-07-21 | Automatic tracing identification system for artificial neural control |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101635834A (en) |
Cited By (38)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101833795A (en) * | 2010-05-11 | 2010-09-15 | 江苏丰东热技术股份有限公司 | Dynamic recorder and recording method for heat treatment parameters |
| CN102348102A (en) * | 2010-07-30 | 2012-02-08 | 鸿富锦精密工业（深圳）有限公司 | Roof safety monitoring system and method thereof |
| CN102436661A (en) * | 2010-09-29 | 2012-05-02 | 新谊整合科技股份有限公司 | Detection method of movable light source and security system of application |
| CN102447881A (en) * | 2010-10-13 | 2012-05-09 | 鸿富锦精密工业（深圳）有限公司 | Camera device and method for monitoring image by camera device |
| CN102447882A (en) * | 2010-10-13 | 2012-05-09 | 鸿富锦精密工业（深圳）有限公司 | TOF (Time of Flight) camera device and method for monitoring image by TOF camera device |
| CN102479319A (en) * | 2010-11-25 | 2012-05-30 | 鸿富锦精密工业（深圳）有限公司 | Image recognition method and control computer thereof |
| CN102650801A (en) * | 2011-02-25 | 2012-08-29 | 鸿富锦精密工业（深圳）有限公司 | Camera and automatic focusing method thereof |
| CN102693634A (en) * | 2011-03-24 | 2012-09-26 | 大连航天金穗科技有限公司 | Television monitoring system for tracking set vehicle and television monitoring method thereof |
| CN102855475A (en) * | 2012-09-17 | 2013-01-02 | 广州杰赛科技股份有限公司 | School bus monitoring method and school bus monitoring system |
| CN103024276A (en) * | 2012-12-17 | 2013-04-03 | 沈阳聚德视频技术有限公司 | Positioning and focusing method of pan-tilt camera |
| CN103136899A (en) * | 2013-01-23 | 2013-06-05 | 宁凯 | Intelligent alarming monitoring method based on Kinect somatosensory equipment |
| CN103501423A (en) * | 2013-09-18 | 2014-01-08 | 苏州景昱医疗器械有限公司 | Video monitoring method and device adopting remote program control |
| CN103795984A (en) * | 2014-02-07 | 2014-05-14 | 彭世藩 | Self-learning omnibearing mobile monitoring system |
| CN103971477A (en) * | 2014-05-19 | 2014-08-06 | 华北水利水电大学 | Anti-theft system based on pattern recognition technology |
| CN104092981A (en) * | 2014-06-30 | 2014-10-08 | 深圳安邦科技有限公司 | Wireless video monitoring system |
| CN104407488A (en) * | 2014-12-19 | 2015-03-11 | 咸洋 | Laser-assisted vehicle information shooting device and shooting method |
| CN104516295A (en) * | 2014-12-15 | 2015-04-15 | 成都凌感科技有限公司 | Device for automatically recognizing human body malignant violent terrorism actions and emitting violence preventing substances |
| CN105245853A (en) * | 2015-10-27 | 2016-01-13 | 太原市公安局 | Video monitoring method |
| CN105260744A (en) * | 2015-10-08 | 2016-01-20 | 北京航空航天大学 | Automatic on-line diagnosis method for freight train coupler tail cotter position faults and system |
| CN105894702A (en) * | 2016-06-21 | 2016-08-24 | 南京工业大学 | Invasion detecting alarming system based on multi-camera data combination and detecting method thereof |
| CN105959653A (en) * | 2016-07-12 | 2016-09-21 | 湖北誉恒科技有限公司 | Video monitoring system for pedestrian crosswalk |
| CN106210467A (en) * | 2016-07-16 | 2016-12-07 | 惠州学院 | Identification system and method to same video object under a kind of different angles |
| CN106815958A (en) * | 2017-02-24 | 2017-06-09 | 上海华崟信息技术有限公司 | Warning system, alarm analysis/display device, alarm analysis/display methods |
| CN107105207A (en) * | 2017-06-09 | 2017-08-29 | 北京深瞐科技有限公司 | Target monitoring method, target monitoring device and video camera |
| CN107277460A (en) * | 2017-07-29 | 2017-10-20 | 安徽博威康信息技术有限公司 | A kind of intellectual monitoring alarm method of identity-based identification |
| CN107277461A (en) * | 2017-07-29 | 2017-10-20 | 安徽博威康信息技术有限公司 | A kind of intellectual monitoring alarm method based on fixation and recognition |
| CN107316463A (en) * | 2017-07-07 | 2017-11-03 | 深圳市诺龙技术股份有限公司 | A kind of method and apparatus of vehicle monitoring |
| CN107688767A (en) * | 2016-08-04 | 2018-02-13 | 惠州学院 | The system and method for human body feature is extracted under a kind of high-noise environment |
| CN109377697A (en) * | 2018-11-07 | 2019-02-22 | 浙江警官职业学院 | Rapid Alarm method of disposal under a kind of intensive camera head environment |
| CN109686032A (en) * | 2019-01-17 | 2019-04-26 | 厦门大学 | A kind of aquaculture organisms theft prevention monitoring method and system |
| CN109740462A (en) * | 2018-12-21 | 2019-05-10 | 北京智行者科技有限公司 | The identification follower method of target |
| CN109816699A (en) * | 2019-01-30 | 2019-05-28 | 华通科技有限公司 | A kind of holder angle computation method inhibiting frame differential method based on background |
| CN110459027A (en) * | 2019-08-15 | 2019-11-15 | 青岛文达通科技股份有限公司 | A kind of Community Safety means of defence and system based on multi-source heterogeneous data fusion |
| CN110662002A (en) * | 2019-10-23 | 2020-01-07 | 徐州丰禾智能科技有限公司 | Factory security system with real-time image recognition and warning functions |
| CN112345079A (en) * | 2020-10-30 | 2021-02-09 | 北京字节跳动网络技术有限公司 | Detection method, device, terminal and storage medium |
| CN114821805A (en) * | 2022-05-18 | 2022-07-29 | 湖北大学 | Dangerous behavior early warning method, device and equipment |
| CN115060734A (en) * | 2022-08-09 | 2022-09-16 | 吉林信息安全测评中心 | Multi-angle image shooting and recording device for industrial vision detection |
| CN116778302A (en) * | 2023-05-19 | 2023-09-19 | 触景无限科技（北京）有限公司 | Object recognition system and portable cloth control box |
-
2008
- 2008-07-21 CN CN200810133598A patent/CN101635834A/en active Pending
Cited By (47)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101833795A (en) * | 2010-05-11 | 2010-09-15 | 江苏丰东热技术股份有限公司 | Dynamic recorder and recording method for heat treatment parameters |
| CN102348102B (en) * | 2010-07-30 | 2015-04-15 | 赛恩倍吉科技顾问（深圳）有限公司 | Roof safety monitoring system and method thereof |
| CN102348102A (en) * | 2010-07-30 | 2012-02-08 | 鸿富锦精密工业（深圳）有限公司 | Roof safety monitoring system and method thereof |
| CN102436661A (en) * | 2010-09-29 | 2012-05-02 | 新谊整合科技股份有限公司 | Detection method of movable light source and security system of application |
| CN102447881A (en) * | 2010-10-13 | 2012-05-09 | 鸿富锦精密工业（深圳）有限公司 | Camera device and method for monitoring image by camera device |
| CN102447882A (en) * | 2010-10-13 | 2012-05-09 | 鸿富锦精密工业（深圳）有限公司 | TOF (Time of Flight) camera device and method for monitoring image by TOF camera device |
| CN102479319A (en) * | 2010-11-25 | 2012-05-30 | 鸿富锦精密工业（深圳）有限公司 | Image recognition method and control computer thereof |
| CN102650801A (en) * | 2011-02-25 | 2012-08-29 | 鸿富锦精密工业（深圳）有限公司 | Camera and automatic focusing method thereof |
| CN102693634A (en) * | 2011-03-24 | 2012-09-26 | 大连航天金穗科技有限公司 | Television monitoring system for tracking set vehicle and television monitoring method thereof |
| CN102693634B (en) * | 2011-03-24 | 2013-11-13 | 大连航天金穗科技有限公司 | Television monitoring system for tracking set vehicle and television monitoring method thereof |
| CN102855475A (en) * | 2012-09-17 | 2013-01-02 | 广州杰赛科技股份有限公司 | School bus monitoring method and school bus monitoring system |
| CN103024276A (en) * | 2012-12-17 | 2013-04-03 | 沈阳聚德视频技术有限公司 | Positioning and focusing method of pan-tilt camera |
| CN103136899A (en) * | 2013-01-23 | 2013-06-05 | 宁凯 | Intelligent alarming monitoring method based on Kinect somatosensory equipment |
| CN103136899B (en) * | 2013-01-23 | 2016-01-20 | 宁凯 | Based on the intelligent alarm method for supervising of Kinect somatosensory device |
| CN103501423A (en) * | 2013-09-18 | 2014-01-08 | 苏州景昱医疗器械有限公司 | Video monitoring method and device adopting remote program control |
| CN103795984A (en) * | 2014-02-07 | 2014-05-14 | 彭世藩 | Self-learning omnibearing mobile monitoring system |
| CN103971477A (en) * | 2014-05-19 | 2014-08-06 | 华北水利水电大学 | Anti-theft system based on pattern recognition technology |
| CN104092981A (en) * | 2014-06-30 | 2014-10-08 | 深圳安邦科技有限公司 | Wireless video monitoring system |
| CN104516295A (en) * | 2014-12-15 | 2015-04-15 | 成都凌感科技有限公司 | Device for automatically recognizing human body malignant violent terrorism actions and emitting violence preventing substances |
| CN104407488A (en) * | 2014-12-19 | 2015-03-11 | 咸洋 | Laser-assisted vehicle information shooting device and shooting method |
| CN105260744A (en) * | 2015-10-08 | 2016-01-20 | 北京航空航天大学 | Automatic on-line diagnosis method for freight train coupler tail cotter position faults and system |
| CN105260744B (en) * | 2015-10-08 | 2018-08-14 | 北京航空航天大学 | The automatic on-line diagnostic method and system of a kind of goods train coupler yoke key position failure |
| CN105245853A (en) * | 2015-10-27 | 2016-01-13 | 太原市公安局 | Video monitoring method |
| CN105894702A (en) * | 2016-06-21 | 2016-08-24 | 南京工业大学 | Invasion detecting alarming system based on multi-camera data combination and detecting method thereof |
| CN105894702B (en) * | 2016-06-21 | 2018-01-16 | 南京工业大学 | A kind of intrusion detection warning system and its detection method based on multiple-camera data fusion |
| CN105959653A (en) * | 2016-07-12 | 2016-09-21 | 湖北誉恒科技有限公司 | Video monitoring system for pedestrian crosswalk |
| CN106210467A (en) * | 2016-07-16 | 2016-12-07 | 惠州学院 | Identification system and method to same video object under a kind of different angles |
| CN107688767A (en) * | 2016-08-04 | 2018-02-13 | 惠州学院 | The system and method for human body feature is extracted under a kind of high-noise environment |
| CN106815958A (en) * | 2017-02-24 | 2017-06-09 | 上海华崟信息技术有限公司 | Warning system, alarm analysis/display device, alarm analysis/display methods |
| CN107105207A (en) * | 2017-06-09 | 2017-08-29 | 北京深瞐科技有限公司 | Target monitoring method, target monitoring device and video camera |
| CN107316463A (en) * | 2017-07-07 | 2017-11-03 | 深圳市诺龙技术股份有限公司 | A kind of method and apparatus of vehicle monitoring |
| CN107277460A (en) * | 2017-07-29 | 2017-10-20 | 安徽博威康信息技术有限公司 | A kind of intellectual monitoring alarm method of identity-based identification |
| CN107277461A (en) * | 2017-07-29 | 2017-10-20 | 安徽博威康信息技术有限公司 | A kind of intellectual monitoring alarm method based on fixation and recognition |
| CN109377697A (en) * | 2018-11-07 | 2019-02-22 | 浙江警官职业学院 | Rapid Alarm method of disposal under a kind of intensive camera head environment |
| CN109740462B (en) * | 2018-12-21 | 2020-10-27 | 北京智行者科技有限公司 | Target identification following method |
| CN109740462A (en) * | 2018-12-21 | 2019-05-10 | 北京智行者科技有限公司 | The identification follower method of target |
| CN109686032A (en) * | 2019-01-17 | 2019-04-26 | 厦门大学 | A kind of aquaculture organisms theft prevention monitoring method and system |
| CN109686032B (en) * | 2019-01-17 | 2021-03-02 | 厦门大学 | Aquaculture organism anti-theft monitoring method and system |
| CN109816699A (en) * | 2019-01-30 | 2019-05-28 | 华通科技有限公司 | A kind of holder angle computation method inhibiting frame differential method based on background |
| CN109816699B (en) * | 2019-01-30 | 2021-07-27 | 国网智能科技股份有限公司 | Holder angle calculation method based on background suppression interframe difference method |
| CN110459027A (en) * | 2019-08-15 | 2019-11-15 | 青岛文达通科技股份有限公司 | A kind of Community Safety means of defence and system based on multi-source heterogeneous data fusion |
| CN110662002A (en) * | 2019-10-23 | 2020-01-07 | 徐州丰禾智能科技有限公司 | Factory security system with real-time image recognition and warning functions |
| CN112345079A (en) * | 2020-10-30 | 2021-02-09 | 北京字节跳动网络技术有限公司 | Detection method, device, terminal and storage medium |
| CN114821805A (en) * | 2022-05-18 | 2022-07-29 | 湖北大学 | Dangerous behavior early warning method, device and equipment |
| CN115060734A (en) * | 2022-08-09 | 2022-09-16 | 吉林信息安全测评中心 | Multi-angle image shooting and recording device for industrial vision detection |
| CN116778302A (en) * | 2023-05-19 | 2023-09-19 | 触景无限科技（北京）有限公司 | Object recognition system and portable cloth control box |
| CN116778302B (en) * | 2023-05-19 | 2024-03-26 | 触景无限科技（北京）有限公司 | Object recognition system and portable cloth control box |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101635834A (en) | Automatic tracing identification system for artificial neural control | |
| US8036425B2 (en) | Neural network-controlled automatic tracking and recognizing system and method | |
| CN101465033B (en) | Automatic tracking recognition system and method | |
| CN105979210B (en) | A kind of pedestrian's identifying system based on the more ball video camera arrays of multiple gun | |
| KR101808587B1 (en) | Intelligent integration visual surveillance control system by object detection and tracking and detecting abnormal behaviors | |
| CN201278180Y (en) | Automatic tracking recognition system | |
| CN102254169B (en) | Multi-camera-based face recognition method and multi-camera-based face recognition system | |
| JP4751442B2 (en) | Video surveillance system | |
| CN110348348B (en) | Quick identification method and early warning system for entrance identities of participants | |
| CN105915847A (en) | Characteristics matching and tracking based video monitoring apparatus and method | |
| CN102414724B (en) | There is at least one for identifying the self-aided terminal of the video camera of crime attempt | |
| JP4357521B2 (en) | Unauthorized passing person detection device and unauthorized passing person recording system using the same | |
| US20140369567A1 (en) | Authorized Access Using Image Capture and Recognition System | |
| CN101814130A (en) | Iris identification device by using camera array and multimodal biometrics identification method | |
| CN107230267B (en) | Intelligence In Baogang Kindergarten based on face recognition algorithms is registered method | |
| KR101961891B1 (en) | Automatic counting method and appratus for human among the human and stuffs entering into automatic immigration check point | |
| CN113223046B (en) | Method and system for identifying prisoner behaviors | |
| CN206272770U (en) | A kind of vehicle and face bayonet system | |
| CN106034217A (en) | Remote face monitoring system and method thereof | |
| JP4617286B2 (en) | Unauthorized passing person detection device and unauthorized passing person recording system using the same | |
| JP3664784B2 (en) | Wide area monitoring device | |
| KR102142315B1 (en) | ATM security system based on image analyses and the method thereof | |
| Wang et al. | A video analysis framework for soft biometry security surveillance | |
| KR20190072323A (en) | Image Monitoring System and Method for Monitoring Image | |
| Kang et al. | Research on intelligent visual surveillance for public security |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20100127 |