CN100585656C - An all-weather intelligent video analysis monitoring method based on a rule - Google Patents
An all-weather intelligent video analysis monitoring method based on a rule Download PDFInfo
- Publication number
- CN100585656C CN100585656C CN200710064387A CN200710064387A CN100585656C CN 100585656 C CN100585656 C CN 100585656C CN 200710064387 A CN200710064387 A CN 200710064387A CN 200710064387 A CN200710064387 A CN 200710064387A CN 100585656 C CN100585656 C CN 100585656C
- Authority
- CN
- China
- Prior art keywords
- rule
- target
- analysis
- classification
- obtains
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000012544 monitoring process Methods 0.000 title claims abstract description 29
- 238000001514 detection method Methods 0.000 claims abstract description 24
- 230000011218 segmentation Effects 0.000 claims abstract description 4
- 206010000117 Abnormal behaviour Diseases 0.000 claims description 23
- 230000033001 locomotion Effects 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 11
- 238000012549 training Methods 0.000 claims description 9
- 206010017577 Gait disturbance Diseases 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 3
- 230000000877 morphologic effect Effects 0.000 claims description 3
- 150000001875 compounds Chemical class 0.000 abstract description 4
- 230000006399 behavior Effects 0.000 description 14
- 230000000694 effects Effects 0.000 description 10
- 230000002547 anomalous effect Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 5
- 230000035945 sensitivity Effects 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 230000005856 abnormality Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000005286 illumination Methods 0.000 description 3
- 238000012546 transfer Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000005352 clarification Methods 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000001960 triggered effect Effects 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 206010003830 Automatism Diseases 0.000 description 1
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 description 1
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000001149 cognitive effect Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Abstract
The invention discloses an all-weather intelligent video analysis monitoring method based on a regulation, which can be used in the intelligent video analysis surveillance method based on a regulation for the indoors, outdoor paths and the outdoors, and the steps includes: carrying out background segmentation with an image sequence acquired by a camera so as to obtain accurate foreground, carrying out object detection with the obtained foreground so as to obtain the object to be monitored, tracking the detected object so as to obtain the trajectory of the object, then carrying out trajectory analysis with the obtained trajectory, meantime carrying out object identification with the detected object so as to obtain the type of the object, then carrying out judgment with the obtained trajectory analysis result and the type of the object according to a preset alarming regulation composed of object type, object action, time and location and a compound regulation composed by simple regulations, so as to obtain the results of whether it alarms or not and alarming in which type.
Description
Technical field
The present invention relates to a kind of method for supervising that abnormal conditions are automatically reported to the police, thus especially under all weather conditions by to camera acquisition to video carry out the method for supervising that rule-based intellectual analysis realizes being provided with warning object, scalable warning sensitivity.
Background technology
Develop rapidly along with modern science and technology, utilize video camera to monitor the every aspect that dynamic scene is widely used in modern society already, particularly those are to the occasion of safety requirements sensitivity, and are along the line etc. as national defence, community, bank, parking lot, military base, cable.The vision monitoring of dynamic scene is the forward position research direction that receives much concern in recent years, and its detection from the video camera sequences of images captured, identification, tracking target are also understood its behavior.Although the present rig camera that extends as human vision ubiquity in commerce is used is not given full play to its initiatively effect of supervision media in real time.Therefore, develop automatism with practical significance, intelligent visual monitor system becomes urgent and necessary day by day.This just requires and can not only replace human eye with video camera, and the general-purpose computers contributor, replaces the people, monitors or control task to finish.
At present, traditional video monitoring system arrives the terminal monitoring chamber with the video transmission that camera acquisition arrives, finally observe and realize the purpose monitored transmitting the video that comes by the security personnel, these traditional visual monitor systems all are a kind of monitoring afterwards, the behavior of various breach securitys are not had the effect of prevention; On the other hand, these supervisory systems mainly rely on the display of security personnel's real-time monitored video camera, and all unscheduled events and abnormal conditions all are to be found by the security personnel, and doing like this needs great amount of manpower and material resources.In this case, people's physiologic factor can greatly influence the effect of monitoring such as notice.Along with the increase of people to security requirement, and the increase of the complicacy of monitor task, these traditional supervisory systems can not adapt to demands of social development.
Thereby the detection from the video camera sequences of images captured of known video monitoring system, identification, tracking target are also carried out the purpose that analysis and understanding reaches monitoring to its behavior, its technical essential comprises: moving object detection and tracking, the classification of target, the motion analysis of people, vehicle and other monitoring objectives, behavior are understood these aspects.Representative product has the SmartCatch of Vidient and the Nextiva of Verint.These method for supervising all are that test is used under indoor or outdoor road environment, can detect target more effectively when illumination is good, can discern people, car, can carry out analysis and understanding to simple behavior in tracking target under the simple scenario.
But a lot of practical application scenes are unsatisfactory.At first, most application request are carried out round-the-clock monitoring to object scene, and can't provide lighting condition as daytime at night.Secondly, national defence, military base, the cable place of needing reliable intelligent video monitoring badly such as along the line mostly is in the field.Be different from indoor and road environment, the field environment complexity, brightness changes greatly, and interference source quantity is many and jamming pattern is complicated.General video monitoring system can't realize detection, identification, tracking target in these cases effectively and analysis and understanding is carried out in its behavior, can not effectively monitor object scene.
Summary of the invention
Can not carry out round-the-clock effective monitoring to object scene in order to overcome existing video monitoring system, can not effectively be applied to the deficiency of complex scenes such as field, the present invention seeks under all weather conditions, not only can monitor common indoor and outdoor road scene, and can be applied to open-air scene and monitor, for this reason, provide a kind of rule-based all-weather intelligent video analysis monitoring method.
In order to realize described purpose, rule-based all-weather intelligent video analysis monitoring method of the present invention, step is as follows:
Background segment step S1: to camera collection to image sequence carry out background segment, be used to obtain correct prospect;
Target detection step S2: the prospect that obtains is carried out target detection, be used to obtain the object that to monitor;
Target following step S3: detected object is followed the tracks of, be used to obtain the track of object;
Trajectory analysis step S4: the track that obtains is carried out trajectory analysis;
Target Recognition step S5: the object that detection is obtained carries out Target Recognition simultaneously, is used to obtain the classification of object;
Abnormal behaviour detects step S6: according to the alarm rule that pre-establishes the trajectory analysis result and the object type that obtain judged, thus the output result who whether is reported to the police and report to the police in which way.
Particularly, its background segment comprises the steps:
Step S11: at first make up background model by the image sequence that collects;
Step S12: compare by current input image and background model, be used for the acquisition prospect;
Step S13: prospect process morphologic filtering and connection component analysis to obtaining are used to obtain the profile segmentation result.
Particularly, its target detection comprises the steps:
Step S21: the result who is obtained by background segment carries out the thresholding processing, is used to obtain candidate target;
Step S22: to the candidate target that obtains, statistical nature and method of motion analysis in the combining image detect target.
Particularly, its target following comprises the steps:
Step S31: at first ask for and detect the gained clarification of objective;
Step S32: adopt the method for statistics that the motion of target is predicted, obtain the motion prediction result of target;
Step S33: to predict the outcome with new image in detect the gained clarification of objective and mate, be used for obtaining the movement locus of image sequence target.
Particularly, its trajectory analysis comprises the steps:
Step S41: the trace information that is obtained by tracking carries out cluster, is used to obtain track classification and sequential relationship;
Step S42: utilize the sequential relationship between the track classification to obtain rule;
Step S43: track is analyzed by the rule that obtains.
Particularly, its Target Recognition comprises the steps:
Step S51: at first will train a group model, the training data that is comprised positive negative sample by the N group obtains a group model storehouse and respective classified device by training algorithm;
Step S52: will detect gained target input category device;
Step S53:, obtain the classification of target according to the model bank classification that training obtains.
Particularly, its output of reporting to the police comprises the steps:
Step S61: by rule-based abnormal behaviour analysis, if find to have abnormal behaviour, then trigger dissimilar warnings according to dissimilar application scenarioss and abnormal behaviour, and export the image and the text description of current warning simultaneously by the strategy that pre-establishes.
Particularly, its alarm rule formulation is that a rule is made of four factors:
The classification that the target classification is obtained by Target Recognition;
Goal behavior is determined through trajectory analysis by the track that target following obtains;
The place comprises: the zone and the line of stumbling;
Time is to preestablish.
Particularly, described alarm rule adopts two or more alarm rule to constitute a combinde alarms rule by sequential relationship.
Particularly, described alarm rule adopts an abnormal behaviour by an alarm rule or a combinde alarms rule description.
Particularly, described alarm rule when an alarm rule or a combinde alarms rule are met, are then judged abnormal behaviour has been taken place.
The invention has the beneficial effects as follows, the present invention is by rule-based intelligent video analysis, cut apart in conjunction with multiple, the detection and tracking technology, the correct differentiation and the identification of various abnormal behaviours have been solved, eliminated by under all weather conditions with complex scene in the interference that brings of bad condition, not only can carry out round-the-clock monitoring to common indoor and outdoor road scene, and can be applied to complex scenes such as field and carry out round-the-clock monitoring, the accurate more and robust of monitoring, can expand the warning object by adding new rule, the warning object can be set, the scalable warning sensitivity, implement, dispose simple and conveniently, can monitor complex behavior.
Description of drawings
Fig. 1 is a system chart of the present invention.
Fig. 2 is the process flow diagram of background segment.
Fig. 3 is the process flow diagram of target following.
Fig. 4 is the process flow diagram of Target Recognition.
Fig. 5 is the user interface of first embodiment.
Fig. 6 is that the algorithm parameter of first embodiment is provided with the interface.
Fig. 7 is the exception rules configuration interface of first embodiment.
Fig. 8 is the warning example of first embodiment.
Embodiment
Below in conjunction with accompanying drawing the present invention is described in detail, be to be noted that described embodiment only is intended to be convenient to the understanding of the present invention, and it is not played any qualification effect.
The method that the present invention uses can install and carry out with the form of software on personal computer, industrial computer and server, also method can be made embedded chip and embody with the form of hardware.
Fig. 1 has provided an embodiment system chart, at first to camera collection to image sequence carry out background segment to obtain correct prospect, then the prospect that obtains is carried out the object of target detection to obtain monitoring, then detected object is followed the tracks of to obtain the track of object, then the track that obtains is carried out trajectory analysis, the object that detection is obtained carries out Target Recognition to obtain the classification of object simultaneously, according to the alarm rule that pre-establishes the trajectory analysis result and the object type that obtain are judged then, thus the result who whether is reported to the police and report to the police in which way.
Fig. 2 has provided the background segment flow process of an embodiment.Background segment among the present invention realizes as follows, at first make up background image based on the multiframe statistical model by the image sequence that collects, then carry out differential ratio to operating and carrying out binary conversion treatment, obtain the profile segmentation result through morphologic filtering and connection component analysis then by current input image and background model.Can weaken effectively under the real-time condition and eliminate satisfying like this because thereby the interference that illumination variation, weather condition etc. are brought obtains robust and profile more accurately.
Target detection among the present invention realizes in the following way.Statistical nature and method of motion analysis that the result who is obtained by background segment carries out in processing of difference thresholding and the combining image detect target.Can effectively utilize the advantage of difference threshold method like this, overcome its deficiency.
Fig. 3 has provided the target following flow process of an embodiment.Target following among the present invention realizes as follows, at first ask for the centroid feature that detects the gained target, then adopt the method for statistics that its motion is predicted, predict by the maximum a posteriori criterion, gained predict the outcome with new image in detect the gained target and carry out outline, obtain the movement locus of target in the image sequence, can get rid of wrong report like this, the stability-of-path that obtains and accurate.
Trajectory analysis among the present invention is realized as follows.At first will train rule, the trace information that is obtained by tracking carries out cluster, and obtains series of rules in conjunction with the sequential relationship between them.Then track is analyzed by the track classification and the sequential relationship thereof that obtain.Can access the high-rise more behavioural information of monitored object by trajectory analysis, be convenient to realize complicated more and abnormal behaviour monitoring accurately.
Fig. 4 has provided the Target Recognition flow process of an embodiment.At first to train a group model, the training data that is comprised positive sample, negative sample by group people (or car, bicycle) obtains a group model storehouse N by training algorithm, then will detect gained target input category device, classify according to the model bank that training obtains, so just obtained the classification of target, N=1,2,3.......By can effectively avoiding some uncontrollable interference such as flying birds to identification of targets, the wrong report that other interference bring etc. are reported to the police for rule-based abnormal behaviour simultaneously condition precedent are provided.
Warning output among the present invention realizes as follows.By rule-based abnormal behaviour analysis, if find to have abnormal behaviour, then trigger for example dissimilar warnings such as sound, light, electricity according to dissimilar application scenarioss and abnormal behaviour by the strategy that pre-establishes, audio alert be can also realize, and the image and the text description of current warning exported simultaneously.Here abnormal conditions refer to break in the specific region, illegally are detained specific region, illegal unidirectional by specific warning line, illegal two-way by specific warning line, illegally trail people or vehicle, illegal moving object.This method combines Flame Image Process, computer vision and mode identification technology.
Fig. 5 is the main interface of an embodiment.Its effect is to be used for showing the current real-time video that collects in each road, and the target detection and the tracking results of each limit of the prohibited area and their inside are drawn out in stack thereon.When not having video, show the sign of Logo.All there is a hurdle tool button in each video window left side, can control the collection and the abnormality detection process of each road video by them.The title of each button is marked by red literal in Fig. 5, is respectively: the beginning video acquisition, algorithm parameter is set, stops/beginning abnormality detection, startup/close rule editor, adjust video parameter, stop video acquisition.The below is warning message tabulation, and its effect is to be used for each bar warning message that display routine writes down in operational process.All Alerts information record is arranged by the backward according to Time To Event, and promptly last detected anomalous event is come the tabulation top.The meaning of each row is as follows in this tabulation:
1. Case Number: since 1, the Case Number that increases progressively of order, delete indivedual warning message records and can not change the numbering of arrival event thereafter, but can reset to it by " emptying tabulation ", thereby the warning message that this operation back is arrived writes down again since 1 serial number.
2. scene numbering: since 1, the scene numbering that increases progressively of order, this numbering is corresponding with each road video, has identified every warning message and has write down and come from which video scene.The video that shows in the video display window in the upper left corner is that No. 2, the lower left corner are that No. 3, the lower right corner are No. 4 corresponding to No. 1 scene, the upper right corner.
3. zone number: since 1, the zone number that increases progressively of order,, identified the anomalous event that every warning message write down and specifically occurred in which forbidden zone in the video scene corresponding to different forbidden zone in the video of same road.When the rule of checking a certain road video correspondence in the abnormality detection rule editor was set, the zone that is presented at the zone list top was No. 1 forbidden zone, is No. 2 forbidden zones under it, and the rest may be inferred for all the other.
4. event time: show the time that every corresponding anomalous event of warning message record takes place, with computer system time at program place be benchmark, its form be " [year]; [moon], [day] [time]: [branch]: [second] ", wherein the time shows with 4 bit digital, all the other are 2, adopt 24 hours systems.
5. anomalous event is described: shows that every corresponding anomalous event of warning message record describes, different forms is arranged corresponding to different rule (regional or stumble line),
1) zone, its general form is " target [' entering '/' leaving '] " [scene description] " scene " [region description] " zone ", wherein " scene description " and " region description " these two needs are by manually being provided with in the exception monitoring rule editor.
2) line of stumbling, its general form is " target [' from left to right '/' from right to left '] " [scene description] " scene " [region description] " line of stumbling ", wherein " scene description " and " region description " these two needs are by manually being provided with in the exception monitoring rule editor.
Fig. 6 is that the algorithm parameter of an embodiment is provided with the interface, adjusts the parameter of algorithm that various piece is used among Fig. 1 by it.The some of them parameter need be adjusted according to practical application:
1. image dimensionality reduction yardstick: in order to save system resource, present embodiment at first carries out down-sampled before video is analyzed to the image that collects according to this parameter value, more detected coordinates of targets be multiply by this value afterwards and obtain the actual coordinate of target in original image.As this parameter value in dialog box shown in Figure 6 is default value 2, suppose that this moment, the video size of input was 320 * 240 pixels, then we are at first down-sampled to 160 * 120 pixels with picture size before processing, if detect at point (80 this moment, 60) having located an area is 100 pixel targets, then the actual coordinate of this target in original image is (160,120), real area is 400 pixels.This parameter value is turned down computational accuracy when can the raising program carrying out target detection, system is easier to detect less moving target thereby make, but also can consume more system resource (especially computational resource), thereby be subjected to noise effect and more wrong report occurs simultaneously also easilier; Heighten this value and then can obviously reduce system resources consumption, reduce the wrong report that causes by picture noise, but can cause the program precise decreasing simultaneously, thereby make it be difficult to find the moving target far away apart from video camera.
2. minimum target size: in current scene, any area (pixel quantity) will be by as noise less than the target of this value.Turn this value down and can make program detect littler target, but also can cause wrong report to increase; Otherwise, also can make program ignore less moving target simultaneously if transfer big this value then can reduce wrong report.Notice that said " target sizes " is meant the size of target in original image here, therefore this parameter be set to less than " image dimensionality reduction yardstick " square value be nonsensical.
3. the maximum frame number of lose objects: frame number that continuously can't be detected when a target is confirmed to be during greater than this parameter value and loses.Transfer big this value to reduce because target is lost the track that causes in individual frame interrupts, but be subjected to noise effect and produce the mistake announcement easilier; Otherwise then help to reduce wrong report, the complete target trajectory that also will be difficult to obtain simultaneously.
4. the initial frame number of target: the frame number that only is detected continuously in scene just is confirmed to be an effective moving target greater than the target of this value.The purpose that this parameter is set is in order to reduce because the wrong report that picture noise causes.Because picture noise randomness spatially, transfer big this value can obviously reduce the wrong report that causes because of picture noise, but the time delay also can increase target detection the time, vice versa.
5. maximum target size: any area (pixel quantity) will be ignored by system equally greater than the target of this value in the current scene.In some applications, some zones are forbidden zones to personnel, but allow vehicle to enter, and can realize this goal this moment by adjusting this value.This parameter also has an important effect: anti-locking system produces a large amount of wrong reports when video scene cataclysm (causing the image that illumination condition changes, the color commentaries on classics ash of video camera brings to change fast as switch lamp).This parameter also is at the size of target in original image.
6. the mutant proportion factor: the effect of this parameter is to misrepresent deliberately in a large number for anti-locking system produces when the video scene cataclysm equally.Its meaning is: if current detection to the ratio of all moving target area sums (pixel quantity) and image area be worth greater than this, then all targets all are left in the basket.
7. Fig. 7 is the regular configuration interface of an embodiment.Rule is formulated as follows.A rule is made of four factors: target classification, goal behavior, place and time.Wherein, the target classification refers to the classification that the Target Recognition by the front obtains; The decision that the track that goal behavior is obtained by target following is made by analysis; The place is pre-set comprising the zone and stumble line; Time is predefined.Two or more rules can constitute a compound rule by sequential relationship.Abnormal behaviour is described by a rule or a compound rule, just thinks when a rule or a compound rule are met abnormal behaviour has taken place.The cognitive pattern that like this description of abnormal behaviour is met the people can the configuration monitoring object, time, place and behavior, can regulate monitoring sensitivity, can robust detect various types of abnormal behaviours, can realize monitoring to complicated abnormal behaviour.Can set rule effective time, the zone or the line of stumbling, object size here.
The rule that breaks in the specific region is set to:
Place: zone; Target classification: people; Time: all; Goal behavior: swarm into;
Illegal rule of being detained the specific region is set at:
Place: zone; Target classification: people; Time: all; Goal behavior: be detained;
Illegal unidirectional rule by specific warning line is set at:
Place: the line of stumbling; Target classification: people; Time: all; Goal behavior: cross over;
Above rule be compounded to form more complex rule.
Fig. 8 has provided the warning example of an embodiment.Present embodiment is to descend with a snow day weather, and the boundary defence of sub-district is reported to the police and is example.Thereby there is a people to attempt to cross enclosure wall and triggered warning in the snow sky.Wherein red area is the pre-alarm zone, and blue region is an alarm region.The top of figure has provided the details of reporting to the police, and comprises Case Number, area type, event time and anomalous event description.Give the video image when reporting to the police among the figure, the time and date of image upper right side display system.Scene is one enclosure wall and by enclosure wall vacant lot divided into two parts among the figure, is provided with two warning lines along edge wall, along enclosure wall side and ground are provided with two warning regions nearby.The warning region that enters the below when suspicious object is just detected and follows the tracks of the pre-alarm of having set out simultaneously in the near future by system, triggered warning so detected by warning line when suspicious object prepares to cross enclosure wall.That computing machine adopts is PIV2.0G, the Dell brand name computer of internal memory 512M.
The above; only be the embodiment among the present invention; but protection scope of the present invention is not limited thereto; anyly be familiar with the people of this technology in the disclosed technical scope of the present invention; can understand conversion or the replacement expected; all should be encompassed in of the present invention comprising within the scope, therefore, protection scope of the present invention should be as the criterion with the protection domain of claims.
Claims (8)
1. rule-based all-weather intelligent video analysis monitoring method may further comprise the steps:
Background segment step S1: to camera collection to image sequence carry out background segment, be used to obtain correct prospect;
Target detection step S2: the prospect that obtains is carried out target detection, be used to obtain the object that to monitor;
Target following step S3: detected object is followed the tracks of, be used to obtain the track of object;
Trajectory analysis step S4: the track that obtains is carried out trajectory analysis;
Target Recognition step S5: the object that when carrying out target following and trajectory analysis detection is obtained carries out Target Recognition, is used to obtain the classification of object;
Abnormal behaviour detects step S6: according to the alarm rule that pre-establishes the trajectory analysis result and the object type that obtain judged, thus the output result who whether is reported to the police and report to the police in which way; It is characterized in that;
Its target detection comprises the steps:
Step S21: the result who is obtained by background segment carries out the thresholding processing, is used to obtain candidate target;
Step S22: to the candidate target that obtains, statistical nature in the combining image and method of motion analysis detect the object that will monitor;
Its trajectory analysis comprises the steps:
Step S41: the trace information that is obtained by tracking carries out cluster, be used to obtain the track classification and the time cun order relation;
Step S42: utilize the sequential relationship between the track classification to obtain rule;
Step S43: track is analyzed by the rule that obtains;
Its Target Recognition comprises the steps:
Step S51: at first will train a group model, the training data that is comprised positive negative sample by the N group obtains a group model storehouse and respective classified device by training algorithm;
Step S52: will detect gained object input category device;
Step S53:, obtain the classification of object according to the model bank classification that training obtains.
2. method for supervising according to claim 1 is characterized in that its background segment comprises the steps:
Step S11: at first make up background model by the image sequence that collects;
Step S12: compare by current input image and background model, be used for the acquisition prospect;
Step S13: prospect process morphologic filtering and connection component analysis to obtaining are used to obtain the profile segmentation result.
3. method for supervising according to claim 1 is characterized in that its target following comprises the steps:
Step S31: at first ask for the feature that detects the gained object;
Step S32: adopt the method for statistics to predict, obtain to detect the gained motion of objects and predict the outcome to detecting the gained motion of objects;
Step S33: to predict the outcome with new image in detect the gained object feature mate, be used for obtaining image sequence and detect gained motion of objects track.
4. method for supervising according to claim 1 is characterized in that, its output of reporting to the police comprises the steps:
Step S61: by abnormal behaviour analysis based on alarm rule, if find to have abnormal behaviour, then trigger dissimilar warnings according to dissimilar application scenarioss and abnormal behaviour, and export the image and the text description of current warning simultaneously by the strategy that pre-establishes.
5. method for supervising according to claim 1 is characterized in that, it is that a rule is made of four factors that its alarm rule is formulated:
The classification that the target classification is obtained by Target Recognition;
Goal behavior is determined through trajectory analysis by the track that target following obtains;
The place comprises: the zone and the line of stumbling;
Time is to preestablish.
6. method for supervising according to claim 5 is characterized in that, adopts two or more alarm rule to constitute a combinde alarms rule by sequential relationship.
7. method for supervising according to claim 6 is characterized in that, adopts an abnormal behaviour by an alarm rule or a combinde alarms rule description.
8. method for supervising according to claim 6 is characterized in that, when an alarm rule or a combinde alarms rule are met, then judge abnormal behaviour has taken place.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN200710064387A CN100585656C (en) | 2007-03-14 | 2007-03-14 | An all-weather intelligent video analysis monitoring method based on a rule |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN200710064387A CN100585656C (en) | 2007-03-14 | 2007-03-14 | An all-weather intelligent video analysis monitoring method based on a rule |
Publications (2)
Publication Number | Publication Date |
---|---|
CN101266710A CN101266710A (en) | 2008-09-17 |
CN100585656C true CN100585656C (en) | 2010-01-27 |
Family
ID=39989098
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN200710064387A Expired - Fee Related CN100585656C (en) | 2007-03-14 | 2007-03-14 | An all-weather intelligent video analysis monitoring method based on a rule |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN100585656C (en) |
Families Citing this family (55)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101765025A (en) * | 2008-12-23 | 2010-06-30 | 北京中星微电子有限公司 | System for abnormal detection of surveillance camera and method thereof |
CN101436337B (en) * | 2008-12-23 | 2013-12-11 | 北京中星微电子有限公司 | Method and apparatus for monitoring event |
CN101448145A (en) * | 2008-12-26 | 2009-06-03 | 北京中星微电子有限公司 | IP camera, video monitor system and signal processing method of IP camera |
CN101799968B (en) * | 2010-01-13 | 2013-06-05 | 李秋华 | Detection method and device for oil well intrusion based on video image intelligent analysis |
CN102236783A (en) * | 2010-04-29 | 2011-11-09 | 索尼公司 | Method and equipment for detecting abnormal actions and method and equipment for generating detector |
CN101883261B (en) * | 2010-05-26 | 2012-12-12 | 中国科学院自动化研究所 | Method and system for abnormal target detection and relay tracking under large-range monitoring scene |
CN102073844A (en) * | 2010-11-10 | 2011-05-25 | 无锡中星微电子有限公司 | Intelligent monitoring system and method |
CN102014274B (en) * | 2010-11-10 | 2016-03-23 | 深圳市海云天科技股份有限公司 | A kind of examination monitoring system and monitor message processing method thereof |
US9226037B2 (en) * | 2010-12-30 | 2015-12-29 | Pelco, Inc. | Inference engine for video analytics metadata-based event detection and forensic search |
CN102547244A (en) * | 2012-01-17 | 2012-07-04 | 深圳辉锐天眼科技有限公司 | Video monitoring method and system |
CN103426180B (en) * | 2012-05-24 | 2017-02-08 | 信帧电子技术(北京)有限公司 | Video wire mixing detection method |
CN103426181A (en) * | 2012-05-25 | 2013-12-04 | 信帧电子技术(北京)有限公司 | Automatic frame rate changing wire-mixing detection method |
CN102938078A (en) * | 2012-11-30 | 2013-02-20 | 北京理工大学 | Unbalanced timing sequence-based monitoring video abnormal event detection method |
CN103856344B (en) * | 2012-12-05 | 2017-09-15 | 中国移动通信集团北京有限公司 | A kind of alarm event information processing method and device |
CN103116959A (en) * | 2013-01-25 | 2013-05-22 | 上海博超科技有限公司 | Analyzing and recognizing method for abnormal behaviors in intelligent videos |
CN104299206B (en) * | 2013-07-16 | 2017-03-22 | 浙江大华技术股份有限公司 | Image processing method and image processing device |
CN103530605B (en) * | 2013-09-29 | 2017-01-25 | 数基科技(北京)有限公司 | Method for detecting abnormal behavior of robustness |
CN103544806B (en) * | 2013-10-31 | 2016-01-06 | 江苏物联网研究发展中心 | The valuable cargo haulage vehicle monitoring early-warning system of line rule of stumbling based on video |
DE102014208524A1 (en) * | 2014-05-07 | 2015-11-12 | Robert Bosch Gmbh | LOCAL TRANSPORTATION ANALYSIS WITH DETECTION OF A TRAFFIC PATH |
CN105450918B (en) | 2014-06-30 | 2019-12-24 | 杭州华为企业通信技术有限公司 | Image processing method and camera |
CN104270608A (en) * | 2014-09-28 | 2015-01-07 | 武汉烽火众智数字技术有限责任公司 | Intelligent video player and playing method thereof |
CN109089077A (en) * | 2014-10-09 | 2018-12-25 | 中控智慧科技股份有限公司 | A kind of method remotely monitored and monitoring client |
CN104469299A (en) * | 2014-12-02 | 2015-03-25 | 柳州市瑞蚨电子科技有限公司 | Network camera shooting device |
CN104519322B (en) * | 2014-12-24 | 2017-08-18 | 芜湖林一电子科技有限公司 | A kind of machine vision Target Tracking System |
CN104866538A (en) * | 2015-04-30 | 2015-08-26 | 北京海尔广科数字技术有限公司 | Method, network and system of dynamic update semantic alarm database |
CN106303403B (en) * | 2015-06-12 | 2019-03-12 | 中国人民公安大学 | Monitoring device presetting bit setting, switching method and system |
KR102334767B1 (en) * | 2015-07-23 | 2021-12-02 | 한화테크윈 주식회사 | Apparatus and method for controlling network camera |
CN105678264A (en) * | 2016-01-07 | 2016-06-15 | 周丽娜 | Parking lot management system excellent in profile identification and filtering performances |
CN109889794A (en) * | 2016-08-30 | 2019-06-14 | 吴玉芳 | Can recognition of face video monitoring intelligence Skynet system and its working method |
CN107066929B (en) * | 2017-01-06 | 2021-06-08 | 重庆大学 | Hierarchical recognition method for parking events of expressway tunnel integrating multiple characteristics |
CN107547865A (en) * | 2017-07-06 | 2018-01-05 | 王连圭 | Trans-regional human body video frequency object tracking intelligent control method |
CN109255360B (en) * | 2017-07-12 | 2021-04-02 | 杭州海康威视数字技术股份有限公司 | Target classification method, device and system |
CN108734910A (en) * | 2018-03-07 | 2018-11-02 | 广州图普网络科技有限公司 | A kind of abnormal behaviour monitoring and warning system and method |
CN108717521A (en) * | 2018-04-17 | 2018-10-30 | 智慧互通科技有限公司 | A kind of parking lot order management method and system based on image |
CN108447208B (en) * | 2018-05-17 | 2019-10-29 | 普联技术有限公司 | Light warning method, equipment, web camera and storage medium |
CN108921086A (en) * | 2018-06-29 | 2018-11-30 | Oppo广东移动通信有限公司 | Image processing method and device, storage medium, electronic equipment |
CN109410496B (en) * | 2018-10-25 | 2022-04-01 | 北京交通大学 | Intrusion early warning method and device and electronic equipment |
CN109712169A (en) * | 2018-11-15 | 2019-05-03 | 上海卫星工程研究所 | Moving-target motion profile prediction technique and method for autonomous tracking based on EO-1 hyperion |
CN109951682A (en) * | 2019-02-25 | 2019-06-28 | 广东协安机电工程有限公司 | A kind of bank safety video monitoring system |
CN110111518A (en) * | 2019-06-06 | 2019-08-09 | 厦门钛尚人工智能科技有限公司 | A kind of dedicated destruction recognizer of venue |
CN110223474A (en) * | 2019-07-04 | 2019-09-10 | 无锡睿勤科技有限公司 | A kind of intelligent monitoring and alarming method, system and storage medium |
CN110490125B (en) * | 2019-08-15 | 2023-04-18 | 成都睿晓科技有限公司 | Oil filling area service quality detection system based on gesture automatic detection |
CN110659627A (en) * | 2019-10-08 | 2020-01-07 | 山东浪潮人工智能研究院有限公司 | Intelligent video monitoring method based on video segmentation |
CN110782639A (en) * | 2019-10-28 | 2020-02-11 | 深圳奇迹智慧网络有限公司 | Abnormal behavior warning method, device, system and storage medium |
CN110929591A (en) * | 2019-11-04 | 2020-03-27 | 广州供电局有限公司 | Transformer substation monitoring method and device, computer equipment and storage medium |
CN111191498A (en) * | 2019-11-07 | 2020-05-22 | 腾讯科技(深圳)有限公司 | Behavior recognition method and related product |
CN111553305B (en) * | 2020-05-09 | 2023-05-05 | 中国石油天然气集团有限公司 | System and method for identifying illegal videos |
CN112052719A (en) * | 2020-07-16 | 2020-12-08 | 国网浙江省电力有限公司 | Personnel behavior analysis method and system based on image feature recognition |
CN111883247B (en) * | 2020-07-29 | 2022-03-15 | 复旦大学 | Analysis system for correlation between behavior data and medical outcome |
CN112084968B (en) * | 2020-09-11 | 2023-05-26 | 清华大学 | Semantic characterization method and system based on air monitoring video and electronic equipment |
CN112329691A (en) * | 2020-11-17 | 2021-02-05 | 珠海大横琴科技发展有限公司 | Monitoring video analysis method and device, electronic equipment and storage medium |
CN112800975A (en) * | 2021-01-29 | 2021-05-14 | 中国船舶重工集团公司第七一三研究所 | Behavior identification method in security check channel based on image processing |
CN113223261B (en) * | 2021-04-30 | 2022-09-30 | 重庆天智慧启科技有限公司 | System and method for deploying defense of residential area perimeter |
CN114160447B (en) * | 2021-11-20 | 2024-02-06 | 中云智慧(北京)科技有限公司 | Early machine inspection system and method |
CN116612594A (en) * | 2023-05-11 | 2023-08-18 | 深圳市云之音科技有限公司 | Intelligent monitoring and outbound system and method based on big data |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1426020A (en) * | 2001-12-13 | 2003-06-25 | 中国科学院自动化研究所 | Far distance identity identifying method based on walk |
CN1738426A (en) * | 2005-09-09 | 2006-02-22 | 南京大学 | Video motion goal division and track method |
US20060078047A1 (en) * | 2004-10-12 | 2006-04-13 | International Business Machines Corporation | Video analysis, archiving and alerting methods and apparatus for a distributed, modular and extensible video surveillance system |
CN1852428A (en) * | 2006-05-25 | 2006-10-25 | 浙江工业大学 | Intelligent tunnel safety monitoring apparatus based on omnibearing computer vision |
-
2007
- 2007-03-14 CN CN200710064387A patent/CN100585656C/en not_active Expired - Fee Related
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1426020A (en) * | 2001-12-13 | 2003-06-25 | 中国科学院自动化研究所 | Far distance identity identifying method based on walk |
US20060078047A1 (en) * | 2004-10-12 | 2006-04-13 | International Business Machines Corporation | Video analysis, archiving and alerting methods and apparatus for a distributed, modular and extensible video surveillance system |
CN1738426A (en) * | 2005-09-09 | 2006-02-22 | 南京大学 | Video motion goal division and track method |
CN1852428A (en) * | 2006-05-25 | 2006-10-25 | 浙江工业大学 | Intelligent tunnel safety monitoring apparatus based on omnibearing computer vision |
Non-Patent Citations (1)
Title |
---|
人运动的视觉分析综述. 王亮,胡卫明,谭铁牛.计算机学报,第25卷第3期. 2002 * |
Also Published As
Publication number | Publication date |
---|---|
CN101266710A (en) | 2008-09-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN100585656C (en) | An all-weather intelligent video analysis monitoring method based on a rule | |
CN105744232B (en) | A kind of method of the transmission line of electricity video external force damage prevention of Behavior-based control analytical technology | |
CN107229894B (en) | Intelligent video monitoring method and system based on computer vision analysis technology | |
CN101587622B (en) | Forest rocket detecting and identifying method and apparatus based on video image intelligent analysis | |
CN201773466U (en) | Video monitoring and pre-warning device for detecting, tracking and identifying object detention/stealing event | |
CN100555348C (en) | Intelligent video monitoring system of bank self-aid apparatus | |
CN101123721A (en) | An intelligent video monitoring system and its monitoring method | |
CN101727672A (en) | Method for detecting, tracking and identifying object abandoning/stealing event | |
US8301577B2 (en) | Intelligent monitoring system for establishing reliable background information in a complex image environment | |
CN102164270A (en) | Intelligent video monitoring method and system capable of exploring abnormal events | |
JP2013218679A (en) | Video-based detection device and notification device for catching short-time parking violation | |
CN201114536Y (en) | A video monitoring system | |
CN101751677A (en) | Target continuous tracking method based on multi-camera | |
CN102665071A (en) | Intelligent processing and search method for social security video monitoring images | |
CN103456024B (en) | A kind of moving target gets over line determination methods | |
CN111523397B (en) | Intelligent lamp post visual identification device, method and system and electronic equipment thereof | |
KR20060031832A (en) | A smart visual security system based on real-time behavior analysis and situation cognizance | |
CN111144291A (en) | Method and device for distinguishing personnel invasion in video monitoring area based on target detection | |
CN110867046A (en) | Intelligent car washer video monitoring and early warning system based on cloud computing | |
CN106355162A (en) | Method for detecting intrusion on basis of video monitoring | |
CN101571982A (en) | Method for judging stolen articles in video monitoring range | |
CN203327163U (en) | Monitoring system | |
Choeychuen | Available car parking space detection from webcam by using adaptive mixing features | |
CN102244769B (en) | Object and key person monitoring system and method thereof | |
CN107657260A (en) | The detection method of ground hidden danger in a kind of transmission line of electricity surrounding enviroment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20100127 Termination date: 20170314 |
|
CF01 | Termination of patent right due to non-payment of annual fee |