CN103810495B - Barrier gate safe anti-hitting method - Google Patents
Barrier gate safe anti-hitting method Download PDFInfo
- Publication number
- CN103810495B CN103810495B CN201310676400.XA CN201310676400A CN103810495B CN 103810495 B CN103810495 B CN 103810495B CN 201310676400 A CN201310676400 A CN 201310676400A CN 103810495 B CN103810495 B CN 103810495B
- Authority
- CN
- China
- Prior art keywords
- point
- background
- feature point
- sift feature
- sift
- 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 50
- 230000004888 barrier function Effects 0.000 title abstract description 13
- 239000013598 vector Substances 0.000 claims abstract description 26
- 238000006243 chemical reaction Methods 0.000 claims description 6
- 238000001914 filtration Methods 0.000 claims description 6
- 238000009825 accumulation Methods 0.000 claims description 5
- 238000001514 detection method Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 description 10
- 230000008859 change Effects 0.000 description 9
- 238000012217 deletion Methods 0.000 description 8
- 230000037430 deletion Effects 0.000 description 8
- 230000000284 resting effect Effects 0.000 description 8
- 238000010168 coupling process Methods 0.000 description 6
- 238000005859 coupling reaction Methods 0.000 description 6
- 230000000903 blocking effect Effects 0.000 description 5
- 238000003860 storage Methods 0.000 description 5
- 230000008878 coupling Effects 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 238000007639 printing Methods 0.000 description 4
- 230000000630 rising effect Effects 0.000 description 4
- 230000003068 static effect Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000000670 limiting effect Effects 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 230000036961 partial effect Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000002829 reductive effect Effects 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000006698 induction Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 239000013589 supplement Substances 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 238000013519 translation Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 description 1
- 239000010931 gold Substances 0.000 description 1
- 229910052737 gold Inorganic materials 0.000 description 1
- 230000035800 maturation Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
Landscapes
- Image Analysis (AREA)
Abstract
The embodiment of the invention discloses a barrier gate safe anti-hitting method and device. The barrier gate safe anti-hitting method and device is used for detecting an obstruction in a target area of a barrier gate and instructs the barrier gate controller to complete corresponding operation according to whether the obstruction exists. The barrier gate safe anti-hitting method comprises obtaining current image features of the target area, wherein the target area is the target area of a barrier gate base; performing SIFT (scale-invariant feature transform) feature point matching on the current image features and background image features and instructing the barrier gate controller to stop barrier gate movement if the SIFT feature point failing in the matching is larger than a first threshold value, wherein the background image features are image features of a target on the barrier gate base and comprise every SIFT feature point in the target area and feature vectors corresponding to the SIFT feature points.
Description
Technical field
The present invention relates to security monitoring field, more particularly to a kind of banister safety pounding-proof method.
Background technology
Even to this day, banister has been widely used in the effective control that parking lot gateway comes in and goes out as vehicle.Banister
Use pattern be all " bar of a car one, car is crossed bar and fallen ".The restrictor bar of banister can exist when carrying out barrier gate have under restrictor bar car or
The situation that person pedestrian passes through.In order to avoid car is accidentally injured with pedestrian, the safety of banister is improved, typically all can be using protection dress
Put.
Traditional electric gate prevents that it is to install ground induction coil by the underface of the restrictor bar in banister to pound car scheme, works as vehicle
Can be detected by ground induction coil into the region, be stopped by gate controller or lift restrictor bar to avoid vehicle from being pounded.But ground
Sense coil cannot detect pedestrian, therefore be easily caused pedestrian and pounded by mistake, reduce safety.
The content of the invention
A kind of banister safety pounding-proof method is embodiments provided, for blocking in the target area for detecting banister
Thing, and according to whether there is shelter indicates that gate controller is operated accordingly.
Banister safety pounding-proof method provided in an embodiment of the present invention, including：
The present image feature of target area is obtained, the target area is the target area of banister base；
The present image feature and background image feature are carried out into scale invariant feature conversion SIFT feature Point matching, if
The SIFT feature point that it fails to match is more than first threshold, it indicates that gate controller stops banister motion；The background image is special
The characteristics of image of the target for banister base is levied, the background image feature includes：Each SIFT feature point in target area,
And the corresponding characteristic vector of the SIFT feature point.
Optionally, if the SIFT feature point that it fails to match is less than or equal to first threshold, it indicates that gate controller is maintained
The normal condition of banister.
Optionally, the present image feature of the acquisition target area, including：
The metric space extreme point of detection target area；
Accurate to determine extreme point position, filtration fraction is located at the characteristic point at edge, obtains SIFT feature point；
For SIFT feature point assigned direction parameter each described, present image feature, the present image feature bag are generated
Include：Each SIFT feature point of present image, and the corresponding characteristic vector of SIFT feature point.
Optionally, it is described that the present image feature and background image feature are carried out into SIFT feature Point matching, including：
Obtain the first SIFT feature point of present image feature；
Two second nearest with the Euclidean distance of the first SIFT feature point are searched in the background image feature
SIFT feature point；
If the second nearest SIFT feature point is less than matching threshold with the ratio of secondary the second near SIFT feature point, currently
The SIFT feature Point matching success of characteristics of image and background image；
If the second nearest SIFT feature point is more than or equal to matching threshold with the ratio of secondary the second near SIFT feature point,
Then the SIFT feature Point matching of present image feature and background image is unsuccessful.
Optionally, methods described also includes：
It is that each SIFT feature point of the background image feature arranges the initial back of the body in initial background characteristics of image
Scape weighted value, the background weight value is used to judge current SIFT feature point if appropriate for as background characteristics point.
Optionally, described each SIFT feature point for the background image feature arrange initial background weight value it
Afterwards, including：
If corresponding SIFT feature Point matching is unsuccessful in the SIFT feature point of present image feature and background image feature,
Then the background weight value of the SIFT feature point in the background image feature is subtracted one, if the SIFT in the background image feature
The current background weight value of characteristic point is less than background threshold, then delete the SIFT feature point in the background image feature；
If corresponding SIFT feature Point matching success in the SIFT feature point of present image feature and background image feature,
The background weight value of the SIFT feature point in the background image feature is added one.
Optionally, after the background weight value of the SIFT feature point by the background image feature adds one, also wrap
Include：
Judge that whether the background weight value after updating is equal to specific threshold, if if so, then SIFT feature Point matching next time
Success, maintains current background weight to be worth constant.
Optionally, described each SIFT feature point for the background image feature arrange initial background weight value it
Afterwards, including：
If the SIFT feature point without corresponding SIFT feature point Yu present image feature in current background image feature
Matched, then the SIFT feature point of present image feature is set to into pre- background characteristics point, and it is special to initialize the pre- background
Background weight value a little is levied, if in the SIFT feature Point matching after, the background weight value of the pre- background characteristics point updates
And reaching stable threshold, it is determined that the pre- background characteristics point is background characteristics point.
Optionally, it is described that the SIFT feature point of present image feature is set to after pre- background characteristics point, including：
The SIFT feature Point matching success or not of the pre- background characteristics point, is not involved in the judgement of the first threshold.
Optionally, methods described also includes：
If the SIFT feature point that it fails to match is more than first threshold, frame number is blocked in increase；If accumulation to block frame number big
In occlusion threshold, then trigger the instruction gate controller and stop banister the step of move.
As can be seen from the above technical solutions, the embodiment of the present invention has advantages below：
In embodiments of the present invention, after the present image feature for obtaining target area, by the present image feature
Scale invariant feature conversion scale invariant feature conversion (Scale-invariant feature are carried out with background image feature
Transform, SIFT) Feature Points Matching, if the SIFT feature point that it fails to match is more than first threshold, it indicates that gate controller
Stop banister motion；Because photographic head is more easily detected shelter compared with ground sensing system, therefore, the banister peace of the embodiment of the present invention
The complete anti-safety for pounding method is higher.
Description of the drawings
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to institute in embodiment
The accompanying drawing that needs are used is briefly described, it should be apparent that, drawings in the following description are only some enforcements of the present invention
Example, for those of ordinary skill in the art, on the premise of not paying creative work, can be being obtained according to these accompanying drawings
Obtain other accompanying drawings.
Fig. 1 is a schematic flow sheet of embodiment of the present invention banister safety pounding-proof method；
Fig. 2 is embodiment of the present invention banister safety pounding-proof one structural representation of device；
Fig. 3 is another schematic flow sheet of embodiment of the present invention banister safety pounding-proof method；
Fig. 4 is another schematic flow sheet of embodiment of the present invention banister safety pounding-proof method；
Fig. 5 is the logical construction schematic diagram of embodiment of the present invention banister safety pounding-proof device.
Specific embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation is described, it is clear that described embodiment is only a part of embodiment of the invention, rather than the embodiment of whole.It is based on
Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under the premise of creative work is not made
Embodiment, belongs to the scope of protection of the invention.
A kind of banister safety pounding-proof method is embodiments provided, for blocking in the target area for detecting banister
Thing, and according to whether there is shelter indicates that gate controller is operated accordingly.
Development and the maturation of Video Analysis Technology with computer vision and artificial intelligence's correlation technique, video point
Analysis technology is a kind of feasible method to solve the problems, such as safety pounding-proof.Whether it is blocked using video camera observation target, so as to
Judge whether someone is a kind of reliable method to monitor area.
Fig. 1 is referred to, one embodiment of the banister safety pounding-proof method in the embodiment of the present invention includes：
101st, the present image feature of target area is obtained；
Banister safety pounding-proof device obtains the present image feature of target area；Optionally, the present image feature bag
Include：Each SIFT feature point of present image, and the corresponding characteristic vector of SIFT feature point.
SIFT is a kind of description for image processing field, and this description has scale invariability, can be in the picture
Key point is detected, is a kind of local description.The matching capacity of SIFT feature matching algorithm is stronger, can process two width images
Between there is translation, rotation, the matching problem in the case of affine transformation, or even to a certain extent to the figure of arbitrarily angled shooting
As also possessing relatively stable characteristic matching ability.
Specifically, in actual applications, banister safety pounding-proof device can be made up of photographic head with video analysis computer.
It is understood that and the banister safety pounding-proof device in the embodiment of the present invention can be single-alone physical equipment,
Can also be the software program being carried in terminal, the description of pattern recognition device is only schematic, for example, image recognition
The division of unit in device, only a kind of division of logic function can have other dividing mode when actually realizing, such as many
Individual unit or component can with reference to or be desirably integrated into another system, or some features can be ignored, or not perform.It is another
Point, shown or discussed coupling each other or direct-coupling or communication connection can be by some interfaces, device or
The INDIRECT COUPLING of unit or communication connection, can be electrical, mechanical or other forms.If the integrated unit is with soft
The form of part functional unit is realized and as independent production marketing or when using, can be stored in an embodied on computer readable and deposit
In storage media.Based on such understanding, the part that technical scheme substantially contributes in other words to prior art
Or all or part of the technical scheme can be embodied in the form of software product, the computer software product is stored in
In one storage medium, including some instructions are used so that computer equipment (can be personal computer, server, or
Person's network equipment etc.) perform all or part of step of each embodiment methods described of the invention.And aforesaid storage medium bag
Include：USB flash disk, portable hard drive, read only memory (ROM, Read-Only Memory), random access memory (RAM, Random
Access Memory), magnetic disc or CD etc. are various can be with the medium of store program codes.
102nd, the present image feature and background image feature are carried out into SIFT feature Point matching；
The present image feature and background image feature are carried out SIFT feature Point matching by banister safety pounding-proof device, if
The SIFT feature point that it fails to match is more than first threshold, then execution step 103, if the SIFT feature point that it fails to match is less than or waits
In first threshold, then execution step 104.The first threshold is the threshold value of a SIFT feature point being blocked, for judging
Whether there is shelter in target area.
The background image is characterized as the characteristics of image of the target of banister base, and the background image feature includes：Target
Each SIFT feature point in region, and the corresponding characteristic vector of SIFT feature point.
In embodiments of the present invention, as shown in Fig. 2 banister safety pounding-proof device is arranged in gate barrier box by one
Target on the base of camera supervised banister, on the base of restrictor bar, two-face printing is for latticed pattern printing.Show
Example property, the banister base that the present invention is used can be a length of base of 1 meter, and the target of grid is scribbled above, optionally,
The pattern of target can also be the obvious pattern of other edge features.
Banister safety pounding-proof device can be acquired in initialization to background image feature, and observe always from target
The characteristic point of extraction is simultaneously matched.
If the SIFT feature point that the 103, it fails to match is more than first threshold, it indicates that gate controller stops banister motion；
If the SIFT feature point that it fails to match is more than first threshold, banister safety pounding-proof device indicates that gate controller is stopped
Only banister motion.
If the SIFT feature point that the 104, it fails to match is less than or equal to first threshold, it indicates that gate controller maintains road
The normal condition of lock.
If the SIFT feature point that it fails to match is less than or equal to first threshold, banister safety pounding-proof device indicates banister control
Device processed maintains the normal condition of banister.
The normal condition for maintaining banister refers to when banister remains static that (e.g., banister is in putting down or lift
The resting state for rising), still maintain the resting state；When banister is kept in motion (e.g., banister moving downward or
Move upwards), it is also old to maintain the kinestate constant.
In embodiments of the present invention, after the present image feature for obtaining target area, by the present image feature
Scale invariant feature conversion SIFT feature Point matching is carried out with background image feature, if the SIFT feature point that it fails to match is more than the
One threshold value, it indicates that gate controller stops banister motion；Because photographic head is more easily detected shelter compared with ground sensing system, because
This, the safety of the banister safety pounding-proof method of the embodiment of the present invention is higher.
In actual applications, because outdoor environment is complicated, monitoring scene changes with time, weather or light etc., Background
As feature can become unreliable, accordingly, it is desirable to provide the accurately update mechanism of background image feature, refers to Fig. 3, the present invention
Another embodiment of banister safety pounding-proof method in embodiment includes：
301st, background image feature is gathered；
Banister safety pounding-proof device can start photographic head in initialization in have no occluder under guaranteeing restrictor bar, carry out
The collection of background image feature.
It is that each SIFT feature point of the background image feature arranges the initial back of the body in initial background characteristics of image
Scape weighted value, the background weight value is used to judge current SIFT feature point if appropriate for as background characteristics point.
Exemplary, the initial background weight value can be 10, and initial assignment 10 is because the spy that initial detecting is arrived
To levy a little be possible to and understand unstable, if initial not assignment, if several frames of beginning are unstable, just easilys lead to SIFT feature point quilt
Delete, initial assignment 10 allows for characteristic point there can be partial frame unstable when incipient, until settling out.
302nd, the present image feature of target area is obtained；
Exemplary, first, banister safety pounding-proof device detects the metric space extreme point of target area；Create DOG
(Difference ofGaussian) pyramid multiscale space, in DOG spaces, the test point of centre is with it with the 8 of yardstick
Totally 26 points compare for individual consecutive points and corresponding 9 × 2 points of neighbouring yardstick, to guarantee in metric space and two dimensional image
Space all detects extreme point.The characteristic point obtained by above method is referred to as candidate feature point.
Secondly, banister safety pounding-proof device accurately determines extreme point position, and filtration fraction is located at the characteristic point at edge, obtains
SIFT feature point；
Again, banister safety pounding-proof device be each described SIFT feature point assigned direction parameter, the model of histogram of gradients
It is 0～360 ° to enclose, and is taken per 10 ° of posts, is always divided into the histogrammic statistical computation of 36 post travel directions, is characterized an imparting side
To parameter；
Finally, coordinate axess are rotated to be banister safety pounding-proof device the direction of characteristic point, to any one characteristic point, at it
The metric space at place takes the neighborhood of 16 × 16 sizes centered on characteristic point, then this neighborhood is evenly divided into into 4 × 4 sons
Every sub-regions are calculated gradient orientation histogram (8 directions) by region.8 histograms of oriented gradients to 4 × 4 sub-regions
Sorted successively according to position, constitute the vector of 4 × 4 × 8=128 dimension.Finally by characteristic vector length normalizing
Change, the vector is exactly SIFT feature point vector.
303rd, the present image feature and background image feature are carried out into SIFT feature Point matching；
Banister safety pounding-proof device obtains the first SIFT feature point of present image feature；The first SIFT feature point is
SIFT feature point in present image feature；
Two second nearest with the Euclidean distance of the first SIFT feature point are searched in the background image feature
SIFT feature point；The second SIFT feature point is the SIFT feature point in background image feature；
If the second nearest SIFT feature point is less than matching threshold with the ratio of secondary the second near SIFT feature point, currently
The SIFT feature Point matching success of characteristics of image and background image；
If the second nearest SIFT feature point is more than or equal to matching threshold with the ratio of secondary the second near SIFT feature point,
Then the SIFT feature Point matching of present image feature and background image is unsuccessful.
Exemplary, the matching threshold can be 0.6.
In the prior art, Euclidean distance regards the similarity degree of signal as.Distance is more near more similar, easier to be mutual
Interference, the bit error rate is higher.
Exemplary, the calculating process of Euclidean distance can be：Certain characteristic point X of present frame, wherein X={ x0,
x1..., x127, i.e., the vector of one 128 dimension, in certain characteristic point Y of background frames, wherein Y={ y0, y1..., y127, that
The Euclidean distance of X and Y can be expressed as：
304th, the renewal of background weight value is carried out according to the result of SIFT feature Point matching；
If corresponding SIFT feature Point matching is unsuccessful in the SIFT feature point of present image feature and background image feature,
Then the background weight value of the SIFT feature point in the background image feature is subtracted one, if the SIFT in the background image feature
The current background weight value of characteristic point is less than background threshold, then delete the SIFT feature point in the background image feature；
If corresponding SIFT feature Point matching success in the SIFT feature point of present image feature and background image feature,
The background weight value of the SIFT feature point in the background image feature is added one.Optionally, banister safety pounding-proof device may be used also
To judge whether the background weight value after updating is equal to specific threshold, if if so, then SIFT feature Point matching success next time, dimension
Hold current background weight value constant.
Exemplary, for each background characteristics point, can all there is a weighted value wi, 0≤i≤n, wiValue change rule
It is then：
A). increase background characteristics point weight, w if the match is successfuli=wi+1；It is not further added by if weight to 100,
It is always maintained at 100.The purpose of the weight of increase be in order that characteristic point increase and deletion process can slowly carry out, it is ensured that it is special
Levy be a little it is stable just increase as background dot, while also true background is protected characteristic point and is slow and unmatches just deletion.
B). background characteristics point weight, w are reduced if it fails to matchi=wi-1；
If c). the weighted value of background characteristics point is less than threshold value, and 0, i.e. w are set in the present embodimenti＜ 0, then delete the back of the body
Scape characteristic point；
D). after have passed through traversal, the characteristic point of present frame, if matching with it without corresponding background characteristics point, increases
For new background characteristics point, and weight is initialized for 10, i.e. wi=10.
The judgement of the SIFT feature point that the 305th, it fails to match in present image feature；
Each corresponding SIFT feature point is all completed after matching in present image feature and background image feature, banister peace
Prevent pounding whether the device judgement SIFT feature point that it fails to match is more than first threshold entirely, then execution step 306, if it fails to match
SIFT feature point is less than or equal to first threshold, then execution step 307.
It is understood that without strict sequential relationship between step 304 and step 305, you can " current to first carry out
The judgement of the SIFT feature point that it fails to match in characteristics of image ", then perform and " background is carried out according to the result of SIFT feature Point matching
The renewal of weighted value ", it is also possible to which two steps are carried out simultaneously, with specific reference to practical situation depending on, be not construed as limiting herein.
If the SIFT feature point that the 306, it fails to match is more than first threshold, it indicates that gate controller stops banister motion；
If the SIFT feature point that it fails to match is more than first threshold, banister safety pounding-proof device indicates that gate controller is stopped
Only banister motion.
If the SIFT feature point that the 307, it fails to match is less than or equal to first threshold, it indicates that gate controller maintains banister
Normal condition.
If the SIFT feature point that it fails to match is less than or equal to first threshold, banister safety pounding-proof device indicates banister control
Device processed maintains the normal condition of banister.
The normal condition for maintaining banister refers to when banister remains static that (e.g., banister is in putting down or lift
The resting state for rising), still maintain the resting state；When banister is kept in motion (e.g., banister moving downward or
Move upwards), it is also old to maintain the kinestate constant.
In embodiments of the present invention, outdoor environment is complicated, and monitoring scene changes with time, weather or light etc., and background is special
Levying a meeting becomes unreliable, it is therefore desirable to checks the reliability of characteristic point and is updated in real time.Used in our algorithm
Whether one weighted value w is reliable come the characteristic point for judging background frames.
In actual applications, after SIFT feature point deletion, can be according to the present image feature constantly updated, to background
SIFT feature point in characteristics of image carries out selection supplement；Also, the change that in actual applications, moment causes may be to blocking
The judgement of thing causes error, the embodiment of the present invention to also provide corresponding solution, specifically refers to Fig. 4, and the present invention is implemented
Another embodiment of banister safety pounding-proof method in example includes：
401st, background image feature is gathered；
Banister safety pounding-proof device can start photographic head in initialization in have no occluder under guaranteeing restrictor bar, carry out
The collection of background image feature.
It is that each SIFT feature point of the background image feature arranges the initial back of the body in initial background characteristics of image
Scape weighted value, the background weight value is used to judge current SIFT feature point if appropriate for as background characteristics point.
Exemplary, the initial background weight value can be 10, and initial assignment 10 is because the spy that initial detecting is arrived
To levy a little be possible to and understand unstable, if initial not assignment, if several frames of beginning are unstable, just easilys lead to SIFT feature point quilt
Delete, initial assignment 10 allows for characteristic point there can be partial frame unstable when incipient, until settling out.
402nd, the present image feature of target area is obtained；
Exemplary, first, banister safety pounding-proof device detects the metric space extreme point of target area；Create DOG
(Difference ofGaussian) pyramid multiscale space, in DOG spaces, the test point of centre is with it with the 8 of yardstick
Totally 26 points compare for individual consecutive points and corresponding 9 × 2 points of neighbouring yardstick, empty in metric space and two dimensional image to guarantee
Between all detect extreme point.The characteristic point obtained by above method is referred to as candidate feature point.
Secondly, banister safety pounding-proof device accurately determines extreme point position, and filtration fraction is located at the characteristic point at edge, obtains
SIFT feature point；
Again, banister safety pounding-proof device be each described SIFT feature point assigned direction parameter, the model of histogram of gradients
It is 0～360 ° to enclose, and is taken per 10 ° of posts, is always divided into the histogrammic statistical computation of 36 post travel directions, is characterized an imparting side
To parameter；
Finally, coordinate axess are rotated to be banister safety pounding-proof device the direction of characteristic point, to any one characteristic point, at it
The metric space at place takes the neighborhood of 16 × 16 sizes centered on characteristic point, then this neighborhood is evenly divided into into 4 × 4 sons
Every sub-regions are calculated gradient orientation histogram (8 directions) by region.8 histograms of oriented gradients to 4 × 4 sub-regions
Sorted successively according to position, constitute the vector of 4 × 4 × 8=128 dimension.Finally by characteristic vector length normalizing
Change, the vector is exactly SIFT feature point vector.
403rd, the present image feature and background image feature are carried out into SIFT feature Point matching；
Banister safety pounding-proof device obtains the first SIFT feature point of present image feature；The first SIFT feature point is
SIFT feature point in present image feature；
Two second nearest with the Euclidean distance of the first SIFT feature point are searched in the background image feature
SIFT feature point；The second SIFT feature point is the SIFT feature point in background image feature；
If the second nearest SIFT feature point is less than matching threshold with the ratio of secondary the second near SIFT feature point, currently
The SIFT feature Point matching success of characteristics of image and background image；
If the second nearest SIFT feature point is more than or equal to matching threshold with the ratio of secondary the second near SIFT feature point,
Then the SIFT feature Point matching of present image feature and background image is unsuccessful.
Exemplary, the matching threshold can be 0.6.
In the prior art, Euclidean distance regards the similarity degree of signal as.Distance is more near more similar, easier to be mutual
Interference, the bit error rate is higher.
Exemplary, the calculating process of Euclidean distance can be：Certain characteristic point X of present frame, wherein X={ x0,
x1..., x127, i.e., the vector of one 128 dimension, in certain characteristic point Y of background frames, wherein Y={ y0, y1..., y127, that
The Euclidean distance of X and Y can be expressed as：
404th, the renewal of background weight value is carried out according to the result of SIFT feature Point matching；
If corresponding SIFT feature Point matching is unsuccessful in the SIFT feature point of present image feature and background image feature,
Then the background weight value of the SIFT feature point in the background image feature is subtracted one, if the SIFT in the background image feature
The current background weight value of characteristic point is less than background threshold, then delete the SIFT feature point in the background image feature；
If corresponding SIFT feature Point matching success in the SIFT feature point of present image feature and background image feature,
The background weight value of the SIFT feature point in the background image feature is added one.Optionally, banister safety pounding-proof device may be used also
To judge whether the background weight value after updating is equal to specific threshold, if if so, then SIFT feature Point matching success next time, dimension
Hold current background weight value constant.
Exemplary, for each background characteristics point, can all there is a weighted value wi, 0≤i≤n, wiValue change rule
It is then：
A). increase background characteristics point weight, w if the match is successfuli=wi+1；It is not further added by if weight to 100,
It is always maintained at 100.The purpose of the weight of increase be in order that characteristic point increase and deletion process can slowly carry out, it is ensured that it is special
Levy be a little it is stable just increase as background dot, while also true background is protected characteristic point and is slow and unmatches just deletion.
B). background characteristics point weight, w are reduced if it fails to matchi=wi-1；
If c). the weighted value of background characteristics point is less than threshold value, and 0, i.e. w are set in the present embodimenti＜ 0, then delete the back of the body
Scape characteristic point；
D). after have passed through traversal, the characteristic point of present frame, if matching with it without corresponding background characteristics point, increases
For new background characteristics point, and weight is initialized for 10, i.e. wi=10.
405th, the SIFT feature point in background image feature is increased；
If the SIFT feature point without corresponding SIFT feature point Yu present image feature in current background image feature
Matched, then the SIFT feature point of present image feature is set to into pre- background characteristics point, and it is special to initialize the pre- background
Background weight value a little is levied, if in the SIFT feature Point matching after, the background weight value of the pre- background characteristics point updates
And reaching stable threshold, it is determined that the pre- background characteristics point is background characteristics point.
Exemplary, the stable threshold in the embodiment of the present invention could be arranged to 30.It is understood that sub- practical application
In, stable threshold can there are other values according to practical situation, therefore, the embodiment of the present invention is not construed as limiting to this.
The judgement of the SIFT feature point that the 406th, it fails to match in present image feature；
Each corresponding SIFT feature point is all completed after matching in present image feature and background image feature, banister peace
Whether the complete anti-device judgement SIFT feature point that it fails to match of pounding then increases and blocks frame number, if the screening of accumulation more than first threshold
Gear frame number is more than occlusion threshold, then trigger execution step 407, if the SIFT feature point that it fails to match is less than or equal to the first threshold
It is worth, then execution step 408.
In embodiments of the present invention, judgement of the change for causing in order to avoid moment to shelter causes error, introduces
The concept of frame is blocked, i.e., when the situation of " the SIFT feature point that it fails to match is more than first threshold " that once occur, it is determined that working as
It is front that blocking for a frame occurs；Continuation carries out SIFT feature Point matching to present image feature and background image feature, if continuous send out
The situation that raw present frame is blocked, it is assumed that occlusion threshold is 3, that is, recur the situation that 3 present frames are blocked, it is determined that mesh
There is shelter in region in mark.
It is understood that without strict sequential relationship between step 404, step 405 and step 406, with specific reference to
Depending on practical situation, it is not construed as limiting herein.
407th, indicate that gate controller stops banister motion；
If the frame number that blocks of accumulation is more than occlusion threshold, banister safety pounding-proof device indicates that gate controller stops banister
Motion.
Exemplary, in embodiments of the present invention, occlusion threshold could be arranged to 3.
408th, indicate that gate controller maintains the normal condition of banister.
If the SIFT feature point that it fails to match is less than or equal to first threshold, banister safety pounding-proof device indicates banister control
Device processed maintains the normal condition of banister.
The normal condition for maintaining banister refers to when banister remains static that (e.g., banister is in putting down or lift
The resting state for rising), still maintain the resting state；When banister is kept in motion (e.g., banister moving downward or
Move upwards), it is also old to maintain the kinestate constant.
During the present invention is implemented, introducing increases the mechanism of background SIFT feature point so that the determination of background image feature is more
Plus accurately, also improve the accuracy of shelter judgement；Meanwhile, the embodiment of the present invention is carrying out introducing screening when shelter judges
The concept of gear frame, it is to avoid the error that the shelter that instantaneous variation causes judges.
Below the banister safety pounding-proof device to realizing said method is described, and Fig. 5 is referred to, in the embodiment of the present invention
One embodiment of banister safety pounding-proof device includes：
Feature acquiring unit 501, for obtaining the present image feature of target area, the target area is banister base
Target area；
Characteristic matching unit 502, for the present image feature and background image feature to be carried out into scale invariant feature
Conversion SIFT feature Point matching, if the SIFT feature point that it fails to match is more than first threshold, it indicates that gate controller stop way
Lock is moved；The background image is characterized as the characteristics of image of the target of banister base, and the background image feature includes：Target area
Each SIFT feature point in domain, and the corresponding characteristic vector of SIFT feature point.
Further, if the characteristic matching unit 502 is additionally operable to the SIFT feature point that it fails to match less than or equal to the
One threshold value, it indicates that gate controller maintains the normal condition of banister.
The feature acquiring unit 502 specifically for：The metric space extreme point of detection target area；It is accurate to determine extreme value
Point position, filtration fraction is located at the characteristic point at edge, obtains SIFT feature point；Join for SIFT feature point assigned direction each described
Number, generates present image feature, and the present image feature includes：Each SIFT feature point of present image, and it is described
The corresponding characteristic vector of SIFT feature point.
The characteristic matching unit 502 is specifically additionally operable to：
Obtain the first SIFT feature point of present image feature；Search in the background image feature and described first
Two nearest the second SIFT feature points of the Euclidean distance of SIFT feature point；If the second nearest SIFT feature point and secondary near the
The ratio of 2 SIFT feature points is less than matching threshold, then the SIFT feature Point matching success of present image feature and background image；
If the second nearest SIFT feature point is more than or equal to matching threshold with the ratio of secondary the second near SIFT feature point, currently scheme
The SIFT feature Point matching of picture feature and background image is unsuccessful.
Optionally, the banister safety pounding-proof device also includes：
Weight setting unit 503, in initial background characteristics of image, being each of the background image feature
SIFT feature point arranges initial background weight value, the background weight value be used to judging current SIFT feature point if appropriate for
As background characteristics point.
Further, the weight setting unit is additionally operable to：
If corresponding SIFT feature Point matching is unsuccessful in the SIFT feature point of present image feature and background image feature,
Then the background weight value of the SIFT feature point in the background image feature is subtracted one, if the SIFT in the background image feature
The current background weight value of characteristic point is less than background threshold, then delete the SIFT feature point in the background image feature；
If corresponding SIFT feature Point matching success in the SIFT feature point of present image feature and background image feature,
The background weight value of the SIFT feature point in the background image feature is added one.
The weight setting unit is additionally operable to：
Judge that whether the background weight value after updating is equal to specific threshold, if if so, then SIFT feature Point matching next time
Success, maintains current background weight to be worth constant.
Optionally, the banister safety pounding-proof device also includes：
Characteristic point updating block 504, if for no corresponding SIFT feature point in current background image feature and working as
The SIFT feature point of front characteristics of image is matched, then the SIFT feature point of present image feature is set to into pre- background characteristics
Point, and the background weight value of the pre- background characteristics point is initialized, if in the SIFT feature Point matching after, the pre- background
The background weight value of characteristic point updates and reaches stable threshold, it is determined that the pre- background characteristics point is background characteristics point.
The characteristic point updating block 504 is additionally operable to：The SIFT feature Point matching success or not of the pre- background characteristics point,
It is not involved in the judgement of the first threshold.
The characteristic matching unit 502 is additionally operable to：
If the SIFT feature point that it fails to match is more than first threshold, frame number is blocked in increase；If accumulation to block frame number big
In occlusion threshold, then trigger the instruction gate controller and stop banister the step of move.
The specific operation process of above-mentioned unit is described below：
Feature acquiring unit 501 obtains the present image feature of target area；Optionally, the present image feature bag
Include：Each SIFT feature point of present image, and the corresponding characteristic vector of SIFT feature point.
SIFT is a kind of description for image processing field, and this description has scale invariability, can be in the picture
Key point is detected, is a kind of local description.The matching capacity of SIFT feature matching algorithm is stronger, can process two width images
Between there is translation, rotation, the matching problem in the case of affine transformation, or even to a certain extent to the figure of arbitrarily angled shooting
As also possessing relatively stable characteristic matching ability.
Specifically, in actual applications, feature acquiring unit 501 can be photographic head.
The present image feature and background image feature are carried out SIFT feature Point matching by characteristic matching unit 502, if
The SIFT feature point that it fails to match is more than first threshold, if the SIFT feature point that then it fails to match is more than first threshold, it indicates that
Gate controller stops banister motion, if the SIFT feature point that it fails to match is less than or equal to first threshold, if it fails to match
SIFT feature point be less than or equal to first threshold, it indicates that gate controller maintain banister normal condition.First threshold
Value is the threshold value of a SIFT feature point being blocked, for judging to whether there is shelter in target area.
The normal condition for maintaining banister refers to when banister remains static that (e.g., banister is in putting down or lift
The resting state for rising), still maintain the resting state；When banister is kept in motion (e.g., banister moving downward or
Move upwards), it is also old to maintain the kinestate constant.
The background image is characterized as the characteristics of image of the target of banister base, and the background image feature includes：Target
Each SIFT feature point in region, and the corresponding characteristic vector of SIFT feature point.
In embodiments of the present invention, as shown in Fig. 2 banister safety pounding-proof device is arranged in gate barrier box by one
Target on the base of camera supervised banister, on the base of restrictor bar, two-face printing is for latticed pattern printing.Show
Example property, the banister base that the present invention is used can be a length of base of 1 meter, and the target of grid is scribbled above, optionally,
The pattern of target can also be the obvious pattern of other edge features.
Banister safety pounding-proof device can be acquired in initialization to background image feature, and observe always from target
The characteristic point of extraction is simultaneously matched.
In actual applications, in initialization, photographic head can be started in have no occluder under guaranteeing restrictor bar, carries out background
The collection of characteristics of image.
In initial background characteristics of image, weight setting unit 503 is special for each SIFT of the background image feature
An initial background weight value of setting is levied, the background weight value is used to judge current SIFT feature point if appropriate for as the back of the body
Scape characteristic point.
Exemplary, the initial background weight value can be 10, and initial assignment 10 is because the spy that initial detecting is arrived
To levy a little be possible to and understand unstable, if initial not assignment, if several frames of beginning are unstable, just easilys lead to SIFT feature point quilt
Delete, initial assignment 10 allows for characteristic point there can be partial frame unstable when incipient, until settling out.
Further, in initialization, feature acquiring unit 501 can gather current picture as present image feature,
Can be specifically：
First, the metric space extreme point of target area is detected；Create DOG (Difference of Gaussian) gold words
Tower multiscale space, in DOG spaces, middle test point is corresponding with neighbouring yardstick with 8 consecutive points of yardstick with it
9 × 2 points totally 26 points compare, to guarantee all to detect extreme point in metric space and two dimensional image space.By the above
The characteristic point that method is obtained is referred to as candidate feature point.
Secondly, accurate to determine extreme point position, filtration fraction is located at the characteristic point at edge, obtains SIFT feature point；
Again, it is each described SIFT feature point assigned direction parameter, the scope of histogram of gradients is 0～360 °, takes every
10 ° of posts, are always divided into the histogrammic statistical computation of 36 post travel directions, are characterized and give a directioin parameter；
Finally, coordinate axess are rotated to be the direction of characteristic point, to any one characteristic point, is taken in its metric space being located
The neighborhood of 16 × 16 sizes centered on characteristic point, then this neighborhood is evenly divided into into 4 × 4 sub-regions, to each sub-district
Domain calculates gradient orientation histogram (8 directions).8 histograms of oriented gradients of 4 × 4 sub-regions are arranged successively according to position
Sequence, constitutes the vector of 4 × 4 × 8=128 dimension.Finally by characteristic vector length normalization method, the vector is exactly
SIFT feature point vector.
Specifically, the process that characteristic matching unit 502 carries out characteristic matching can be：
Obtain the first SIFT feature point of present image feature；The first SIFT feature point is in present image feature
SIFT feature point；
Two second nearest with the Euclidean distance of the first SIFT feature point are searched in the background image feature
SIFT feature point；The second SIFT feature point is the SIFT feature point in background image feature；
If the second nearest SIFT feature point is less than matching threshold with the ratio of secondary the second near SIFT feature point, currently
The SIFT feature Point matching success of characteristics of image and background image；
If the second nearest SIFT feature point is more than or equal to matching threshold with the ratio of secondary the second near SIFT feature point,
Then the SIFT feature Point matching of present image feature and background image is unsuccessful.
Exemplary, the matching threshold can be 0.6.
In the prior art, Euclidean distance regards the similarity degree of signal as.Distance is more near more similar, easier to be mutual
Interference, the bit error rate is higher.
Exemplary, the calculating process of Euclidean distance can be：Certain characteristic point X of present frame, wherein X={ x0,
x1..., x127, i.e., the vector of one 128 dimension, in certain characteristic point Y of background frames, wherein Y={ y0, y1..., y127, that
The Euclidean distance of X and Y can be expressed as：
In actual applications, if corresponding SIFT feature in the SIFT feature point and background image feature of present image feature
Point matching is unsuccessful, then subtract one by the background weight value of the SIFT feature point in the background image feature, if the Background
As the current background weight value of the SIFT feature point in feature is less than background threshold, then in deleting the background image feature
SIFT feature point；
If corresponding SIFT feature Point matching success in the SIFT feature point of present image feature and background image feature,
The background weight value of the SIFT feature point in the background image feature is added one by weight setting unit 503.Optionally, banister peace
Complete anti-device of pounding can also judge whether the background weight value after updating is equal to specific threshold, if if so, then next time SIFT is special
Point matching success is levied, maintains current background weight to be worth constant.
Exemplary, for each background characteristics point, can all there is a weighted value wi, 0≤i≤n, wiValue change rule
It is then：
A). increase background characteristics point weight, w if the match is successfuli=wi+1；It is not further added by if weight to 100,
It is always maintained at 100.The purpose of the weight of increase be in order that characteristic point increase and deletion process can slowly carry out, it is ensured that it is special
Levy be a little it is stable just increase as background dot, while also true background is protected characteristic point and is slow and unmatches just deletion.
B). background characteristics point weight, w are reduced if it fails to matchi=wi-1；
If c). the weighted value of background characteristics point is less than threshold value, and 0, i.e. w are set in the present embodimenti＜ 0, then delete the back of the body
Scape characteristic point；
D). after have passed through traversal, the characteristic point of present frame, if matching with it without corresponding background characteristics point, increases
For new background characteristics point, and weight is initialized for 10, i.e. wi=10.
In embodiments of the present invention, outdoor environment is complicated, and monitoring scene changes with time, weather or light etc., and background is special
Levying a meeting becomes unreliable, it is therefore desirable to checks the reliability of characteristic point and is updated in real time.Used in our algorithm
Whether one weighted value w is reliable come the characteristic point for judging background frames.
After SIFT feature point deletion, can be according to the present image feature constantly updated, in background image feature
SIFT feature point carries out selection supplement；Also, the change that in actual applications, moment causes may be caused to the judgement of shelter
Error, the banister safety pounding-proof device in the embodiment of the present invention can pass through to introduce the mechanism for increasing background SIFT feature point,
So that the determination of background image feature is more accurate, the accuracy of shelter judgement is also improved；Meanwhile, the embodiment of the present invention exists
Carry out introducing the concept for blocking frame when shelter judges, it is to avoid the error that the shelter that instantaneous variation causes judges.
In several embodiments provided herein, it should be understood that disclosed apparatus and method can pass through it
Its mode is realized.For example, device embodiment described above is only schematic, for example, the division of the unit, and only
Only a kind of division of logic function, can there is other dividing mode when actually realizing, such as multiple units or component can be tied
Close or be desirably integrated into another system, or some features can be ignored, or do not perform.It is another, it is shown or discussed
Coupling each other or direct-coupling or communication connection can be the INDIRECT COUPLINGs by some interfaces, device or unit or logical
Letter connection, can be electrical, mechanical or other forms.
The unit as separating component explanation can be or may not be it is physically separate, it is aobvious as unit
The part for showing can be or may not be physical location, you can with positioned at a place, or can also be distributed to multiple
On NE.Some or all of unit therein can according to the actual needs be selected to realize the mesh of this embodiment scheme
's.
In addition, each functional unit in each embodiment of the invention can be integrated in a processing unit, it is also possible to
It is that unit is individually physically present, it is also possible to which two or more units are integrated in a unit.Above-mentioned integrated list
Unit both can be realized in the form of hardware, it would however also be possible to employ the form of SFU software functional unit is realized.
If the integrated unit is realized using in the form of SFU software functional unit and as independent production marketing or used
When, during a computer read/write memory medium can be stored in.Based on such understanding, technical scheme is substantially
The part for contributing to prior art in other words or all or part of the technical scheme can be in the form of software products
Embody, the computer software product is stored in a storage medium, including some instructions are used so that a computer
Equipment (can be personal computer, server, or network equipment etc.) performs each embodiment methods described of the invention
All or part of step.And aforesaid storage medium includes：USB flash disk, portable hard drive, read only memory (ROM, Read-Only
Memory), random access memory (RAM, Random Access Memory), magnetic disc or CD etc. are various can store journey
The medium of sequence code.
The above, the only specific embodiment of the present invention, but protection scope of the present invention is not limited thereto, any
Those familiar with the art the invention discloses technical scope in, change or replacement can be readily occurred in, all should contain
Cover within protection scope of the present invention.Therefore, protection scope of the present invention described should be defined by scope of the claims.
Claims (9)
1. a kind of banister safety pounding-proof method, it is characterised in that include：
The present image feature of target area is obtained, the target area is the target area of banister base；
The present image feature and background image feature are carried out into scale invariant feature conversion SIFT feature Point matching, if matching
The SIFT feature point of failure is more than first threshold, it indicates that gate controller stops banister motion；The background image is characterized as
The characteristics of image of the target of banister base, the background image feature includes：Each SIFT feature point in target area, and
The corresponding characteristic vector of the SIFT feature point；
It is described that the present image feature and background image feature are carried out into SIFT feature Point matching, including：
Obtain the first SIFT feature point of present image feature；
Two twoth SIFTs nearest with the Euclidean distance of the first SIFT feature point are searched in the background image feature
Characteristic point；
If the second nearest SIFT feature point is less than matching threshold, present image with the ratio of secondary the second near SIFT feature point
The SIFT feature Point matching success of feature and background image；
If the second nearest SIFT feature point is more than or equal to matching threshold with the ratio of secondary the second near SIFT feature point, when
The SIFT feature Point matching of front characteristics of image and background image is unsuccessful.
2. method according to claim 1, it is characterised in that if the SIFT feature point that it fails to match is less than or equal to first
Threshold value, it indicates that gate controller maintains the normal condition of banister.
3. method according to claim 1, it is characterised in that the present image feature of the acquisition target area, including：
The metric space extreme point of detection target area；
Accurate to determine extreme point position, filtration fraction is located at the characteristic point at edge, obtains SIFT feature point；
For SIFT feature point assigned direction parameter each described, present image feature is generated, the present image feature includes：When
Each SIFT feature point of front image, and the corresponding characteristic vector of SIFT feature point.
4. method according to claim 1, it is characterised in that methods described also includes：
It is that each SIFT feature point of the background image feature arranges initial background power in initial background characteristics of image
Weight values, the background weight value is used to judge current SIFT feature point if appropriate for as background characteristics point.
5. method according to claim 4, it is characterised in that described each SIFT feature for the background image feature
Point is arranged after initial background weight value, including：
If corresponding SIFT feature Point matching is unsuccessful in the SIFT feature point of present image feature and background image feature, will
The background weight value of the SIFT feature point in the background image feature subtracts one, if the SIFT feature in the background image feature
The current background weight value of point is less than background threshold, then delete the SIFT feature point in the background image feature；
If corresponding SIFT feature Point matching success in the SIFT feature point of present image feature and background image feature, by institute
The background weight value for stating the point of the SIFT feature in background image feature plus one.
6. method according to claim 5, it is characterised in that the SIFT feature point by the background image feature
Background weight value add one after, also include：
Judge whether the background weight value after updating is equal to specific threshold, if if so, then SIFT feature Point matching success next time,
Current background weight is maintained to be worth constant.
7. method according to claim 4, it is characterised in that described each SIFT feature for the background image feature
Point is arranged after initial background weight value, including：
If carrying out with the SIFT feature point of present image feature without corresponding SIFT feature point in current background image feature
Matching, then be set to pre- background characteristics point by the SIFT feature point of present image feature, and initializes the pre- background characteristics point
Background weight value, if in the SIFT feature Point matching after, the background weight value of the pre- background characteristics point updates and reaches
To stable threshold, it is determined that the pre- background characteristics point is background characteristics point.
8. method according to claim 7, it is characterised in that described to be set to the SIFT feature point of present image feature
After pre- background characteristics point, including：
The SIFT feature Point matching success or not of the pre- background characteristics point, is not involved in the judgement of the first threshold.
9. method according to claim 1, it is characterised in that methods described also includes：
If the SIFT feature point that it fails to match is more than first threshold, frame number is blocked in increase；If the frame number that blocks of accumulation is more than screening
Gear threshold value, then trigger the instruction gate controller and stop banister the step of move.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310676400.XA CN103810495B (en) | 2013-12-11 | 2013-12-11 | Barrier gate safe anti-hitting method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310676400.XA CN103810495B (en) | 2013-12-11 | 2013-12-11 | Barrier gate safe anti-hitting method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103810495A CN103810495A (en) | 2014-05-21 |
| CN103810495B true CN103810495B (en) | 2017-05-10 |
Family
ID=50707236
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310676400.XA Expired - Fee Related CN103810495B (en) | 2013-12-11 | 2013-12-11 | Barrier gate safe anti-hitting method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103810495B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106677094B (en) * | 2017-03-27 | 2019-11-19 | 深圳市捷顺科技实业股份有限公司 | A kind of banister control method and device |
| CN108086194A (en) * | 2017-11-22 | 2018-05-29 | 苏州爱开客信息技术有限公司 | Banister self-detection system and method for detecting |
| CN110244313A (en) * | 2019-05-24 | 2019-09-17 | 中国科学院合肥物质科学研究院 | A kind of collision avoidance system and method based on TOF camera |
| CN112364865B (en) * | 2020-11-12 | 2022-09-23 | 郑州大学 | Method for detecting small moving target in complex scene |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102147835A (en) * | 2010-11-26 | 2011-08-10 | 中华人民共和国深圳出入境检验检疫局 | Driver body temperature automatic measurement system applied to port lanes and implementation method thereof |
| CN102867350A (en) * | 2012-08-29 | 2013-01-09 | 广东铂亚信息技术股份有限公司 | Prison AB door safety management method based on face recognition technology |
| CN203034398U (en) * | 2013-01-04 | 2013-07-03 | 齐齐哈尔大学 | Image-processing-based intelligent barrier gate system |
| CN103337098A (en) * | 2013-06-14 | 2013-10-02 | 中国电子科技集团公司第三十八研究所 | Electronic toll collection system based on video vehicle license plate recognition technology and method thereof |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8942481B2 (en) * | 2012-03-11 | 2015-01-27 | Universidad De Santiago De Compostela | Three dimensional CMOS image processor for feature detection |
-
2013
- 2013-12-11 CN CN201310676400.XA patent/CN103810495B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102147835A (en) * | 2010-11-26 | 2011-08-10 | 中华人民共和国深圳出入境检验检疫局 | Driver body temperature automatic measurement system applied to port lanes and implementation method thereof |
| CN102867350A (en) * | 2012-08-29 | 2013-01-09 | 广东铂亚信息技术股份有限公司 | Prison AB door safety management method based on face recognition technology |
| CN203034398U (en) * | 2013-01-04 | 2013-07-03 | 齐齐哈尔大学 | Image-processing-based intelligent barrier gate system |
| CN103337098A (en) * | 2013-06-14 | 2013-10-02 | 中国电子科技集团公司第三十八研究所 | Electronic toll collection system based on video vehicle license plate recognition technology and method thereof |
Non-Patent Citations (2)
| Title |
|---|
| "四川职业技术学院智能停车管理系统设计与实现";谭娟;《中国优秀硕士学位论文全文数据库信息科技辑》;20130115;正文第第1段-第42页第13段，第65页，附图4.2、4.3、7.3、7.4、7.5 * |
| "基于Adaboost与SIFT的人脸识别与跟踪系统的研究与设计";黄兴;《中国优秀硕士学位论文全文数据库信息科技辑》;20130415;正文第35页第2段-第36页第4段、第38页第2段-第48页第4段，附图4.2-4.7 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103810495A (en) | 2014-05-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107644204B (en) | Human body identification and tracking method for security system | |
| CN110807385B (en) | Target detection method, target detection device, electronic equipment and storage medium | |
| CN109255322B (en) | A kind of human face in-vivo detection method and device | |
| CN103605971B (en) | Method and device for capturing face images | |
| CN108038176B (en) | Method and device for establishing passerby library, electronic equipment and medium | |
| CN103810495B (en) | Barrier gate safe anti-hitting method | |
| KR20200088219A (en) | Item identification method and system, electronic equipment | |
| KR101557376B1 (en) | Method for Counting People and Apparatus Therefor | |
| CN108985162A (en) | Object real-time tracking method, apparatus, computer equipment and storage medium | |
| CN104123544A (en) | Video analysis based abnormal behavior detection method and system | |
| CN103679118A (en) | Human face in-vivo detection method and system | |
| JP2007334623A (en) | Face authentication device, face authentication method, and access control device | |
| CN107563281A (en) | A kind of construction site personal security hidden danger monitoring method based on deep learning | |
| CN103810711A (en) | Keyframe extracting method and system for monitoring system videos | |
| CN102945366A (en) | Method and device for face recognition | |
| JP6941966B2 (en) | Person authentication device | |
| CN104599287A (en) | Object tracking method and device and object recognition method and device | |
| CN107292917A (en) | A kind of Dragging recognition methods based on Computer Vision | |
| CN103632427A (en) | Gate cracking protection method and gate control system | |
| CN110781844A (en) | Security patrol monitoring method and device | |
| CN110322472A (en) | A kind of multi-object tracking method and terminal device | |
| CN113628245A (en) | Multi-target tracking method, device, electronic equipment and storage medium | |
| KR101406334B1 (en) | System and method for tracking multiple object using reliability and delayed decision | |
| CN114219829A (en) | Vehicle tracking method, computer equipment and storage device | |
| CN111126257A (en) | Behavior detection method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170510 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |