CN102289820B - Intelligent tracking ball machine and tracking method thereof - Google Patents

Intelligent tracking ball machine and tracking method thereof Download PDF

Info

Publication number
CN102289820B
CN102289820B CN201110233861.0A CN201110233861A CN102289820B CN 102289820 B CN102289820 B CN 102289820B CN 201110233861 A CN201110233861 A CN 201110233861A CN 102289820 B CN102289820 B CN 102289820B
Authority
CN
China
Prior art keywords
tracking
distance
ball machine
intelligent
mobile object
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.)
Active
Application number
CN201110233861.0A
Other languages
Chinese (zh)
Other versions
CN102289820A (en
Inventor
吴占伟
郭海训
全晓臣
刘志宇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hangzhou Hikvision Digital Technology Co Ltd
Original Assignee
Hangzhou Hikvision Digital Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Hangzhou Hikvision Digital Technology Co Ltd filed Critical Hangzhou Hikvision Digital Technology Co Ltd
Priority to CN201110233861.0A priority Critical patent/CN102289820B/en
Publication of CN102289820A publication Critical patent/CN102289820A/en
Priority to PCT/CN2012/076223 priority patent/WO2013023474A1/en
Application granted granted Critical
Publication of CN102289820B publication Critical patent/CN102289820B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/20—Analysis of motion
    • G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10016—Video; Image sequence
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30196—Human being; Person
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30232—Surveillance
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30241—Trajectory

Abstract

The invention relates to the field of video monitoring, and discloses an intelligent tracking ball machine and a tracking method thereof. In the invention, after a mobile object is detected, the distance from the mobile object to a specified point is calculated, whether the tracking is triggered is determined according to the distance, the calculated amount can be reduced, and the traced triggering is more accurate. Distance information at any point in the current picture can be measured, and the obtained distance information is more complete and effective and has stronger practicability. Outside interference factors can be effectively eliminated without depending on the traditional coordinate system or the relationship between focusing and zooming, thus strong adaptability is obtained. When an equidistant curve is drawn and the triggered warning wire is tracked, a user more visually judges the warning wire. An object to be tracked is always in the center position of the picture, and the tracking continuity and effectiveness can be ensured.

Description

Intelligent-tracking ball machine and tracking thereof
Technical field
The present invention relates to field of video monitoring, particularly use intelligent-tracking ball machine to carry out the security and guard technology of video monitoring.
Background technology
Intelligent video monitoring is to utilize computer vision technique that vision signal is processed, analyzed and understands; in the situation that not needing human intervention; by to sequence image automatic analysis; variation in monitoring scene is positioned, identified and follows the tracks of; and analyze on this basis and judge the behavior of target; can when abnormal conditions occur, give the alarm in time or useful information is provided, effectively assisting Security Officer to process crisis, and reduce to greatest extent wrong report and fail to report phenomenon.
At present, automatic tracking ball machine on market its to move detection be mainly to process and the ultimate principle of pattern-recognition based on image, by the special image/video processing unit of the compositions such as difference, optical flow method, carry out the detection of mobile object, the triggering object detecting is carried out to the tracking of ball machine thereupon.
The present inventor finds, the major defect of this kind of method existence is that algorithm process amount is large on the one hand, need to go out current movement locus of object with algorithm simulation and complete triggering judgement, moreover be to be easily disturbed, the movement environment of inspected object is had to stricter requirement, be finally visual pattern not, can not reflect the effective informations such as the movement velocity of current object and true altitude.So, need a kind of new Video Supervision Technique badly, less algorithm complex and the dependence to scene largely, and the tracking triggering is more accurate and effective, directly perceived.
Summary of the invention
The object of the present invention is to provide a kind of intelligent-tracking ball machine and tracking thereof, can reduce calculated amount and the dependence to scene, effectively improve accuracy, real-time and the intuitive from motion tracking.
For solving the problems of the technologies described above, embodiments of the present invention provide a kind of tracking of intelligent-tracking ball machine, comprise the following steps:
According to the picture of video acquisition unit gained, detect mobile object;
If mobile object detected, calculate this mobile object to the distance of specified point;
If periodically calculate the distance of gained, comprise predefined alarm distance, this mobile object is followed the tracks of.
Embodiments of the present invention also provide a kind of intelligent-tracking ball machine, comprising:
Video acquisition unit;
Detecting unit, for detecting mobile object according to the picture of video acquisition unit gained;
Range cells, for when detecting unit detects mobile object, calculates mobile object to the distance of specified point;
Tracking cell, for comprising predefined alarm distance in the situation that range cells is periodically calculated the distance of gained, follows the tracks of this mobile object.
Compared with prior art, the key distinction and effect thereof are embodiment of the present invention:
Detect after mobile object, calculate mobile object to the distance of specified point, according to this, apart from determining whether trigger tracking, can reduce calculated amount, and the triggering of following the tracks of is more accurate.
Further, distance-finding method based on video acquisition unit image-forming principle and coordinate conversion principle, can record the range information of any point in current picture, not only can measure object under test to the distance of video acquisition unit, can also measure in picture the distance between any two points on the height of object on ground level or ground level, more comprehensively effectively, practicality is stronger for the range information obtaining.Need to or not depend on the relation between focusing, zoom by means of existing coordinate system, can effectively get rid of external interference factor, strong adaptability.
Further, drawing equidistant curve can make user judge relatively having more intuitively of alarming line with the alarming line of following the tracks of triggering.
Further, the present invention is by calculating distance, rather than use associated picture processing and Video Detection Algorithm based on monitored picture to judge whether trigger alarm line, largely, reduce algorithm complex and the dependence to scene, can effectively improve accuracy, real-time, intuitive from motion tracking.
Further, according to this speed of moving body, shrink the tracking multiplying power of video acquisition unit or the following rate of raising ball machine, can guarantee like this continuation and the validity of following the tracks of.
Further, while being interrupted according to quilt, according to by the just height change of tracking object change tracking multiplying power before and after being interrupted, guarantee that tracking object is all the time in picture center, thereby effectively follow the tracks of.
Accompanying drawing explanation
Fig. 1 is the schematic flow sheet of a kind of tracking of intelligent-tracking ball machine in first embodiment of the invention;
Fig. 2 is the schematic flow sheet of a kind of tracking of intelligent-tracking ball machine in second embodiment of the invention;
Fig. 3 is that tracking ball machine is overlooked illustraton of model;
Fig. 4 is the distance range schematic diagram of the current picture of tracking ball machine;
Fig. 5 is equidistant curve synoptic diagram;
Fig. 6 triggers equidistant curve schematic diagram;
Fig. 7 is that video acquisition unit is demarcated geometric model figure;
Fig. 8 is the schematic diagram of mobile object trigger alarm curve;
Fig. 9 is the structural representation of a kind of intelligent-tracking ball machine in third embodiment of the invention;
Figure 10 is the structural representation of a kind of intelligent-tracking ball machine in four embodiment of the invention.
Embodiment
In the following description, in order to make reader understand the application better, many ins and outs have been proposed.But, persons of ordinary skill in the art may appreciate that even without these ins and outs and the many variations based on following embodiment and modification, also can realize each claim of the application technical scheme required for protection.
For making the object, technical solutions and advantages of the present invention clearer, below in conjunction with accompanying drawing, embodiments of the present invention are described in further detail.
First embodiment of the invention relates to a kind of tracking of intelligent-tracking ball machine.Fig. 1 is the schematic flow sheet of the tracking of this intelligent-tracking ball machine.
Specifically, as shown in Figure 1, the tracking of this intelligent-tracking ball machine mainly comprises the following steps:
In step 101, according to the picture of video acquisition unit gained, detect mobile object.In each embodiment of the present invention, video acquisition unit refers to the parts that obtain video pictures in ball machine, also can have other title, as camera, and picture catching module etc.
After this enter step 102, judge whether to detect mobile object.
If so, enter step 103; If not, again get back to step 102.
If mobile object detected, calculate this mobile object to the distance of specified point.
In present embodiment, preferentially specified point is intelligent-tracking ball machine itself.Be appreciated that in some other examples of the present invention, specified point can be also the point outside intelligent-tracking ball machine, for example, can be the location point at the target place that will protect, the central point of door for example, the central point of showpiece, seat of VIP etc.
In step 103, calculate mobile object to the distance of specified point.
After this enter step 104, the distance that gained is periodically calculated in judgement comprises predefined alarm distance.For example, periodically calculate distance, the N time distance is X, the N+1 time distance is Z, and alarm distance is Y, X > Y > Z wherein, is judged to be the distance that periodically calculates gained and comprises predefined alarm distance.
If so, enter step 105; If not, again get back to step 104.
If periodically calculate the distance of gained, comprise predefined alarm distance, this mobile object is followed the tracks of.
In step 105, follow the tracks of mobile object.After this process ends.
Detect after mobile object, calculate mobile object to the distance of specified point, according to this, apart from determining whether trigger tracking, can reduce calculated amount, and the triggering of following the tracks of is more accurate.
Second embodiment of the invention relates to a kind of tracking of intelligent-tracking ball machine.Fig. 2 is the schematic flow sheet of the tracking of this intelligent-tracking ball machine.The second embodiment improves on the basis of the first embodiment, and main improvements are: by following formula
X = ( Z - H ) · u v cos Φ - f sin Φ
Y = Z ( v sin Φ + f cos Φ ) - H · v / sin Φ v cos Φ - f sin Φ - - - ( 3 )
Z＝Z
Carry out the coordinate of computed image coordinate system under world coordinate system, then calculate mobile object to the distance of specified point according to the coordinate under world coordinate system.
Wherein f is intrinsic parameter, refers to the focal length of video acquisition unit, H, and Φ is outer parameter, refer to respectively the height of video acquisition unit hypothesis and the vertical angle of video acquisition unit and ground level, (u, v) fastens point coordinate for image coordinate, (X, Y, Z) fastens coordinate a little for world coordinates.
Specifically, as shown in Figure 2, the tracking of this intelligent-tracking ball machine mainly comprises the following steps:
In step 201, tracking ball machine is installed, obtain ball machine height H.
After this enter step 202, formulate tracking strategy, comprise and determine triggering tracking mode and trigger tracking range.
After this enter step 203, determine maximum distance and the near distance spot in current picture, obtain distance range and draw equidistant curve.
When tracking ball moves to a certain position, now need to obtain the distance range of current picture.The distance range of current picture depends on apart from the distance of ball machine solstics and closest approach, is analyzed as follows:
Fig. 3 is that tracking ball machine is overlooked illustraton of model.Because the pixel in not all image all has projection on surface level, must find the position that has the critical point of intersection point with ground level.Middle d is approximately picture at video acquisition unit Charge Coupled Device (CCD) (Charge Coupled Device as shown, abbreviation " CCD ") projection in vertical direction, f is focal length, Φ is the horizontal sextant angle of ball machine, and d=f*tg (Φ), suppose that video acquisition unit charge coupled device ccd vertical dimension is D, can draw the image coordinate in image space solstics in video acquisition unit charge coupled device ccd plane, according to the relation of d and D
If f*tg (Φ) > is D, normalized image coordinate in solstics, position is A (0.5,0.5) and A ' (0.5,0.5) middle distance person far away;
Otherwise solstics, position normalized image coordinate is B (0.5, f*tg (Φ)/D) and B ' (0.5, f*tg (Φ)/D) middle distance person far away.
Apart from the nearest person of ball machine, be lower-left angle point C (0.5 ,-0.5) and C ' (0.5 ,-0.5) medium machine close together person.As shown in Figure 4, A, A ', B, B ', C, C ' coordinate can be obtained in substitution range finding formula (3) respectively current picture farthest and minimum distance.
According to the distance range of current picture, according to equidistant principle counter pushing away of formula (3) of finding range, draw equidistant series of points, thereby draw some equidistant curves at current picture, wherein the point on arbitrary curve all equates to the distance of ball machine, as shown in Figure 5, wherein A, B, C, D, E are equidistant curve, and every adjacent two curve spacings are what determine.
After this enter step 204, in current picture, draw according to actual needs the equidistant curve that triggers tracking range.
After completing, equidistant curve can in picture, draw visually triggering equidistant curve according to triggering distance requirement, as shown in Figure 6,
Wherein red curve is for triggering equidistant curve.
Drawing equidistant curve and following the tracks of the alarming line triggering to make user judge relatively having more intuitively of alarming line.
In addition, being appreciated that distance is the distance that mobile object arrives video acquisition unit, can be also that mobile object is to the distance of specified point.Equidistant curve is booster action, also can not draw.Also can not draw according to actual needs and follow the tracks of the alarming line triggering.Can first draw equidistant curve and draw again alarming line, also can directly draw alarming line.
After this enter step 205, start moving Object Detection algorithm, obtain the change of distance information of all mobile objects in current picture.
Specifically, as shown in Figure 7, in the drawings, there is image coordinate system UO ' V, video acquisition unit coordinate system XcYcOcZc, world coordinate system XYOZ.The optical axis that wherein O ' Oc is video acquisition unit, the focus that Oc is video acquisition unit, O ' is the intersection point of optical axis and CCD imaging target surface, O is the intersection point of optical axis and ground level.Also have in the drawings the inside and outside parameter of video acquisition unit, wherein f is intrinsic parameter, refers to the focal length of video acquisition unit, H, Φ is outer parameter, refers to respectively the height of video acquisition unit hypothesis and the vertical angle of video acquisition unit and ground level, if it is (u that image coordinate is fastened point coordinate, v), the point coordinate on video acquisition unit coordinate system is (Xc, Yc, Zc), the coordinate that world coordinates is fastened is a little (X, Y, Z).
According to the image-forming principle of video acquisition unit, the transforming relationship that obtains image coordinate system and video acquisition unit coordinate system is:
Xc = - Zc f u
Yc = - Zc f v - - - ( 1 )
Zc＝Zc
U wherein, v, f is for through the normalized coordinate of unit and focal length value.
Separately, according to the translation rotation transforming relationship between three-dimensional system of coordinate, can obtain the three-dimensional transforming relationship of video acquisition unit coordinate system and world coordinate system, that is:
X c＝X
Y c＝ZcosΦ+YsinΦ (2)
Z c＝Zsin-Ycos-H/sinΦ
Formula (1) and formula (2) combination can be obtained to some coordinate under world coordinate system in image coordinate system,
X = ( Z - H ) · u v cos Φ - f sin Φ
Y = Z ( v sin Φ + f cos Φ ) - H · v / sin Φ v cos Φ - f sin Φ - - - ( 3 ) , So just having obtained image coordinate system and the world sits
Z＝Z
The transforming relationship formula of mark system.
Distance-finding method based on video acquisition unit image-forming principle and coordinate conversion principle, can record the range information of any point in current picture, not only can measure object under test to the distance of video acquisition unit, can also measure in picture the distance between any two points on the height of object on ground level or ground level, more comprehensively effectively, practicality is stronger for the range information obtaining.Need to or not depend on the relation between focusing, zoom by means of existing coordinate system, can effectively get rid of external interference factor, strong adaptability.
In addition, be appreciated that in some other examples of the present invention, also can find range by the conversion between other coordinate system, such as by polar coordinates etc., corresponding computing formula also can be different.
After this enter step 206, judged whether object cross-line.
If so, enter step 207; If not, again get back to step 206.
In step 207, start track algorithm cross-line object is followed the tracks of.
Periodically calculate mobile object apart from the distance of ball machine, if mobile object is crossed over alarming line, follow the tracks of.Here the cycle can be the time interval of 1 second, 2 seconds, 3 seconds etc. satisfying the demand arbitrarily.
Specifically, as shown in Figure 8, as t constantly exists a mobile object at position C in picture, at this object of the t ' moment, move to position C ', algorithm detects the variation of this object space and by this object marker, is trigger alarm curve object is followed the tracks of.
The present invention is by calculating distance, rather than use associated picture processing and Video Detection Algorithm based on monitored picture to judge whether trigger alarm line, largely, reduce algorithm complex and the dependence to scene, can effectively improve accuracy, real-time, intuitive from motion tracking.
After this enter step 208, judge that whether tracked object ratio in picture is suitable.
If so, enter step 209; If not, enter step 210.
In step 209, keep suitable multiplying power to follow the tracks of.After this process ends.
In step 210, Zoom lens obtains suitable tracking multiplying power.After this again get back to step 208.
In each embodiment of the present invention, " multiplying power " or " tracking multiplying power " refers to the convergent-divergent multiplying power of camera lens to video image, is also " Zoom " function.
Once object detected, touch line, start automatic track algorithm and carry out the tracking to this object.In reality is followed the tracks of, having important problem is that how to confirm is followed the tracks of multiplying power and tracking velocity, and this point also can utilize location algorithm to solve.
1. according to the distance of detecting object distance ball machine, can obtain this object of which movement trend and movement velocity, suitably adjust following rate guarantee continuation and the validity of following the tracks of according to the variation of trend and speed, concrete steps are as follows:
First ball machine its horizontal and vertical direction under a certain multiplying power respectively has best tracking velocity Va, a Vb, and the performances such as the motor that this speed is selected with ball machine motion parts, Timing Belt are relevant, can be used as empirical value and use in this patent.
After starting the tracking of object, can the current mobile object of Real-time Obtaining apart from the distance of ball machine, according to the variation of object distance ball machine distance in the unit interval, can obtain the approximate uniform motion speed of this object.
The variation of putting with variation and this time period endosphere seat in the plane of ball machine distance within the unit interval according to tracked object, the speed of tracked object is decomposed into the direction and the direction that is parallel to equal space line of vertical equal space line, obtain respectively the speed of this both direction, if these two speed are respectively Vh and Vv, and be translated into movement velocity Vha, the Vvb of ball machine horizontal direction and vertical direction.
Best tracking velocity Va, Vb under Vha, Vvb and this multiplying power are contrasted, if gap within the acceptable range, ball organizational security is held existing speed and multiplying power is followed the tracks of, if speed is higher, dwindle ball machine multiplying power, the ball machine multiplying power that increases as on the low side in speed, dwindles and the multiplying power that increases and the variation near-linear dependency of tracking object movement velocity.
According to the difference of detecting the distance of object distance ball machine in the unit interval, calculate speed of moving body, according to speed, shrink the multiplying power of video acquisition unit or the following rate of raising ball machine, can guarantee like this continuation and the validity of following the tracks of, the unit interval can be 0.5 second, 1 second, 2 seconds etc.
2. according to the computing method of object height, can obtain the true altitude information of current tracking object, and in conjunction with this object shared picture in picture, recently adjust the zoom speed of ball machine, guarantee that tracking object is all the time in picture center, thereby effectively follow the tracks of, this algorithm is mainly used in to be followed the tracks of in situation about being interrupted, and concrete steps are as follows:
When using track algorithm to follow the tracks of object, the true altitude of current object can obtain in real time, if followed the tracks of, is not interrupted, and significant change should not occur the height of object, and now this algorithm does not change tracking multiplying power;
As followed the tracks of, interrupted, algorithm judgement interrupts the height change of the tracked object in front and back, if tracking object step-down increases and follows the tracks of multiplying power; If tracking object uprises, dwindle tracking multiplying power; If tracking object height does not change, keep following the tracks of multiplying power.
According to the computing method of object height, obtain the true altitude information of current tracking object, and in conjunction with this object shared picture in picture, recently adjust the zoom speed of ball machine.Guarantee that tracking object is all the time in picture center, thereby effectively follow the tracks of.
In addition, be appreciated that this algorithm is mainly used in to follow the tracks of in situation about being interrupted, also can be with in other cases.
Each method embodiment of the present invention all can be realized in modes such as software, hardware, firmwares.No matter the present invention realizes with software, hardware or firmware mode, instruction code can be stored in the storer of computer-accessible of any type (for example permanent or revisable, volatibility or non-volatile, solid-state or non-solid-state, fixing or removable medium etc.).Equally, storer can be for example programmable logic array (Prog rammable Array Logic, be called for short " PAL "), random access memory (Random Access Memory, be called for short " RAM "), programmable read only memory (Programmable Read Only Memory, be called for short " PROM "), ROM (read-only memory) (Read-Only Memory, be called for short " ROM "), Electrically Erasable Read Only Memory (Electrically Erasable Prog rammable ROM, be called for short " EEPROM "), disk, CD, digital versatile disc (Digital Versatile Disc, be called for short " DVD ") etc.
Third embodiment of the invention relates to a kind of intelligent-tracking ball machine.Fig. 9 is the structural representation of this intelligent-tracking ball machine.As shown in Figure 9, this intelligent-tracking ball owner will comprise:
Video acquisition unit.
Detecting unit, for detecting mobile object according to the picture of video acquisition unit gained.
Range cells, for when detecting unit detects mobile object, calculates mobile object to the distance of specified point.
Tracking cell, for comprising predefined alarm distance in the situation that range cells is periodically calculated the distance of gained, follows the tracks of this mobile object.
The first embodiment is the method embodiment corresponding with present embodiment, present embodiment can with the enforcement of working in coordination of the first embodiment.The correlation technique details of mentioning in the first embodiment is still effective in the present embodiment, in order to reduce repetition, repeats no more here.Correspondingly, the correlation technique details of mentioning in present embodiment also can be applicable in the first embodiment.
Four embodiment of the invention relates to a kind of intelligent-tracking ball machine.Fig. 4 is the structural representation of this intelligent-tracking ball machine.
The 4th embodiment improves on the basis of the 3rd embodiment, and main improvements are: range cells is first by following formula
X = ( Z - H ) · u v cos Φ - f sin Φ
Y = Z ( v sin Φ + f cos Φ ) - H · v / sin Φ v cos Φ - f sin Φ - - - ( 3 )
Z＝Z
Carry out the coordinate of computed image coordinate system under world coordinate system.And then according to the coordinate under world coordinate system, calculate mobile object to the distance of specified point.
Wherein f is intrinsic parameter, refers to the focal length of video acquisition unit, H, Φ is outer parameter, refer to respectively the height of video acquisition unit hypothesis and the vertical angle of video acquisition unit and ground level, establish image coordinate and fasten point coordinate for (u, v), point coordinate on video acquisition unit coordinate system is (Xc, Yc, Zc), the coordinate that world coordinates is fastened is a little (X, Y, Z).
Formula 3 is a kind of preferred range finding formula in the present invention, also can use other range finding formula without form, for example, can in polar coordinates, in conjunction with the derivation mode that is similar to formula 1 and formula 2 mentioned above, derive the range finding formula of another kind of form.
Figure 10 is the structural representation of this intelligent-tracking ball machine.As shown in figure 10, this intelligent-tracking ball machine has also increased with lower unit on the basis of the 3rd embodiment:
Distance range acquiring unit, for determining most distant point and the near distance spot of current picture, obtains the distance range of current picture.
Setting-out unit, draws equidistant curve in the distance range for the current picture that obtains at distance range acquiring unit, and draws and follow the tracks of the alarming line triggering according to predefined alarm distance.
The first ball machine regulon, calculates speed of moving body for calculate the difference of the distance of gained according to range cells in the unit interval, according to speed, shrinks the tracking multiplying power of video acquisition unit or the following rate of raising ball machine.
The second ball machine regulon, for according to the computing method of object height, obtains the true altitude information of current object, and binding object shared picture in picture is recently adjusted the zoom speed of ball machine.
The second embodiment is the method embodiment corresponding with present embodiment, present embodiment can with the enforcement of working in coordination of the second embodiment.The correlation technique details of mentioning in the second embodiment is still effective in the present embodiment, in order to reduce repetition, repeats no more here.Correspondingly, the correlation technique details of mentioning in present embodiment also can be applicable in the second embodiment.
It should be noted that, each unit of mentioning in each intelligent-tracking ball machine of the present invention is all logical block, physically, a logical block can be a physical location, also can be a part for a physical location, can also realize with the combination of a plurality of physical locations, the physics realization mode of these logical blocks itself is not most important, and the combination of the function that these logical blocks realize is the key that just solves technical matters proposed by the invention.In addition, for outstanding innovation part of the present invention, above-mentioned each intelligent-tracking ball machine of the present invention is not introduced the unit not too close with solving technical matters relation proposed by the invention, and this does not show that above-mentioned intelligent-tracking ball machine does not exist other unit.
Although pass through with reference to some of the preferred embodiment of the invention, the present invention is illustrated and described, but those of ordinary skill in the art should be understood that and can do various changes to it in the form and details, and without departing from the spirit and scope of the present invention.

Claims (11)

1. a tracking for intelligent-tracking ball machine, is characterized in that, comprises the following steps:
According to the picture of video acquisition unit gained, detect mobile object;
If mobile object detected, according to following formula
X = ( Z - H ) · u v cos Φ - f sin Φ
Y = Z ( v sin Φ + f cos Φ ) - H · v / sin Φ v cos Φ - f sin Φ
Z＝Z
The coordinate of mobile object in computed image coordinate system under world coordinate system, and calculate mobile object to the distance of specified point according to the coordinate under world coordinate system, wherein f is intrinsic parameter, refers to the focal length of video acquisition unit, H, Φ is outer parameter, refers to respectively the height of video acquisition unit hypothesis and the vertical angle of video acquisition unit and ground level, (u, v) for image coordinate is fastened point coordinate, (X, Y, Z) fastens coordinate a little for world coordinates;
If periodically calculate the distance of gained, comprise predefined alarm distance, this mobile object is followed the tracks of.
2. the tracking of intelligent-tracking ball machine according to claim 1, is characterized in that, described specified point is this intelligent-tracking ball machine.
3. according to the tracking of the intelligent-tracking ball machine described in any one in claim 1 to 2, it is characterized in that, the described picture according to video acquisition unit gained detects in the step of mobile object, comprises following sub-step:
Determine most distant point and the near distance spot in current picture, obtain the distance range of current picture;
According to described distance range, calculate and draw equidistant curve in the display that shows current picture;
In described display, calculate and draw the alarming line of following the tracks of triggering.
4. the tracking of intelligent-tracking ball machine according to claim 3, is characterized in that, at described most distant point and the near distance spot of determining in current picture, obtains in the step of distance range of current picture, also comprises following sub-step:
If f*tg (Φ) is >D, normalized image coordinate in solstics, position is A(-0.5,0.5) and A'(0.5,0.5) middle distance person far away;
Otherwise solstics, position normalized image coordinate is B(-0.5, f*tg (Φ)/D) and B'(0.5, f*tg (Φ)/D);
Wherein f is focal length, and Φ is the horizontal sextant angle of ball machine, and D is the vertical dimension of video acquisition unit charge-coupled image sensor.
5. the tracking of intelligent-tracking ball machine according to claim 4, is characterized in that, if the described distance that periodically calculates gained comprises predefined alarm distance, the step that this mobile object is followed the tracks of, comprises following sub-step:
Periodically calculate mobile object apart from the distance of ball machine, if mobile object is crossed over alarming line, follow the tracks of.
6. the tracking of intelligent-tracking ball machine according to claim 5, is characterized in that, in the step of described tracking mobile object, further comprising the steps of:
According to the difference of detecting the distance of object distance ball machine in the unit interval, calculate speed of moving body, according to this speed of moving body, change the tracking multiplying power of video acquisition unit or the following rate of change ball machine.
7. the tracking of intelligent-tracking ball machine according to claim 5, is characterized in that, in the step of described tracking mobile object, further comprising the steps of:
As followed the tracks of, interrupted, judgement interrupts the height change of the tracked object in front and back, if tracked object step-down, increase and follow the tracks of multiplying power, if tracked object uprises, dwindle tracking multiplying power, if tracked object height does not change, keep following the tracks of multiplying power.
8. an intelligent-tracking ball machine, is characterized in that, comprising:
Video acquisition unit;
Detecting unit, for detecting mobile object according to the picture of video acquisition unit gained;
Range cells, for when described detecting unit detects mobile object, according to following formula
X = ( Z - H ) · u v cos Φ - f sin Φ
Y = Z ( v sin Φ + f cos Φ ) - H · v / sin Φ v cos Φ - f sin Φ - - - ( 3 )
Z＝Z
The coordinate of computed image coordinate system under world coordinate system, and calculate mobile object to the distance of specified point according to the coordinate under world coordinate system, wherein f is intrinsic parameter, refers to the focal length of video acquisition unit, H, Φ is outer parameter, refers to respectively the height of video acquisition unit hypothesis and the vertical angle of video acquisition unit and ground level, (u, v) for image coordinate is fastened point coordinate, (X, Y, Z) fastens coordinate a little for world coordinates;
Tracking cell, for comprising predefined alarm distance in the situation that described range cells is periodically calculated the distance of gained, follows the tracks of this mobile object.
9. intelligent-tracking ball machine according to claim 8, is characterized in that, also comprises:
Distance range acquiring unit, for determining most distant point and the near distance spot of current picture, obtains the distance range of current picture;
Setting-out unit, for drawing equidistant curve in the distance range of the current picture that obtains at described distance range acquiring unit on the display that shows current picture, and draws and follows the tracks of the alarming line triggering according to predefined alarm distance.
10. the intelligent-tracking ball machine described in any one according to Claim 8 to 9, is characterized in that, also comprises:
The first ball machine regulon, calculates speed of moving body for calculate the difference of the distance of gained according to described range cells in the unit interval, according to this speed of moving body, changes the tracking multiplying power of described video acquisition unit or the following rate of change ball machine.
Intelligent-tracking ball machine in 11. according to Claim 8 to 9 described in any one, is characterized in that, also comprises:
The second ball machine regulon, the height change of tracked object before and after interrupting for judgement when tracking is interrupted, if tracked object step-down, increase and follow the tracks of multiplying power, if tracked object uprises, dwindle tracking multiplying power, if tracked object height does not change, keep following the tracks of multiplying power.
CN201110233861.0A 2011-08-16 2011-08-16 Intelligent tracking ball machine and tracking method thereof Active CN102289820B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201110233861.0A CN102289820B (en) 2011-08-16 2011-08-16 Intelligent tracking ball machine and tracking method thereof
PCT/CN2012/076223 WO2013023474A1 (en) 2011-08-16 2012-05-29 Intelligent tracking dome camera and tracking method of same

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201110233861.0A CN102289820B (en) 2011-08-16 2011-08-16 Intelligent tracking ball machine and tracking method thereof

Publications (2)

Publication Number Publication Date
CN102289820A CN102289820A (en) 2011-12-21
CN102289820B true CN102289820B (en) 2014-04-02

Family

ID=45336209

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201110233861.0A Active CN102289820B (en) 2011-08-16 2011-08-16 Intelligent tracking ball machine and tracking method thereof

Country Status (2)

Country Link
CN (1) CN102289820B (en)
WO (1) WO2013023474A1 (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102289820B (en) * 2011-08-16 2014-04-02 杭州海康威视数字技术股份有限公司 Intelligent tracking ball machine and tracking method thereof
CN107846549A (en) * 2016-09-21 2018-03-27 杭州海康威视数字技术股份有限公司 A kind of method for tracking target, apparatus and system
CN107659790A (en) * 2017-10-23 2018-02-02 上海集光安防科技股份有限公司 A kind of method of ball machine automatically track target
CN114071362A (en) * 2021-11-05 2022-02-18 国网江苏省电力有限公司电力科学研究院 Multi-target dynamic monitoring method, device, equipment and medium

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1725266A (en) * 2004-07-21 2006-01-25 上海高德威智能交通系统有限公司 Vehicle intelligent monitoring recording syste mand method based on video trigger and speed measuring
CN1977563A (en) * 2004-06-30 2007-06-06 Sk电信有限公司 Method of providing alarm service upon movement out of safety zone

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101098465A (en) * 2007-07-20 2008-01-02 哈尔滨工程大学 Moving object detecting and tracing method in video monitor
US8761434B2 (en) * 2008-12-17 2014-06-24 Sony Computer Entertainment Inc. Tracking system calibration by reconciling inertial data with computed acceleration of a tracked object in the three-dimensional coordinate system
CN101616307B (en) * 2009-07-08 2013-03-20 宝鸡市公安局 Video target tracking selection output method and system
CN101969548B (en) * 2010-10-15 2012-05-23 中国人民解放军国防科学技术大学 Active video acquiring method and device based on binocular camera shooting
CN102289820B (en) * 2011-08-16 2014-04-02 杭州海康威视数字技术股份有限公司 Intelligent tracking ball machine and tracking method thereof

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1977563A (en) * 2004-06-30 2007-06-06 Sk电信有限公司 Method of providing alarm service upon movement out of safety zone
CN1725266A (en) * 2004-07-21 2006-01-25 上海高德威智能交通系统有限公司 Vehicle intelligent monitoring recording syste mand method based on video trigger and speed measuring

Also Published As

Publication number Publication date
CN102289820A (en) 2011-12-21
WO2013023474A1 (en) 2013-02-21

Similar Documents

Publication Publication Date Title
Kilicarslan et al. Predict vehicle collision by TTC from motion using a single video camera
EP2948927B1 (en) A method of detecting structural parts of a scene
US9547795B2 (en) Image processing method for detecting objects using relative motion
Held et al. Precision tracking with sparse 3d and dense color 2d data
CN104902246A (en) Video monitoring method and device
CN104378582A (en) Intelligent video analysis system and method based on PTZ video camera cruising
CN102542843A (en) Early warning method for preventing vehicle collision and device
CN102289820B (en) Intelligent tracking ball machine and tracking method thereof
CN104331901A (en) TLD-based multi-view target tracking device and method
CN105374208B (en) Road conditions are reminded and camera detection method and device thereof
CN104966062A (en) Video monitoring method and device
CN105306892A (en) Method for generating and displaying ship video in evidence chain form
CN104021571A (en) Moving target automatic tracking method and device for improving user experience
JPH09226490A (en) Detector for crossing object
JP2011513876A (en) Method and system for characterizing the motion of an object
CN105975923A (en) Method and system for tracking human object
CN205680289U (en) A kind of parking detecting system
CN104063863A (en) Pitch-down type binocular vision system for watercourse monitoring and image processing method
EP2587462A1 (en) Image monitoring system and method with false alarm rate reduction
CN105141887A (en) Submarine cable area video alarming method based on thermal imaging
CN101782386B (en) Non-visual geometric camera array video positioning method and system
Ren et al. Lane Detection on the iPhone
CN109977796A (en) Trail current detection method and device
KR101032098B1 (en) Stand-alone environment traffic detecting system using thermal infra-red
Kilicarslan et al. Direct vehicle collision detection from motion in driving video

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