CN108876821A - Across camera lens multi-object tracking method and system - Google Patents
Across camera lens multi-object tracking method and system Download PDFInfo
- Publication number
- CN108876821A CN108876821A CN201810729547.3A CN201810729547A CN108876821A CN 108876821 A CN108876821 A CN 108876821A CN 201810729547 A CN201810729547 A CN 201810729547A CN 108876821 A CN108876821 A CN 108876821A
- Authority
- CN
- China
- Prior art keywords
- target
- camera
- picture
- current picture
- dimensional coordinate
- 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.)
- Granted
Links
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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
- H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
Abstract
This disclosure relates to which a kind of across camera lens multi-object tracking method for football, includes the following steps:1, multiple cameras respectively shoot the picture at current time, form multiple current pictures;2, the target occurred in each picture of the multiple current picture is detected respectively, records two-dimensional coordinate of the target in each current picture;3, two-dimensional coordinate of each target in each current picture is converted into public three-dimensional coordinate;4, all targets in the current picture of the multiple camera are matched with the target of previous picture;5, for each target after matching, multiple three-dimensional coordinates according to determined by the multiple camera determine the final three-dimensional coordinate of the target;6, the picture of each camera shooting subsequent time, as new current picture, and repeats 2 to 5, obtains multiple three-dimensional coordinates of the target in the continuous moment, form the three-dimensional track of the target.
Description
Technical field
The disclosure is originally related to computer image processing technology field, more particularly, to across the camera lens multiple target tracking of one kind
Method and system.
Background technique
Video frequency object tracking refers to the initial position to set the goal in video, then exports the target in video every
The position at a moment.Object tracking is an important problem in computer vision, usually the first step of video analysis processing.
Currently, the research for having a large amount of scholars to be engaged in object tracking, the algorithm of numerous effective object trackings is suggested.?
Under some monitoring scenes, needs under a complicated scene while tracking multiple objects.Mutually blocking between multiple objects
The difficulty of object tracking is increased, this point often occurs in the tracking of pedestrian.When a stack of people appears in picture pick-up device simultaneously
It is overlapped so that its physical location can not be obtained accurately between everyone when in picture.
Multi-target tracking method is broadly divided into two classes:Multi-target tracking based on single camera and based on more camera shooting cameras
Multi-target tracking method.
Multi-target tracking method based on multi-cam is primarily upon the data fusion for how carrying out multi-cam at present,
Mainly there is the method for method and characteristic matching based on picture pick-up device calibration.Method based on picture pick-up device calibration mainly utilizes
Picture pick-up device projection matrix, will be in different picture pick-up device image projections to the same picture.For the side based on characteristic matching
Method mainly improves matching result by finding efficient appearance features and space time information.The tracing problem of more picture pick-up devices
Due to having biggish illumination and visual angle difference between different camera lenses, compared to the tracking problem of picture pick-up device, there is bigger challenge
Property.Due to the complex nature of the problem, existing across camera tracking has comparable error due to various reasons.
Football/match video is analyzed and is handled, be the application of above-mentioned multi-target tracking method scene it
One, since football has extensive audient, many business and practical value can be brought.For example, from the angle of spectators' vision
Degree, relay need to be added various visual effects to meet the vision requirement of spectators;From angle is researched and analysed, team's exchange can be mentioned
The related data of team member in match video is taken out to assist to carry out technical-tactics analyzing and research, team is helped to promote competitiveness;This
Outside, the copyright side and sponsor etc. of sports cast are also required to fully excavate out the business opportunity contained in sports event broadcast.
However, further include football other than sportsman/referee since the moving target in football match is more, in addition
The influence of illumination in weather and field proposes huge challenge to the tracking speed and tracking accuracy of all kinds of track algorithms, if
It is not able to satisfy the demand of real-time and control errors, then is difficult to realize business application.
Summary of the invention
In view of the above problem of the prior art, inventor is made that the present invention, passes through the corner cloth in football pitch
If multiple video cameras, using across camera correlating method, come the position of the target (sportsman and ball) in real-time tracking court.This hair
It is bright to have used exclusive camera calibration and coordinate conversion, enable two-dimensional tracking position to be directly presented on three-dimensional space, and
It can directly correspond in unified coordinate system, across the camera association after convenience, and use people's ball based on deep learning
The testing result of single camera target following is connected into higher tracking of confidence level by detection method.
According to an embodiment of the invention, providing a kind of across camera lens multi-object tracking method for football, wherein
Different location in football place is equipped with multiple cameras, and the visual field of the multiple camera can collectively cover whole
A football place, wherein across the camera lens multi-object tracking method includes the following steps:Step 1, the multiple camera
The respectively picture at shooting current time, forms multiple current pictures;Step 2 is detected respectively in the every of the multiple current picture
The target occurred in a picture records two-dimensional coordinate of the target in each current picture;Step 3 exists each target
Two-dimensional coordinate in each current picture is converted to public three-dimensional coordinate;Step 4, by the current picture of the multiple camera
In all targets matched with the target of previous picture;Step 5, for matching after each target, according to the multiple
Multiple three-dimensional coordinates determined by camera determine the final three-dimensional coordinate of the target;Step 6, in tracking cycle, it is next
Moment to as current time, repeats step 1 to 5, thus obtains multiple three-dimensionals of the target in the continuous moment
Coordinate forms the three-dimensional track of the target.
Beneficial effects of the present invention essentially consist in that:
1) in being associated with across camera, while in view of target appearance and three-dimensional geometry relationship carry out object matching, in this way
Result it is more acurrate, target exchange a possibility that be also reduced to it is minimum;
2) when considering target position confidence level, target depth, visual ratio and the information apart from edge distance are combined
Carry out associated data, keeps the unified goal position finally calculated more reasonable;
3) for the tracking of football, judge football whether in the sky using the variance of the target position estimation between camera
Flight removes the biggish aerial football position of error, and which greatly enhances the flexibilities of people's ball analysis.
Detailed description of the invention
Fig. 1 and 2 is the part flow diagram across camera lens multi-object tracking method according to the embodiment of the present invention;
Fig. 3 is the functional block diagram across camera lens multiple-target system according to the embodiment of the present invention;
Fig. 4 is the schematic diagram according to the running environment of the system for being mounted with application program of the embodiment of the present invention.
Specific embodiment
In the following, being described in further detail in conjunction with attached drawing to the implementation of technical solution.
It will be appreciated by those of skill in the art that although the following description is related to many of embodiment for the present invention
Technical detail, but be only for not meaning that any restrictions for illustrating the example of the principle of the present invention.The present invention can be applicable in
In the occasion being different from except technical detail exemplified below, without departing from the principle and spirit of the invention.
It, may be to can be in description in the present specification in addition, tedious in order to avoid being limited to the description of this specification
The portion of techniques details obtained in prior art data has carried out the processing such as omission, simplification, accommodation, this technology for this field
It will be understood by for personnel, and this will not influence the open adequacy of this specification.
Hereinafter, description is used to carry out the embodiment of the present invention.Note that by description is provided with following order:1, it sends out
The summary of bright design;2, across camera lens multi-object tracking method (Fig. 1,2);3, across camera lens multiple-target system (Fig. 3);4, basis
The system (Fig. 4) for being mounted with application program of the embodiment of the present invention.
1, the summary of inventive concept
It is used to realize working environment of the invention firstly, illustrating, that is, exploitation and running environment are as follows:
1) in terms of exploitation, entire program can be built under Linux environment using C Plus Plus, can directly be carried out with g++
Compiling;
2) in terms of operation, common operating system, such as Windows, Linux is supported, is reached in the case where CPU is used only
To real-time operation.And consider target detection when, can also have GPU support in the case where real-time perfoming.
Firstly, in order to realize the present invention, needs that multiple cameras are installed in court and (such as taken the photograph positioned at four of each corner flag
As head), after being corrected to each camera and going deformation, the two-dimensional coordinate for the picture that each camera is respectively shot
It can be reflected and be mapped to the respective three-dimensional system of coordinate of each camera.Wherein, the conversion of above-mentioned 2 d-to-3 d is also referred to as anti-
Mapping, can be used all kinds of existing methods.
In addition, each respective three-dimensional system of coordinate of camera can be switched to a public three-dimensional system of coordinate (overall situation respectively
Coordinate system, or referred to as world coordinate system).
As a result, by above-mentioned two layers conversion, the two-dimensional coordinate for the picture that each camera is respectively shot can be converted
To public three-dimensional coordinate.
Next, carry out feature identification by shooting picture to each camera, filter out as sportsman target and
As the target (optionally, target of the exclusion as referee) of football, (tracking is tracked respectively to these two types of targets
It will be described in more detail below).
Specifically, when some target (sportsman or football) do not appeared in due to blocking some/certain several camera shooting
When in head, by may continue to the target that tracking is blocked using the complementarity between multiple camera shooting angle, thus real
Existing target is continuously tracked.
In the following, in conjunction with the embodiments come illustrate foregoing invention design realization.
2, across camera lens multi-object tracking method
Fig. 1 and 2 is the overall procedure schematic diagram across camera lens multi-object tracking method according to the embodiment of the present invention.
The embodiment provides a kind of across camera lens multi-object tracking methods for football, wherein in foot
The different location on ball athletic ground is equipped with multiple cameras, wherein the visual field of the multiple camera can collectively cover
Entire football place, the multiple camera are commonly connected to central processing unit, wherein target to be identified includes foot
The personnel on the scene on ball athletic ground and football,
It the described method comprises the following steps:
Step S100, the multiple camera respectively shoots the picture at current time, forms multiple current pictures;
Step S200, the target occurred in each picture of the multiple current picture is detected respectively, records the mesh
The two-dimensional coordinate being marked in each current picture;
Step S300, two-dimensional coordinate of each target in each current picture is converted into public three-dimensional coordinate;
Step S400, by the target progress of all targets and previous picture in the current picture of the multiple camera
Match;
Step S500, for each target after matching, multiple three-dimensional coordinates according to determined by the multiple camera,
Determine the final three-dimensional coordinate of the target;
Step S600, the picture of each camera shooting subsequent time, and step S100 to S500 is repeated, obtain institute
Multiple three-dimensional coordinates of the target in the continuous moment are stated, the three-dimensional track of the target is formed.
Optionally, before step S100, position correction is carried out to each camera and removes distortion correction, position correction is used
In making the visual field of camera all fall in effective coverage as far as possible, the invalid visual field of elimination/reduction, removing distortion correction includes that flake is gone to become
Shape, keystone, etc..
Wherein, the target includes football and sportsman, may also include referee.After the step S200, this method is also
May include:
Step S210, according to the resemblance of target, the target detected can be distinguished, for example, being retained as foot
The target of ball and sportsman can exclude the target as referee;
Wherein, all kinds of common Image Feature Matching methods can be used in the target identification in two-dimensional picture;
For the target as sportsman, step S400 may include:
Step S410, for current picture, first the target in the first camera of the multiple camera is numbered,
The distance between each target in other cameras and three-dimensional coordinate of target in the first camera are calculated again, are taken the photograph other
As the target discrimination nearest with the target range in the first camera in head is same target, and target designation is unified;
Step S420, the matching degree of each target in current picture and some target in previous picture is successively calculated,
The calculation of the matching degree is as follows:
Matching degree=(the Batachelia distance for the histogram that 1.0- is made of the viewable portion of target)/(target is three-dimensional
The Euclidean distance of pin point/current goal depth)
Wherein, midpoint of the three-dimensional pin point expression target projection to ground;
Wherein, target depth indicate target to camera floor projection distance, that is, target pin point to camera just under
Euclidean distance just), the i.e. distance of target to camera,
It should be noted that it is appreciated that the number of the target detected in the current picture of some camera, possible
It is different from the number of target detected in previous picture, but this will not influence the tracking of target.
Specifically, some target may stop being tracked because of being blocked or leaving picture or be not detected, but
The target can at least be detected by least one of other cameras, and the three-dimensional information between these cameras is all shared
, so can be inferred according to shared three-dimensional information later when previous camera has the indefinite target of ID (number) to occur
Whether this is to stop the target of tracking before, and restore its ID.
Step S430, will be with some highest target identification of object matching degree in previous picture in current picture:
With in previous picture described in the identical target of some target;
For the target as sportsman, step S500 may include:
Step S510, for the current picture of each camera, the tracking confidence level of each target is calculated separately, it is as follows:
Confidence level=(target is at a distance from nearest edge by (square of 1/ target depth) * (square of target visual ratio) *
Square),
Wherein, target visual ratio be according to the mutual circumstance of occlusion of target and the target of determination appears in camera view
Area (that is, not by the region area of other target occlusions) account for the ratio of target itself gross area, i.e.,:
Effective area/target the gross area in the visual field of the target visual ratio=target in the visual field,
Wherein, nearest edge refer to four of rectangular image while mid-range objectives it is nearest while,
Above-mentioned distance refers to the corresponding distance of number of pixels, that is, the numerical value after coordinate conversion,
By above formula as it can be seen that confidence level declines with the increase of the depth of target, with target visual ratio
Decline and decline, as target declines close to edge.
Step S520, for each target in current picture, the target position of the highest camera of confidence level is selected, is made
For the final position of target (public three-dimensional coordinate).
In addition, for the target as football the matching and tracking mode different from sportsman can be carried out, for example, step
S400 may include:
Step S440, the standard deviation between the three-dimensional coordinate of the target determined by each camera is calculated;
If step S450, the described standard deviation is more than predetermined threshold, the target position of current picture, Zhi Houzhuan are not recorded
To step S600;
If step S460, the described standard deviation is less than predetermined threshold, the target position of the multiple camera can use
Average value goes to step S600 as the target position of current picture later;
The purpose of the step is as follows:Since football flies in the sky, in each camera, two-dimensional position is counter projects ground
Three-dimensional position on face is very big by difference, will generate large error, therefore a football is counter to project three-dimensional position by being added
Variance threshold value, to judge whether football flies in the sky.When football flight in the sky, rail that football is flown in the sky
Mark removes, and (is fitted/is inserted that is, passing through come the skyborne position of approximate insertion football with the position of flight path both ends on the ground
The mode of value supplies the tracing point of missing), to improve the accuracy of football tracking.
3, the system for being mounted with application program of embodiment according to the present invention
Fig. 3 is the functional block diagram across camera lens multiple-target system according to the embodiment of the present invention.
The embodiment provides a kind of across camera lens multiple-target system for football, it is used to execute
Above-mentioned across camera lens multi-object tracking method, across the camera lens multiple-target system is connected to multiple cameras, described across camera lens
Multiple-target system mainly includes that shooting picture obtains module, module of target detection, coordinate transferring, target identification mould
Block, object matching module, target tracking module,
Wherein, the shooting picture obtains the shooting picture that module is used to respectively obtain current time from the multiple camera
Face forms multiple current pictures;
The module of target detection is used to detect the target occurred in each picture of the multiple current picture respectively,
Record two-dimensional coordinate of the target in each current picture;
The coordinate transferring is public for being converted to two-dimensional coordinate of each target in each current picture
Three-dimensional coordinate;
The object matching module be used for by the current picture of the multiple camera all targets and previous picture
Target matched;
The target tracking module is used for:For each target after matching, according to determined by the multiple camera
Multiple three-dimensional coordinates determine the final three-dimensional coordinate of the target;
The target identification module is used for according to the resemblance of target, the target that will test out divide into football, sportsman,
Referee.
In addition, different embodiments of the invention by software module or can also be stored in one or more computer-readable
The mode of computer-readable instruction on medium is realized, wherein the computer-readable instruction is when by processor or equipment group
When part executes, different embodiment of the present invention is executed.Similarly, software module, computer-readable medium and Hardware Subdivision
Any combination of part is all expected from the present invention.The software module can be stored in any type of computer-readable storage
On medium, such as RAM, EPROM, EEPROM, flash memory, register, hard disk, CD-ROM, DVD etc..
4, the system for being mounted with application program of embodiment according to the present invention
Referring to Fig. 4, it illustrates the running environment of the system according to an embodiment of the present invention for being mounted with application program.
In the present embodiment, the system of the installation application program is installed and is run in electronic device.The electronics
Device can be desktop PC, notebook, palm PC and server etc. and calculate equipment.The electronic device may include but not
It is limited to memory, processor and display.This Figure only shows the electronic devices with said modules, it should be understood that
It is not required for implementing all components shown, the implementation that can be substituted is more or less component.
The memory can be the internal storage unit of the electronic device, such as electronics dress in some embodiments
The hard disk or memory set.The memory is also possible to the External memory equipment of the electronic device in further embodiments,
Such as the plug-in type hard disk being equipped on the electronic device, intelligent memory card (Smart Media Card, SMC), secure digital
(Secure Digital, SD) card, flash card (Flash Card) etc..Further, the memory can also both include institute
The internal storage unit for stating electronic device also includes External memory equipment.The memory is installed on the electronics dress for storing
The application software and Various types of data set, such as the program code etc. of the system for installing application program.The memory may be used also
For temporarily storing the data that has exported or will export.
The processor can be in some embodiments central processing unit (Central Processing Unit,
CPU), microprocessor or other data processing chips, for running the program code stored in the memory or processing data,
Such as execute the system etc. of the installation application program.
The display can be in some embodiments light-emitting diode display, liquid crystal display, touch-control liquid crystal display with
And OLED (Organic Light-Emitting Diode, Organic Light Emitting Diode) touches device etc..The display is for showing
Show the information handled in the electronic device and for showing visual customer interface, such as application menu interface, answers
With icon interface etc..The component of the electronic device is in communication with each other by system bus.
Through the above description of the embodiments, those skilled in the art is it will be clearly understood that above embodiment
In method can realize by means of software and necessary general hardware platform, naturally it is also possible to realized by hardware,
But the former is more preferably embodiment in many cases.Based on this understanding, the technical solution of the application of the present invention is substantially
The part that contributes to existing technology can be embodied in the form of Software Commodities in other words, which deposits
Storage in a storage medium (such as ROM/RAM, magnetic disk, CD), including some instructions are used so that terminal device (can be with
It is mobile phone, computer, server, air conditioner or the network equipment etc.) execute side described in each embodiment of the application of the present invention
Method.
That is, according to an embodiment of the invention, additionally provide a kind of computer readable storage medium, the computer
The program of the method for executing embodiment according to the present invention is stored on readable storage medium storing program for executing, described program is processed
When device executes, each step of the method is executed.
By upper, it will be appreciated that for illustrative purposes, specific embodiments of the present invention are described herein, still, can make
Each modification, without departing from the scope of the present invention.It will be apparent to one skilled in the art that drawn in flow chart step or this
In the operation that describes and routine can be varied in many ways.More specifically, the order of step can be rearranged, step can be executed parallel
Suddenly, step can be omitted, it may include other steps can make the various combinations or omission of routine.Thus, the present invention is only by appended power
Benefit requires limitation.
Claims (9)
1. a kind of across camera lens multi-object tracking method for football, wherein the different location in football place is pacified
Equipped with multiple cameras, the visual field of the multiple camera can collectively cover entire football place,
Wherein, across the camera lens multi-object tracking method includes the following steps:
Step 1, the multiple camera respectively shoot the picture at current time, form multiple current pictures;
Step 2 detects the target occurred in each picture of the multiple current picture respectively, records the target each
Two-dimensional coordinate in current picture;
Two-dimensional coordinate of each target in each current picture is converted to public three-dimensional coordinate by step 3;
Step 4 matches all targets in the current picture of the multiple camera with the target of previous picture;
Step 5, for each target after matching, multiple three-dimensional coordinates according to determined by the multiple camera determine institute
State the final three-dimensional coordinate of target;
Step 6, in tracking cycle, subsequent time to step 1 as current time, is repeated to 5, thus obtain described in
Multiple three-dimensional coordinates of the target in the continuous moment, form the three-dimensional track of the target.
2. across camera lens multi-object tracking method according to claim 1, wherein the target includes football and sportsman,
After the step 2, the method also includes:
Step 2-1, according to the resemblance of the target, can the target divide into football and sportsman.
3. across camera lens multi-object tracking method according to claim 2, wherein for the target as sportsman, step
Rapid 4 include:
Step 4-1, for current picture, first the target in the first camera of the multiple camera is numbered, then presses
In turn according to the target designation in the first camera, each target in other cameras and the mesh in the first camera are calculated
The distance between target three-dimensional coordinate, by the target discrimination nearest with the target range in the first camera in other cameras
For same target, and target designation is unified,
Wherein, the target numbers detected in first camera are most in all cameras.
4. across camera lens multi-object tracking method according to claim 3, wherein for the target as sportsman, step
Rapid 4 further include:
Step 4-2, the matching degree of each target in current picture and some target in previous picture is successively calculated;
Step 4-3, will be with some highest target identification of object matching degree in previous picture in current picture with it is previous
The identical target of some described target in picture,
Wherein, the calculation of the matching degree is as follows:
Matching degree=(the Batachelia distance for the histogram that 1.0- is made of the viewable portion of target)/(target three-dimensional pin point
Euclidean distance/current goal depth),
Wherein, three-dimensional pin point indicate target projection to ground midpoint,
Wherein, floor projection distance of the target depth expression target to camera, that is, immediately below target pin point to camera)
The distance of Euclidean distance, i.e. target to camera.
5. across camera lens multi-object tracking method according to claim 4, wherein for the target as sportsman, step
Rapid 5 further include:
Step 5-1, for the current picture of each camera, the tracking confidence level of each target is calculated separately;
Step 5-2, for each target in current picture, the target position of the highest camera of confidence level is selected, as mesh
Final position is marked,
Wherein, the calculation formula of the confidence level is as follows:
Confidence level=(square of 1/ target depth) × (square of target visual ratio) × (target is at a distance from nearest edge
Square),
Wherein, target visual ratio be according to the mutual circumstance of occlusion of target and the target of determination appears in the face in camera view
Product accounts for the ratio of target expected area in the visual field, i.e.,:
Expection area of effective area/target of the target visual ratio=target in the visual field in the visual field,
Wherein, nearest edge refer to four of rectangular image while mid-range objectives it is nearest while.
6. across camera lens multi-object tracking method according to claim 2, wherein for the target as football, step
Rapid 4 further include:
Step 4-4, the standard deviation between the three-dimensional coordinate of the target determined by each camera is calculated;
If step 4-5, the described standard deviation is more than predetermined threshold, determine that football for state of flight, does not record current picture
Target position goes to step 6 later;
If step 4-6, the described standard deviation is less than predetermined threshold, being averaged for the target position of the multiple camera can use
Value, as the target position of current picture, goes to step 6 later.
7. across camera lens multi-object tracking method according to claim 6, wherein for the target as sportsman, step
Suddenly 4-5 further includes:
Step 4-5-1, Unrecorded target position is supplied by way of linear fit or interpolation.
8. it is a kind of for executing across the camera lens multiple-target system to method described in any of 7 according to claim 1,
It is connected to the multiple camera, and across the camera lens multiple-target system includes that shooting picture obtains module, target detection
Module, coordinate transferring, target identification module, object matching module, target tracking module,
Wherein, the shooting picture obtains the shooting picture that module is used to respectively obtain current time from the multiple camera,
Form multiple current pictures;
The module of target detection is recorded for detecting the target occurred in each picture of the multiple current picture respectively
Two-dimensional coordinate of the target in each current picture;
The coordinate transferring is used to two-dimensional coordinate of each target in each current picture being converted to public three-dimensional
Coordinate;
The object matching module is used for the mesh of all targets and previous picture in the current picture of the multiple camera
Mark is matched;
The target tracking module is used for:It is multiple according to determined by the multiple camera for each target after matching
Three-dimensional coordinate determines the final three-dimensional coordinate of the target;
The target identification module is used to divide into football, sportsman, judge according to the resemblance of target, the target that will test out
Member.
9. a kind of computer readable storage medium, which is characterized in that be stored on the computer readable storage medium for holding
Row is according to claim 1 to the program of method described in any of 7, when described program is executed by processor, described in execution
The step of method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810729547.3A CN108876821B (en) | 2018-07-05 | 2018-07-05 | Across camera lens multi-object tracking method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810729547.3A CN108876821B (en) | 2018-07-05 | 2018-07-05 | Across camera lens multi-object tracking method and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108876821A true CN108876821A (en) | 2018-11-23 |
CN108876821B CN108876821B (en) | 2019-06-07 |
Family
ID=64298956
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810729547.3A Active CN108876821B (en) | 2018-07-05 | 2018-07-05 | Across camera lens multi-object tracking method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108876821B (en) |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110288051A (en) * | 2019-07-03 | 2019-09-27 | 电子科技大学 | A kind of polyphaser multiple target matching process based on distance |
CN110349109A (en) * | 2019-07-12 | 2019-10-18 | 创新奇智(重庆)科技有限公司 | Based on flake distortion correction method and its system, electronic equipment |
CN110602449A (en) * | 2019-09-01 | 2019-12-20 | 天津大学 | Intelligent construction safety monitoring system method in large scene based on vision |
CN110674746A (en) * | 2019-09-24 | 2020-01-10 | 视云融聚(广州)科技有限公司 | Method and device for realizing high-precision cross-mirror tracking by using video spatial relationship assistance, computer equipment and storage medium |
CN111145213A (en) * | 2019-12-10 | 2020-05-12 | 中国银联股份有限公司 | Target tracking method, device and system and computer readable storage medium |
CN111242987A (en) * | 2020-01-08 | 2020-06-05 | 北京天正聚合科技有限公司 | Target tracking method and device, electronic equipment and storage medium |
CN111263955A (en) * | 2019-02-28 | 2020-06-09 | 深圳市大疆创新科技有限公司 | Method and device for determining movement track of target object |
CN111325772A (en) * | 2020-02-19 | 2020-06-23 | 四川九洲电器集团有限责任公司 | Low-slow small target detection device and method based on photoelectric equipment array |
CN111372051A (en) * | 2020-03-17 | 2020-07-03 | 三一重工股份有限公司 | Multi-camera linkage blind area detection method and device and electronic equipment |
CN111951598A (en) * | 2019-05-17 | 2020-11-17 | 杭州海康威视数字技术股份有限公司 | Vehicle tracking monitoring method, device and system |
WO2021026758A1 (en) * | 2019-08-13 | 2021-02-18 | Intel Corporation | Technology to automatically locate ball-holding players in multi-camera video feeds |
CN112686178A (en) * | 2020-12-30 | 2021-04-20 | 中国电子科技集团公司信息科学研究院 | Multi-view target track generation method and device and electronic equipment |
CN113450389A (en) * | 2021-09-01 | 2021-09-28 | 智道网联科技(北京)有限公司 | Target tracking method and device and electronic equipment |
CN116528062A (en) * | 2023-07-05 | 2023-08-01 | 合肥中科类脑智能技术有限公司 | Multi-target tracking method |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102036054A (en) * | 2010-10-19 | 2011-04-27 | 北京硅盾安全技术有限公司 | Intelligent video monitoring system based on three-dimensional virtual scene |
CN102142147A (en) * | 2010-01-29 | 2011-08-03 | 索尼公司 | Device and method for analyzing site content as well as device and method for detecting and tracking target |
CN102214301A (en) * | 2011-07-26 | 2011-10-12 | 西南交通大学 | Multi-target tracking method for associated cooperation of adaptive motion |
CN103886605A (en) * | 2014-03-31 | 2014-06-25 | 江南大学 | Method for predicting and tracking moving object based on center of curvature |
CN104715471A (en) * | 2014-01-03 | 2015-06-17 | 杭州海康威视数字技术股份有限公司 | Target positioning and tracking method and device |
CN105701828A (en) * | 2016-01-14 | 2016-06-22 | 广州视睿电子科技有限公司 | Image-processing method and device |
CN105741261A (en) * | 2014-12-11 | 2016-07-06 | 北京大唐高鸿数据网络技术有限公司 | Planar multi-target positioning method based on four cameras |
CN106803912A (en) * | 2017-03-10 | 2017-06-06 | 武汉东信同邦信息技术有限公司 | It is a kind of from motion tracking Camcording system and method |
CN107481270A (en) * | 2017-08-10 | 2017-12-15 | 上海体育学院 | Table tennis target following and trajectory predictions method, apparatus, storage medium and computer equipment |
CN108090922A (en) * | 2016-11-21 | 2018-05-29 | 中国科学院沈阳计算技术研究所有限公司 | Intelligent Target pursuit path recording method |
-
2018
- 2018-07-05 CN CN201810729547.3A patent/CN108876821B/en active Active
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102142147A (en) * | 2010-01-29 | 2011-08-03 | 索尼公司 | Device and method for analyzing site content as well as device and method for detecting and tracking target |
CN102036054A (en) * | 2010-10-19 | 2011-04-27 | 北京硅盾安全技术有限公司 | Intelligent video monitoring system based on three-dimensional virtual scene |
CN102214301A (en) * | 2011-07-26 | 2011-10-12 | 西南交通大学 | Multi-target tracking method for associated cooperation of adaptive motion |
CN104715471A (en) * | 2014-01-03 | 2015-06-17 | 杭州海康威视数字技术股份有限公司 | Target positioning and tracking method and device |
CN103886605A (en) * | 2014-03-31 | 2014-06-25 | 江南大学 | Method for predicting and tracking moving object based on center of curvature |
CN105741261A (en) * | 2014-12-11 | 2016-07-06 | 北京大唐高鸿数据网络技术有限公司 | Planar multi-target positioning method based on four cameras |
CN105701828A (en) * | 2016-01-14 | 2016-06-22 | 广州视睿电子科技有限公司 | Image-processing method and device |
CN108090922A (en) * | 2016-11-21 | 2018-05-29 | 中国科学院沈阳计算技术研究所有限公司 | Intelligent Target pursuit path recording method |
CN106803912A (en) * | 2017-03-10 | 2017-06-06 | 武汉东信同邦信息技术有限公司 | It is a kind of from motion tracking Camcording system and method |
CN107481270A (en) * | 2017-08-10 | 2017-12-15 | 上海体育学院 | Table tennis target following and trajectory predictions method, apparatus, storage medium and computer equipment |
Cited By (25)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111263955A (en) * | 2019-02-28 | 2020-06-09 | 深圳市大疆创新科技有限公司 | Method and device for determining movement track of target object |
WO2020172870A1 (en) * | 2019-02-28 | 2020-09-03 | 深圳市大疆创新科技有限公司 | Method and apparatus for determining motion trajectory of target object |
CN111951598A (en) * | 2019-05-17 | 2020-11-17 | 杭州海康威视数字技术股份有限公司 | Vehicle tracking monitoring method, device and system |
CN110288051B (en) * | 2019-07-03 | 2022-04-22 | 电子科技大学 | Multi-camera multi-target matching method based on distance |
CN110288051A (en) * | 2019-07-03 | 2019-09-27 | 电子科技大学 | A kind of polyphaser multiple target matching process based on distance |
CN110349109A (en) * | 2019-07-12 | 2019-10-18 | 创新奇智(重庆)科技有限公司 | Based on flake distortion correction method and its system, electronic equipment |
CN110349109B (en) * | 2019-07-12 | 2023-04-21 | 创新奇智(重庆)科技有限公司 | Fisheye distortion correction method and system and electronic equipment thereof |
US11861907B2 (en) | 2019-08-13 | 2024-01-02 | Intel Corporation | Technology to automatically locate ball-holding players in multi-camera video feeds |
WO2021026758A1 (en) * | 2019-08-13 | 2021-02-18 | Intel Corporation | Technology to automatically locate ball-holding players in multi-camera video feeds |
CN110602449A (en) * | 2019-09-01 | 2019-12-20 | 天津大学 | Intelligent construction safety monitoring system method in large scene based on vision |
CN110674746A (en) * | 2019-09-24 | 2020-01-10 | 视云融聚(广州)科技有限公司 | Method and device for realizing high-precision cross-mirror tracking by using video spatial relationship assistance, computer equipment and storage medium |
CN110674746B (en) * | 2019-09-24 | 2022-05-31 | 视云融聚(广州)科技有限公司 | Method and device for realizing high-precision cross-mirror tracking by using video spatial relationship assistance, computer equipment and storage medium |
TWI795667B (en) * | 2019-12-10 | 2023-03-11 | 中國銀聯股份有限公司 | A target tracking method, device, system, and computer accessible storage medium |
WO2021114702A1 (en) * | 2019-12-10 | 2021-06-17 | 中国银联股份有限公司 | Target tracking method, apparatus and system, and computer-readable storage medium |
CN111145213A (en) * | 2019-12-10 | 2020-05-12 | 中国银联股份有限公司 | Target tracking method, device and system and computer readable storage medium |
CN111242987A (en) * | 2020-01-08 | 2020-06-05 | 北京天正聚合科技有限公司 | Target tracking method and device, electronic equipment and storage medium |
CN111242987B (en) * | 2020-01-08 | 2024-02-06 | 北京天正聚合科技有限公司 | Target tracking method and device, electronic equipment and storage medium |
CN111325772A (en) * | 2020-02-19 | 2020-06-23 | 四川九洲电器集团有限责任公司 | Low-slow small target detection device and method based on photoelectric equipment array |
CN111372051A (en) * | 2020-03-17 | 2020-07-03 | 三一重工股份有限公司 | Multi-camera linkage blind area detection method and device and electronic equipment |
CN112686178A (en) * | 2020-12-30 | 2021-04-20 | 中国电子科技集团公司信息科学研究院 | Multi-view target track generation method and device and electronic equipment |
CN112686178B (en) * | 2020-12-30 | 2024-04-16 | 中国电子科技集团公司信息科学研究院 | Multi-view target track generation method and device and electronic equipment |
CN113450389A (en) * | 2021-09-01 | 2021-09-28 | 智道网联科技(北京)有限公司 | Target tracking method and device and electronic equipment |
CN113450389B (en) * | 2021-09-01 | 2021-12-10 | 智道网联科技(北京)有限公司 | Target tracking method and device and electronic equipment |
CN116528062A (en) * | 2023-07-05 | 2023-08-01 | 合肥中科类脑智能技术有限公司 | Multi-target tracking method |
CN116528062B (en) * | 2023-07-05 | 2023-09-15 | 合肥中科类脑智能技术有限公司 | Multi-target tracking method |
Also Published As
Publication number | Publication date |
---|---|
CN108876821B (en) | 2019-06-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108876821B (en) | Across camera lens multi-object tracking method and system | |
JP7004017B2 (en) | Object tracking system, object tracking method, program | |
US9684928B2 (en) | Foot tracking | |
CN102640185B (en) | The method and apparatus of the combined tracking that object represents in real time in image sequence | |
Xu et al. | Architecture and algorithms for tracking football players with multiple cameras | |
US9633265B2 (en) | Method for improving tracking in crowded situations using rival compensation | |
US7957557B2 (en) | Tracking apparatus and tracking method | |
Gao et al. | Robust RGB-D simultaneous localization and mapping using planar point features | |
Ren et al. | Tracking the soccer ball using multiple fixed cameras | |
Ren et al. | Multi-camera video surveillance for real-time analysis and reconstruction of soccer games | |
Calderara et al. | Bayesian-competitive consistent labeling for people surveillance | |
US10769810B2 (en) | Apparatus, systems and methods for shadow assisted object recognition and tracking | |
WO2014136623A1 (en) | Method for detecting and tracking objects in sequence of images of scene acquired by stationary camera | |
US11776213B2 (en) | Pose generation apparatus, generation method, and storage medium | |
Yang et al. | A new hybrid synthetic aperture imaging model for tracking and seeing people through occlusion | |
EP2946360B1 (en) | Foot tracking | |
Brouwers et al. | Automatic calibration of stationary surveillance cameras in the wild | |
Wientapper et al. | Composing the feature map retrieval process for robust and ready-to-use monocular tracking | |
Li et al. | Robust object tracking in crowd dynamic scenes using explicit stereo depth | |
Chen et al. | Player trajectory reconstruction from broadcast basketball video | |
Takahashi et al. | Real-time ball position measurement for football games based on ball's appearance and motion features | |
Fei et al. | A registration method based on nature feature with KLT tracking algorithm for wearable computers | |
Farhat et al. | A new model based approach for tennis court tracking in real time | |
Desai et al. | Determine absolute soccer ball location in broadcast video using SYBA descriptor | |
Velipasalar et al. | Frame-level temporal calibration of video sequences from unsynchronized cameras |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |