CN108055501A - A kind of target detection and the video monitoring system and method for tracking - Google Patents
A kind of target detection and the video monitoring system and method for tracking Download PDFInfo
- Publication number
- CN108055501A CN108055501A CN201711174398.0A CN201711174398A CN108055501A CN 108055501 A CN108055501 A CN 108055501A CN 201711174398 A CN201711174398 A CN 201711174398A CN 108055501 A CN108055501 A CN 108055501A
- Authority
- CN
- China
- Prior art keywords
- target
- target object
- data
- video
- tracking
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 62
- 238000000034 method Methods 0.000 title claims abstract description 23
- 238000012544 monitoring process Methods 0.000 title claims abstract description 21
- 230000033001 locomotion Effects 0.000 claims abstract description 47
- 230000010365 information processing Effects 0.000 claims abstract description 9
- 238000007689 inspection Methods 0.000 claims description 12
- 238000013136 deep learning model Methods 0.000 claims description 11
- 238000000205 computational method Methods 0.000 claims description 5
- 238000013135 deep learning Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 238000004891 communication Methods 0.000 claims description 4
- 230000000007 visual effect Effects 0.000 claims description 4
- 101000606504 Drosophila melanogaster Tyrosine-protein kinase-like otk Proteins 0.000 claims description 3
- 238000007619 statistical method Methods 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- 230000005540 biological transmission Effects 0.000 abstract description 4
- 238000013528 artificial neural network Methods 0.000 abstract description 2
- 230000009286 beneficial effect Effects 0.000 abstract 2
- 241000894007 species Species 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 241001465754 Metazoa Species 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000007667 floating Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000013011 mating Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
Classifications
-
- 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
-
- 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
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N5/00—Details of television systems
- H04N5/14—Picture signal circuitry for video frequency region
- H04N5/144—Movement detection
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N5/00—Details of television systems
- H04N5/14—Picture signal circuitry for video frequency region
- H04N5/144—Movement detection
- H04N5/145—Movement estimation
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The present invention relates to a kind of target detection and the video monitoring systems and method of tracking, including video acquisition, target detection, information processing, information transmission and image capture device, wherein image capture device is made of front-end image collecting device and video decoding apparatus, obtains target object live video stream;Target object location and size in target detection analysis video image;Information processing is handled in real time for the target object relevant information in continuous time, obtains movement velocity, track and directional information and anticipation target object quantity, the position and size information of target object;Target object relevant information is sent to image capture device by information transmission;Image capture device control makes target object always in picture intermediate region, to the real-time tracking of target object.The present invention, beneficial to the tracking to numerous Small objects, improves discrimination, is beneficial to the good operation of track algorithm, realizes the real-time tracking to target object according to the deep neural network of multiple features.
Description
Technical field
The invention belongs to technical field of video monitoring, the video monitoring system of more particularly to a kind of target detection and tracking and
Method.
Background technology
With the development of Video Supervision Technique, intelligent security guard has been had been applied in all trades and professions of society, wherein mesh
The "smart" products application scenarios in mark detection and tracking field are also increasingly wider, including traffic, road monitoring, public domain, spy
Determine area monitoring etc..
At present, existing object detecting and tracking system mainly identifies target object using traditional image processing algorithm,
Such as background subtraction, frame differential method, optical flow method etc..Its principle is the pixel information by analyzing image, and comparison is current
Image and background image or previous frame image, obtain required dynamic object information.
Above-mentioned traditional image processing algorithm then can not generally directed to multiple target object when object pixel accounting is smaller
Correct identification target.In addition, during changing for dynamic scene, when being disturbed such as light variation with external unrelated thing
It can identify mistake, and then target object is caused to be lost, tracking failure.
The content of the invention
It is an object of the invention to overcome the deficiencies of the prior art and provide a kind of target detection of precision and regarding for tracking
Frequency monitoring system and method.
The scheme of the invention is be achieved in that:
A kind of target detection and the video monitoring system of tracking, including image capture device, embedded unit, interchanger with
And PC machine, the interchanger are connected respectively with PC machine, image capture device and embedded unit, the interchanger and Image Acquisition
Modular converter is connected between equipment;
The modular converter turns RS485 devices for USB;
The embedded unit includes module of target detection and message processing module.
Moreover, what the network gunlock that described image collecting device is network monopod video camera, holder carries, holder carried
The video camera module that USB camera or holder carry.
Moreover, the embedded unit is card computer.
A kind of method of target detection and the video monitoring system of tracking, includes the following steps:
Step 1: video acquisition
Video acquisition is made of front-end image collecting device and video decoding apparatus, and acquisition includes the real-time of target object
Video flowing;
Step 2: target detection
By analyzing the live video stream obtained, target object species, quantity, position and size information are obtained;
Step 3: information processing
It is handled in real time according to the target object relevant information in continuous time;
Step 4: information is transmitted
Anticipation position, size and the velocity information of target object are sent to image capture device;
Also the user operation instruction that image capture device is sent is returned into embedded unit, realizes embedded unit and figure
As the both-way communication between collecting device;
Step 5, the control of image capture device
According to anticipation position, size and the velocity information of target object, judge position of the target object in picture and
And the deviation of central area, image capture device perform the 3D algorithms of corresponding speed, target object are moved to field of view center area
Domain.
Moreover, in the step 2, target detection mainly includes the following steps:
(1) real-time video is gathered
By the way that OpenCV storehouses or FFMPEG is called to decode storehouse, video acquisition sub-line journey is opened up, obtains live video stream, is somebody's turn to do
Thread can work always, until EP (end of program) or exception exit;
(2) it is loaded into deep learning model
The model be according to actual scene and particular detection target, it is pre- using the yolo frames based on deep learning algorithm
First train deep learning model;
(3) fiducial probability of object space, size, classification and generic is calculated
The deep learning model is loaded into the memory of embedded unit, it, will be every while real-time video is gathered
Frame video flowing inputs the model and is calculated, and obtains the fiducial probability of target object location, size, classification and generic;
The flase drop missing inspection target in the picture found during the test is gathered, these flase drop missing inspection target images are added in
Training set is trained again, draws the highest model of accuracy rate under the scene;
(4) fiducial probability is compared with given threshold
After target object fiducial probability is obtained, and compared with given threshold:More than given threshold, then program continues to judge
Target object;If less than given threshold, then give up group detection data, return to acquisition real-time video, start next frame target
Detection;
(5) judge target whether in effective coverage
Judged whether according to target object location in effective coverage, if in effective coverage, program continues to judge mesh
Mark object;If do not give up group detection data in effective coverage, return to acquisition real-time video, start the inspection of next frame target
It surveys;
(6) whether target classification is particular detection target
, it is necessary to judge whether the species of target meets the spy of user interface input after threshold decision and effective coverage judge
Determine species, if meeting particular types, for effective target, continued with into information handling step;If invalid targets, then
Detection data are rejected, and return to acquisition real-time video.
Moreover, in the step 3, the flase drop missing inspection target is for the less Small object of pixel accounting or in various light
Line changes the target of flase drop missing inspection under the interference with external unrelated things.
Moreover, in the step 3, information processing mainly includes the following steps:
(1) present frame target data is obtained
Live video stream can export present frame specific objective related data, including target species after target detection step
Class, position and size information etc.;
(2) target data of certain amount successive frame is gathered
, it is necessary to gather the target object data value of certain continuous data frame in information process, if acquiring N group mesh
Data are marked, after present frame target data is obtained, can upgrade in time last group i.e. target data of N groups, according to this N group number
According to the operations such as the anticipation coordinate for analyzing the movement locus of target and providing target direction of advance can be continued;
(3) invalid target data is rejected
In the target data of continuous N frames, it may appear that the situation that target is temporarily lost, this group of target data is nothing at this time
Data are imitated, it is necessary to be rejected to these data, it is (0,0) to reject invalid data to refer to reject target location coordinate point in data group
Coordinate points;
(4) target object movement locus, the target data for rejecting off-track are judged
The target object direction of motion, movement velocity are estimated, target object movement locus is drawn, refers to according to target set of data
The coordinate points (x, y) of interior data, sequentially in time do x-axis and y-axis in direction and speed calculates, and obtain target object in x-axis
With the movement locus and movement tendency of y-axis;Statistical analysis is carried out to data point on target trajectory afterwards, calculates object
The credibility interval of body related data, that is, target location, target direction of motion and target speed, reject data group in deviate can
Believe the data value in section;
(5) target object next frame movement position is prejudged
Refer to after obtaining target direction of motion, according to the effective target coordinate points position of last frame and target object
The direction of motion, movement velocity are obtained per frame target moving distance.
Moreover, in the step (5), the target object direction of motion to the left, then target object next frame target location
Computational methods are:Prejudge target location coordinate point (x, y)=nearest frame valid frame coordinate points (x, y)-per the movement of frame target away from
From (△ x, △ y);If the target object direction of motion is to the right, target object next frame target location computational methods are:Anticipation
Target location coordinate point (x, y)=nearest frame valid frame coordinate points (x, y)+per frame target moving distance (△ x, △ y).
Moreover, in the step 5, the translational speed according to coordinate position according to the far and near adjust automatically of center, and
Accounting for the corresponding zoom action of screen proportion progress according to target object makes target object be of moderate size in visual field, and target object accounts for
/ 3rd of screen proportion.
The advantages and positive effects of the present invention are:
1st, the present invention, based on the target detection technique of deep learning algorithm, is adjusted using current advanced artificial intelligence technology
With by repeatedly trained deep learning model, the Small object in picture and multiple target can be identified correctly and accurately fixed
The accuracy of target detection is improved in position.
2nd, the present invention can train the deep learning model of specific objective, exclude various according to system practical application scene
Light changes and the interference of external unrelated things, and target object is accurately detected and is accurately positioned.
3rd, the present invention is conducive to the tracking to numerous Small objects, effectively improves knowledge according to the deep neural network of multiple features
Not rate is also advantageous for the good operation of track algorithm, realizes the real-time tracking to target object.
Description of the drawings
Fig. 1 is the connection block diagram of present system;
Fig. 2 is the flow chart of the method for the present invention;
Fig. 3 is the flow chart of target detection step in Fig. 2;
Fig. 4 is the flow chart of information handling step in Fig. 2.
Specific embodiment
Below in conjunction with the accompanying drawings and pass through specific embodiment the invention will be further described.
A kind of target detection and the video monitoring system of tracking, as shown in Figure 1, including image capture device, embedded single
Member, interchanger and PC machine, the interchanger are connected respectively with PC machine, image capture device and embedded unit, the exchange
Modular converter is connected between machine and image capture device, i.e., is connected image capture device, embedded unit, PC machine by interchanger
It picks up and, achieve the purpose that network share.Wherein, image capture device output H264 video informations, embedded unit pass through solution
Code rtsp streams obtain image information, and PC machine can monitor and record real-time video.
The modular converter turns RS485 devices for USB.
The embedded unit is card computer.
Communication connection:After embedded unit obtains the image information that image capture device is sent, by the anticipation of target object
Position, size and velocity information send image capture device to, and image capture device performs 3D algorithms and rotating image acquisition is set
It is standby, detection target is made to be in always in the image capture device visual field.In addition, embedded unit can also receive image capture device
The functions such as the user interface sent instructs, realization region detection, and tracking target determines.Connection mode is communicated as USB turns 485, it is embedding
Enter formula unit and connect USB ends, monopod video camera connects 485 ends.
Power supply connects:Described image collecting device, embedded unit, interchanger are equipped with mating power supply, and correctly connect.
A kind of method of target detection and the video monitoring system of tracking, as shown in Fig. 2, including the following steps:
Step 1: video acquisition
Real time video collection is made of front-end image collecting device and video decoding apparatus.In actual use, Ke Yitong
It crosses following four kinds of modes and obtains the live video stream including target object.
The first, using high-definition network monopod video camera as front-end image collecting device, then live video stream passes through net
Network exports, and video decoding apparatus obtains realtime graphic by obtaining rtsp video flowings and decoding;
Second, the network gunlock carried using holder is as front-end collection equipment, then live video stream is defeated by network
Go out, video decoding process is the same as the first;
The third, using the USB camera that holder carries as front-end image collecting device, then live video stream passes through
USB is exported, and video decoding apparatus obtains realtime graphic by reading USB interface data;
4th kind, using the video camera module that holder carries as front-end image collecting device, then live video stream passes through
MIPI interfaces or the output of other video interfaces, video decoding apparatus obtain realtime graphic by serioparallel exchange or digital-to-analogue conversion.
Step 2: target detection
Target detection can be run on embedded unit, refer mainly to, by analyzing the live video stream obtained, obtain mesh
Mark kind of object, quantity, position and size information.Target object can be people, vehicle, ship, animal etc. any one or a few.Mesh
Mark detection uses the yolo algorithms based on deep learning, extracts the great amount of samples data and special scenes for including target object
The image data easily misidentified trains suitable deep learning model.In actual use, embedded unit or card electricity
Brain is loaded into this learning model, judges whether there is the credible threshold value of target object and target object in single-frame images.More than judgement
During threshold value, represent this time to judge effectively, to be used;It is on the contrary then abandon this group of data.
Step 3: information processing
Information processing can be run on embedded unit, be referred mainly to according to the target object relevant information in continuous time
It is handled in real time.Firstly, it is necessary to which the target object species and object detection area that are selected according to user select, weed out
Other objects and not in the invalid source location of detection zone.Then according to the position of target object between successive video frames
It is different, it can be determined that go out the movement velocity and directional information of target object, and draw the movement locus of target object.Afterwards, root
Prejudge out movement position of target object next frame etc. according to the motion conditions of target object in actual scene, and by target object
Anticipation position, size and velocity information are timely transmitted to image capture device.Why transmission anticipation target location, be because
There is corresponding time loss during target detection, delay can be reduced by sending anticipation position, and target is made to be in image always
Detection range in.
Step 4: information is transmitted
Information transmission turns RS422 modules by USB or USB RS 232 modules form, by the anticipation position of target object, greatly
Small and velocity information is sent to image capture device.In addition, also the user operation instruction that image capture device is sent is returned to
Embedded unit realizes the both-way communication between embedded unit and image capture device.
Step 5, the control of image capture device
According to anticipation position, size and the velocity information of target object, judge position of the target object in picture and
And the deviation of central area, image capture device performs the 3D algorithms of corresponding speed, also known as direct position control algolithm, by target
Object is moved to field of view center region.Translational speed can according to coordinate position according to the far and near adjust automatically of center, and according to
Target object accounts for screen proportion and carries out corresponding zoom action target object is made to be of moderate size in visual field (accounting for screen proportion
1/3rd).In this way, target object can be made to be in the intermediate region of picture, the real-time tracking to target object is realized.
In the present embodiment, as shown in figure 3, in step 2, target detection mainly includes the following steps:
(4) real-time video is gathered
By the way that OpenCV storehouses or FFMPEG is called to decode storehouse, video acquisition sub-line journey is opened up, slave device interface obtains
Live video stream, the thread can work always, until EP (end of program) or exception exit.
(5) it is loaded into deep learning model
The model be according to actual scene and particular detection target, it is pre- using the yolo frames based on deep learning algorithm
First train deep learning model.
(6) fiducial probability of object space, size, classification and generic is calculated
In actual use, which can be loaded into the memory of embedded unit, is regarded in real time in acquisition
While frequency, it will be calculated per frame video flowing input model, putting for target object location, size, classification and generic can be obtained
Believe probability.
The flase drop missing inspection target in the picture found during the test, such as the Small object that pixel accounting is less are gathered,
Or various light variation and external unrelated things interference under flase drop missing inspection target, by these flase drop missing inspection target images
It adds in training set to train again, draws the highest model of accuracy rate under the scene.
(4) fiducial probability is compared with given threshold
Since under different weather or different scenes, the fiducial probability of same target object has floating, therefore needs
Before system brings into operation, the operative scenario of the system is inputted by user, system can automatically update generic under the scene
Given threshold.It, can be compared with given threshold after target object fiducial probability is obtained:More than given threshold, then program continues to sentence
Disconnected target object;If less than given threshold, then give up group detection data, start next frame target detection.
(5) judge target whether in effective coverage
When the system, which is applied, to be monitored in specific region, client may require that some regions in only detection picture, become effective
Detection zone, other regions that are ignored are known as inactive area.Therefore, it is necessary to be according to target object location judgement after threshold decision
It is no in effective coverage, if in effective coverage, program continues to judge target object;If do not give up in effective coverage
The group detects data, starts next frame target detection;
(6) whether target classification is particular detection target
, it is necessary to judge whether the species of target meets the spy of user interface input after threshold decision and effective coverage judge
Determine species.If meeting particular types, for effective target, program enters message processing module and continues with;If invalid mesh
Mark, then detect data and be rejected.
Preferably, in above-mentioned steps three, information processing mainly includes the following steps:
(2) present frame target data is obtained
Live video stream can export present frame specific objective related data, including target species after target detection step
Class, position and size information etc..
(2) target data of certain amount successive frame is gathered
It is false in the present embodiment, it is necessary to gather the target object data value of certain continuous data frame in information process
If acquiring N group target datas, after present frame target data is obtained, can upgrade in time last group i.e. number of targets of N groups
According to according to this N group data, can continuing to analyze the movement locus of target and provide the anticipation coordinate of target direction of advance
Deng operation.
(3) invalid target data is rejected
In the target data of continuous N frames, it is present with the situation that target is temporarily lost certainly, this group of target data is at this time
For invalid data, it is necessary to be rejected to these data.Specifically, invalid data is rejected to refer to reject target location coordinate in data group
Point is the coordinate points of (0,0).
(4) target object movement locus, the target data for rejecting off-track are judged
The target object direction of motion, movement velocity are estimated, target object movement locus is drawn, refers to according to target set of data
The coordinate points (x, y) of interior data, sequentially in time do x-axis and y-axis in direction and speed calculates, and obtain target object in x-axis
With the movement locus and movement tendency of y-axis.Statistical analysis is carried out to data point on target trajectory afterwards, calculates object
The credibility interval of body related data, that is, target location, target direction of motion and target speed, reject data group in deviate can
Believe the data value in section.
(6) target object next frame movement position is prejudged
Refer to after obtaining target direction of motion, according to the effective target coordinate points position of last frame and target object
The direction of motion, movement velocity are obtained per frame target moving distance.If the direction of motion is to the left, target object next frame target
Position calculating method is:Target location coordinate point (x, y)=nearest frame valid frame coordinate points (x, y)-every frame target is prejudged to move
Dynamic distance (△ x, △ y);If the direction of motion is to the right, target object next frame target location computational methods are:Prejudge target
Position coordinates point (x, y)=nearest frame valid frame coordinate points (x, y)+per frame target moving distance (△ x, △ y).
It is emphasized that embodiment of the present invention is illustrative rather than limited, therefore present invention bag
The embodiment being not limited to described in specific embodiment is included, it is every by those skilled in the art's technique according to the invention scheme
The other embodiment drawn, also belongs to the scope of protection of the invention.
Claims (9)
1. a kind of target detection and the video monitoring system of tracking, it is characterised in that:Including image capture device, embedded single
Member, interchanger and PC machine, the interchanger are connected respectively with PC machine, image capture device and embedded unit, the exchange
Modular converter is connected between machine and image capture device;
The modular converter turns RS485 devices for USB;
The embedded unit includes module of target detection and message processing module.
2. target detection according to claim 1 and the video monitoring system of tracking, it is characterised in that:Described image gathers
The camera shooting of equipment is network monopod video camera, holder carries network gunlock, the USB camera that holder carries or holder carrying
Machine module.
3. target detection according to claim 1 and the video monitoring system of tracking, it is characterised in that:The embedded single
Member is card computer.
4. the method for target detection according to any one of claims 1 to 3 and the video monitoring system of tracking, feature exist
In:Include the following steps:
Step 1: video acquisition
Video acquisition is made of front-end image collecting device and video decoding apparatus, and obtains the real-time video for including target object
Stream;
Step 2: target detection
By analyzing the live video stream obtained, target object species, quantity, position and size information are obtained;
Step 3: information processing
It is handled in real time according to the target object relevant information in continuous time;
Step 4: information is transmitted
Anticipation position, size and the velocity information of target object are sent to image capture device;
Also the user operation instruction that image capture device is sent is returned into embedded unit, realizes that embedded unit is adopted with image
Collect the both-way communication between equipment;
Step 5, the control of image capture device
According to anticipation position, size and the velocity information of target object, position of the target object in picture is judged and in
The deviation in heart district domain, image capture device perform the 3D algorithms of corresponding speed, target object are moved to field of view center region.
5. the method for target detection according to claim 4 and the video monitoring system of tracking, it is characterised in that:The step
In rapid two, target detection mainly includes the following steps:
(1) real-time video is gathered
By the way that OpenCV storehouses or FFMPEG is called to decode storehouse, video acquisition sub-line journey is opened up, obtains live video stream, the thread
It can work always, until EP (end of program) or exception exit;
(2) it is loaded into deep learning model
The model is according to actual scene and particular detection target, is instructed in advance using the yolo frames based on deep learning algorithm
Practise deep learning model;
(3) fiducial probability of object space, size, classification and generic is calculated
The deep learning model is loaded into the memory of embedded unit, while real-time video is gathered, will be regarded per frame
Frequency stream inputs the model and is calculated, and obtains the fiducial probability of target object location, size, classification and generic;
The flase drop missing inspection target in the picture found during the test is gathered, these flase drop missing inspection target images are added in and are trained
Collection is trained again, draws the highest model of accuracy rate under the scene;
(4) fiducial probability is compared with given threshold
After target object fiducial probability is obtained, and compared with given threshold:More than given threshold, then program continues to judge target
Object;If less than given threshold, then give up group detection data, return to acquisition real-time video, start next frame target detection;
(5) judge target whether in effective coverage
Judged whether according to target object location in effective coverage, if in effective coverage, program continues to judge object
Body;If do not give up group detection data in effective coverage, return to acquisition real-time video, start next frame target detection;
(6) whether target classification is particular detection target
, it is necessary to judge whether the species of target meets specific kind of user interface input after threshold decision and effective coverage judge
Class if meeting particular types, for effective target, is continued with into information handling step;If invalid targets, then detect
Data are rejected, and return to acquisition real-time video.
6. the method for target detection according to claim 5 and the video monitoring system of tracking, it is characterised in that:The step
In rapid 3, the flase drop missing inspection target is for the less Small object of pixel accounting or in the variation of various light and external unrelated things
Interference under flase drop missing inspection target.
7. the method for target detection according to claim 4 and the video monitoring system of tracking, it is characterised in that:The step
In rapid three, information processing mainly includes the following steps:
(1) present frame target data is obtained
Live video stream can export present frame specific objective related data, including targeted species, position after target detection step
It puts and size information etc.;
(2) target data of certain amount successive frame is gathered
, it is necessary to gather the target object data value of certain continuous data frame in information process, if acquiring N group number of targets
According to, after present frame target data is obtained, the target data for last i.e. N group of group that can upgrade in time, according to this N group data,
It can continue the operations such as the anticipation coordinate for analyzing the movement locus of target and providing target direction of advance;
(3) invalid target data is rejected
In the target data of continuous N frames, it may appear that the situation that target is temporarily lost, this group of target data is invalid number at this time
According to, it is necessary to be rejected to these data, it is the coordinate of (0,0) to reject invalid data to refer to reject target location coordinate point in data group
Point;
(4) target object movement locus, the target data for rejecting off-track are judged
The target object direction of motion, movement velocity are estimated, draws target object movement locus, is referred to according to number in target set of data
According to coordinate points (x, y), direction and speed are done to x-axis and y-axis sequentially in time and is calculated, obtains target object in x-axis and y-axis
Movement locus and movement tendency;Statistical analysis is carried out to data point on target trajectory afterwards, calculates target object phase
The credibility interval of data, that is, target location, target direction of motion and target speed is closed, rejects in data group and deviates confidence region
Between data value;
(5) target object next frame movement position is prejudged
Refer to after obtaining target direction of motion, moved according to the effective target coordinate points position of last frame and target object
Direction, movement velocity are obtained per frame target moving distance.
8. the method for target detection according to claim 7 and the video monitoring system of tracking, it is characterised in that:The step
Suddenly in (5), to the left, then target object next frame target location computational methods are the target object direction of motion:Prejudge target
Position coordinates point (x, y)=nearest frame valid frame coordinate points (x, y)-per frame target moving distance (△ x, △ y);If target
To the right, then target object next frame target location computational methods are movement direction of object:Anticipation target location coordinate point (x, y)=
Nearest frame valid frame coordinate points (x, y)+per frame target moving distance (△ x, △ y).
9. the method for target detection according to claim 4 and the video monitoring system of tracking, it is characterised in that:The step
In rapid five, the translational speed, according to the far and near adjust automatically of center, and accounts for screen according to coordinate position according to target object
Ratio, which carries out corresponding zoom action, makes target object be of moderate size in visual field, target object account for screen proportion three/
One.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711174398.0A CN108055501A (en) | 2017-11-22 | 2017-11-22 | A kind of target detection and the video monitoring system and method for tracking |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711174398.0A CN108055501A (en) | 2017-11-22 | 2017-11-22 | A kind of target detection and the video monitoring system and method for tracking |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108055501A true CN108055501A (en) | 2018-05-18 |
Family
ID=62119101
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711174398.0A Pending CN108055501A (en) | 2017-11-22 | 2017-11-22 | A kind of target detection and the video monitoring system and method for tracking |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108055501A (en) |
Cited By (43)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108876823A (en) * | 2018-07-02 | 2018-11-23 | 晋建志 | Based on across the camera multi-targets recognition locating and tracking method of time and space continuity monocular |
CN108895756A (en) * | 2018-06-29 | 2018-11-27 | 青岛海尔股份有限公司 | The method and apparatus of the calorie value of displaying food for refrigerator |
CN108898169A (en) * | 2018-06-19 | 2018-11-27 | Oppo广东移动通信有限公司 | Image processing method, picture processing unit and terminal device |
CN108898587A (en) * | 2018-06-19 | 2018-11-27 | Oppo广东移动通信有限公司 | Image processing method, picture processing unit and terminal device |
CN108961310A (en) * | 2018-06-13 | 2018-12-07 | 北京清瑞维航技术发展有限公司 | Device, system, the medium, calculating device and method that automaticidentifying& tracking is realized |
CN108985169A (en) * | 2018-06-15 | 2018-12-11 | 浙江工业大学 | Across the door operation detection method in shop based on deep learning target detection and dynamic background modeling |
CN108981277A (en) * | 2018-06-29 | 2018-12-11 | 青岛海尔股份有限公司 | The method and apparatus of modification beverage shelf-life for refrigerator |
CN109001833A (en) * | 2018-06-22 | 2018-12-14 | 天和防务技术(北京)有限公司 | A kind of Terahertz hazardous material detection method based on deep learning |
CN109241938A (en) * | 2018-09-27 | 2019-01-18 | 深圳金三立视频科技股份有限公司 | Congestion in road detection method and terminal |
CN109308693A (en) * | 2018-08-29 | 2019-02-05 | 北京航空航天大学 | By the target detection and pose measurement list binocular vision system of a ptz camera building |
CN109410496A (en) * | 2018-10-25 | 2019-03-01 | 北京交通大学 | Attack early warning method, apparatus and electronic equipment |
CN109544595A (en) * | 2018-10-29 | 2019-03-29 | 苏宁易购集团股份有限公司 | A kind of customer's route tracing method and system |
CN109587445A (en) * | 2018-11-18 | 2019-04-05 | 中设设计集团股份有限公司 | A kind of security system of defense based on lamp networking |
CN109800714A (en) * | 2019-01-22 | 2019-05-24 | 浙江华是科技股份有限公司 | A kind of ship detecting system and method based on artificial intelligence |
CN109816310A (en) * | 2019-01-22 | 2019-05-28 | 浙江华是科技股份有限公司 | A kind of navigation channel Ship IC Card port management system and its management method based on artificial intelligence |
CN109873987A (en) * | 2019-03-04 | 2019-06-11 | 深圳市梦网百科信息技术有限公司 | A kind of Target Searching Method and system based on monitor video |
CN109977783A (en) * | 2019-02-28 | 2019-07-05 | 浙江新再灵科技股份有限公司 | Method based on the independent boarding detection of vertical ladder scene perambulator |
CN110135377A (en) * | 2019-05-21 | 2019-08-16 | 北京百度网讯科技有限公司 | Object moving state detection method, device, server and computer-readable medium |
CN110290351A (en) * | 2019-06-26 | 2019-09-27 | 广东康云科技有限公司 | A kind of video target tracking method, system, device and storage medium |
CN110490834A (en) * | 2018-06-16 | 2019-11-22 | 南京理工大学 | The object recognition and detection method of foreign matter device is removed based on laser |
CN110488312A (en) * | 2018-06-16 | 2019-11-22 | 南京理工大学 | The target prediction method of foreign matter device is removed based on reflection-type laser |
CN110480180A (en) * | 2018-06-16 | 2019-11-22 | 南京理工大学 | The object recognition and detection method of foreign matter device is removed based on reflection-type laser |
CN110488309A (en) * | 2018-06-16 | 2019-11-22 | 南京理工大学 | The target Calibration Method of foreign matter device is removed based on laser |
WO2020057346A1 (en) * | 2018-09-21 | 2020-03-26 | 深圳市九洲电器有限公司 | Video monitoring method and apparatus, monitoring server and video monitoring system |
CN110933318A (en) * | 2019-12-12 | 2020-03-27 | 天地伟业技术有限公司 | Moving object snapshot method |
CN110939351A (en) * | 2019-10-28 | 2020-03-31 | 优创嘉(大连)科技有限公司 | Visual intelligent control method and visual intelligent control door |
CN110991283A (en) * | 2019-11-21 | 2020-04-10 | 北京格灵深瞳信息技术有限公司 | Re-recognition and training data acquisition method and device, electronic equipment and storage medium |
CN111079638A (en) * | 2019-12-13 | 2020-04-28 | 河北爱尔工业互联网科技有限公司 | Target detection model training method, device and medium based on convolutional neural network |
CN111191507A (en) * | 2019-11-26 | 2020-05-22 | 恒大智慧科技有限公司 | Safety early warning analysis method and system for smart community |
CN111339977A (en) * | 2020-03-03 | 2020-06-26 | 河南中光学集团有限公司 | Small target intelligent identification system based on remote video monitoring and identification method thereof |
CN112330650A (en) * | 2020-11-12 | 2021-02-05 | 李庆春 | Retrieval video quality evaluation method |
CN112785639A (en) * | 2021-01-06 | 2021-05-11 | 鞍钢矿业爆破有限公司 | Video analysis-based method for metering bagged materials into and out of warehouse |
CN112804487A (en) * | 2020-12-30 | 2021-05-14 | 济南和普威视光电技术有限公司 | Double-path video target tracking method and system |
CN112866558A (en) * | 2020-11-04 | 2021-05-28 | 苏州臻迪智能科技有限公司 | Operation method of electronic equipment, control method of holder and holder system |
CN113344965A (en) * | 2021-08-05 | 2021-09-03 | 鑫安利中(北京)科技有限公司 | Safety management system based on image acquisition |
CN113408518A (en) * | 2021-07-06 | 2021-09-17 | 世邦通信股份有限公司 | Audio and video acquisition equipment control method and device, electronic equipment and storage medium |
CN113723841A (en) * | 2021-09-03 | 2021-11-30 | 安徽大学绿色产业创新研究院 | Online detection method for tool missing in assembly type prefabricated part |
CN113743380A (en) * | 2021-11-03 | 2021-12-03 | 江苏博子岛智能产业技术研究院有限公司 | Active tracking method based on video image dynamic monitoring |
CN114067575A (en) * | 2021-11-23 | 2022-02-18 | 安徽富煌科技股份有限公司 | Traffic hub region safety analysis device based on 3D structured light detection |
CN114220077A (en) * | 2022-02-21 | 2022-03-22 | 金叶仪器(山东)有限公司 | Method for realizing object quantity statistics and moving direction monitoring based on monitoring equipment |
CN115063452A (en) * | 2022-06-13 | 2022-09-16 | 中国船舶重工集团公司第七0七研究所九江分部 | Cloud deck camera tracking method for offshore target |
CN115914582A (en) * | 2023-01-05 | 2023-04-04 | 百鸟数据科技(北京)有限责任公司 | Small object detection optimization method based on fusion time sequence information |
CN117649537A (en) * | 2024-01-30 | 2024-03-05 | 浙江省公众信息产业有限公司 | Monitoring video object identification tracking method, system, electronic equipment and storage medium |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104244113A (en) * | 2014-10-08 | 2014-12-24 | 中国科学院自动化研究所 | Method for generating video abstract on basis of deep learning technology |
JP2015002553A (en) * | 2013-06-18 | 2015-01-05 | キヤノン株式会社 | Information system and control method thereof |
CN105160313A (en) * | 2014-09-15 | 2015-12-16 | 中国科学院重庆绿色智能技术研究院 | Method and apparatus for crowd behavior analysis in video monitoring |
CN204929077U (en) * | 2015-08-18 | 2015-12-30 | 江苏捷诚车载电子信息工程有限公司 | Embedded video monitor system |
CN105678338A (en) * | 2016-01-13 | 2016-06-15 | 华南农业大学 | Target tracking method based on local feature learning |
CN106127802A (en) * | 2016-06-16 | 2016-11-16 | 南京邮电大学盐城大数据研究院有限公司 | A kind of movement objective orbit method for tracing |
US20170132468A1 (en) * | 2015-11-06 | 2017-05-11 | The Boeing Company | Systems and methods for object tracking and classification |
CN107122439A (en) * | 2017-04-21 | 2017-09-01 | 图麟信息科技(深圳)有限公司 | A kind of video segment querying method and device |
CN107197199A (en) * | 2017-05-22 | 2017-09-22 | 哈尔滨工程大学 | A kind of intelligent monitoring and controlling device and method for tracking target |
CN107330920A (en) * | 2017-06-28 | 2017-11-07 | 华中科技大学 | A kind of monitor video multi-target tracking method based on deep learning |
-
2017
- 2017-11-22 CN CN201711174398.0A patent/CN108055501A/en active Pending
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2015002553A (en) * | 2013-06-18 | 2015-01-05 | キヤノン株式会社 | Information system and control method thereof |
CN105160313A (en) * | 2014-09-15 | 2015-12-16 | 中国科学院重庆绿色智能技术研究院 | Method and apparatus for crowd behavior analysis in video monitoring |
CN104244113A (en) * | 2014-10-08 | 2014-12-24 | 中国科学院自动化研究所 | Method for generating video abstract on basis of deep learning technology |
CN204929077U (en) * | 2015-08-18 | 2015-12-30 | 江苏捷诚车载电子信息工程有限公司 | Embedded video monitor system |
US20170132468A1 (en) * | 2015-11-06 | 2017-05-11 | The Boeing Company | Systems and methods for object tracking and classification |
CN105678338A (en) * | 2016-01-13 | 2016-06-15 | 华南农业大学 | Target tracking method based on local feature learning |
CN106127802A (en) * | 2016-06-16 | 2016-11-16 | 南京邮电大学盐城大数据研究院有限公司 | A kind of movement objective orbit method for tracing |
CN107122439A (en) * | 2017-04-21 | 2017-09-01 | 图麟信息科技(深圳)有限公司 | A kind of video segment querying method and device |
CN107197199A (en) * | 2017-05-22 | 2017-09-22 | 哈尔滨工程大学 | A kind of intelligent monitoring and controlling device and method for tracking target |
CN107330920A (en) * | 2017-06-28 | 2017-11-07 | 华中科技大学 | A kind of monitor video multi-target tracking method based on deep learning |
Cited By (52)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108961310A (en) * | 2018-06-13 | 2018-12-07 | 北京清瑞维航技术发展有限公司 | Device, system, the medium, calculating device and method that automaticidentifying& tracking is realized |
CN108985169B (en) * | 2018-06-15 | 2020-12-11 | 浙江工业大学 | Shop cross-door operation detection method based on deep learning target detection and dynamic background modeling |
CN108985169A (en) * | 2018-06-15 | 2018-12-11 | 浙江工业大学 | Across the door operation detection method in shop based on deep learning target detection and dynamic background modeling |
CN110490834A (en) * | 2018-06-16 | 2019-11-22 | 南京理工大学 | The object recognition and detection method of foreign matter device is removed based on laser |
CN110488309A (en) * | 2018-06-16 | 2019-11-22 | 南京理工大学 | The target Calibration Method of foreign matter device is removed based on laser |
CN110480180A (en) * | 2018-06-16 | 2019-11-22 | 南京理工大学 | The object recognition and detection method of foreign matter device is removed based on reflection-type laser |
CN110488312A (en) * | 2018-06-16 | 2019-11-22 | 南京理工大学 | The target prediction method of foreign matter device is removed based on reflection-type laser |
CN108898169A (en) * | 2018-06-19 | 2018-11-27 | Oppo广东移动通信有限公司 | Image processing method, picture processing unit and terminal device |
CN108898587A (en) * | 2018-06-19 | 2018-11-27 | Oppo广东移动通信有限公司 | Image processing method, picture processing unit and terminal device |
CN109001833A (en) * | 2018-06-22 | 2018-12-14 | 天和防务技术(北京)有限公司 | A kind of Terahertz hazardous material detection method based on deep learning |
CN108981277B (en) * | 2018-06-29 | 2021-07-23 | 海尔智家股份有限公司 | Method and apparatus for modifying beverage shelf life for a refrigerator |
CN108981277A (en) * | 2018-06-29 | 2018-12-11 | 青岛海尔股份有限公司 | The method and apparatus of modification beverage shelf-life for refrigerator |
CN108895756B (en) * | 2018-06-29 | 2020-11-20 | 海尔智家股份有限公司 | Method and apparatus for displaying caloric value of food for refrigerator |
CN108895756A (en) * | 2018-06-29 | 2018-11-27 | 青岛海尔股份有限公司 | The method and apparatus of the calorie value of displaying food for refrigerator |
CN108876823B (en) * | 2018-07-02 | 2022-05-17 | 晋建志 | Monocular cross-camera multi-target recognition, positioning and tracking device and method based on space-time continuity |
CN108876823A (en) * | 2018-07-02 | 2018-11-23 | 晋建志 | Based on across the camera multi-targets recognition locating and tracking method of time and space continuity monocular |
CN109308693A (en) * | 2018-08-29 | 2019-02-05 | 北京航空航天大学 | By the target detection and pose measurement list binocular vision system of a ptz camera building |
WO2020057346A1 (en) * | 2018-09-21 | 2020-03-26 | 深圳市九洲电器有限公司 | Video monitoring method and apparatus, monitoring server and video monitoring system |
CN109241938A (en) * | 2018-09-27 | 2019-01-18 | 深圳金三立视频科技股份有限公司 | Congestion in road detection method and terminal |
CN109410496A (en) * | 2018-10-25 | 2019-03-01 | 北京交通大学 | Attack early warning method, apparatus and electronic equipment |
CN109544595B (en) * | 2018-10-29 | 2020-06-16 | 苏宁易购集团股份有限公司 | Customer path tracking method and system |
CN109544595A (en) * | 2018-10-29 | 2019-03-29 | 苏宁易购集团股份有限公司 | A kind of customer's route tracing method and system |
CN109587445A (en) * | 2018-11-18 | 2019-04-05 | 中设设计集团股份有限公司 | A kind of security system of defense based on lamp networking |
CN109816310A (en) * | 2019-01-22 | 2019-05-28 | 浙江华是科技股份有限公司 | A kind of navigation channel Ship IC Card port management system and its management method based on artificial intelligence |
CN109800714A (en) * | 2019-01-22 | 2019-05-24 | 浙江华是科技股份有限公司 | A kind of ship detecting system and method based on artificial intelligence |
CN109977783A (en) * | 2019-02-28 | 2019-07-05 | 浙江新再灵科技股份有限公司 | Method based on the independent boarding detection of vertical ladder scene perambulator |
CN109873987A (en) * | 2019-03-04 | 2019-06-11 | 深圳市梦网百科信息技术有限公司 | A kind of Target Searching Method and system based on monitor video |
CN109873987B (en) * | 2019-03-04 | 2020-12-29 | 深圳市梦网视讯有限公司 | Target searching method and system based on monitoring video |
CN110135377A (en) * | 2019-05-21 | 2019-08-16 | 北京百度网讯科技有限公司 | Object moving state detection method, device, server and computer-readable medium |
CN110290351A (en) * | 2019-06-26 | 2019-09-27 | 广东康云科技有限公司 | A kind of video target tracking method, system, device and storage medium |
CN110939351A (en) * | 2019-10-28 | 2020-03-31 | 优创嘉(大连)科技有限公司 | Visual intelligent control method and visual intelligent control door |
CN110991283A (en) * | 2019-11-21 | 2020-04-10 | 北京格灵深瞳信息技术有限公司 | Re-recognition and training data acquisition method and device, electronic equipment and storage medium |
CN111191507A (en) * | 2019-11-26 | 2020-05-22 | 恒大智慧科技有限公司 | Safety early warning analysis method and system for smart community |
CN110933318A (en) * | 2019-12-12 | 2020-03-27 | 天地伟业技术有限公司 | Moving object snapshot method |
CN111079638A (en) * | 2019-12-13 | 2020-04-28 | 河北爱尔工业互联网科技有限公司 | Target detection model training method, device and medium based on convolutional neural network |
CN111339977A (en) * | 2020-03-03 | 2020-06-26 | 河南中光学集团有限公司 | Small target intelligent identification system based on remote video monitoring and identification method thereof |
CN112866558A (en) * | 2020-11-04 | 2021-05-28 | 苏州臻迪智能科技有限公司 | Operation method of electronic equipment, control method of holder and holder system |
CN112330650A (en) * | 2020-11-12 | 2021-02-05 | 李庆春 | Retrieval video quality evaluation method |
CN112804487A (en) * | 2020-12-30 | 2021-05-14 | 济南和普威视光电技术有限公司 | Double-path video target tracking method and system |
CN112785639A (en) * | 2021-01-06 | 2021-05-11 | 鞍钢矿业爆破有限公司 | Video analysis-based method for metering bagged materials into and out of warehouse |
CN113408518A (en) * | 2021-07-06 | 2021-09-17 | 世邦通信股份有限公司 | Audio and video acquisition equipment control method and device, electronic equipment and storage medium |
CN113344965A (en) * | 2021-08-05 | 2021-09-03 | 鑫安利中(北京)科技有限公司 | Safety management system based on image acquisition |
CN113723841A (en) * | 2021-09-03 | 2021-11-30 | 安徽大学绿色产业创新研究院 | Online detection method for tool missing in assembly type prefabricated part |
CN113743380A (en) * | 2021-11-03 | 2021-12-03 | 江苏博子岛智能产业技术研究院有限公司 | Active tracking method based on video image dynamic monitoring |
CN114067575A (en) * | 2021-11-23 | 2022-02-18 | 安徽富煌科技股份有限公司 | Traffic hub region safety analysis device based on 3D structured light detection |
CN114220077A (en) * | 2022-02-21 | 2022-03-22 | 金叶仪器(山东)有限公司 | Method for realizing object quantity statistics and moving direction monitoring based on monitoring equipment |
CN115063452A (en) * | 2022-06-13 | 2022-09-16 | 中国船舶重工集团公司第七0七研究所九江分部 | Cloud deck camera tracking method for offshore target |
CN115063452B (en) * | 2022-06-13 | 2024-03-26 | 中国船舶重工集团公司第七0七研究所九江分部 | Cloud deck camera tracking method for offshore targets |
CN115914582A (en) * | 2023-01-05 | 2023-04-04 | 百鸟数据科技(北京)有限责任公司 | Small object detection optimization method based on fusion time sequence information |
CN115914582B (en) * | 2023-01-05 | 2023-04-28 | 百鸟数据科技(北京)有限责任公司 | Small object detection optimization method based on fusion time sequence information |
CN117649537A (en) * | 2024-01-30 | 2024-03-05 | 浙江省公众信息产业有限公司 | Monitoring video object identification tracking method, system, electronic equipment and storage medium |
CN117649537B (en) * | 2024-01-30 | 2024-04-26 | 浙江省公众信息产业有限公司 | Monitoring video object identification tracking method, system, electronic equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108055501A (en) | A kind of target detection and the video monitoring system and method for tracking | |
CN109657575B (en) | Intelligent video tracking algorithm for outdoor constructors | |
CN106707296B (en) | It is a kind of based on the unmanned machine testing of Based on Dual-Aperture photo electric imaging system and recognition methods | |
CN105744232B (en) | A kind of method of the transmission line of electricity video external force damage prevention of Behavior-based control analytical technology | |
CN101751677B (en) | Target continuous tracking method based on multi-camera | |
CN103927878B (en) | A kind of automatic shooting device for parking offense and automatically grasp shoot method | |
CN101587606B (en) | Method and system for detecting staff flow in exhibition venue | |
CN107145851A (en) | Constructions work area dangerous matter sources intelligent identifying system | |
CN102447835A (en) | Non-blind-area multi-target cooperative tracking method and system | |
CN104637058B (en) | A kind of volume of the flow of passengers identify statistical methods based on image information | |
CN102426785B (en) | Traffic flow information perception method based on contour and local characteristic point and system thereof | |
CN103246896B (en) | A kind of real-time detection and tracking method of robustness vehicle | |
CN112396658B (en) | Indoor personnel positioning method and system based on video | |
CN108447091A (en) | Object localization method, device, electronic equipment and storage medium | |
CN108229407A (en) | A kind of behavioral value method and system in video analysis | |
CN111325048B (en) | Personnel gathering detection method and device | |
CN103997624A (en) | Overlapped domain dual-camera target tracking system and method | |
CN106558224B (en) | A kind of traffic intelligent monitoring and managing method based on computer vision | |
CN101537618A (en) | Visual system for ball picking robot in stadium | |
CN102799935A (en) | Human flow counting method based on video analysis technology | |
CN103581614A (en) | Method and system for tracking targets in video based on PTZ | |
CN105185118A (en) | Vehicle peccancy detection device and method based on video detection | |
CN109445453A (en) | A kind of unmanned plane Real Time Compression tracking based on OpenCV | |
CN109376584A (en) | A kind of poultry quantity statistics system and method for animal husbandry | |
CN105957300B (en) | A kind of wisdom gold eyeball identification is suspicious to put up masking alarm method and device |
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 | ||
AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20231208 |
|
AD01 | Patent right deemed abandoned |