CN107563961A - A kind of system and method for the moving-target detection based on camera sensor - Google Patents
A kind of system and method for the moving-target detection based on camera sensor Download PDFInfo
- Publication number
- CN107563961A CN107563961A CN201710780857.3A CN201710780857A CN107563961A CN 107563961 A CN107563961 A CN 107563961A CN 201710780857 A CN201710780857 A CN 201710780857A CN 107563961 A CN107563961 A CN 107563961A
- Authority
- CN
- China
- Prior art keywords
- moving
- camera
- image
- target
- sensor
- 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
Landscapes
- Studio Devices (AREA)
Abstract
The present invention discloses a kind of system and method for the moving-target detection based on camera sensor.The present invention includes camera, Data transfer system, satellited system, camera includes optical machine main body and camera controller, and optical machine main body includes imaging focal plane component, and camera control machine includes image processing circuit, focal plane subassembly is three imaging sensor focal plane subassemblies, completes photoelectric signal transformation imaging function;Image processing circuit is embedded in moving-target detection algorithm, completes image on-board processing;Satellited system with realizing Satellite operation management, autonomous control, information transmission and star big circuit operation, Data transfer system complete on star data to the transmission on ground;Camera controller completes the functions such as picture signal processing, bus communication, focusing control.Target reduces target missing inspection to effective detection at a slow speed, suppresses the false-alarm caused by parallax, environmental change etc., improves moving-target detection integrity degree, reduces the false alarm rate of moving-target detection.
Description
Technical field
The present invention relates to moving-target detection field, the system of especially a kind of moving-target detection based on camera sensor and side
Method.
Background technology
Moving-target detection is to extract moving target from background based on sequence image, the detection of moving target for
Later stage target classification, tracking and behavioral study are particularly significant.Due to the factor such as weather, illumination, shade image background, there is also change
Change so that moving-target detection is relatively difficult.Existing moving-target detection is based on monitor video, with China's remote sensing application technology
Continuous improvement and the acceleration of national economic development promote, the data source of moving-target detection is also from Video Expansion to remote sensing image.
The moving-target detection coverage of remote sensing is wider compared with the conventional moving-target detection based on monitor video, obtains information
It is more rich.
The moving-target detection of remote sensing image is main including SAR and multispectral, and SAR can utilize Doppler effect to realize to target
The detection of movable information, multispectral image mainly produce time delay to realize by different spectral coverage to the time irreversibility of image objects
Detection to high-speed object, as the panchromatic remote sensing of Main Means, there is single-sensor satellite video moving-target to be detected and used double
Sensor obtains the two width sequence image precession target detections at specified time interval.
Therefore prior art has the following disadvantages:
Monitor video moving-target detects:Scope is small, and it is few to obtain information.
SAR moving-targets detect:Moving-target is submerged in main clutter at a slow speed, and detection integrity degree reduces.
Multispectral image moving-target detects:Resolution ratio is low, and time interval is small, and target detection is relatively more tired to Small object and at a slow speed
Difficulty, moving-target detection integrity degree are relatively low.
Single-sensor satellite video moving-target detects:Multi-purpose background modeling method, because remote sensing satellite range of video is wide, data
Amount is big, and background is complicated, causes background modeling complicated, operation efficiency is relatively low.
The width sequential images moving-target of dual sensor two detects:It is very few to obtain consecutive image, two frame difference methods, nothing can only be used
Method breaks through the shortcomings that two frame differences are intrinsic, changes because the factor image such as weather, illumination, shade background is present, testing result is empty
Alert rate is high.
The content of the invention
It is an object of the invention to provide a kind of system and method for the moving-target detection based on camera sensor, arrangement is used
The spaceborne imaging sensor focal plane subassembly device of three pieces of independent sub-controls on focal plane carries out moving-target detection, and effective detection is slow
Fast target reduces target missing inspection, suppresses the false-alarm caused by parallax, environmental change etc., improves moving-target detection integrity degree, drop
The false alarm rate of low moving-target detection.
To achieve the above object, the invention provides following scheme:
A kind of system of the moving-target detection based on camera sensor, including camera, Data transfer system, satellited system, camera
Including optical machine main body and camera controller, optical machine main body includes imaging focal plane component, and camera control machine includes image procossing electricity
Road, focal plane subassembly are connected with image processing circuit, and image processing circuit is connected with data transmission system, satellited system;Focal plane group
Part is three imaging sensor focal plane subassemblies, completes photoelectric signal transformation imaging function;Image processing circuit insertion moving-target detection
Algorithm, complete image on-board processing;Satellited system is with realizing Satellite operation management, autonomous control, information transmission and star big to return
Dataway operation, Data transfer system complete star on data to ground transmission;Camera controller complete picture signal processing, bus communication,
The functions such as focusing control.
Optionally, in addition to power supply-distribution system, heat control system, camera are connected with distribution system, heat control system, and thermal control provides
Required heating circuit simultaneously carries out temperature control.
Optionally, camera also includes shading shade assembly, camera thermal control component, and camera main-body is supported, shading shade assembly and light
Owner's body is connected, and thermal control component is connected with optical machine main body, thermal control, and camera main-body is supported to be connected with optical machine main body.
Optionally, optical machine main body also include primary mirror component, secondary mirror assembly, correction mirror assembly, focus adjusting mechanism, primary mirror component with
Shading shade assembly, the connection of secondary mirror assembly, secondary mirror assembly are connected with correction component, and correction component is connected with imaging focal plane component;Phase
Machine controller is also included for matching somebody with somebody circuit, and autofocus circuit, for being connected with circuit with image processing circuit, power supply-distribution system, focusing is electric
Road is connected with supplying with circuit, focus adjusting mechanism.
Optionally, image processing circuit includes master control FPGA, the dsp chip and DDR of embedded moving platen detection algorithm
SDRAM memory is formed, and master control FPGA is completed to optical machine main body and the drive control of camera controller, memory read/write is accessed
Control and reception and transmission to data;Dsp chip is realized to be detected to the moving-target of acquired sequential images;DDR SDRAM are used
In the view data of sensing system of the caching from optical machine main body.
A kind of method of the moving-target detection based on camera sensor, it is characterised in that including step:
(1) focal plane structure imaging;Three image sensor arrays combined by CCD or CMOS on focal plane according to
Secondary arrangement, imaging sensor are separated by certain not imaging region between any two;Each combination sensor independence sub-control imaging, exposure
Order is the first combination sensor, the second combination sensor, the 3rd combination sensor, controls the adjacent combination exposure sensor moment
Interval time t, ensure that imaging gained image overlap area size and target moving displacement are sufficient for moving-target detection three times;Three
Times exposure image pushes away sweep in the y-direction, and the k-1 moment only has first sensor imaging, and the k moment after the t times only has the second sensing
Device is imaged, then only has 3rd sensor to be imaged at the k+1 moment after the t times, middle all three sensors of t times not into
Picture, the image obtained at different moments three times have certain overlapping region and detected for moving-target;
(2) moving-target detects;Moving-target is detected using three frame difference methods, after three width images of acquisition are registered, carries out two
Two difference, detection intermediate time k moving target;
1. the image registration based on SURF algorithm;The image on the basis of k time chart pictures, respectively to k-1 time charts picture and k+2
Moment image carries out registration;
2. three frame difference method moving-targets detect
K-1 moment images after subtracting registration with k moment images, motion change image is obtained, then subtracted with k moment images
K+1 moment images after registration, obtain motion change image, and modified-image carries out "AND" and moving object detection knot is calculated
Fruit;With computational methods be:If same position is being identified as moving target with, this position is in k moment images
Moving target, testing result are binary map, be worth for 1 be the k moment moving target;Final unit passes output campaign binary map.
According to specific embodiment provided by the invention, the invention discloses following technique effect:
The present invention obtains image using three pieces of image sensor apparatus being arranged on focal plane and carries out moving-target detection, carries
Operation efficiency is risen, improves moving-target detection integrity degree, reduces the false alarm rate of moving-target detection.It is imaged with every specified time, to dynamic
Target detection is more sensitive, and testing result integrity degree is high;It is poor that three exclusive sensor constructions obtain the frame of three width time-series image three
Method detects moving-target, and false alarm rate is low, and computing is simple;All detection process are all completed on star, transmit binary map, are reduced data and are passed
Defeated pressure.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to institute in embodiment
The accompanying drawing needed to use is briefly described, it should be apparent that, drawings in the following description are only some implementations of the present invention
Example, for those of ordinary skill in the art, without having to pay creative labor, can also be according to these accompanying drawings
Obtain other accompanying drawings.
Fig. 1 is a kind of structural representation of the system of the moving-target detection based on camera sensor of the present invention;
Fig. 2 is a kind of image processing circuit structural representation of the system of the moving-target detection based on camera sensor of the present invention
Figure;
Fig. 3 is a kind of device focal plane layout drawing of the method for the moving-target detection based on camera sensor of the present invention;
Fig. 4 is a kind of imaging schematic diagram three times of the method for the moving-target detection based on camera sensor of the present invention;
Fig. 5 is a kind of moving-target detection algorithm flow of the method for the moving-target detection based on camera sensor of the present invention
Figure;
Fig. 6 is a kind of tank filters target figure of the method for the moving-target detection based on camera sensor of the present invention.
In figure, 1- cameras, 2- Data transfer systems, 3- satellited systems, 4- power supply-distribution systems, 5- heat control systems, 11- ray machine masters
Body, 12- camera controllers, 13- shading shade assemblies, 14- camera thermal control components, the support of 15- camera main-bodies, 111- imaging focal plane groups
Part, 112- primary mirrors component, 113- mirror assembly, 114 correction mirror assemblies, 115- focus adjusting mechanisms, 121- image processing circuits, 122-
For with circuit, 123- autofocus circuits.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, rather than whole embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other under the premise of creative work is not made
Embodiment, belong to the scope of protection of the invention.
It is an object of the invention to provide a kind of system and method for the moving-target detection based on camera sensor, arrangement is used
The spaceborne imaging sensor focal plane subassembly device of three pieces of independent sub-controls on focal plane carries out moving-target detection, and effective detection is slow
Fast target reduces target missing inspection, suppresses the false-alarm caused by parallax, environmental change etc., improves moving-target detection integrity degree, drop
The false alarm rate of low moving-target detection.
In order to facilitate the understanding of the purposes, features and advantages of the present invention, it is below in conjunction with the accompanying drawings and specific real
Applying mode, the present invention is further detailed explanation.
Embodiment:
As shown in figure 1, a kind of system of the moving-target detection based on camera sensor, including camera 1, Data transfer system 2, star
Business system 3, camera 1 include optical machine main body 11 and camera controller 12, and optical machine main body 11 includes imaging focal plane component 111, camera
Control machine 12 includes image processing circuit 121, and focal plane subassembly 111 is connected with image processing circuit 121, image processing circuit
121 are connected 3 with data transmission system 2, satellited system;Focal plane subassembly 111 is three imaging sensor focal plane subassemblies, completes optical telecommunications
Number conversion imaging function;Image processing circuit 121 is embedded in moving-target detection algorithm, completes image on-board processing;Satellited system is real
The big circuit operation in existing Satellite operation management, autonomous control, information transmission and star ground, Data transfer system 2 complete on star data to
The transmission in face;Camera controller 12 completes picture signal processing, bus communication, focusing control function.Also include power supply-distribution system
4, heat control system 5, camera 1 is connected with distribution system 4, heat control system 5, and heat control system 5 provides required heating circuit and progress
Temperature control.Camera 1 also includes shading shade assembly 13, camera thermal control component 14, camera main-body support 15, shading shade assembly 13 and ray machine
Main body connection 11, camera thermal control component 14 is connected with optical machine main body 11, heat control system 5, and camera main-body supports 15 and optical machine main body
11 connections.Optical machine main body 11 also includes primary mirror component 112, secondary mirror assembly 113, correction mirror assembly 114, focus adjusting mechanism 115, primary mirror
Component 112 is connected 113 with shading shade assembly 13, secondary mirror assembly, and secondary mirror assembly 113 is connected with correction component 114, corrects component
114 are connected with imaging focal plane component 111;Camera controller 12 is also included for matching somebody with somebody circuit 122, autofocus circuit 123, for circuit
122 are connected with image processing circuit 121, power supply-distribution system 4, and autofocus circuit 123 connects with supplying with circuit 122, focus adjusting mechanism 115
Connect.
Image processing circuit 121 includes master control FPGA, the dsp chip of embedded moving platen detection algorithm and DDR SDRAM and deposited
Reservoir, master control FPGA completes to optical machine main body 11 and the drive control of camera controller 12, to memory read/write access control and
Reception and transmission to data;Dsp chip is realized to be detected to the moving-target of acquired sequential images;DDR SDRAM are used to cache
The view data of sensing system from optical machine main body 11.
A kind of method of the moving-target detection based on camera sensor, including step:
(1) focal plane structure imaging;Three image sensor arrays combined by CCD or CMOS on focal plane according to
Secondary arrangement, imaging sensor are separated by certain not imaging region between any two;Each combination sensor independence sub-control imaging, exposure
Order is the first combination sensor, the second combination sensor, the 3rd combination sensor, controls the adjacent combination exposure sensor moment
Interval time t, ensure that imaging gained image overlap area size and target moving displacement are sufficient for moving-target detection three times;Three
Times exposure image pushes away sweep in the y-direction, and the k-1 moment only has first sensor imaging, and the k moment after the t times only has the second sensing
Device is imaged, then only has 3rd sensor to be imaged at the k+1 moment after the t times, middle all three sensors of t times not into
Picture, the image obtained at different moments three times have certain overlapping region and detected for moving-target;
(2) moving-target detects;Moving-target is detected using three frame difference methods, after three width images of acquisition are registered, carries out two
Two difference, detection intermediate time k moving target;
1. the image registration based on SURF algorithm;The image on the basis of k time chart pictures, respectively to k-1 time charts picture and k+2
Moment image carries out registration;
2. three frame difference method moving-targets detect
K-1 moment images after subtracting registration with k moment images, motion change image is obtained, then subtracted with k moment images
K+1 moment images after registration, obtain motion change image, and modified-image carries out "AND" and moving object detection knot is calculated
Fruit;With computational methods be:If same position is being identified as moving target with, this position is in k moment images
Moving target, testing result are binary map, be worth for 1 be the k moment moving target;Final unit passes output campaign binary map.
(1) device focal plane structure and imaging method
Three image sensor arrays combined by CCD or CMOS are arranged in order on focal plane, to ensure to be imaged
Time interval, it is separated by certain not imaging region between any two.Such as arrangement of the imaging sensor on focal plane that Fig. 3 is device
Schematic diagram.
Each combination sensor independence sub-control imaging, exposure order is combination sensor I → combination sensor II → combination
Sensor III, adjacent combination exposure sensor moment interval time t is controlled, ensure that imaging gained image overlap area is big three times
Small and target moving displacement is sufficient for moving-target detection.Three times exposure image schematic diagram such as Fig. 4.Three times exposure images are along y side
To be swept to pushing away, the k-1 moment only has sensor I imagings, and the k moment after the t times only has sensor II imagings, then after the t times only
There is sensor III to be imaged at the k+1 moment, middle all three sensors of t times are not imaged, and are obtained at different moments three times
Image has certain overlapping region and detected for moving-target.
(2) device moving target detection method
Moving target detection method is three frame difference methods, after three width images of acquisition are registered, carries out difference two-by-two, among detection
Moment k moving target.Algorithm flow chart such as Fig. 5.
1. the image registration based on SURF algorithm
The image on the basis of k time chart pictures in the present invention, matches somebody with somebody to k-1 time charts picture and k+2 moment images respectively
It is accurate.Method for registering images based on SURF algorithm, it is based on Hessian operator extractions reference images and image subject to registration first
Characteristic point, then description is generated with Haar small echos, then carry out characteristic matching and reject error matching points using RANSAC algorithms,
Affine transformation Matching Model is finally established with the correct match point filtered out, registering image is obtained through model resampling.
Selection of the SURF algorithm to characteristic point uses Hessian operators.If X=(x, y) is a point on image I,
Hessian is in point X=(x, y), definition when yardstick is δ:
In formula:Lxx(X, δ) is gaussian filtering second dervativeWith the result of image I (x, y) convolution,Lxy(X, δ), LyyThe implication of (X, δ) is similar.Typically now two are replaced using tank filters
Rank gaussian filtering.
According to the requirement of SURF algorithm, extreme point is found on different scale, it is necessary to establish the metric space of image.Box-shaped
Wave filter replaces second order Gauss filtering to be used for establishing metric space.On the original image, formed not by expanding the size of box
With the image pyramid of yardstick.Box filtering filters for second order Gauss introduces ω proportionality coefficients, uses Dxx、DyyAnd DxyRepresent mould
Plate with the value after image convolution, Hessian determinants of a matrix can approximation be reduced to Fig. 6:
Det (H)=DxxDyy-(ωDxy)2
Haar wavelet transform calculates description operator:To ensure rotational invariance, first centered on characteristic point, calculating radius is 6s
Haar small echo (the Haar small echo length of side take 4s) of the point in x, y direction in the neighborhood of (s is characterized a scale-value at place) responds,
And assign Gauss weight coefficient to these responses so that the response contribution close to characteristic point is big, and the response tribute away from characteristic point
Offer small, more meet objective reality;Secondly the response in the range of 60 ° is summed to form new vector, travels through whole border circular areas,
The direction of most long vector is selected as the principal direction of this feature point.Then with principal direction generation description of all characteristic points, with spy
Centered on sign point, reference axis is rotated into principal direction first, the square area that the length of side is 20s is chosen according to principal direction, by this
Window area is divided into 4 × 4 subregion, in each sub-regions, calculates small in the range of 5s × 5s (sampling step length takes s)
Ripple responds, and is denoted as dx, dy respectively relative to the Haar small echos response of the level of principal direction, vertical direction, equally assigns response
With weight coefficient, to increase the robustness to geometric transformation;Then by the absolute value phase of the response of every sub-regions and response
Add to form ∑ dx, ∑ dy, ∑ | dx |, ∑ | dx |.So, every sub-regions formed four-dimensional component vector V=(∑ dx, ∑ |
Dx |, ∑ dy, ∑ | dy |), therefore, to each characteristic point, then the description vectors of 4 × (4 × 4)=64 dimension are formed, then enter row vector
Normalization.
Matching based on distance:The matching of characteristic point is using selected characteristic vector Euclidean distance as being closed in two field pictures
The similarity determination measurement of key point, the characteristic vector Euclidean distance of two characteristic points are then considered as matching double points within the specific limits.
The thought of RANSAC algorithms is:Three matching double points are randomly chosen first, and estimate model parameter, by what is obtained
The error of other match points of model parameter calculation;When error is less than given threshold value, then the match point supports current model
Parameter;Point counting number to meeting model;Randomly choose match point again in addition, after the certain number of iteration, selection is supported current
The most set of matches of match point of model parameter recalculate model parameter as last model.Affine Transform Model
RANSAC algorithm steps are:
A. 3 points are randomly selected to establishing equation group from n group candidate matches characteristic point centerings, solves 6 of transformation matrix M
Unknown parameter.
B. remaining n-3 characteristic point is calculated after transformation matrix M conversion, with the distance between its candidate matches point.
If c. distance is less than the threshold value of setting, this feature point is interior point, is otherwise exterior point.
D. the quantity of the interior point under the transformation matrix is counted.
E. 3 pairs of candidate matches points pair again again, repeat a~d, several times after, selection is most with interior quantity
Set is final matching point set.
After determining model parameter, resampling is carried out to image subject to registration using affine transformation, obtains registering image.
2. three frame difference method moving-targets detect
With k moment images fk(x, y) subtracts the k-1 moment images f after registrationk-1(x, y), obtain motion change image g1
(x, y), then the k+1 moment images f after subtracting registration with k moment imagesk+1(x, y), obtain motion change image g2(x, y), become
Change image progress "AND" and moving object detection result D (x, y) is calculated.
g1(x, y) and g2The computational methods of (x, y) are:
"AND" is calculated as:
If same position is in g1(x, y) and g2Moving target is identified as in (x, y), then this position is in k moment images
Moving target, testing result D (x, y) is binary map, be worth for 1 be the k moment moving target.Final unit passes output campaign
Binary map.
Specific case used herein is set forth to the principle and embodiment of the present invention, and above example is said
It is bright to be only intended to help the method and its core concept for understanding the present invention;Meanwhile for those of ordinary skill in the art, foundation
The thought of the present invention, in specific embodiments and applications there will be changes.In summary, this specification content is not
It is interpreted as limitation of the present invention.
Claims (6)
1. a kind of system of the moving-target detection based on camera sensor, it is characterised in that including camera, Data transfer system, Star Service
System, the camera include optical machine main body and camera controller, and the optical machine main body includes imaging focal plane component, the camera control
Machine processed includes image processing circuit, and the focal plane subassembly is connected with described image process circuit, described image process circuit with
The data transmission system, satellited system connection;The focal plane subassembly is three imaging sensor focal plane subassemblies, completes photoelectricity
Signal conversion imaging function;Described image process circuit is embedded in moving-target detection algorithm, completes image on-board processing;The Star Service
Complete star on realizing Satellite operation management, autonomous control, information transmission and star by big circuit operation, the Data transfer system for system
Transmission of the data to ground;The camera controller completes picture signal processing, bus communication, focusing control function.
2. a kind of system of the moving-target detection based on camera sensor according to claim 1, it is characterised in that also include
Power supply-distribution system, heat control system, the camera are connected with the distribution system, the heat control system, needed for the thermal control offer
Heating circuit and carry out temperature control.
A kind of 3. system of the moving-target detection based on camera sensor according to claim 2, it is characterised in that the phase
Machine also includes shading shade assembly, camera thermal control component, and camera main-body supports that the shading shade assembly connects with the optical machine main body
Connect, the thermal control component is connected with the optical machine main body, the thermal control, and the camera main-body is supported to connect with the optical machine main body
Connect.
A kind of 4. system of the moving-target detection based on camera sensor according to claim 3, it is characterised in that the light
Owner's body also includes primary mirror component, secondary mirror assembly, correction mirror assembly, focus adjusting mechanism, the primary mirror component and the light shield group
Part, the secondary mirror assembly connection, the secondary mirror assembly are connected with the correction component, the correction component and the imaging focal plane
Component connects;The camera controller also includes supplying to match somebody with somebody circuit, and autofocus circuit is described for electric with circuit and described image processing
Road, power supply-distribution system connection, the autofocus circuit is with described for being connected with circuit, the focus adjusting mechanism.
A kind of 5. system of the moving-target detection based on camera sensor according to claim 1, it is characterised in that the figure
Formed as process circuit includes master control FPGA, the dsp chip of embedded moving platen detection algorithm and DDR SDRAM memories, it is described
Master control FPGA is completed to the optical machine main body and the drive control of the camera controller, to the memory read/write access control
With the reception and transmission to data;The dsp chip is realized to be detected to the moving-target of acquired sequential images;The DDR
SDRAM is used for the view data for caching the sensing system from the optical machine main body.
A kind of 6. method of the moving-target detection based on camera sensor, it is characterised in that including step:
(1) focal plane structure imaging;Three image sensor arrays combined by CCD or CMOS are arranged successively on focal plane
Row, described image sensor are separated by certain not imaging region between any two;Each combination sensor independence sub-control imaging, exposure
Order is the first combination sensor, the second combination sensor, the 3rd combination sensor, controls the adjacent combination exposure sensor moment
Interval time t, ensure that imaging gained image overlap area size and target moving displacement are sufficient for moving-target detection three times;Three
Times exposure image pushes away sweep in the y-direction, and the k-1 moment only has first sensor imaging, and the k moment after the t times only has the second sensing
Device is imaged, then only has 3rd sensor to be imaged at the k+1 moment after the t times, middle all three sensors of t times not into
Picture, the image obtained at different moments three times have certain overlapping region and detected for moving-target;
(2) moving-target detects;Moving-target is detected using three frame difference methods, after three width images of acquisition are registered, it is poor two-by-two to carry out
Point, detection intermediate time k moving target;
1. the image registration based on SURF algorithm;The image on the basis of the k time charts picture, respectively to the k-1 time charts picture
Registration is carried out with k+2 moment image;
2. three frame difference method moving-targets detect
K-1 moment images after subtracting registration with the k moment image, motion change image is obtained, then subtracted with k moment images
K+1 moment images after registration, obtain motion change image, and modified-image carries out "AND" and moving object detection knot is calculated
Fruit;With computational methods be:If same position is being identified as moving target with, this position is in k moment images
Moving target, testing result are binary map, be worth for 1 be the k moment moving target;Final unit passes output campaign binary map.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710780857.3A CN107563961A (en) | 2017-09-01 | 2017-09-01 | A kind of system and method for the moving-target detection based on camera sensor |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710780857.3A CN107563961A (en) | 2017-09-01 | 2017-09-01 | A kind of system and method for the moving-target detection based on camera sensor |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107563961A true CN107563961A (en) | 2018-01-09 |
Family
ID=60977810
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710780857.3A Pending CN107563961A (en) | 2017-09-01 | 2017-09-01 | A kind of system and method for the moving-target detection based on camera sensor |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107563961A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109035306A (en) * | 2018-09-12 | 2018-12-18 | 首都师范大学 | Moving-target automatic testing method and device |
CN109087378A (en) * | 2018-09-11 | 2018-12-25 | 首都师范大学 | Image processing method and system |
CN109165628A (en) * | 2018-09-12 | 2019-01-08 | 首都师范大学 | Improve method, apparatus, electronic equipment and the storage medium of moving-target detection accuracy |
CN109255353A (en) * | 2018-09-12 | 2019-01-22 | 首都师范大学 | A kind of moving target detection method, device, electronic equipment and storage medium |
CN109284707A (en) * | 2018-09-12 | 2019-01-29 | 首都师范大学 | Moving target detection method and device |
CN110378216A (en) * | 2019-06-13 | 2019-10-25 | 浙江大华技术股份有限公司 | Object detection method, device, picture pick-up device and storage medium |
CN110877067A (en) * | 2018-09-05 | 2020-03-13 | Blm有限公司 | Pipe machining machine with device for detecting any slipping of the machined pipe |
CN111929717A (en) * | 2020-07-24 | 2020-11-13 | 北京航空航天大学 | Satellite-borne image processor and processing method for remote sensing image target characteristic identification |
CN112514373A (en) * | 2018-08-14 | 2021-03-16 | 华为技术有限公司 | Image processing apparatus and method for feature extraction |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102184550A (en) * | 2011-05-04 | 2011-09-14 | 华中科技大学 | Mobile platform ground movement object detection method |
CN102509307A (en) * | 2011-10-12 | 2012-06-20 | 西安理工大学 | Method for searching moving target based on longitude and latitude location and image registration |
CN103325112A (en) * | 2013-06-07 | 2013-09-25 | 中国民航大学 | Quick detecting method for moving objects in dynamic scene |
-
2017
- 2017-09-01 CN CN201710780857.3A patent/CN107563961A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102184550A (en) * | 2011-05-04 | 2011-09-14 | 华中科技大学 | Mobile platform ground movement object detection method |
CN102509307A (en) * | 2011-10-12 | 2012-06-20 | 西安理工大学 | Method for searching moving target based on longitude and latitude location and image registration |
CN103325112A (en) * | 2013-06-07 | 2013-09-25 | 中国民航大学 | Quick detecting method for moving objects in dynamic scene |
Non-Patent Citations (2)
Title |
---|
徐鸿伟: "分布式多摄像机协同的运动目标检测", 《中国优秀硕士学位论文全文数据库-信息科技辑》 * |
王孝艳: "运动目标检测的三帧差法算法研究", 《沈阳理工大学学报》 * |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112514373A (en) * | 2018-08-14 | 2021-03-16 | 华为技术有限公司 | Image processing apparatus and method for feature extraction |
CN112514373B (en) * | 2018-08-14 | 2023-09-15 | 华为技术有限公司 | Image processing apparatus and method for feature extraction |
CN110877067A (en) * | 2018-09-05 | 2020-03-13 | Blm有限公司 | Pipe machining machine with device for detecting any slipping of the machined pipe |
CN109087378A (en) * | 2018-09-11 | 2018-12-25 | 首都师范大学 | Image processing method and system |
CN109255353A (en) * | 2018-09-12 | 2019-01-22 | 首都师范大学 | A kind of moving target detection method, device, electronic equipment and storage medium |
CN109284707A (en) * | 2018-09-12 | 2019-01-29 | 首都师范大学 | Moving target detection method and device |
CN109035306A (en) * | 2018-09-12 | 2018-12-18 | 首都师范大学 | Moving-target automatic testing method and device |
CN109035306B (en) * | 2018-09-12 | 2020-12-15 | 首都师范大学 | Moving target automatic detection method and device |
CN109165628A (en) * | 2018-09-12 | 2019-01-08 | 首都师范大学 | Improve method, apparatus, electronic equipment and the storage medium of moving-target detection accuracy |
CN109284707B (en) * | 2018-09-12 | 2021-07-20 | 首都师范大学 | Moving target detection method and device |
CN110378216A (en) * | 2019-06-13 | 2019-10-25 | 浙江大华技术股份有限公司 | Object detection method, device, picture pick-up device and storage medium |
CN110378216B (en) * | 2019-06-13 | 2022-03-01 | 浙江大华技术股份有限公司 | Target detection method, target detection device, image pickup apparatus, and storage medium |
CN111929717A (en) * | 2020-07-24 | 2020-11-13 | 北京航空航天大学 | Satellite-borne image processor and processing method for remote sensing image target characteristic identification |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107563961A (en) | A kind of system and method for the moving-target detection based on camera sensor | |
CN107253485B (en) | Foreign matter invades detection method and foreign matter invades detection device | |
CN104574393B (en) | A kind of three-dimensional pavement crack pattern picture generates system and method | |
CN107993258B (en) | Image registration method and device | |
CN104599258B (en) | A kind of image split-joint method based on anisotropic character descriptor | |
CN110517325B (en) | Coordinate transformation and method and system for positioning objects around vehicle body through coordinate transformation | |
US10909395B2 (en) | Object detection apparatus | |
US11398053B2 (en) | Multispectral camera external parameter self-calibration algorithm based on edge features | |
CN107560592B (en) | Precise distance measurement method for photoelectric tracker linkage target | |
CN106970024B (en) | Distance measuring method and system are detected based on the limit of camera and controllable stroboscopic light source | |
CN111080709B (en) | Multispectral stereo camera self-calibration algorithm based on track feature registration | |
CN105608671A (en) | Image connection method based on SURF algorithm | |
JP6174104B2 (en) | Method, apparatus and system for generating indoor 2D plan view | |
Beekmans et al. | Cloud photogrammetry with dense stereo for fisheye cameras | |
CN102901489A (en) | Pavement water accumulation and ice accumulation detection method and apparatus thereof | |
CN107025660A (en) | A kind of method and apparatus for determining binocular dynamic visual sensor image parallactic | |
CN104168430B (en) | A kind of TDI CCD focal planes code super-resolution imaging device and method | |
Nagy et al. | Online targetless end-to-end camera-LiDAR self-calibration | |
CN113408454B (en) | Traffic target detection method, device, electronic equipment and detection system | |
CN115937288A (en) | Three-dimensional scene model construction method for transformer substation | |
CN115063477A (en) | Infrared and visible light double-channel synchronous imaging real-time registration fusion acquisition method and device | |
Wang et al. | Corners positioning for binocular ultra-wide angle long-wave infrared camera calibration | |
CN114298151A (en) | 3D target detection method based on point cloud data and image data fusion | |
Ma et al. | A subpixel matching method for stereovision of narrow baseline remotely sensed imagery | |
Liu et al. | Vehicle video surveillance system based on image fusion and parallel computing |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180109 |
|
RJ01 | Rejection of invention patent application after publication |