CN102968801A - Moving target tracking method based on photoelectric mixing combination transformation correlation - Google Patents
Moving target tracking method based on photoelectric mixing combination transformation correlation Download PDFInfo
- Publication number
- CN102968801A CN102968801A CN2012103579996A CN201210357999A CN102968801A CN 102968801 A CN102968801 A CN 102968801A CN 2012103579996 A CN2012103579996 A CN 2012103579996A CN 201210357999 A CN201210357999 A CN 201210357999A CN 102968801 A CN102968801 A CN 102968801A
- Authority
- CN
- China
- Prior art keywords
- target
- image
- processing
- correlation
- light modulator
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Abstract
The invention belongs to the technical field of light information processing and digital image processing and particularly relates to a moving target tracking method based on photoelectric mixing combination transformation correlation. The method includes utilizing the quick real-time optical correlation principle to conduct optical correlation operation on image sequence collected by a high speed charge coupled device (CCD) to further obtain mutual correlation peaks of moving images, extracting mass center coordinates of the correlation peaks to obtain relative moving displacement between adjacent image frames. The moving target tracking method adopts an optical correlation processing method, and utilizes light speed to operate. The operation speed is improved by hundreds of times compared with the pure digital processing, and the method lays a foundation for real-time tracking. The method is high in mass center position information detection accuracy. The method has the advantages of high speed and large capacity of optical image processing and flexibility, accuracy and programmability in parallel processing and digital processing technology, thereby being high in tracking accuracy and good in real time performance.
Description
The present invention relates to a kind of motion target tracking method, relate in particular to the relevant motion target tracking method of a kind of optic-electronic hybrid joint transform, its purpose is to solve the difficult problem of real time tracking motion target under the complex background, belongs to optical information processing and digital image processing techniques field.
Background technology
Target following is fused images processing, pattern-recognition, a probability multi-disciplinary complicated problems such as stochastic process of touching upon, and it is the research emphasis of computer vision field always, is widely used in the fields such as Aero-Space, missile guidance, safety monitoring.But because the variations such as imaging visual angle, time cause the variation of illumination; The relative motion of different time internal cause causes that target imaging rotates, the variation of convergent-divergent, deformation; The reason such as it is image blurring that the target high-speed motion causes is so that moving object real-time tracking is very difficult.The method of at present target following mainly contains based on characteristic matching, based on detection, based on three major types methods such as filter forecastings.Wherein more outstanding have methods such as SIFT characteristic matching, Meanshiff, particle filter.Wherein the SIFT characteristic matching has good robustness to target rotation, convergent-divergent, distortion etc., but its calculated amount is large, and the storage space that needs is also large; Meanshiff is the average drifting method, uses the nuclear histogram can resist part edge as object module and blocks, but lose robustness to all blocking; Particle filter method adopts the probability theory method future position, and shortcoming is that calculated amount is large, and bearing accuracy is not high.Above-mentioned method commonly used is pure digi-tal and processes, and calculated amount is large, causes the real-time follow-up difficulty.Therefore, the motion target tracking method that a kind of optic-electronic hybrid joint transform of the present invention is relevant, utilize photoelectricity to mix the method for the combination that combines, have dirigibility, accuracy, the programmable advantage of the high speed of optical imagery processing, large capacity, parallel processing and digital processing technology concurrently, therefore, tracking accuracy is high, time property is good.
Summary of the invention
Calculated amount is large in the motion target tracking, tracking is difficult in order to overcome, the poor difficult problem of real-time performance of tracking, the motion target tracking method that a kind of optic-electronic hybrid joint transform of the present invention is relevant, its purpose is to solve the difficult problem of real time tracking motion target under the complex background.
The motion target tracking method that a kind of optic-electronic hybrid joint transform of the present invention is relevant, technical characterstic are the mixed processing methods that adopts Optical Correlation Processing to combine with digital processing.For achieving the above object, design of the present invention is: the image sequence that high-speed CCD is collected carries out the optical joint conversion, then obtain the simple crosscorrelation peak of two width of cloth flanking sequence images, extract the center-of-mass coordinate at simple crosscorrelation peak by centroid algorithm, it is the position of adjacent target image, just can know the moving displacement vector between the adjacent target image, and then can calculate the movement locus of tracking time internal object, realize the high precision real-time follow-up.
According to the foregoing invention design, the motion target tracking method that a kind of optic-electronic hybrid joint transform is relevant comprises following steps:
Step (1) at first utilizes the high-speed CCD imageing sensor to obtain the image sequence of target, it is stored in the computing machine 1, and adjacent image is input in the spatial light modulator 1 side by side, wherein present frame is reference picture, next frame is input picture (target image).
Step (2) is shone spatial light modulator 1 through the laser (laser) of lens 1 collimation, and by Fourier transform lens 2, at the joint transform power spectrum of its fourier lense output face with CCD1 detector record reference picture and target image.
Step (3) is input to this power spectrum in the computing machine 2, is input in the spatial light modulator 2 through digital processor processes again, through another road Ear Mucosa Treated by He Ne Laser Irradiation spatial light modulator 2 and Fourier transform lens 3, records its relevant peaks by the CCD2 detector.
The center-of-mass coordinate at simple crosscorrelation peak is surveyed and processed to step (4), and namely the center-of-mass coordinate of target image and reference picture namely can be known the relative motion vectors between adjacent two frames.
Step (5) constantly replaces present frame as the reference image with next frame, detects the centroid position of sequence image, can process out the movement locus of target, realizes real-time follow-up.
The method has dirigibility, accuracy, the programmable advantage of the high speed (basically being undertaken by the light velocity) of optical imagery processing, large capacity, parallel processing and digital processing technology simultaneously concurrently, so target tracking accuracy is high, real-time is good.
The motion target tracking method that a kind of optic-electronic hybrid joint transform is relevant, its mainly comprise by: high-speed CCD, CCD1, CCD2, laser instrument (laser), lens 1, lens 2, lens 3, spatial light modulator 1, spatial light modulator 2, computing machine 1, computing machine 2, digital processing unit form.
Described high-speed CCD is used for obtaining target image sequence.
Described detector C CD1, CCD2 are used for surveying joint image power spectrum and relevant peaks.
Described digital processing is used for digital processing and goes out the target centroid position.
Described spatial light modulator 1 and 2 is used for showing image.
Described laser is as the light source of system.
Description of drawings
Fig. 1 is ultimate principle figure of the present invention.
Fig. 2 is target following time-displacement curve figure.
Embodiment
Describe in detail below in conjunction with drawings and Examples, but the present invention is absolutely not only for the embodiment that introduces.
As shown in Figure 1, the motion target tracking method that a kind of optic-electronic hybrid joint transform is relevant may further comprise the steps:
Step (1) at first utilizes the high-speed CCD imageing sensor to obtain the image sequence of target, be stored in the computing machine 1, and adjacent image is input in the spatial light modulator 1 side by side, wherein present frame is reference picture r (x, y), and next frame is input picture (target image) t (x, y), so joint image is r (x, y)+t (x, y).
The Ear Mucosa Treated by He Ne Laser Irradiation spatial light modulator 1 that step (2) will collimate through lens 1, and by Fourier transform lens 2, at the joint transform power spectrum of its fourier lense output face with CCD1 detector record reference picture and target image, its joint power spectrum S (u, v) Mathematical representation is
|S(u,v)|
2=|R(u,v)|
2+|T(u,v)|
2
+R(u,v)T
*(u,v)exp[-2iπuΔx-2iπv(2a+Δy)]
+T(u,v)R
*(u,v)exp[2iπuΔx+2iπv(2a+Δy)]
S (u, v) in the formula, R (u, v), T (u, v) represents respectively s (x, y), r (x, y) and t (x, y) Fourier transform, (u, v) represents respectively x and the volume coordinate of y direction on Fourior plane, wherein x=λ fu, y=λ fv, λ and f represent respectively the operation wavelength of laser and the focal length of lens 2.
Step (3) then is input to this power spectrum in the computing machine 2, be input to again in the spatial light modulator 2 through digital processor processes, through another road Ear Mucosa Treated by He Ne Laser Irradiation spatial light modulator 2 and Fourier transform lens 3, record its relevant peaks by the CCD2 detector, its relevant peaks c (x, y) is output as
In the formula
Represent associative operation, δ represents dirac and meets, and Δ x, Δ y represent respectively the relative displacement vector that adjacent two two field pictures produce because of motion, and it represents positional information.
The center-of-mass coordinate position at simple crosscorrelation peak is surveyed and processed out to step (4), and namely the center-of-mass coordinate of target image and reference picture namely can be known the motion vector between adjacent two frames.Specific formula for calculation is:
I in the formula
IjBe the simple crosscorrelation peak image in the subwindow, X
IjBe the line number of current pixel or row number.
Step (5) constantly replaces present frame as the reference image with next frame, detects the centroid position of sequence image, can process out the movement locus of target, realizes the high precision real-time follow-up.
The present invention utilizes quick real time optical correlation principle that the target image sequence of high-speed CCD collection is carried out the optical correlation computing, realizes the photodetection of picture position information, and processing speed improves hundreds of times than pure digi-tal electronic processing.And then for real-time follow-up provide the basis.The method has dirigibility, accuracy, the programmable advantage of the high speed (basically being undertaken by the light velocity) of optical imagery processing, large capacity, parallel processing and digital processing technology concurrently simultaneously, and therefore, precision is high, real-time is good.
Claims (4)
1. the relevant motion target tracking method of a kind of optic-electronic hybrid joint transform of the present invention is characterized in that may further comprise the steps:
(1) at first utilize the high-speed CCD imageing sensor to obtain the image sequence of target, and adjacent image is input in the spatial light modulator 1 side by side, wherein present frame is reference picture, and next frame is input picture (target image);
(2) the Ear Mucosa Treated by He Ne Laser Irradiation spatial light modulator 1 that collimates through lens 1, and by Fourier transform lens 2, at the joint transform power spectrum of its fourier lense output face with CCD1 detector record reference picture and target image;
(3) this power spectrum is input in the computing machine 2, treated being input to again in the spatial light modulator 2 through another road Ear Mucosa Treated by He Ne Laser Irradiation spatial light modulator 2 and Fourier transform lens 3, recorded its relevant peaks by the CCD2 detector;
(4) survey and process out the centroid position coordinate at simple crosscorrelation peak, i.e. the center-of-mass coordinate of target image and reference picture;
(5) constantly replace present frame as the reference image with next frame, detect the centroid position of sequence image, can process out the movement locus of target, realize real-time follow-up.
2. require describedly according to right 1, it is characterized in that: the target center-of-mass coordinate in the step (4) is based on the basis at optical correlation peak to extract by centroid algorithm and obtains.
3. require describedly according to right 1, it is characterized in that: target following has utilized the method for optical processing, i.e. the optical correlation principle.
4. require describedly according to right 1, it is characterized in that: the method is to adopt the tracking of photoelectricity hybrid processing.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210357999.6A CN102968801B (en) | 2012-09-12 | 2012-09-12 | The motion target tracking method that a kind of optic-electronic hybrid joint transform is relevant |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210357999.6A CN102968801B (en) | 2012-09-12 | 2012-09-12 | The motion target tracking method that a kind of optic-electronic hybrid joint transform is relevant |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102968801A true CN102968801A (en) | 2013-03-13 |
CN102968801B CN102968801B (en) | 2016-01-06 |
Family
ID=47798927
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210357999.6A Expired - Fee Related CN102968801B (en) | 2012-09-12 | 2012-09-12 | The motion target tracking method that a kind of optic-electronic hybrid joint transform is relevant |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102968801B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104100256A (en) * | 2013-04-15 | 2014-10-15 | 西安科技大学 | Method for measuring coal mine underground drilling depth based on image processing technology |
CN107578388A (en) * | 2017-09-11 | 2018-01-12 | 陕西美星恒祺新能源科技有限公司 | A kind of image deblurring precision methods of the raising based on electrical combined Transform Correlator |
CN111158010A (en) * | 2020-01-06 | 2020-05-15 | 航天金鹏科技装备(北京)有限公司 | Laser active tracking system and tracking method |
WO2021228235A1 (en) * | 2020-05-15 | 2021-11-18 | 中国科学院合肥物质科学研究院 | Photoelectric detection and collection system and centroid detection method based on single pixel detector |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20100080329A1 (en) * | 2008-09-29 | 2010-04-01 | Hidetoshi Kawauchi | Information processor and information processing method, display device, and program |
CN102509277A (en) * | 2011-09-14 | 2012-06-20 | 浙江师范大学 | Real-time motion blurred image restoration method for photoelectric hybrid joint transform correlation |
-
2012
- 2012-09-12 CN CN201210357999.6A patent/CN102968801B/en not_active Expired - Fee Related
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20100080329A1 (en) * | 2008-09-29 | 2010-04-01 | Hidetoshi Kawauchi | Information processor and information processing method, display device, and program |
CN102509277A (en) * | 2011-09-14 | 2012-06-20 | 浙江师范大学 | Real-time motion blurred image restoration method for photoelectric hybrid joint transform correlation |
Non-Patent Citations (1)
Title |
---|
樊超等: "基于光学相关的空间相机像移测量方法研究", 《光子学报》, vol. 37, no. 6, 30 June 2008 (2008-06-30), pages 1213 - 1216 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104100256A (en) * | 2013-04-15 | 2014-10-15 | 西安科技大学 | Method for measuring coal mine underground drilling depth based on image processing technology |
CN104100256B (en) * | 2013-04-15 | 2017-04-12 | 西安科技大学 | Method for measuring coal mine underground drilling depth based on image processing technology |
CN107578388A (en) * | 2017-09-11 | 2018-01-12 | 陕西美星恒祺新能源科技有限公司 | A kind of image deblurring precision methods of the raising based on electrical combined Transform Correlator |
CN111158010A (en) * | 2020-01-06 | 2020-05-15 | 航天金鹏科技装备(北京)有限公司 | Laser active tracking system and tracking method |
CN111158010B (en) * | 2020-01-06 | 2022-06-24 | 航天金鹏科技装备(北京)有限公司 | Laser active tracking system and tracking method |
WO2021228235A1 (en) * | 2020-05-15 | 2021-11-18 | 中国科学院合肥物质科学研究院 | Photoelectric detection and collection system and centroid detection method based on single pixel detector |
Also Published As
Publication number | Publication date |
---|---|
CN102968801B (en) | 2016-01-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Qiu et al. | Deeplidar: Deep surface normal guided depth prediction for outdoor scene from sparse lidar data and single color image | |
Hee Lee et al. | Motion estimation for self-driving cars with a generalized camera | |
CN103325112B (en) | Moving target method for quick in dynamic scene | |
US10366480B2 (en) | Super-resolution systems and methods | |
US20120201477A1 (en) | Kinetic Super-Resolution Imaging | |
Huang et al. | Correlation and local feature based cloud motion estimation | |
CN102968801A (en) | Moving target tracking method based on photoelectric mixing combination transformation correlation | |
EP3182372B1 (en) | Method and system for estimating the position of a projection of a chief ray on a sensor of a light-field acquisition device | |
AliAkbarpour et al. | Parallax-tolerant aerial image georegistration and efficient camera pose refinement—without piecewise homographies | |
Cao | Applying image registration algorithm combined with CNN model to video image stitching | |
Fitzgerald et al. | Toward deep-space object detection in persistent wide field of view camera arrays | |
Neumann et al. | Eyes from eyes: New cameras for structure from motion | |
US10373321B2 (en) | Sun position tracking method based on black sun effect of image sensor, computer program stored in medium, and solar sensor using the same | |
Sheinin et al. | Diffraction line imaging | |
Hartung et al. | Improvement of persistent tracking in wide area motion imagery by CNN-based motion detections | |
Saetchnikov et al. | Robust SY-biLSTM object tracking method for on-road objects shoot from an unmanned aerial vehicle | |
Yue et al. | An intelligent identification and acquisition system for UAVs based on edge computing using in the transmission line inspection | |
Zuo et al. | Algorithm for unmanned aerial vehicle aerial different-source image matching | |
CN102509277A (en) | Real-time motion blurred image restoration method for photoelectric hybrid joint transform correlation | |
Karaca et al. | Ground-based panoramic stereo hyperspectral imaging system with multiband stereo matching | |
Cai et al. | A target tracking and location robot system based on omnistereo vision | |
US11448483B1 (en) | Projectile tracking and 3D traceback method | |
Lee et al. | Instance segmentation in urban scenes using inverse perspective mapping of 3D point clouds and 2D images | |
Xie et al. | Real-time reconstruction of unstructured scenes based on binocular vision depth | |
Saovana et al. | A Quantitative Effect Evaluation of the Unwanted Features Removal of Infrastructure Digital Images |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160106 Termination date: 20160912 |