CN105405111A - Hybrid distortion image optimization method based on decoupling control - Google Patents

Hybrid distortion image optimization method based on decoupling control Download PDF

Info

Publication number
CN105405111A
CN105405111A CN201510855824.1A CN201510855824A CN105405111A CN 105405111 A CN105405111 A CN 105405111A CN 201510855824 A CN201510855824 A CN 201510855824A CN 105405111 A CN105405111 A CN 105405111A
Authority
CN
China
Prior art keywords
distortion
model
image
mixing
signal
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
Application number
CN201510855824.1A
Other languages
Chinese (zh)
Inventor
陈勇
帅锋
刘焕淋
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chongqing University of Post and Telecommunications
Original Assignee
Chongqing University of Post and Telecommunications
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Chongqing University of Post and Telecommunications filed Critical Chongqing University of Post and Telecommunications
Priority to CN201510855824.1A priority Critical patent/CN105405111A/en
Publication of CN105405111A publication Critical patent/CN105405111A/en
Pending legal-status Critical Current

Links

Classifications

    • G06T5/70

Landscapes

  • Image Processing (AREA)

Abstract

The present invention relates to a hybrid distortion image optimization method based on decoupling control, and belongs to the technical field of an image optimization method. Aiming at the hybrid distortion image optimization problem, the method comprises: firstly, characterizing a distortion signal by introducing a distortion detection method; secondly, finding a relationship between a different input and a different output in a nonlinear fitting manner; and finally, building a system model according to obtained characteristics and the relationship, and obtaining a decoupling model, thereby fulfilling the aim that one optimization scheme is only efficient for one distortion and cannot cause change for other distortions.

Description

Based on the mixing distorted image optimization method of uneoupled control
Technical field
The invention belongs to technical field of image processing, particularly image optimization method field, relate to a kind of mixing distorted image optimization method based on uneoupled control.
Background technology
Existing image optimization method is generally optimized for a kind of picture signal of certain distortion, as: noise distortion, contrast distortion, fuzzy distortion, JPEG2K distortion and JPEG distortion etc.But the image of everyday exposure, often and not only by the impact of single distortion, acts on while being also likely subject to two or more distortion.For this kind of image by mixing distortion effect, also do not have a kind of good optimization method can process the type of distortion of mixing at present simultaneously.
At present, the prioritization scheme for mixing distorted image is all generally process respectively for certain class distortion of image, namely processes a kind of reprocessing another kind of.But this disposal route there will be again new problem, often have strong coupled relation between different disposal routes, namely when processing a certain distortion, can increase another kind of distortion affects picture quality.Such as: piece image is not only containing noise distortion but also containing contrast distortion, can have an impact to picture signal when denoising is carried out to image, cause contrast distortion more serious, but and the contrast of image is optimized time may strengthen the noise signal of image.
Summary of the invention
In view of this, the object of the present invention is to provide a kind of mixing distorted image optimization method based on uneoupled control, by designing a Decoupled Model, to offset the coupling when processing distortion between different control signal, making when when processing a certain type of distortion of this image, the impact that other distortions increase the weight of picture quality can not be caused.
For achieving the above object, the invention provides following technical scheme:
A kind of mixing distorted image optimization method based on uneoupled control, in the method, by adding Decoupled Model in systems in which, the coupling when being optimized mixing distorted image between each distortion optimizing method being eliminated, to promote effect of optimization, improving picture quality.
Further, the method specifically comprises the following steps:
1), judge that whether image fault type is known, if known, skip step 2) perform step 3), otherwise continue to perform step 2);
2), by analyzing the feature of different distortion, setting up a distortion disaggregated model, its mixing type of distortion need be obtained in image input model to be processed;
3), distorted signal is characterized by introducing distortion detection method;
4) relation of different constrained input, is found out by the mode of nonlinear fitting;
5), according to striked sign and relation set up system model and try to achieve Decoupled Model, realizing a kind of prioritization scheme only to a kind of distortion effectively and the object of change can not be caused to other distortions.
Beneficial effect of the present invention is: the method for the invention is by design Decoupled Model, to offset the coupling when processing distortion between different control signal, making when when processing a certain type of distortion of this image, the impact that other distortions increase the weight of picture quality can not be caused.
Accompanying drawing explanation
In order to make object of the present invention, technical scheme and beneficial effect clearly, the invention provides following accompanying drawing and being described:
Fig. 1 is decoupled system frame diagram;
Fig. 2 has the decoupled system block diagram from coupling;
Fig. 3 has the decoupled system simplified block diagram from coupling;
Fig. 4 is decoupling zero implementing procedure figure.
Embodiment
Below in conjunction with accompanying drawing, the preferred embodiments of the present invention are described in detail.
The object of the invention is to: an image width being subject to simultaneously to multiple distortion effect, design a Decoupled Model, to offset the coupling when processing distortion between different control signal, making when when processing a certain type of distortion of this image, the impact that other distortions increase the weight of picture quality can not be caused.
Fig. 4 is decoupling zero implementing procedure figure, and as shown in the figure, this method specifically comprises the following steps: 1), judge that whether image fault type is known, if known, skip step 2) perform step 3), otherwise continue to perform step 2); 2), by analyzing the feature of different distortion, setting up a distortion disaggregated model, its mixing type of distortion need be obtained in image input model to be processed; 3), distorted signal is characterized by introducing distortion detection method; 4) relation of different constrained input, is found out by the mode of nonlinear fitting; 5), according to striked sign and relation set up system model and try to achieve Decoupled Model, realizing a kind of prioritization scheme only to a kind of distortion effectively and the object of change can not be caused to other distortions.
In the present embodiment, to contain the mixing distortion that two kinds of distortions are formed in image.As shown in Figure 1, x1 (s) represents the control signal processed type of distortion one, x 2s () represents the control signal be optimized the type of distortion two of image; y 1the s picture output signal that type of distortion one was processed that () represents, y 2what s () represented exports the picture signal that type of distortion two processed; G 11s () represents the optimal control signal x through type of distortion one 1s impact that () applies type of distortion one, G 21s () represents the optimal control signal x through type of distortion one 1s impact that () applies the image affected by type of distortion two; G 22s () represents the optimal control signal x through type of distortion two 2s impact that () applies type of distortion two, G 12s () represents the optimal control signal x through type of distortion two 2s impact that () applies the image affected by type of distortion one.D 12(s) and D 21s () is the Decoupled Model that the present invention designs respectively.Following relation can be obtained according to Fig. 1:
G 11 ( s ) x 1 ( s ) - D 21 ( s ) G 12 ( s ) x 1 ( s ) + G 12 ( s ) x 2 ( s ) - D 12 ( s ) G 11 ( s ) x 2 ( s ) = y 1 ( s ) G 22 ( s ) x 2 ( s ) - D 12 ( s ) G 21 ( s ) x 2 ( s ) + G 21 ( s ) x 1 ( s ) - D 21 ( s ) G 22 ( s ) x 1 ( s ) = y 2 ( s ) - - - ( 1 )
In formula (1), output signal y 1s () can be analyzed to: with control signal x 1s matched signal G that () is corresponding 11(s) x 1(s), from coupled signal D 21(s) G 12(s) x 1(s), control signal x 2s undesired signal G that () produces 12(s) x2 (s), for decoupling zero introduce decoupling signal D 12(s) G 11(s) x 2(s).If coupled signal and decoupling signal cancel each other, then relieve control signal x 1s () is to control signal x 2the coupling of (s).Equally, decoupling signal D 21(s) G 22(s) x 1(s) and undesired signal G 21(s) x 1s () is cancelled out each other, can remove control signal x 2s () is to control signal x 1the coupling of (s).
Can obtain (1) formula abbreviation:
[ G 11 ( s ) - D 21 ( s ) G 12 ( s ) ] · x 1 ( s ) + [ G 12 ( s ) - D 12 ( s ) G 11 ( s ) ] · x 2 ( s ) = y 1 ( s ) [ G 22 ( s ) - D 12 ( s ) G 21 ( s ) ] · x 2 ( s ) + [ G 21 ( s ) - D 21 ( s ) G 22 ( s ) ] · x 1 ( s ) = y 2 ( s ) - - - ( 2 )
In order to make y 1s () is only by x 1s () determines, y 2s () is only by x 2s () determines, control single object exported to reach single input, then the condition that signal is full decoupled is:
G 12 ( s ) - D 12 ( s ) G 11 ( s ) = 0 G 21 ( s ) - D 21 ( s ) G 22 ( s ) = 0 - - - ( 3 )
I.e. demand fulfillment:
D 12 ( s ) = G 12 ( s ) G 11 ( s ) D 21 ( s ) = G 21 ( s ) G 22 ( s ) - - - ( 4 )
Wushu (4) substitutes in formula (2) and can obtain:
[ 1 - D 21 ( s ) D 12 ( s ) ] - 1 · x 1 ( s ) · G 11 ( s ) = y 1 ( s ) [ 1 - D 12 ( s ) D 21 ( s ) ] · x 2 ( s ) · G 22 ( s ) = y 2 ( s ) - - - ( 5 )
Wherein, D in formula (5) 21(s) D 12s () is from coupling ring joint, be at two distortion processing control signals x 1(s) and x 2s lower two transport functions produced of () effect are identical, but independent of each other from coupled signal.If connect a compensation tache [1-D in systems in which 21(s) D 12(s)] -1, then can carry out from compensating for coupling from coupled signal, have from the decoupled system of compensating for coupling as shown in Figure 2.
In fig. 2, due to decoupling zero branch road with there is in compensation feedback loop identical transport function, therefore decoupling signal can take out from feedback compensation ring, and system can simplify as shown in Figure 3.
In Fig. 3, positive feedback loop is the equalizing network from coupled signal, and relevant with the feature formula of object, decoupling signal is taken from the feedback element of equalizing network, and realizes signal decoupling zero in the control signal that is added to.In Fig. 3, signal decoupler only has two decoupling zero transport function D 12(s) and D 21s (), is made up of object model completely formula (4) these two decoupling zero transport functions known.As long as therefore determine the relation between input and output, just device D can be compensated 12(s), D 21s they are added in system the object that just can reach the decoupling zero exported with an input control one by (), namely a kind of optimization method only can not change the impact of other type of distortion on image for a kind of type of distortion.For two or more mixing distorted images, although system model is more complicated, and along with mixing the increase of distortion kind, need the distorted signal of sign also more, system model complexity is caused to increase, but the principle of Modling model is consistent with thought and the present invention, the suitable Decoupled Model that multiple distortion is coupled just can be set up according to principle of the present invention.
What finally illustrate is, above preferred embodiment is only in order to illustrate technical scheme of the present invention and unrestricted, although by above preferred embodiment to invention has been detailed description, but those skilled in the art are to be understood that, various change can be made to it in the form and details, and not depart from claims of the present invention limited range.

Claims (2)

1. the mixing distorted image optimization method based on uneoupled control, it is characterized in that: in the method, by adding Decoupled Model in systems in which, the coupling when being optimized mixing distorted image between each distortion optimizing method is eliminated, to promote effect of optimization, improve picture quality.
2. a kind of mixing distorted image optimization method based on uneoupled control according to claim 1, is characterized in that: the method specifically comprises the following steps:
1), judge that whether image fault type is known, if known, skip step 2) perform step 3), otherwise continue to perform step 2);
2), by analyzing the feature of different distortion, setting up a distortion disaggregated model, its mixing type of distortion need be obtained in image input model to be processed;
3), distorted signal is characterized by introducing distortion detection method;
4) relation of different constrained input, is found out by the mode of nonlinear fitting;
5), according to striked sign and relation set up system model and try to achieve Decoupled Model, realizing a kind of prioritization scheme only to a kind of distortion effectively and the object of change can not be caused to other distortions.
CN201510855824.1A 2015-11-30 2015-11-30 Hybrid distortion image optimization method based on decoupling control Pending CN105405111A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510855824.1A CN105405111A (en) 2015-11-30 2015-11-30 Hybrid distortion image optimization method based on decoupling control

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510855824.1A CN105405111A (en) 2015-11-30 2015-11-30 Hybrid distortion image optimization method based on decoupling control

Publications (1)

Publication Number Publication Date
CN105405111A true CN105405111A (en) 2016-03-16

Family

ID=55470578

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510855824.1A Pending CN105405111A (en) 2015-11-30 2015-11-30 Hybrid distortion image optimization method based on decoupling control

Country Status (1)

Country Link
CN (1) CN105405111A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106780446A (en) * 2016-12-01 2017-05-31 重庆邮电大学 It is a kind of to mix distorted image quality evaluating method without reference
CN107247965A (en) * 2017-05-31 2017-10-13 安徽四创电子股份有限公司 A kind of distorted image processing method and system based on Adaptive matching and study

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102788978A (en) * 2012-07-20 2012-11-21 电子科技大学 Squint spaceborne/airborne hybrid bistatic synthetic aperture radar imaging method
CN103064430A (en) * 2012-12-18 2013-04-24 湖南华南光电(集团)有限责任公司 Mechanical and electrical integration type image stabilization device

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102788978A (en) * 2012-07-20 2012-11-21 电子科技大学 Squint spaceborne/airborne hybrid bistatic synthetic aperture radar imaging method
CN103064430A (en) * 2012-12-18 2013-04-24 湖南华南光电(集团)有限责任公司 Mechanical and electrical integration type image stabilization device

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
吴鹏松 等: "基于信号叠加原理的信号解耦方法", 《化工自动化及仪表》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106780446A (en) * 2016-12-01 2017-05-31 重庆邮电大学 It is a kind of to mix distorted image quality evaluating method without reference
CN107247965A (en) * 2017-05-31 2017-10-13 安徽四创电子股份有限公司 A kind of distorted image processing method and system based on Adaptive matching and study

Similar Documents

Publication Publication Date Title
Li et al. Adaptive event-triggered output-feedback controller for uncertain nonlinear systems
KR101356598B1 (en) Method and apparatus for compensating for transceiver impairments
US8670501B2 (en) Digital pre-distortion of non-linear systems with reduced bandwidth feedback
CN101441764B (en) MTFC remote sensing image restoration method
US20140212132A1 (en) Adaptive equalizer tap coefficient correction method and optical receiver
US8855233B2 (en) Method and apparatus for implementing digital baseband predistortion
CN103942758A (en) Dark channel prior image dehazing method based on multiscale fusion
CN101425176A (en) Image wavelet de-noising method based on median filter
CN105405111A (en) Hybrid distortion image optimization method based on decoupling control
CN105593927A (en) Instability detection and correction in sinusoidal active noise reduction systems
CN102567955A (en) Method and system for inpainting images
US9525475B1 (en) Adaptive dual polarized MIMO for dynamically moving transmitter and receiver
CN102968765A (en) Method for correcting infrared focal plane heterogeneity based on sigma filter
CN101872471B (en) Image restoration method based on aerial TDI-CCD (Time Delay and Integration-Charge Coupled Device) imaging error vibration model
Farina et al. An output feedback distributed predictive control algorithm
Duan et al. Finite frequency filter design for nonlinear 2-D continuous systems in T–S form
US10739388B2 (en) Harmonic distortion separation method, nonlinear character determination method and apparatus and system
CN111934694B (en) Distortion compensation device of broadband zero intermediate frequency transceiving system
CN110971197A (en) Digital predistortion device and method
CN101039152B (en) Method and system for realizing electrical dispersion compensation
CN106910606A (en) A kind of transformer active noise denoising device and control method
US6633606B1 (en) Blind channel equalizer
CN106817336A (en) A kind of I/Q data calibration method and its device based on constant envelope signal
CN109756205B (en) Phasor calculation compensation method and system for median filtering flat-top distortion
CN105590456A (en) Geomagnetic threshold value adjusting method and geomagnetic threshold value adjusting system

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication

Application publication date: 20160316

RJ01 Rejection of invention patent application after publication