CN105809650A - Bidirectional iteration optimization based image integrating method - Google Patents
Bidirectional iteration optimization based image integrating method Download PDFInfo
- Publication number
- CN105809650A CN105809650A CN201610122602.3A CN201610122602A CN105809650A CN 105809650 A CN105809650 A CN 105809650A CN 201610122602 A CN201610122602 A CN 201610122602A CN 105809650 A CN105809650 A CN 105809650A
- Authority
- CN
- China
- Prior art keywords
- image
- iteration
- matrix
- source images
- fusion
- 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
- 238000005457 optimization Methods 0.000 title claims abstract description 49
- 238000000034 method Methods 0.000 title claims abstract description 47
- 230000002457 bidirectional effect Effects 0.000 title claims abstract description 26
- 230000000007 visual effect Effects 0.000 claims abstract description 8
- 238000012545 processing Methods 0.000 claims abstract description 6
- 239000011159 matrix material Substances 0.000 claims description 95
- 230000004927 fusion Effects 0.000 claims description 65
- 230000000694 effects Effects 0.000 claims description 12
- 238000002591 computed tomography Methods 0.000 claims description 7
- 238000004458 analytical method Methods 0.000 claims description 6
- 238000013461 design Methods 0.000 claims description 4
- 230000005540 biological transmission Effects 0.000 claims description 3
- 230000015572 biosynthetic process Effects 0.000 claims description 3
- 238000007499 fusion processing Methods 0.000 claims description 3
- 238000003384 imaging method Methods 0.000 claims description 3
- 238000004519 manufacturing process Methods 0.000 claims description 3
- 238000001228 spectrum Methods 0.000 claims description 3
- 238000012804 iterative process Methods 0.000 claims description 2
- 230000010354 integration Effects 0.000 abstract 4
- 230000006870 function Effects 0.000 description 12
- 230000009466 transformation Effects 0.000 description 6
- 238000011160 research Methods 0.000 description 4
- 238000000354 decomposition reaction Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000007500 overflow downdraw method Methods 0.000 description 3
- 230000007812 deficiency Effects 0.000 description 2
- 230000002401 inhibitory effect Effects 0.000 description 2
- 238000002156 mixing Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000016776 visual perception Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
-
- 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
- G06T2207/10048—Infrared image
-
- 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
- G06T2207/10072—Tomographic images
- G06T2207/10081—Computed x-ray tomography [CT]
-
- 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
- G06T2207/10072—Tomographic images
- G06T2207/10088—Magnetic resonance imaging [MRI]
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The invention provides a bidirectional iteration optimization based image integrating method, comprising the following steps: firstly, inputting already captured visible light and infrared images, computed tomoscaning medical science images and magnetic resonance medical science images into a computer; secondly, processing the visible light and infrared images, the computed tomoscaning medical science images and the magnetic resonance medical science images to calculate correlation matrixes and distance matrixes among the original images and to further set iteration step lengths; thirdly, seeking constrained conditions for iteration based on the image integration determining standards; obtaining an optimized objective function through the variables from the second step and the third step as well as the constrained conditions to realize images after the integration process; With the adoption of the above steps, the bidirectional iteration optimization based image integrating method is capable of achieving good visual effects on the integration of both visible light and infrared images and medical images. Further, with the method, more image information can be maintained in the process of integration; practical problems are properly handled; therefore, the method promises wide use in the market.
Description
Technical field
The present invention provides a kind of image interfusion method optimized based on bidirectional iteration, particularly to the digital image processing techniques of visible ray with infrared image, the fusion of medical image, belongs to digital image processing techniques field.
Background technology
Along with the development of sensor technology, single visible mode develops into multiple sensors pattern gradually.Single imageing sensor tends not to enough enough information of extracting from scene, to such an extent as to is difficult to even cannot independently obtain the comprehensive description to a width scene.This is accomplished by research multi-source image and merges.Image co-registration can work in coordination with the multiple sensors image information utilizing Same Scene, exports a width and is more suitable for human visual perception or computer processes and the fusion image analyzed further.It can significantly improve the deficiency of single-sensor, improves definition and the information amount of comprising of result images, be conducive to more accurately, more reliably, more fully obtain the information of target or scene.Therefore, the research of image fusion technology is an important topic having important theory and using value.
For image co-registration problem, Chinese scholars conducts extensive research, it is proposed that a series of more effective methods, for instance Weighted Fusion algorithm, pyramid blending algorithm, Wavelet Transform Fusion algorithm etc..
It is that the pixel value to two original images takes certain weights that Weighted Fusion calculates ratio juris, is then weighted obtaining the pixel value of fusion image.But, the shortcoming that this algorithm exists the edge of fused image and profile thickens, when image intensity value difference is bigger, it is easy to occur comparatively significantly splicing vestige.
First each image to be fused is done pyramid decomposition by pyramid blending algorithm, selects afterwards to constitute fusion pyramid, finally by the image that inverse transformation can generate after being merged from image pyramid decomposition coefficient to be fused.But, owing to pyramid fusion method merges on different decomposition layer, inevitably introduce bigger noise, cause that the effect of fusion image is bad, and after pyramid transform method is decomposed, each layer data has redundancy, adds amount of calculation.
Wavelet Transform Fusion algorithm is algorithm that is comparatively classical and that be used widely.Wavelet transformation essence is a kind of high-pass filtering, adopts different wavelet basiss will produce different filter effects.Original image can be resolved into a series of subimage with different spatial resolutions and frequency domain characteristic by wavelet transformation, and the wavelet coefficient for different frequency bands subimage is combined, and forms the wavelet coefficient of fusion image.But, choosing of domain transformation window size is had very big restriction by Wavelet Transform Fusion algorithm in image co-registration, which results in the scope of application of a kind of Wavelet Transform Fusion algorithm relatively small.
All in all, all there is some problems in traditional Image Fusion, therefore a kind of Digital Image Fusion method effective, general, quick of research has important practical significance.On the basis of these thoughts, adopt bidirectional iteration optimization that image is merged herein, and describe the ultimate principle of this algorithm.This algorithm goes for visible ray and the several scenes such as infrared image, medical image.
Summary of the invention
(1) purpose of the present invention
It is an object of the invention to provide a kind of image interfusion method optimized based on bidirectional iteration, the method is the improvement to traditional images fusion method, by finding a suitable object function and optimizing constraint, design corresponding inverse problem Restoration model, realize merging, overcome the deficiency that traditional method exists, not only remain in source images total CONSTRUCTED SPECIFICATION, and characteristic information exclusive in comprehensive source images well.
(2) technical scheme
A kind of image interfusion method optimized based on bidirectional iteration of the present invention, it includes step in detail below:
Step one, the visible ray and the infrared image that obtain, CT Scan medical image (i.e. CT medical image) and magnetic resonance medical image (i.e. MRI medical image) are input in computer；
Visible ray is obtained by digital imaging apparatus with MRI medical image with infrared image, CT medical image, is read into respectively in computer by generation source images, and these data messages are by the basis of image co-registration；
Step 2, process visible ray and infrared image, CT medical image and MRI medical image calculate the correlation matrix and distance matrix that obtain between source images, and set iteration step length；
It is respectively calculated for visible ray in step one and infrared image, CT medical image and MRI medical image (hereafter we be referred to as source images image S and image T), is divided into following step to realize:
(1) correlation matrix that visible ray and infrared image, CT medical image and MRI medical image are respectively calculated between source images is utilized.Correlation matrix is to be made up of the correlation coefficient between matrix respectively arranges, and the element of correlation matrix the i-th row jth row is the correlation coefficient that source images matrix i-th arranges and jth arranges.It is to say, correlation matrix takes the normalized covariance matrix of source images, each element of covariance matrix is corresponding to the covariance of column vector any in two images, and correlation matrix cov (S, T) is such as following formula:
Wherein cov (S, T) is the correlation matrix between image S and image T,For column vector SiWith TjBetween covariance, SiWith TjFor the column vector of source images S and T, n is the dimension of source images,WithFor on average row amount,Maximum for covariance；
Based on correlation matrix cov (S, T) and the cov (T, S) covariance between the pixel value of two source images matrixes, represent the dependency between two source images respectively, illustrate the tolerance of the concrete image detail of fusion image；
(2) distance matrix that visible ray and infrared image, CT medical image and MRI medical image are respectively calculated between source images is utilized；Distance matrix is the matrix of a pixel value distance between any two comprising correspondence position pixel, and in this article, we go out the foundation of fusion image using distance matrix as two original image iteration, and distance matrix form is as follows:
Wherein, Δ (S, T) and Δ (T, S) represent the difference of the pixel value between picture element matrix S and picture element matrix T respectively；
Effect with correlation matrix is different, and distance matrix here represents the difference of the pixel value between image, it is determined that the fusion direction of image co-registration, essentially dictates the fusion of image overall profile, serves critical effect；The present invention has selected correlation matrix and distance matrix, and employs bidirectional iteration optimization, is the syncretizing effect in order to reach optimum on contour direction and concrete image detail simultaneously；
Iteration step length is set, the impact that the size of iteration step length can be certain on the entropy production of fusion image；
Step 3, foundation image co-registration criterion obtain iterative constrained condition
By the analysis to image co-registration criterion, the constraints obtaining iteration is as follows:
min||F(S)-F(T)||(3)
Wherein, the image that F (S) and F (T) produces for iteration；
Step 4, utilize the variable and constraints that obtain in step 2 and step 3, obtain optimization object function, adopt the thinking that iteration updates to merge source images, it is achieved image co-registration process, the image after being merged；
(1) design optimization object function, adopts the thinking that iteration updates, and using correlation matrix and distance matrix as the foundation updating iteration variable, utilizes iteration optimization model to carry out alternative optimization computing.Adopt fusion rules standard as constraints, based on constraints, devise inverse problem Restoration model, seek the fusion image of optimum.For source images S and source images T, optimization object function is following formula such as:
Wherein, F (S) and the F (T) image for producing in iterative process, m and n is iteration step length, the difference that Δ (S, T) and Δ (T, S) are image pixel value, cov (S, T) referring to that source images S points to the correlation matrix of T, cov (T, S) refers to that source images T points to the correlation matrix of S.
(2) image co-registration process is realized according to optimization object function, the image after being merged.
Wherein, " visible images " described in step one, it refers to by the image of the Spectrum Formation of human readable." infrared image " described in step one, it refers to that infrared remote sensor receives the infrared ray of clutter reflections or its own transmission and the image that formed；" CT medical image " described in step one, it refers to CT Scan medical image；" MRI medical image " described in step one, it refers to magnetic resonance medical image.
Wherein, " by the analysis to image co-registration criterion, obtain the constraints of iteration " described in step 3, it is as follows that it follows principle:
Fusion image should comprise all of useful information in each source images；Fusion image and source images should not introduce other artificial information, so we need the similar of picture structure is made constraint；Method needs have real-time.
Wherein, " the realizing image co-registration process according to optimization object function, the image after being merged " described in step 4, its practice is as follows:
Optimization aim equation adopts the thinking that iteration updates, and using correlation matrix and distance matrix as the foundation updating iteration variable, utilizes iteration optimization model to carry out alternative optimization computing.First, by processing source images with optimization aim equation, obtain the renewal of two iteration variable F (S) and F (T) simultaneously.Secondly, utilizing iteration variable to carry out fusion image, image co-registration correlation matrix and distance matrix determine fusion direction and tolerance degree, illustrate the degree of accuracy in fusion process with iteration step length.Finally, when adjudicating iterative constrained condition and being optimum, finally constantly iteration goes out optimal solution.
Pass through above step, the bidirectional iteration optimization in present invention fusion results on visible ray with infrared image, medical image can obtain good visual effect, in image co-registration, it is possible to make target object feature in fusion image become apparent from, edge is apparent, merges the used time less.Compared with traditional Laplacian algorithm, average weighted algorithm, Haar Wavelet Transformation Algorithm, the entropy of the image that bidirectional iteration optimization merges is improved, and can retain more image information.The present invention can be used for the fusion of image, fusion results on visible ray with infrared image, medical image all achieves good visual effect, for the fusion important in inhibiting of visible ray with infrared image, medical image, there is wide market prospect and actual application value.
(3) beneficial effect
The present invention adopts the method based on bidirectional iteration optimization to carry out image co-registration, make full use of multiple correlation coefficient between bidirectional iteration optimization and source images and carry out the fusion of image, the accuracy of fusion image, reliability, information content are greatly improved by the deep redundancy utilizing multiple sensor to provide fully and complementary information, fusion image has higher visual quality, feature becomes apparent from, and edge is apparent；By the optimization of correlation matrix so that fusion image has had the tolerance of concrete image detail；Illustrated the difference of pixel value between image by distance matrix, it is determined that the fusion direction of image, mainly image overall profile is served critical effect；Adopting relatively simple computing formula and algorithm flow, time complexity is less, performs speed, has higher efficiency, it is possible to requirement of real time.Algorithm fusion results on visible ray with infrared image, medical image all achieves good visual effect and good objective quality performance evaluation, has wide using value and market prospect.
Accompanying drawing explanation
Fig. 1: the method for the invention flow chart.
Detailed description of the invention
In order to be more fully understood that technical scheme, the present invention is discussed in detail below in conjunction with the drawings and the specific embodiments.
A kind of image interfusion method optimized based on bidirectional iteration of the present invention, as it is shown in figure 1, the method mainly includes following step:
1. the visible ray obtained is input in computer with magnetic resonance medical image (i.e. MRI medical image) with infrared image, CT Scan medical image (i.e. CT medical image).
2. process visible ray and infrared image, CT medical image and MRI medical image calculate the correlation matrix and distance matrix that obtain between source images, and set iteration step length.
3. obtain iterative constrained condition according to image co-registration criterion.
4. utilize the variable and constraints that obtain in step 2 and step 3, obtain optimization object function, adopt the thinking that iteration updates to merge source images, it is achieved image co-registration process, the image after being merged.
The present invention implement flow process as it is shown in figure 1, each several part to be embodied as details as follows:
1. the visible ray obtained is input in computer with magnetic resonance medical image (i.e. MRI medical image) with infrared image, CT Scan medical image (i.e. CT medical image).
Visible ray is obtained by digital imaging apparatus with MRI medical image with infrared image, CT medical image, is read into respectively in computer by generation source images, and these data messages are by the basis of image co-registration.
2. process visible ray and infrared image, CT medical image and MRI medical image calculate the correlation matrix and distance matrix that obtain between source images, and set iteration step length.
For visible ray in step 1 and infrared image, CT medical image and MRI medical image (hereafter source images is referred to as image S and image T and is respectively calculated by us, is divided into following step to realize:
(1) correlation matrix that visible ray and infrared image, CT medical image and MRI medical image are respectively calculated between source images is utilized.Correlation matrix is to be made up of the correlation coefficient between matrix respectively arranges, and the element of correlation matrix the i-th row jth row is the correlation coefficient that source images matrix i-th arranges and jth arranges.It is to say, correlation matrix takes the normalized covariance matrix of source images, each element of covariance matrix is corresponding to the covariance of column vector any in two images, correlation matrix cov (S, T) such as formula (1).
Based on correlation matrix cov (S, T) and the cov (T, S) covariance between the pixel value of two source images matrixes, represent the dependency between two source images respectively, illustrate the tolerance of the concrete image detail of fusion image.
(2) distance matrix that visible ray and infrared image, CT medical image and MRI medical image are respectively calculated between source images is utilized.Distance matrix is the matrix of a pixel value distance between any two comprising correspondence position pixel.In this article, we go out the foundation of fusion image using distance matrix as two original image iteration.Distance matrix form such as formula (2).
Effect with correlation matrix is different, and distance matrix here represents the difference of the pixel value between image, it is determined that the fusion direction of image co-registration, essentially dictates the fusion of image overall profile, serves critical effect.Select correlation matrix and distance matrix herein, and employed bidirectional iteration optimization, be the syncretizing effect in order to reach optimum on contour direction and concrete image detail simultaneously.
(3) iteration step length is set, the impact that the size of iteration step length can be certain on the entropy production of fusion image.
3. obtain iterative constrained condition according to image co-registration criterion.
By the analysis to image co-registration criterion, the constraints of iteration such as formula (3).
4. utilize the variable and constraints that obtain in step 2 and step 3, obtain optimization object function, adopt the thinking that iteration updates to merge source images, it is achieved image co-registration process, the image after being merged
(1) design optimization object function, adopts the thinking that iteration updates, and using correlation matrix and distance matrix as the foundation updating iteration variable, utilizes iteration optimization model to carry out alternative optimization computing.Adopt fusion rules standard as constraints, based on constraints, devise inverse problem Restoration model, seek the fusion image of optimum.For source images S and source images T, optimization object function such as formula (4).
(2) image co-registration process is realized according to optimization object function, the image after being merged.
Wherein, " visible images " described in step 1, it refers to by the image of the Spectrum Formation of human readable." infrared image " described in step 1, it refers to that infrared remote sensor receives the infrared ray of clutter reflections or its own transmission and the image that formed；" CT medical image " described in step 1, it refers to CT Scan medical image；" MRI medical image " described in step 1, it refers to magnetic resonance medical image.
Wherein, " by the analysis to image co-registration criterion, obtain the constraints of iteration " described in 3, it is as follows that it follows principle:
Fusion image should comprise all of useful information in each source images；Fusion image and source images should not introduce other artificial information, so we need the similar of picture structure is made constraint；Method needs have real-time.
Wherein, " the realizing image co-registration process according to optimization object function, the image after being merged " described in 4, its practice is as follows:
Optimization aim equation adopts the thinking that iteration updates, and using correlation matrix and distance matrix as the foundation updating iteration variable, utilizes iteration optimization model to carry out alternative optimization computing.First, by processing source images with optimization aim equation, obtain the renewal of two iteration variable F (S) and F (T) simultaneously.Secondly, utilizing iteration variable to carry out fusion image, image co-registration correlation matrix and distance matrix determine fusion direction and tolerance degree, illustrate the degree of accuracy in fusion process with iteration step length.Finally, when adjudicating iterative constrained condition and being optimum, finally constantly iteration goes out optimal solution.
Pass through above step, the bidirectional iteration optimization in present invention fusion results on visible ray with infrared image, medical image can obtain good visual effect, in image co-registration, it is possible to make target object feature in fusion image become apparent from, edge is apparent, merges the used time less.Compared with traditional Laplacian algorithm, average weighted algorithm, Haar Wavelet Transformation Algorithm, the entropy of the image that bidirectional iteration optimization merges is improved, and can retain more image information.The present invention can be used for the fusion of image, fusion results on visible ray with infrared image, medical image all achieves good visual effect, for the fusion important in inhibiting of visible ray with infrared image, medical image, there is wide market prospect and actual application value.
Claims (7)
1. the image interfusion method optimized based on bidirectional iteration, it is characterised in that: it includes step in detail below:
Step one, the visible ray of acquisition and infrared image, CT Scan medical image and CT medical image are input in computer with magnetic resonance medical image and MRI medical image；
Visible ray is obtained by digital imaging apparatus with MRI medical image with infrared image, CT medical image, is read into respectively in computer by generation source images, and these data messages are by the basis of image co-registration；
Step 2, process visible ray and infrared image, CT medical image and MRI medical image calculate the correlation matrix and distance matrix that obtain between source images, and set iteration step length；
It is respectively calculated with MRI medical image with infrared image, CT medical image for visible ray in step one, is divided into following step to realize: source images is referred to as image S and image T by us
(1) correlation matrix that visible ray and infrared image, CT medical image and MRI medical image are respectively calculated between source images is utilized, correlation matrix is to be made up of the correlation coefficient between matrix respectively arranges, and the element of correlation matrix the i-th row jth row is the correlation coefficient that source images matrix i-th arranges and jth arranges；It is to say, correlation matrix takes the normalized covariance matrix of source images, each element of covariance matrix is corresponding to the covariance of column vector any in two images, and correlation matrix cov (S, T) is such as following formula:
Wherein cov (S, T) is the correlation matrix between image S and image T,For column vector SiWith TjBetween covariance, SiWith TjFor the column vector of source images S and T, n is the dimension of source images,WithFor on average row amount,Maximum for covariance；
Based on correlation matrix cov (S, T) and the cov (T, S) covariance between the pixel value of two source images matrixes, represent the dependency between two source images respectively, illustrate the tolerance of the concrete image detail of fusion image；
(2) distance matrix that visible ray and infrared image, CT medical image and MRI medical image are respectively calculated between source images is utilized；Distance matrix is the matrix of a pixel value distance between any two comprising correspondence position pixel, and we go out the foundation of fusion image using distance matrix as two original image iteration, and distance matrix form is as follows:
Wherein, Δ (S, T) and Δ (T, S) represent the difference of the pixel value between picture element matrix S and picture element matrix T respectively；
Effect with correlation matrix is different, and distance matrix here represents the difference of the pixel value between image, it is determined that the fusion direction of image co-registration, essentially dictates the fusion of image overall profile, serves critical effect；The present invention has selected correlation matrix and distance matrix, and employs bidirectional iteration optimization, is the syncretizing effect in order to reach optimum on contour direction and concrete image detail simultaneously；
Iteration step length is set, the impact that the size of iteration step length can be certain on the entropy production of fusion image；
Step 3, foundation image co-registration criterion obtain iterative constrained condition
By the analysis to image co-registration criterion, the constraints obtaining iteration is as follows:
min||F(S)-F(T)||(3)
Wherein, the image that F (S) and F (T) produces for iteration；
Step 4, utilize the variable and constraints that obtain in step 2 and step 3, obtain optimization object function, adopt the thinking that iteration updates to merge source images, it is achieved image co-registration process, the image after being merged；
(1) design optimization object function, adopts the thinking that iteration updates, and using correlation matrix and distance matrix as the foundation updating iteration variable, utilizes iteration optimization model to carry out alternative optimization computing；Adopt fusion rules standard as constraints, based on constraints, devise inverse problem Restoration model, seek the fusion image of optimum；For source images S and source images T, optimization object function is following formula such as:
Wherein, F (S) and the F (T) image for producing in iterative process, m and n is iteration step length, the difference that Δ (S, T) and Δ (T, S) are image pixel value, cov (S, T) referring to that source images S points to the correlation matrix of T, cov (T, S) refers to that source images T points to the correlation matrix of S；
(2) image co-registration process is realized according to optimization object function, the image after being merged
Pass through above step, the bidirectional iteration optimization in present invention fusion results on visible ray with infrared image, medical image all obtains good visual effect, target object feature in fusion image is made to become apparent from, edge is apparent, merge the used time less, remain more image information, adequate solution actual application problem.
2. a kind of image interfusion method optimized based on bidirectional iteration according to claim 1, it is characterised in that: " visible images " described in step one, it refers to by the image of the Spectrum Formation of human readable.
3. a kind of image interfusion method optimized based on bidirectional iteration according to claim 1, it is characterised in that: " infrared image " described in step one, it refers to that infrared remote sensor receives the infrared ray of clutter reflections or its own transmission and the image that formed.
4. a kind of image interfusion method optimized based on bidirectional iteration according to claim 1, it is characterised in that: " the CT medical image " described in step one, it refers to CT Scan medical image.
5. a kind of image interfusion method optimized based on bidirectional iteration according to claim 1, it is characterised in that: " the MRI medical image " described in step one, it refers to magnetic resonance medical image.
6. a kind of image interfusion method optimized based on bidirectional iteration according to claim 1, it is characterised in that: " by the analysis to image co-registration criterion, obtain the constraints of iteration " described in step 3, it is as follows that it follows principle:
Fusion image should comprise all of useful information in each source images；Fusion image and source images should not introduce other artificial information, so we need the similar of picture structure is made constraint；Method needs have real-time.
7. a kind of image interfusion method optimized based on bidirectional iteration according to claim 1, it is characterised in that: " the realizing image co-registration process according to optimization object function, the image after being merged " described in step 4, its practice is as follows:
Optimization aim equation adopts the thinking that iteration updates, and using correlation matrix and distance matrix as the foundation updating iteration variable, utilizes iteration optimization model to carry out alternative optimization computing；First, by processing source images with optimization aim equation, obtain the renewal of two iteration variable F (S) and F (T) simultaneously；Secondly, utilizing iteration variable to carry out fusion image, image co-registration correlation matrix and distance matrix determine fusion direction and tolerance degree, illustrate the degree of accuracy in fusion process with iteration step length；Finally, when adjudicating iterative constrained condition and being optimum, finally constantly iteration goes out optimal solution.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610122602.3A CN105809650B (en) | 2016-03-04 | 2016-03-04 | A kind of image interfusion method based on bidirectional iteration optimization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610122602.3A CN105809650B (en) | 2016-03-04 | 2016-03-04 | A kind of image interfusion method based on bidirectional iteration optimization |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105809650A true CN105809650A (en) | 2016-07-27 |
| CN105809650B CN105809650B (en) | 2018-06-29 |
Family
ID=56466678
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610122602.3A Active CN105809650B (en) | 2016-03-04 | 2016-03-04 | A kind of image interfusion method based on bidirectional iteration optimization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105809650B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107784606A (en) * | 2016-08-30 | 2018-03-09 | 源渠（上海）信息技术有限公司 | A kind of training organization's commending system and method |
| CN109767410A (en) * | 2018-12-22 | 2019-05-17 | 复旦大学 | A kind of lung CT and MRI image blending algorithm |
| CN109883550A (en) * | 2019-03-14 | 2019-06-14 | 安徽慧视金瞳科技有限公司 | A kind of update method and equipment of spatial light spectrometer instruments scanning function |
| CN111353967A (en) * | 2020-03-06 | 2020-06-30 | 浙江杜比医疗科技有限公司 | Image acquisition method and device, electronic equipment and readable storage medium |
| CN116363252A (en) * | 2023-06-02 | 2023-06-30 | 南京诺源医疗器械有限公司 | Target imaging method and system |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102789638B (en) * | 2012-07-16 | 2016-01-06 | 北京市遥感信息研究所 | Based on the image interfusion method of gradient fields and Scale-space theory |
| CN104021537A (en) * | 2014-06-23 | 2014-09-03 | 西北工业大学 | Infrared and visible image fusion method based on sparse representation |
-
2016
- 2016-03-04 CN CN201610122602.3A patent/CN105809650B/en active Active
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107784606A (en) * | 2016-08-30 | 2018-03-09 | 源渠（上海）信息技术有限公司 | A kind of training organization's commending system and method |
| CN109767410A (en) * | 2018-12-22 | 2019-05-17 | 复旦大学 | A kind of lung CT and MRI image blending algorithm |
| CN109767410B (en) * | 2018-12-22 | 2024-03-08 | 复旦大学 | Lung CT and MRI image fusion algorithm |
| CN109883550A (en) * | 2019-03-14 | 2019-06-14 | 安徽慧视金瞳科技有限公司 | A kind of update method and equipment of spatial light spectrometer instruments scanning function |
| CN111353967A (en) * | 2020-03-06 | 2020-06-30 | 浙江杜比医疗科技有限公司 | Image acquisition method and device, electronic equipment and readable storage medium |
| CN116363252A (en) * | 2023-06-02 | 2023-06-30 | 南京诺源医疗器械有限公司 | Target imaging method and system |
| CN116363252B (en) * | 2023-06-02 | 2023-08-04 | 南京诺源医疗器械有限公司 | Target imaging method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105809650B (en) | 2018-06-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Jin et al. | A survey of infrared and visual image fusion methods | |
| CN104809734B (en) | A method of the infrared image based on guiding filtering and visual image fusion | |
| CN105809650A (en) | Bidirectional iteration optimization based image integrating method | |
| CN107341786A (en) | The infrared and visible light image fusion method that wavelet transformation represents with joint sparse | |
| CN109035188A (en) | A kind of intelligent image fusion method based on target signature driving | |
| CN103020933B (en) | A kind of multisource image anastomosing method based on bionic visual mechanism | |
| CN104408700A (en) | Morphology and PCA (principal component analysis) based contourlet fusion method for infrared and visible light images | |
| CN106327459A (en) | Visible light and infrared image fusion algorithm based on UDCT (Uniform Discrete Curvelet Transform) and PCNN (Pulse Coupled Neural Network) | |
| CN105761214A (en) | Remote sensing image fusion method based on contourlet transform and guided filter | |
| CN108230260A (en) | A kind of fusion method of new infrared image and twilight image | |
| CN105809640A (en) | Multi-sensor fusion low-illumination video image enhancement method | |
| CN103971329A (en) | Cellular nerve network with genetic algorithm (GACNN)-based multisource image fusion method | |
| CN110097617B (en) | Image fusion method based on convolutional neural network and significance weight | |
| CN104008536A (en) | Multi-focus noise image fusion method based on CS-CHMT and IDPCNN | |
| CN104766290B (en) | A kind of Pixel Information estimation fusion method based on quick NSCT | |
| Liu et al. | An effective wavelet-based scheme for multi-focus image fusion | |
| Li et al. | Infrared and visible fusion imaging via double-layer fusion denoising neural network | |
| Jian et al. | A multi-source image fusion algorithm based on gradient regularized convolution sparse representation | |
| Patel et al. | A review on infrared and visible image fusion techniques | |
| CN103985104A (en) | Multi-focusing image fusion method based on higher-order singular value decomposition and fuzzy inference | |
| CN103198456B (en) | Remote sensing image fusion method based on directionlet domain hidden Markov tree (HMT) model | |
| CN103400360A (en) | Multi-source image fusing method based on Wedgelet and NSCT (Non Subsampled Contourlet Transform) | |
| Luo et al. | Infrared and visible image fusion based on VPDE model and VGG network | |
| Li et al. | Multi-sensor image fusion by NSCT-PCNN transform | |
| CN116051444A (en) | Effective infrared and visible light image self-adaptive fusion method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |