CN103489172A - Image fusion method for image chain - Google Patents
Image fusion method for image chain Download PDFInfo
- Publication number
- CN103489172A CN103489172A CN201310431594.7A CN201310431594A CN103489172A CN 103489172 A CN103489172 A CN 103489172A CN 201310431594 A CN201310431594 A CN 201310431594A CN 103489172 A CN103489172 A CN 103489172A
- Authority
- CN
- China
- Prior art keywords
- image
- individuality
- fusion method
- value
- fitness
- 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
Images
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The invention relates to an image fusion method for an image chain. In the process of image fusion, the fitness value of each individual in a species is calculated, the individuals are decoded to space to participate in the image fusion process, selection, intersection and mutation operators are utilized, the optimal individual is obtained under the control of a fitness evaluation function, a corresponding fusion image is generated, panchromatic information and color information do not need to be abandoned, and the high-quality image can be obtained.
Description
Technical field
The present invention relates to the computing machine process field, relate in particular to a kind of image interfusion method of image chain.
Background technology
The image chain, together with imaging sensor, refer to a kind of process with light path, comprises and utilize sensing data to carry out reconstructed image, and the color of matching image, noise reduction and sharpening etc., to form the picture of a panel height quality.The quality that strengthens the image of image chain is the study hotspot of this area always, at medical domain, usually need to be extracted by two width under different obtain manners or same obtain manner different parameters or the effective information of many good fortune image, in conjunction with advantage separately, be combined into the image that a width is new, and the quality that improves this class image normally realizes by image co-registration, so-called image co-registration is exactly by the full-colour image of high spatial resolution, low frequency spectrum resolution, and the multispectral image of high frequency spectrum resolution, low spatial resolution is merged.But in current all kinds of fusion methods, the two is contradiction, can not gets both, abandon a part of full-color information if need the more how general information of acquisition to mean, obtain more texture information and must abandon a part of color information, cause that image quality after fusion is not high, image is clear not, details is obvious not.
Summary of the invention
The present invention has overcome the deficiencies in the prior art, provides a kind of clear picture of image chain, details obvious, the image interfusion method that image quality is high.
For achieving the above object, the technical solution used in the present invention is: a kind of picture fusion method of image chain is characterized in that comprising step:
1), determine desired parameters according to present image, the selection coded system;
2), by the coding parametric solution of generation is mapped to space encoder, produce at random initial population;
3), enter the chromosome population iterative loop;
4), calculate solution numerical value corresponding to each chromosome of colony, this is separated to numerical value for image co-registration, and usings the objective evaluation criteria of result images as this chromosomal fitness value;
5), application choice, crossover and mutation operator, produce group of new generation and plant;
6), judge and stopping criterion if do not met, be back to step 3);
7), the output optimized individual, the chromosome coding of optimized individual is decoded, obtain optimum solution, usining this optimum solution is merged as parameter, generates fused images, calculates its objective evaluation functional value, as this chromosomal fitness function.
In a preferred embodiment of the present invention, further comprise in described step 1) and select the binary coding mode to be encoded to the desired parameters of present image.
In a preferred embodiment of the present invention, further comprise in described step 5) and to adopt while selecting operator, take each individual fitness value of population as selecting foundation, the probability that the individuality that fitness value is dominant participates in breeding of future generation is larger, the individuality that fitness value is less, participate in number of evolving of future generation less even superseded, obtain an intermediate solution more approaching with optimum solution.
In a preferred embodiment of the present invention, further comprise that in described step 5), two of random selections are carried out message exchange for breeding some same positions of follow-on individuality on this position, generate the new assortment of genes.
In a preferred embodiment of the present invention, further comprise that described step 5) is in mutation operator operates, negated in the variation position, produce new individuality.
In a preferred embodiment of the present invention, the functional value that further comprises fitness function in described step 7) is more than or equal to 0, utilize the functional value of fitness function measure each individuality in colony in evolutionary computation may close to or contribute to find the good degree of optimum solution.
The invention solves the defect existed in background technology, the image interfusion method of image chain provided by the invention is in the image co-registration process, calculate the group and plant each individual fitness value, individuality is decoded to space, participates in the image co-registration process, utilized selection, intersection, mutation operator, under controlling, fitness function tries to achieve optimized individual, generate a corresponding width fused images, without abandoning full-color information and color information, can obtain high-quality image.
The accompanying drawing explanation
Below in conjunction with drawings and Examples, the present invention is further described.
Fig. 1 is the process flow diagram of the image interfusion method of image chain provided by the invention.
Embodiment
The present invention is further detailed explanation in conjunction with the accompanying drawings and embodiments now, and these accompanying drawings are the schematic diagram of simplification, basic structure of the present invention only is described in a schematic way, so it only shows the formation relevant with the present invention.
As shown in Figure 1, a kind of picture fusion method of image chain, realize by following steps:
Step 1, according to present image, determine desired parameters, select coded system.Coding is the problem that at first will solve, and coding method shows and determined technique of expression and the gene meaning of chromosome in the chromosome space, and the coding/decoding method that is converted to common variable format, adopts binary coding method in the present invention, easy to operate fast.
Step 2, by the coding parametric solution of generation is mapped to space encoder, produce at random initial population.Colony's operation of image interfusion method of the present invention is started by a group initial population, therefore need to prepare an initial population formed by some initial solutions, also cry initial generation, each individuality in initial generation produces by random generation method, and optimum solution of the present invention will be evolved and obtain by these original hypothesis solutions.
Step 3, enter the chromosome population iterative loop.
Step 4, calculate solution numerical value corresponding to each chromosome of colony, this is separated to numerical value for image co-registration, and using the objective evaluation criteria of result images as this chromosomal fitness value.Because subjective evaluation method is not a kind of comprehensive method, there is certain one-sidedness, when observation condition changes, visual results may have certain difference, and therefore, the present invention adopts the objective evaluation function to be estimated, and avoids acceptor's viewing to ring.Objective evaluation be a kind of based on the image statistics value method of quantitative analyzing and processing image, to the input and output image before and after fusion treatment, some variation inevitably occurs in its statistic and quantity of information, so, adopt input and output image statistics value and quantity of information, the attribute of objective quantification image fully.
Step 5, application choice, crossover and mutation operator, produce group of new generation and plant.When adopting the selection operator, take each individual fitness value of population as selecting foundation, and the probability that the individuality that fitness value is dominant participates in breeding of future generation is larger, the individuality that fitness value is less, participate in number of evolving of future generation less even superseded, obtain an intermediate solution more approaching with optimum solution; Select at random two for breeding some same positions of follow-on individuality, crossover location carries out message exchange on this position, generates the new assortment of genes; In the mutation operator operation, negated in the variation position, produce new individuality, mutation operator can guarantee that algorithmic procedure can not produce the single-population that can't evolve, because when all individualities are the same, intersection is to produce new individuality, can only produce new individuality by variation, therefore, variation has increased the speciality of global optimization.
Step 6, judgement stopping criterion, if do not met optimum solution, be back to step 3.
Step 7, output optimized individual, decoded the chromosome coding of optimized individual, obtains optimum solution, usings this optimum solution to be merged as parameter, generates fused images, calculates its objective evaluation functional value, as this chromosomal fitness function.The functional value of fitness function is more than or equal to 0, utilize the functional value of fitness function measure each individuality in colony in evolutionary computation may close to or contribute to find the good degree of optimum solution.When the fitness of optimum individual reaches given threshold values, when perhaps the fitness of optimum individual and colony's fitness no longer rise, the iterative process of algorithm convergence, algorithm finishes, otherwise, with through the resulting colony of new generation of selecting, intersect, make a variation, replacing previous generation colony, and return to iteration and start place and select operation place to continue the circulation execution.
The present invention is in the image co-registration process, calculate the group and plant each individual fitness value, individuality is decoded to space, participate in the image co-registration process, utilize selection, intersection, mutation operator, under fitness function is controlled, tried to achieve optimized individual, generated a corresponding width fused images, without abandoning full-color information and color information, can obtain high-quality image.
Above foundation desirable embodiment of the present invention is enlightenment, and by above-mentioned description, the related personnel can, in the scope that does not depart from this invention technological thought, carry out various change and modification fully.The technical scope of this invention is not limited to the content on instructions, must determine technical scope according to the claim scope.
Claims (6)
1. the picture fusion method of an image chain is characterized in that comprising step:
1), determine desired parameters according to present image, the selection coded system;
2), by the coding parametric solution of generation is mapped to space encoder, produce at random initial population;
3), enter the chromosome population iterative loop;
4), calculate solution numerical value corresponding to each chromosome of colony, this is separated to numerical value for image co-registration, and usings the objective evaluation criteria of result images as this chromosomal fitness value;
5), application choice, crossover and mutation operator, produce group of new generation and plant;
6), judge and stopping criterion if do not met, be back to step 3);
7), the output optimized individual, the chromosome coding of optimized individual is decoded, obtain optimum solution, usining this optimum solution is merged as parameter, generates fused images, calculates its objective evaluation functional value, as this chromosomal fitness function.
2. the picture fusion method of a kind of image chain according to claim 1, is characterized in that: in described step 1), select the binary coding mode to be encoded to the desired parameters of present image.
3. the picture fusion method of a kind of image chain according to claim 1, it is characterized in that: in described step 5) when adopt selecting operator, take each individual fitness value of population as selecting foundation, the probability that the individuality that fitness value is dominant participates in breeding of future generation is larger, the individuality that fitness value is less, participate in number of evolving of future generation less even superseded, obtain an intermediate solution more approaching with optimum solution.
4. the picture fusion method of a kind of image chain according to claim 3, it is characterized in that: select at random two in described step 5) for breeding some same positions of follow-on individuality, carry out message exchange on this position, generate the new assortment of genes.
5. the picture fusion method of a kind of image chain according to claim 4 is characterized in that: described step 5), in the mutation operator operation, is negated to the variation position, produces new individuality.
6. the picture fusion method of a kind of image chain according to claim 1, it is characterized in that: in described step 7), the functional value of fitness function is more than or equal to 0, utilize the functional value of fitness function measure each individuality in colony in evolutionary computation may close to or contribute to find the good degree of optimum solution.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310431594.7A CN103489172A (en) | 2013-09-22 | 2013-09-22 | Image fusion method for image chain |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310431594.7A CN103489172A (en) | 2013-09-22 | 2013-09-22 | Image fusion method for image chain |
Publications (1)
Publication Number | Publication Date |
---|---|
CN103489172A true CN103489172A (en) | 2014-01-01 |
Family
ID=49829371
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310431594.7A Pending CN103489172A (en) | 2013-09-22 | 2013-09-22 | Image fusion method for image chain |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103489172A (en) |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040141659A1 (en) * | 2003-01-17 | 2004-07-22 | Yun Zhang | System and method for image fusion |
CN1619593A (en) * | 2004-12-09 | 2005-05-25 | 上海交通大学 | Video frequency motion target adaptive tracking method based on multicharacteristic information fusion |
-
2013
- 2013-09-22 CN CN201310431594.7A patent/CN103489172A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040141659A1 (en) * | 2003-01-17 | 2004-07-22 | Yun Zhang | System and method for image fusion |
CN1619593A (en) * | 2004-12-09 | 2005-05-25 | 上海交通大学 | Video frequency motion target adaptive tracking method based on multicharacteristic information fusion |
Non-Patent Citations (4)
Title |
---|
余旭初 等: "《高光谱影像分析与应用》", 1 May 2013, article "高光谱影像分析与应用", pages: 223-226 * |
张谦 等: "基于平台/插件软件架构的多源遥感影像融合系统设计", 《遥感技术与应用》, vol. 25, no. 3, 30 June 2010 (2010-06-30), pages 394 - 398 * |
彭开: "基于遗传算法的遥感图像融合方法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》, no. 1, 15 January 2011 (2011-01-15), pages 140 - 675 * |
李婷 等: "基于EI优化权值的多聚焦图像融合方法", 《计算机工程与应用》, vol. 44, no. 15, 31 December 2008 (2008-12-31), pages 192 - 194 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Djurisic et al. | Extremely simple activation shaping for out-of-distribution detection | |
CN107679502B (en) | A kind of Population size estimation method based on the segmentation of deep learning image, semantic | |
Seo et al. | A novel just-noticeable-difference-based saliency-channel attention residual network for full-reference image quality predictions | |
CN104517110B (en) | The binarization method and system of a kind of image in 2 D code | |
Hole et al. | Application of genetic algorithm for image enhancement and segmentation | |
CN104243992B (en) | Image processing apparatus and image processing method | |
CN105913408A (en) | Multi-focusing-image fusion method based on bilateral filter | |
CN113111716B (en) | Remote sensing image semiautomatic labeling method and device based on deep learning | |
CN113705490B (en) | Anomaly detection method based on reconstruction and prediction | |
KR20220023323A (en) | Automatic multi-organ and tumor contouring system based on artificial intelligence for radiation treatment planning | |
Babu et al. | Contrast enhancement using real coded genetic algorithm based modified histogram equalization for gray scale images | |
CN111652822A (en) | Single image shadow removing method and system based on generation countermeasure network | |
CN116630464A (en) | Image style migration method and device based on stable diffusion | |
CN116664605A (en) | Medical image tumor segmentation method based on diffusion model and multi-mode fusion | |
CN115457359A (en) | PET-MRI image fusion method based on adaptive countermeasure generation network | |
Rippel et al. | GAN-based defect synthesis for anomaly detection in fabrics | |
CN108665436A (en) | A kind of multi-focus image fusing method and system based on gray average reference | |
CN112785540B (en) | Diffusion weighted image generation system and method | |
CN109741315B (en) | Non-reference image quality objective evaluation method based on deep reinforcement learning | |
CN105913451B (en) | A kind of natural image superpixel segmentation method based on graph model | |
Lin et al. | Quality quantification in deep convolutional neural networks for skin lesion segmentation using fuzzy uncertainty measurement | |
Cui et al. | Film effect optimization by deep learning and virtual reality technology in new media environment | |
CN103489172A (en) | Image fusion method for image chain | |
CN106303545A (en) | Complement vector mutually | |
CN110838120A (en) | Weighting quality evaluation method of asymmetric distortion three-dimensional video based on space-time information |
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: 20140101 |
|
RJ01 | Rejection of invention patent application after publication |