CN107833182A - The infrared image super resolution ratio reconstruction method of feature based extraction - Google Patents
The infrared image super resolution ratio reconstruction method of feature based extraction Download PDFInfo
- Publication number
- CN107833182A CN107833182A CN201711158675.9A CN201711158675A CN107833182A CN 107833182 A CN107833182 A CN 107833182A CN 201711158675 A CN201711158675 A CN 201711158675A CN 107833182 A CN107833182 A CN 107833182A
- Authority
- CN
- China
- Prior art keywords
- image
- infrared image
- resolution ratio
- characteristic pattern
- feature
- 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
- 238000000034 method Methods 0.000 title claims abstract description 63
- 238000000605 extraction Methods 0.000 title claims abstract description 33
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 10
- 230000000877 morphologic effect Effects 0.000 claims abstract description 4
- 239000011159 matrix material Substances 0.000 claims description 7
- 230000004927 fusion Effects 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 230000003628 erosive effect Effects 0.000 claims description 5
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims 1
- 238000005728 strengthening Methods 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 7
- 238000005516 engineering process Methods 0.000 description 4
- 238000003384 imaging method Methods 0.000 description 4
- 238000001914 filtration Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000006740 morphological transformation Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000003321 amplification Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000003331 infrared imaging Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000000155 melt Substances 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004660 morphological change Effects 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 239000000047 product Substances 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 230000000007 visual effect Effects 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
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4053—Scaling of whole images or parts thereof, e.g. expanding or contracting based on super-resolution, i.e. the output image resolution being higher than the sensor resolution
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
-
- 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
- G06T2207/20036—Morphological image processing
-
- 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
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The present invention discloses a kind of infrared image super resolution ratio reconstruction method of feature based extraction, including:The first step, nonsingular decomposition is carried out to the original infrared image of collection, filters out the noise jamming in original infrared image;Second step, image after denoising is used feature extraction is carried out based on Gabor filter, and the feature after extraction is further enhanced with based on morphologic method;3rd step, enhanced characteristic pattern is subjected to Its Sparse Decomposition, sparse coefficient is solved using method of Lagrange multipliers, and coefficient reconstruction required by use goes out the super-resolution image of original infrared image.The present invention can effectively avoid picture noise when carrying out super-resolution rebuilding to infrared image to rebuilding the influence of effect, meanwhile, because the characteristic pattern that is extracted has preferably openness, also further improve the quality of reconstruction image.
Description
Technical field
The invention belongs to technical field of image processing, more particularly to a kind of super-resolution rebuilding side for infrared image
Method.
Background technology
Infrared imagery technique because its with perceive object radiation infrared ray degree so that be imaged the characteristics of, can effectively overcome because
Misty rain, block, the drawbacks of objective factor such as illumination deficiency causes that target can not be observed, can carry out having for round-the-clock 24 hours
Effect monitoring, therefore be used widely.
In imaging process, image can because of optical lens torsional deformation, atmosphere fuzzy, sensor be fuzzy, optical dimming and
A series of influence of disturbing factors such as motion blur and noise, can only obtain comparatively low resolution image.Tradition improves infrared
The method of picture quality is limited in by raising contrast more makes image detail visible.However, as digital picture is more and more
Be applied to every field, demand more and more higher of the people to high-definition picture.Such as, with medical image, satellite remote sensing, video
Each application field based on monitoring etc..Picture element density in high-definition picture is higher, and low-resolution image is compared under identical size
More details can be provided, these are indispensable in many practical applications, and these details are only to improve not obtaining under contrast
.However, because of the limitation of technical matters level, development cost and volume requirement etc., cause red from hardware aspect raising
Outer image resolution ratio is subject to many limitations, and the resolution ratio of existing passive type infrared detector is also universal relatively low.Therefore, should be from software
Angle, which is set out, improves infrared image resolution ratio, while is also one of important channel for improving infrared image quality.
Super-resolution rebuilding concept is proposed the sixties in last century by Harris and Goodman.Its meaning is by soft
The mode of part improves image resolution ratio, breaks through image intrinsic information and system hardware condition limitation, is set not replacing existing imaging
Spatial resolution that is standby, not increasing raising initial data under the premise of system cost, overcomes because of image degradation and image discretization institute
Caused resolution ratio distortion, the deficiency of original image spatial resolution is made up, improves the visual effect of image, is advantageous to image
Further analysis, processing and identification.Therefore, super-resolution rebuilding technology is applied in the imaging process of infrared system to have
The level of hardware limitation of existing infrared imaging system is broken through on effect ground, improves the spatial resolution of scope imaging, strengthens image
The characteristic informations such as edge, scope are further improved to recognition capability and accuracy of identification, favorably it is in military and civilian
The application in each field.
From after Harris in 1964 sets up the theoretical foundation of super-resolution rebuilding technology, domestic and foreign scholars carry in the field
Go out all multi-methods.
2006, F.Sroubek and J.Flusser propose by MAP estimate based on method, using multiframe low resolution
Image carries out fusion and realizes super-resolution rebuilding, overcomes the defect information of the low-resolution image collected and interframe unpredictable
Error.
2007, Carlos Miravet et al. proposed the super resolution ratio reconstruction method based on study, designed a kind of nerve
The structured training sampled images that network and genetic method are combined, then linear filtering obtain high-definition picture, this method
Efficiency is greatly improved compared with the method based on MAP.
2010, Liu Yang et al. proposed the Super-resolution Image Restoration algorithm mixed based on Wiener filtering and POCS, effectively
The defects of suppressing to be also easy to produce ringing effect and amplification noise in classical POCS algorithms.
2013, Yuzhang Chen et al. proposed to be based on MAP methods for submarine target observation, while by point spread function
Number and regularization are combined, and further improve image resolution ratio and the robustness of algorithm.
2015, the method that Kwok-Wai Hung et al. propose the Wiener filtering based on FIR, only to single frames low resolution
Image carries out super-resolution rebuilding, although parameter uses empirical value in algorithm, application in a variety of contexts can obtain well
Effect.
2016, nonredundancy sub-pixs of the Y.Tao et al. based on original image proposed GPT methods, and is applied to remote sensing
Resolution ratio is improved five times by image.
But the super-resolution rebuilding technical research for infrared image in numerous theoretical methods is less and to be reconstructed
Image run-of-the-mill it is preferable.In recent years, the method for reconstructing development based on study is very fast, and this method is believed to supplement original
The information of low-resolution image missing.Wherein, using rarefaction representation as representative, by establishing the super complete of height-low-resolution image
Wordbook, realize the super-resolution rebuilding to image.But in process of reconstruction because using complete dictionary, also easily because of information redundancy
Influence to rebuild effect.
The content of the invention
It is above-mentioned existing to overcome it is an object of the invention to provide a kind of super resolution ratio reconstruction method of feature based extraction
The problem of infrared image is easily by noise jamming and bad image quality in technology.The inventive method utilizes feature extraction and sparse table
Show the mode being combined, by extracting characteristics of image, improve influence that is openness, reducing redundancy, improve infrared image
Super-resolution rebuilding effect.
To achieve the above object, the present invention uses following technical scheme:
The infrared image super resolution ratio reconstruction method of feature based extraction, comprises the following steps:
The first step, unusual decomposition is carried out to the original infrared image of collection, filters out the noise jamming in original infrared image;
Second step, image after denoising is used feature extraction is carried out based on Gabor filter, and with being based on morphologic side
Method further enhances to the feature after extraction;
3rd step, enhanced characteristic pattern is subjected to Its Sparse Decomposition, sparse coefficient is solved using method of Lagrange multipliers, and
Go out the super-resolution image of original infrared image with required coefficient reconstruction.
Further, in the first step, original infrared image Y sizes are m × n;Obtained after singular value decomposition (SVD) by L
Singular value matrix ∑=the diag (σ 1, σ 2 ..., σ L) for the diagonal form that individual singular value is formed, wherein L are figure Y order, and unusual
It is worth descending arrangement;Given threshold σth, the singular value that will be less than threshold value casts out, with the singular value that remains and corresponding strange
Incorgruous amount rebuilds image, realizes denoising;
Further, adaptive threshold σthChoosing method be
Reconstruction image after denoising is Y 'm×n≈Um×m∑′m×nVn×n T, wherein singular value matrix be ∑ 'm×n=diag
(σ1,...σth)。
Further, in second step, when carrying out feature extraction using Gabor filter, Orientation Features extraction is carried out;
Meanwhile the information that each characteristic pattern is covered is considered when being merged to these characteristic patterns, merged using different weight coefficients.
Further, two-dimentional gabor wave filters are
Wherein, (i, j) is space pixel, σiAnd σjIt is standard deviation, f0It is spatial domain frequency;If convert former coordinate system for i '=
Icos θ+jsin θ, j '=- isin θ+jcos θ then obtain multiple directions gabor wave filters by adjusting angle parameter theta.
Further, 6 different angle direction structure gabor wave filters are chosen, and are rolled up respectively with image after denoising
Product computing, obtains series of features figure F (i, j)={ Fθ(i, j) | θ=0 °, 30 °, 60 °, 90 °, 120 °, 150 ° };
Wherein, Fθ(i, j)=Y'(i, j) * gθ(i, j), gθIt is the gabor wave filters that angle is θ;6 different angle difference
For 0 °, 30 °, 60 °, 90 °, 120 ° and 150 °.
Further, in second step, the characteristics of image contained by the characteristic pattern under different directions is different, melts carrying out feature
During conjunction, using different weight coefficients;Coefficient is
Wherein, M (Fθ(i, j)) it is characterized figure FθThe mean square deviation of (i, j);
Characteristic pattern fusion results are
Further, in second step, image is handled using expansion and erosion operation, method is
Wherein,Expansion, erosion operation are represented with Θ;B is morphology operations template, and u and v are template interior elements;Meanwhile
Gradient information is considered during morphological transformation, wherein F " (i-u, j-v)=(F+ Δs F ') (i-u, j-v) is in artwork
Increase gradient information;F " (i+u, j+v)=(F- Δs F ') (i+u, j+v) is that gradient information is subtracted in artwork;
It is F " to obtain enhanced characteristic pattern.
Further, the 3rd step specifically includes:
To enhanced characteristic pattern F " in low-resolution dictionary DlIn decomposed;Using L1Norm solves, such as following formula:
min||αl||1 s.t.||Dlαl-F”||≤ξ
Wherein, αlFor the sparse coding obtained from low resolution characteristic pattern;S.t. constraint is represented;ξ is the first of setting
Arithmetic number.
To ensure the continuity and uniformity of image after rebuilding, above formula is optimized, obtained
min||α||1 s.t.||Dlα-F”||≤ξ1
||PDhα-Q||≤ξ2
Wherein, Q is and the previous equitant part of reconstruction image block;P is that extraction is adapted with lap size
Dictionary size;α is sparse coding to be asked;ξ1For the second arithmetic number of setting;ξ2For the 3rd arithmetic number of setting;DhFor high-resolution
Rate dictionary, introduce method of Lagrange multipliers and carry out solving to obtain following formula
Wherein,λ=0.1 is relaxation factor;
Factor alpha is solved, through x=Dhα rebuilds high-definition picture block;Image block is combined again, forms complete high-resolution
Rate image.
Relative to prior art, the invention has the advantages that:
The present invention extracts original low-resolution infrared image using single-frame images as input, using the method for view-based access control model mechanism
Feature, improve image sparse;Meanwhile the characteristic pattern of extraction is strengthened using the method for morphology combination gradient, further
Key message in prominent image.It is that input is trained collection study to enhanced characteristic pattern, reduced superfluous in complete dictionary
Influence of the remaining information in process of reconstruction, overcome lack priori while, without additional samples storehouse, improve high after rebuilding
Image in different resolution quality.
Brief description of the drawings
The super resolution ratio reconstruction method overall flow figure of Fig. 1 feature baseds extraction;
The super resolution ratio reconstruction method particular flow sheet of Fig. 2 feature baseds extraction.
Embodiment
The method combination accompanying drawing that reference will now be made in detail the present invention is described in further detail to embodiments of the invention.
Refer to shown in Fig. 1 and Fig. 2, a kind of infrared image super resolution ratio reconstruction method of feature based extraction of the present invention,
Including:
The first step, to the original Infrared Image Denoising of collection:The original infrared image collected often contains much noise,
First to image denoising.
Singular value decomposition (SVD) is used as a kind of matrix disassembling method in linear algebra, can reflect the inherent algebraically sheet of image
Matter, can be as the important method of pattern classification.
Original noisy low resolution infrared image Y is decomposed into Y through SVDm×n=Um×m∑m×nVn×n T。
Wherein, image Y sizes are m × n;The diagonal form for obtaining being made up of L singular value after singular value decomposition it is strange
Different value matrix ∑=diag (σ 1, σ 2 ..., σ L), wherein L are figure Y order, and the descending arrangement of singular value.Singular value it is big
The heterogeneity and feature of small reflection image, larger singular value are related to signal in image;By contrast, less singular value
Noise in correspondence image.Pass through given threshold σth, the singular value that will be less than threshold value casts out, with the singular value that remains and right
The singular vector answered rebuilds image, realizes denoising.Compared with other denoising methods, this method can be compared with while denoising
The edge feature of good reservation image information.
Adaptive threshold σthChoosing method be
Remain larger than and be equal to σthSingular value, reconstruction image realizes denoising;Reconstruction image after denoising is Y 'm×n≈Um×m
∑′m×nVn×n T, wherein singular value matrix be ∑ 'm×n=diag (σ1,...σth)。
Second step, feature extraction is carried out to original low-resolution image after denoising.The present invention uses and combines Gabor filter
The process is realized with the method for morphological change;Specifically include:
2.1), the feature of Gabor filter and visual perception are similar, to extract characteristics of image.
Two-dimentional gabor wave filters are
Wherein, (i, j) is space pixel, σiAnd σjIt is standard deviation, f0It is spatial domain frequency.If convert former coordinate system for i '=
Icos θ+jsin θ, j '=- isin θ+jcos θ then can obtain multiple directions gabor wave filters by adjusting angle parameter theta.
The present invention chooses 6 different angle direction structure gabor wave filters, and carries out convolution fortune with image after denoising respectively
Calculate, obtain series of features figure F (i, j)={ Fθ(i, j) | θ=0 °, 30 °, 60 °, 90 °, 120 °, 150 ° }.
Wherein Fθ(i, j)=Y'(i, j) * gθ(i, j), gθThe gabor wave filters for being θ for angle.
Characteristic pattern under different directions is merged, using different weight coefficients.
Coefficient is
Wherein, M (Fθ(i, j)) it is characterized figure FθThe mean square deviation of (i, j).
Characteristic pattern fusion results are
2.2), combining form method further enhances to the characteristic pattern after extraction.
Characteristic pattern is further enhanced using based on morphological method.
Method is
Wherein,Expansion, erosion operation are represented with Θ;B is morphology operations template, and u and v are template interior elements;Meanwhile
Gradient information is considered during morphological transformation, wherein F " (i-u, j-v)=(F+ Δs F ') (i-u, j-v) is in artwork
Increase gradient information;F " (i+u, j+v)=(F- Δs F ') (i+u, j+v) is that gradient information is subtracted in artwork;
It is F " to obtain enhanced characteristic pattern.
3rd step, sparse expression coefficient is drawn with the characteristic pattern training of low-resolution image, passes through excessively complete dictionary training
Study obtains high-definition picture.
To from enhanced characteristic pattern F " in low-resolution dictionary DlIn decomposed.Because characteristic pattern F " has enough
It is openness, therefore use L1Norm solves, such as following formula.
min||αl||1 s.t.||Dlαl-F”||≤ξ
Wherein, αlFor the sparse coding obtained from low resolution characteristic pattern;S.t. constraint is represented;ξ is the first of setting
Arithmetic number.
To ensure the continuity and uniformity of image after rebuilding, above formula is optimized, obtained
min||α||1 s.t.||Dlα-F”||≤ξ1
||PDhα-Q||≤ξ2
Wherein, Q is and the previous equitant part of reconstruction image block;P is that extraction is adapted with lap size
Dictionary size;α is sparse coding to be asked;ξ1For the second arithmetic number of setting;ξ2For the 3rd arithmetic number of setting;DhFor high-resolution
Rate dictionary, introduce method of Lagrange multipliers and carry out solving to obtain following formula
Wherein,λ=0.1 is relaxation factor;
Factor alpha is solved, through x=Dhα rebuilds high-definition picture block;Image block is combined again, forms complete high-resolution
Rate image.
Above-described specific descriptions, the purpose, technical scheme and beneficial effect of invention are carried out further specifically
It is bright, it should be understood that the specific embodiment that the foregoing is only the present invention, the protection model being not intended to limit the present invention
Enclose, within the spirit and principles of the invention, any modification, equivalent substitution and improvements done etc., should be included in the present invention
Protection domain within.
Claims (8)
1. the infrared image super resolution ratio reconstruction method of feature based extraction, it is characterised in that comprise the following steps:
The first step, unusual decomposition is carried out to the original infrared image of collection, filters out the noise jamming in original infrared image;
Second step, image after denoising is used feature extraction is carried out based on Gabor filter, and with being based on morphologic method pair
Feature after extraction further enhances;
3rd step, Its Sparse Decomposition is carried out by enhanced characteristic pattern, solves sparse coefficient using method of Lagrange multipliers, and use institute
Coefficient reconstruction is asked to go out the super-resolution image of original infrared image.
2. the infrared image super resolution ratio reconstruction method of feature based extraction according to claim 1, it is characterised in that:The
In one step, original infrared image Y sizes are m × n;The diagonal form for obtaining being made up of L singular value after singular value decomposition
Singular value matrix ∑=diag (σ 1, σ 2 ..., σ L), wherein L are figure Y order, and the descending arrangement of singular value;Given threshold
σth, the singular value that will be less than threshold value casts out, and image is rebuild with the singular value and corresponding singular vector that remain, realizes
Denoising;
Adaptive threshold σthChoosing method be
Reconstruction image after denoising is Y 'm×n≈Um×m∑′m×nVn×n T, wherein singular value matrix be ∑ 'm×n=diag (σ1,...
σth)。
3. the infrared image super resolution ratio reconstruction method of feature based extraction according to claim 1, it is characterised in that:The
In two steps, when carrying out feature extraction using Gabor filter, Orientation Features extraction is carried out;Meanwhile these characteristic patterns are entered
Consider the information that each characteristic pattern is covered during row fusion, merged using different weight coefficients.
4. the infrared image super resolution ratio reconstruction method of feature based extraction according to claim 3, it is characterised in that:Two
Tieing up gabor wave filters is
Wherein, (i, j) is space pixel, σiAnd σjIt is standard deviation, f0It is spatial domain frequency;If it is i '=icos θ to convert former coordinate system
+ jsin θ, j '=- isin θ+jcos θ then obtain multiple directions gabor wave filters by adjusting angle parameter theta.
5. the infrared image super resolution ratio reconstruction method of feature based extraction according to claim 3, it is characterised in that:Choosing
6 different angle direction structure gabor wave filters are taken, and carry out convolution algorithm with image after denoising respectively, obtain a series of spies
Sign figure F (i, j)={ Fθ(i, j) | θ=0 °, 30 °, 60 °, 90 °, 120 °, 150 ° };
Wherein, Fθ(i, j)=Y'(i, j) * gθ(i,j);gθIt is the gabor wave filters that angle is θ;6 different angles are respectively
0 °, 30 °, 60 °, 90 °, 120 ° and 150 °.
6. the infrared image super resolution ratio reconstruction method of feature based extraction according to claim 3, it is characterised in that:The
In two steps, the characteristics of image contained by the characteristic pattern under different directions is different, when carrying out Fusion Features, using different weights
Coefficient, coefficient are
Wherein, M (Fθ(i, j)) it is characterized figure FθThe mean square deviation of (i, j);
Characteristic pattern fusion results are
7. the infrared image super resolution ratio reconstruction method of feature based extraction according to claim 1, it is characterised in that:The
In two steps, image is handled using expansion and erosion operation, method is
Wherein, ⊕ and Θ represents expansion, erosion operation;B is morphology operations template, and u and v are template interior elements;Meanwhile in shape
Gradient information is considered in state conversion process, wherein F " (i-u, j-v)=(F+ Δs F ') (i-u, j-v) is to increase in artwork
Gradient information;F " (i+u, j+v)=(F- Δs F ') (i+u, j+v) is that gradient information is subtracted in artwork;
F " is the characteristic pattern obtained after strengthening.
8. the infrared image super resolution ratio reconstruction method of feature based extraction according to claim 1, it is characterised in that:The
Three steps specifically include:
To enhanced characteristic pattern F " in low-resolution dictionary DlIn decomposed;Using L1Norm solves, such as following formula:
min||αl||1s.t.||Dlαl-F”||≤ξ
Wherein, αlFor the sparse coding obtained from low resolution characteristic pattern;S.t. constraint is represented;ξ is that first set is positive real
Number;
To ensure the continuity and uniformity of image after rebuilding, above formula is optimized, obtained
min||α||1 s.t.||Dlα-F”||≤ξ1
||PDhα-Q||≤ξ2
Wherein, Q is and the previous equitant part of reconstruction image block;P is the dictionary that extraction is adapted with lap size
Size;α is sparse coding to be asked;ξ1For the second arithmetic number of setting;ξ2For the 3rd arithmetic number of setting;DhFor high-resolution word
Allusion quotation, introduce method of Lagrange multipliers and carry out solving to obtain following formula:
<mrow>
<munder>
<mi>min</mi>
<mi>&alpha;</mi>
</munder>
<mo>|</mo>
<mo>|</mo>
<msup>
<mi>D</mi>
<mo>&prime;</mo>
</msup>
<mi>&alpha;</mi>
<mo>-</mo>
<mi>G</mi>
<mo>|</mo>
<msubsup>
<mo>|</mo>
<mn>2</mn>
<mn>2</mn>
</msubsup>
<mo>+</mo>
<mi>&lambda;</mi>
<mo>|</mo>
<mo>|</mo>
<mi>&alpha;</mi>
<mo>|</mo>
<msub>
<mo>|</mo>
<mn>1</mn>
</msub>
<mo>;</mo>
</mrow>
Wherein,λ=0.1 is relaxation factor;
Factor alpha is solved, through x=Dhα rebuilds high-definition picture block;Image block is combined again, forms complete high resolution graphics
Picture.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711158675.9A CN107833182A (en) | 2017-11-20 | 2017-11-20 | The infrared image super resolution ratio reconstruction method of feature based extraction |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711158675.9A CN107833182A (en) | 2017-11-20 | 2017-11-20 | The infrared image super resolution ratio reconstruction method of feature based extraction |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107833182A true CN107833182A (en) | 2018-03-23 |
Family
ID=61652238
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711158675.9A Pending CN107833182A (en) | 2017-11-20 | 2017-11-20 | The infrared image super resolution ratio reconstruction method of feature based extraction |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107833182A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108648143A (en) * | 2018-04-17 | 2018-10-12 | 中国科学院光电技术研究所 | A kind of image resolution ratio Enhancement Method using sequence image |
CN108898557A (en) * | 2018-05-30 | 2018-11-27 | 商汤集团有限公司 | Image recovery method and device, electronic equipment, computer program and storage medium |
CN109214989A (en) * | 2018-09-04 | 2019-01-15 | 四川大学 | Single image super resolution ratio reconstruction method based on Orientation Features prediction priori |
CN109697691A (en) * | 2018-12-27 | 2019-04-30 | 重庆大学 | A kind of limited view projection method for reconstructing based on the optimization of the biregular item of L0 norm and singular value threshold decomposition |
CN112508786A (en) * | 2020-12-03 | 2021-03-16 | 武汉大学 | Satellite image-oriented arbitrary-scale super-resolution reconstruction method and system |
US11346938B2 (en) | 2019-03-15 | 2022-05-31 | Msa Technology, Llc | Safety device for providing output to an individual associated with a hazardous environment |
CN116205806A (en) * | 2023-01-28 | 2023-06-02 | 荣耀终端有限公司 | Image enhancement method and electronic equipment |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7809155B2 (en) * | 2004-06-30 | 2010-10-05 | Intel Corporation | Computing a higher resolution image from multiple lower resolution images using model-base, robust Bayesian estimation |
CN102722865A (en) * | 2012-05-22 | 2012-10-10 | 北京工业大学 | Super-resolution sparse representation method |
CN104063856A (en) * | 2014-05-28 | 2014-09-24 | 北京大学深圳研究生院 | Rapid super-resolution image reconstruction method and device |
CN104766273A (en) * | 2015-04-20 | 2015-07-08 | 重庆大学 | Infrared image super-resolution reestablishing method based on compressed sensing theory |
JP2016035512A (en) * | 2014-08-04 | 2016-03-17 | 国立研究開発法人物質・材料研究機構 | Optical super-resolution medium, hyper lens, method for producing the same and hyper lens array |
CN106934766A (en) * | 2017-03-15 | 2017-07-07 | 西安理工大学 | A kind of infrared image super resolution ratio reconstruction method based on rarefaction representation |
-
2017
- 2017-11-20 CN CN201711158675.9A patent/CN107833182A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7809155B2 (en) * | 2004-06-30 | 2010-10-05 | Intel Corporation | Computing a higher resolution image from multiple lower resolution images using model-base, robust Bayesian estimation |
CN102722865A (en) * | 2012-05-22 | 2012-10-10 | 北京工业大学 | Super-resolution sparse representation method |
CN104063856A (en) * | 2014-05-28 | 2014-09-24 | 北京大学深圳研究生院 | Rapid super-resolution image reconstruction method and device |
JP2016035512A (en) * | 2014-08-04 | 2016-03-17 | 国立研究開発法人物質・材料研究機構 | Optical super-resolution medium, hyper lens, method for producing the same and hyper lens array |
CN104766273A (en) * | 2015-04-20 | 2015-07-08 | 重庆大学 | Infrared image super-resolution reestablishing method based on compressed sensing theory |
CN106934766A (en) * | 2017-03-15 | 2017-07-07 | 西安理工大学 | A kind of infrared image super resolution ratio reconstruction method based on rarefaction representation |
Non-Patent Citations (2)
Title |
---|
ZENG, JIAN 等: "The infrared image closely spaced objects super resolution method based on sparse reconstruction under the noise environment", 《FOURTH INTERNATIONAL CONFERENCE ON OPTICAL AND PHOTONICS ENGINEERING》 * |
刘春生: "人脸特征检测与疲劳状态识别研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108648143A (en) * | 2018-04-17 | 2018-10-12 | 中国科学院光电技术研究所 | A kind of image resolution ratio Enhancement Method using sequence image |
CN108898557A (en) * | 2018-05-30 | 2018-11-27 | 商汤集团有限公司 | Image recovery method and device, electronic equipment, computer program and storage medium |
CN109214989A (en) * | 2018-09-04 | 2019-01-15 | 四川大学 | Single image super resolution ratio reconstruction method based on Orientation Features prediction priori |
CN109214989B (en) * | 2018-09-04 | 2019-08-13 | 四川大学 | Single image super resolution ratio reconstruction method based on Orientation Features prediction priori |
CN109697691A (en) * | 2018-12-27 | 2019-04-30 | 重庆大学 | A kind of limited view projection method for reconstructing based on the optimization of the biregular item of L0 norm and singular value threshold decomposition |
CN109697691B (en) * | 2018-12-27 | 2022-11-25 | 重庆大学 | Dual-regularization-term-optimized finite-angle projection reconstruction method based on L0 norm and singular value threshold decomposition |
US11346938B2 (en) | 2019-03-15 | 2022-05-31 | Msa Technology, Llc | Safety device for providing output to an individual associated with a hazardous environment |
CN112508786A (en) * | 2020-12-03 | 2021-03-16 | 武汉大学 | Satellite image-oriented arbitrary-scale super-resolution reconstruction method and system |
CN112508786B (en) * | 2020-12-03 | 2022-04-29 | 武汉大学 | Satellite image-oriented arbitrary-scale super-resolution reconstruction method and system |
CN116205806A (en) * | 2023-01-28 | 2023-06-02 | 荣耀终端有限公司 | Image enhancement method and electronic equipment |
CN116205806B (en) * | 2023-01-28 | 2023-09-19 | 荣耀终端有限公司 | Image enhancement method and electronic equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Wang et al. | An experiment-based review of low-light image enhancement methods | |
CN107833182A (en) | The infrared image super resolution ratio reconstruction method of feature based extraction | |
CN109741256A (en) | Image super-resolution rebuilding method based on rarefaction representation and deep learning | |
CN111861961B (en) | Single image super-resolution multi-scale residual error fusion model and restoration method thereof | |
CN105243670B (en) | A kind of sparse and accurate extracting method of video foreground object of low-rank Combined expression | |
CN106709875A (en) | Compressed low-resolution image restoration method based on combined deep network | |
CN108921786A (en) | Image super-resolution reconstructing method based on residual error convolutional neural networks | |
CN108537754B (en) | Face image restoration system based on deformation guide picture | |
CN106339998A (en) | Multi-focus image fusion method based on contrast pyramid transformation | |
CN113516601B (en) | Image recovery method based on deep convolutional neural network and compressed sensing | |
CN105825472A (en) | Rapid tone mapping system and method based on multi-scale Gauss filters | |
CN102982520B (en) | Robustness face super-resolution processing method based on contour inspection | |
CN112001843B (en) | Infrared image super-resolution reconstruction method based on deep learning | |
CN106920214A (en) | Spatial target images super resolution ratio reconstruction method | |
CN102243711A (en) | Neighbor embedding-based image super-resolution reconstruction method | |
CN104299193B (en) | Image super-resolution reconstruction method based on high-frequency information and medium-frequency information | |
CN112163998A (en) | Single-image super-resolution analysis method matched with natural degradation conditions | |
CN114820408A (en) | Infrared and visible light image fusion method based on self-attention and convolutional neural network | |
CN114187203A (en) | Attention-optimized deep codec defogging generation countermeasure network | |
CN104021523A (en) | Novel method for image super-resolution amplification based on edge classification | |
CN115222614A (en) | Priori-guided multi-degradation-characteristic night light remote sensing image quality improving method | |
CN108492252A (en) | Face image super-resolution reconstruction method based on secondary reconstruction | |
CN111553856A (en) | Image defogging method based on depth estimation assistance | |
CN109064394B (en) | Image super-resolution reconstruction method based on convolutional neural network | |
CN112132757B (en) | General image restoration method based on neural network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180323 |
|
WD01 | Invention patent application deemed withdrawn after publication |