CN107316274A - A kind of Infrared image reconstruction method that edge is kept - Google Patents
A kind of Infrared image reconstruction method that edge is kept Download PDFInfo
- Publication number
- CN107316274A CN107316274A CN201710326271.XA CN201710326271A CN107316274A CN 107316274 A CN107316274 A CN 107316274A CN 201710326271 A CN201710326271 A CN 201710326271A CN 107316274 A CN107316274 A CN 107316274A
- Authority
- CN
- China
- Prior art keywords
- image
- reconstruction
- resolution
- frequency
- original
- 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 55
- 230000008569 process Effects 0.000 claims abstract description 16
- 239000011159 matrix material Substances 0.000 claims description 11
- 238000005070 sampling Methods 0.000 claims description 7
- 239000000654 additive Substances 0.000 claims description 4
- 230000000996 additive effect Effects 0.000 claims description 4
- 230000009466 transformation Effects 0.000 claims description 4
- 230000004927 fusion Effects 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 230000002146 bilateral effect Effects 0.000 claims description 2
- 230000007423 decrease Effects 0.000 claims description 2
- 238000001914 filtration Methods 0.000 claims description 2
- 238000004260 weight control Methods 0.000 claims description 2
- 230000000007 visual effect Effects 0.000 abstract description 10
- 238000012545 processing Methods 0.000 abstract description 8
- 238000000605 extraction Methods 0.000 abstract description 4
- 238000000926 separation method Methods 0.000 abstract description 3
- 230000001737 promoting effect Effects 0.000 abstract 1
- 238000003384 imaging method Methods 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 6
- 230000008859 change Effects 0.000 description 4
- 230000007547 defect Effects 0.000 description 4
- 238000003331 infrared imaging Methods 0.000 description 4
- 230000007812 deficiency Effects 0.000 description 3
- 238000002156 mixing Methods 0.000 description 3
- 230000003412 degenerative effect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000001931 thermography Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000011549 displacement method Methods 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000020169 heat generation Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/40—Scaling the whole image or part thereof
- G06T3/4053—Super resolution, i.e. output image resolution higher than sensor resolution
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/10—Image enhancement or restoration by non-spatial domain filtering
-
- 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
- G06T2207/20028—Bilateral filtering
-
- 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
- G06T2207/20172—Image enhancement details
- G06T2207/20192—Edge enhancement; Edge preservation
Abstract
The infrared image super resolution ratio reconstruction method that a kind of edge is kept is claimed in the present invention, belongs to image reconstruction process field.Based on traditional regularization super-resolution image reconstruction algorithm, the present invention takes into full account that infrared image edge is fuzzy, the low feature of contrast, in process of reconstruction, it assign the separation and Extraction of high price value information in low-resolution image and processing as a ring important in super-resolution rebuilding, the detailed information such as edge, texture by individually focusing on processing reconstructed image, reach that edge is kept, improve the purpose of reconstruction image visual effect.The inventive method implementation process is simple, operand is small, rebuild infrared image edge keeps working well, with higher application and promotional value.
Description
Technical field
The invention belongs to image reconstruction process technical field, the Infrared image reconstruction method that specifically a kind of edge is kept.
Background technology
Infrared thermal imaging has that resolution ratio is low, contrast is low, image blur and the features such as obvious noise, in infrared imaging
In, the information content that the higher representative image of resolution ratio is carried is more, it is meant that the detailed information such as texture, edge of image is abundanter,
Visual effect is better.
The fast development of science and technology and the huge advance of society, the thing followed is each field to high-resolution Thermo-imaging system
Active demand.Highdensity focal plane infrared detector is to obtain the most direct effective manner of high-resolution Thermo-imaging system, however,
Current semiconductor fabrication process level is limited, and existing detector array density is often relatively low, and imaging resolution is low, it is impossible to meet high
Resolution requirements.For example in military investigation field, by detecting, target range is remote and infrared detector array density is limited, very
Difficulty gets high-resolution Thermo-imaging system, and the recognition capability and identification precision struck target is limited；In medical domain, low point
The infrared human body thermography of resolution can not provide accurate lesion sites of heat generation, the diagnosis not details of morbidity；Public
Security fields, when abnormal conditions occur, the critical information, video such as personage figure, license plate number that infrared imaging is extracted
Limited resolution, identifiability is low, it is impossible to provide help etc. for the detection of case.Therefore, Technique of Super-resolution Image Construction should
Transport and give birth to.
Super-resolution rebuilding technology is that low-resolution image is utilized in the case where not changing detector hardware present situation
Rebuilding series obtain the technology of high-definition picture.The technology is combined using optical imagery mechanism with computer disposal
Mode, under relatively low cost price, realize high-resolution infrared imaging, be the heat of current image processing field
Point.U.S. congruent people is accounted for exist《Total variation regularization super-resolution image reconstruction based on L1 norms》Middle use L1 norms pair
Reconstruction image fidelity enters row constraint, and the pathosis of Problems of Reconstruction is overcome using total variation regularization, and cost function isAlgorithm not only has good anti-noise
Acoustic performance, and amount of calculation is small, calculates simple, but the details high-frequency information such as edge, the texture of reconstruction image is significantly put down
Sliding to suppress, reconstruction image is fuzzy, gloomy not clear；Reconstruction cost function based on L2 norms fidelity and total variation regularization term
ForBecause L2 norms think all
The value of image information for being used to reconstruct be identical, in the reconstruction image finally produced, the contribution of image information is also
Pixel singular value in identical, therefore image is propagated in process of reconstruction, the edge details of the reconstruction image based on L2 norms
Reduced etc. information, but there is the obvious defect of noise, the signal to noise ratio of image is not high after reconstruction, and visual effect is undesirable；Examine
Consider exclusive use L1 norms estimation operator or L2 norms estimation operator all existing defects and deficiency, then Li Yin brightness et al. exists
《The sequence image super-resolution reconstruction of normal form is mixed based on L1 and L2》Propose to filter out asking for Gaussian noise and impulsive noise simultaneously
Topic, rebuilding cost function is
Reconstruction is improved using the image reconstruction algorithm of mixing normal form to a certain extent while data fit term evaluated error is reduced
The anti-noise ability of algorithm, but during applied to the Infrared image reconstruction that contrast is low, edge detail information is not enriched, it is difficult to do
To the edge detail information for keeping infrared image, reconstructed results image border texture is lacked, and visual effect is undesirable.
In summary, the photosensitive unit density of existing infrared detector array can not meet modern military, industry, doctor
The each side such as treatment, life obtain the demand of high-resolution Thermo-imaging system.On the premise of detector photosensitive unit density is not changed,
Super resolution ratio reconstruction method is on acquisition high-definition picture is realized, with the dual affirmative in theoretical and experimental results.But should
Technology is still immature applied to infrared imaging field at present, implements difficult, it is necessary to which we more deeply and carefully grind
Study carefully, therefore be of great significance in the super-resolution image reconstruction method research tool that infrared regime carries out edge holding.
The content of the invention
Present invention seek to address that above problem of the prior art.Propose a kind of edge, texture of protection reconstruction image etc.
Detailed information, improves the Infrared image reconstruction method that the edge of reconstruction image visual effect is kept.Technical scheme is such as
Under：
A kind of Infrared image reconstruction method that edge is kept, it comprises the following steps：
1) image high-frequency information composition, is extracted in the original low-resolution image sequence information for reconstruction
Yk(high-frequency)；
2), the high-frequency information composition Y to extractingk(high-frequency)Estimate based on L2 norms the super-resolution of operator
Algorithm for reconstructing processing, obtains the high-definition picture X corresponding to high-frequency information compositionhigh-freqency；
3), to original low-resolution image sequence information based on L1 norms estimate the super-resolution algorithms weight of operator
Build, obtain the high-definition picture X of original reconstructionoriginal；
4), by step 2) high-frequency information rebuild Xhigh-freqencyWith step 3) original reconstruction high-definition picture
XoriginalThe method being directly superimposed using gray value carries out information fusion, obtains final reconstruction high-definition picture.
Further, the step 1) middle extraction image high-frequency information composition Yk(high-frequency)Use using three times
The interpolation sampling factor obtains the high-frequency information in image sequence, and specific formula is as follows,
Ylow-frequency=(YkBkD ↓) D ↑ D ↑,
For a frame low-resolution image Yk, its corresponding high-resolution
Rate image is X, and this mapping relations are represented by, Yk=(D ↓) BkWkX+nk, k=1,2 ..., K, wherein,Represent
Original image is up-sampled, Ylow-frequencyExpression low frequency component information, the D ↑ up-sampling factor, D ↓ it is the down-sampling factor, Bk
Represent fuzzy matrix, WkRepresent geometry motion matrix, nkThe additive noise in image acquisition process is represented, K represents kth frame image,
The high-frequency information components finally given are Yk(high-frequency)。
Further, the step 2) in based on L2 norms estimate operator super-resolution rebuilding be using gradient decline it is excellent
Change method is solved, and high-frequency information reconstruction image is obtained when convergence.
Further, it is described to estimate that the super-resolution rebuilding of operator is specifically included based on L2 norms：Adopt
The full variation regular terms of image border can be kept with regularization term use, the cost function that radio-frequency component is rebuild is, Represent bilateral filtering operator power
Weight control coefrficient,Horizontal direction transformation matrix is represented,Represent vertical direction transformation matrix, XhRepresent original high resolution figure
The radio-frequency component of picture.
Further, the step 3) in original low-resolution image sequence information carry out based on L1 norms estimate operator
Super-resolution algorithms rebuild and use gradient descent method optimization and solved, obtain the high-resolution of original reconstruction when convergence
Rate image Xoriginal。
Further, the step 3) use can keep the full variation regular terms of image border, what original image was rebuild
Cost function is,
Represent the image after being up-sampled to low-resolution image.
Advantages of the present invention and have the beneficial effect that：
The infrared image super resolution ratio reconstruction method that edge designed by the present invention is kept to existing regularization by rebuilding
Method is improved, and this method assign the separation and Extraction of high price value information in low-resolution image and processing as super-resolution rebuilding
In an important ring, reached the detailed information such as holding edge, texture, improved the Expected Results of reconstruction image visual effect.Examine
Consider in exclusive use L1 normal forms estimation operator or L2 estimations operator all existing defects and deficiency, and real imaging circumstances, it is all kinds of
Noise is present, and Gaussian noise should be considered during super-resolution rebuilding, again consider impulsive noise presence, then it is existing just
Then change the image rebuilding method rebuild using norm is mixed based on L1 and L2, take into account using the image reconstruction of mixing normal form
Each noise like in reality reconstruction, algorithm adaptability is improved, certain journey while data fit term evaluated error is reduced
The anti-noise ability of algorithm for reconstructing is improved on degree.But applied to contrast is low, edge detail information not abundant infrared figure
During as rebuilding, it is difficult to accomplish to keep the edge detail information of infrared image, reconstructed results image border texture missing, visual effect
It is undesirable.
Brief description of the drawings
The flow framework for the Infrared image reconstruction method that the edge that Fig. 1 present invention provides the preferred embodiment present invention is kept
Figure.
Fig. 2 is traditional regularization reconstruction method and method for reconstructing reconstructed results comparison diagram of the present invention, wherein Fig. 2A left hand views
As being traditional algorithm reconstructed results, Fig. 2 B image rights are the inventive method reconstructed results.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, detailed
Carefully describe.Described embodiment is only a part of embodiment of the present invention.
The present invention solve above-mentioned technical problem technical scheme be：
Understanding for the ease of the public to technical scheme, it is right before technical scheme is described in detail
Regularization is rebuild principle and is illustrated.
In the acquisition process of low-resolution image, the quality that the low-resolution image of acquisition shows varying degree is moved back
Change, image deterioration model exactly expresses the degenerative process between original image and collection image.Degenerative process is image acquisition procedures,
But it is even more the foundation for rebuilding reduction real scene information.It is the inverse process rebuild to degrade, and the accurate understanding model that degrades is that backstepping is obtained
To the basic assurance for being enough to reflect real scene image, the model that degrades can be expressed as follows,
Yk=HkX+nk,1≤k≤N
In formula, YkKth frame observed image is represented, X represents high-resolution original scene, HkRepresent degenerate matrix, nkRepresent
Additive noise.
Regularization, which is rebuild, sets up on MAP frame foundations, is the reverse temperature intensity process of image degradation model, is converted into
It is the most frequently used solution mode that the minimization problem of object function, which is solved, according to the cost equation of object function under MAP frameworks, generation
Valency function can be expressed as follows,
Cost function is rebuild in regularization includes two parts, Part I A (X, Yk)=ρ (Yk,DBkWkX) it is data fidelity
Degree of approximation between item, i.e. data fit term, the image that the original high-resolution scene of expression and estimation are obtained；Part II
B (X) is regular terms, the prior information comprising image reconstruction；λ is regularization parameter, can balance super-resolution rebuilding fidelity
With the ratio of regular terms.Then formula is rebuild cost function and can be rewritten as,
P is related to the noise model employed in imaging process in above formula, during using laplace model, and p=1 intends
Item is closed using L1 norms to be estimated；Correspondingly, when using Gauss model, p=2, corresponding to L2 norms.
In view of L1 normal forms estimation operator or L2 estimations operator all existing defects and deficiency, and reality imaging ring is used alone
In border, each noise like is present, and Gaussian noise should be considered during super-resolution rebuilding, and the presence of impulsive noise is considered again, in
It is that existing regularization is rebuild using the image rebuilding method that norm is mixed based on L1 and L2, cost function such as following formula,
Each noise like during reality is rebuild is take into account using the image reconstruction of mixing normal form, algorithm adaptability is carried
Height, improves the anti-noise ability of algorithm for reconstructing to a certain extent while data fit term evaluated error is reduced.But application
When the Infrared image reconstruction that contrast is low, edge detail information is not enriched, it is difficult to accomplish to keep the edge of infrared image thin
Information is saved, reconstructed results image border texture missing, visual effect is undesirable.
The thinking of the present invention is to assign the separation and Extraction of high price value information in low-resolution image and processing as super-resolution
An important ring in reconstruction.The detailed information such as edge, texture by individually focusing on processing reconstructed image, reach that edge is kept,
Improve the effect of reconstruction image visual effect.
Specifically, the inventive method, as shown in Figure 1, image reconstruction is carried out according to procedure below：
The present invention devises a kind of Infrared image reconstruction method that edge is kept, including step in detail below：
Step (1)：Image high-frequency information composition is extracted in the sequence of low resolution pictures information for reconstruction
Yk(high-frequency)。
Step (2)：The high-frequency information composition extracted is carried out to estimate that the super-resolution rebuilding of operator is calculated based on L2 norms
Method processing, obtains the high-definition picture X corresponding to high-frequency information compositionhigh-freqency。
Step (3)：Original low-resolution image sequence information is carried out to estimate that the super-resolution of operator is calculated based on L1 norms
Method is rebuild, and obtains the high-definition picture X of original reconstructionoriginal。
Step (4)：The X that high-frequency information is rebuildhigh-freqencyWith original reconstruction high-definition picture XoriginalBelieved
Breath fusion, obtains final reconstruction high-definition picture.
Further, in the step (1), for a frame low-resolution image Yk, its corresponding high-definition picture is
X, this mapping relations are represented by, Yk=(D ↓) BkWkX+nk, k=1,2 ..., K, wherein, D ↓ it is the down-sampling factor, BkRepresent
Fuzzy matrix, WkRepresent geometry motion matrix, nkRepresent the additive noise in image acquisition process.Extract the high frequency letter in image
During breath, because cubic interpolation algorithm has the effect of low-frequency information composition of being effectively maintained, therefore using cubic interpolation sampling because
Son obtains the high-frequency information in image sequence, and specific formula is as follows,
Ylow-frequency=(YkBkD ↓) D ↑ D ↑,
The high-frequency information components finally given are Yk(high-frequency)。
Further, in the step (2), regularization term use can keep the full variation regular terms of image border, high
Frequency composition rebuild cost function be,
Further, in the step (3), regularization term use can keep the full variation regular terms of image border, former
The cost function of beginning image reconstruction is,
Further, in the step (4), by merging high-frequency reconstruction image and original reconstruction image, obtain finally
High-resolution Thermo-imaging system, formula is, X=Xh+Xoriginal。
The Infrared image reconstruction method kept according to edge proposed by the present invention, from the figure that a width size is 320 × 240
As being tested, and this image is translated, obscured, down-sampled 4 width low-resolution images are generated.Between translational motion is used etc.
Away from displacement method；Ambiguity function preferred dimension is 3 × 3, and variance is 0.01 Gaussian Blur；Down-sampled use interlacing is every the side of row
Formula.
Using traditional regularization reconstruction method and the method for reconstructing result of edge holding of the invention as shown in Figure 2.By
Accompanying drawing 2 is as can be seen that the reconstruction image visual effect of the inventive method is substantially better than traditional algorithm.
The above embodiment is interpreted as being merely to illustrate the present invention rather than limited the scope of the invention.
After the content for the record for having read the present invention, technical staff can make various changes or modifications to the present invention, these equivalent changes
Change and modification equally falls into the scope of the claims in the present invention.
Claims (6)
1. a kind of Infrared image reconstruction method that edge is kept, it is characterised in that comprise the following steps：
1) image high-frequency information composition, is extracted in the original low-resolution image sequence information for reconstruction
Yk(high-frequency)；
2), the high-frequency information composition Y to extractingk(high-frequency)Estimate based on L2 norms the super-resolution rebuilding of operator
Algorithm process, obtains the high-definition picture X corresponding to high-frequency information compositionhigh-freqency；
3), original low-resolution image sequence information is carried out to estimate that the super-resolution algorithms of operator are rebuild based on L1 norms, obtained
To the high-definition picture X of original reconstructionoriginal；
4), by step 2) high-frequency information rebuild Xhigh-freqencyWith step 3) original reconstruction high-definition picture Xoriginal
The method being directly superimposed using gray value carries out information fusion, obtains final reconstruction high-definition picture.
2. the Infrared image reconstruction method that edge according to claim 1 is kept, it is characterised in that the step 1) in carry
Take image high-frequency information composition Yk(high-frequency)Use using cubic interpolation decimation factor to obtain in image sequence
High-frequency information, specific formula is as follows,
<mrow>
<msub>
<mover>
<mi>Y</mi>
<mo>~</mo>
</mover>
<mrow>
<mi>l</mi>
<mi>o</mi>
<mi>w</mi>
<mo>-</mo>
<mi>f</mi>
<mi>r</mi>
<mi>e</mi>
<mi>q</mi>
<mi>u</mi>
<mi>e</mi>
<mi>n</mi>
<mi>c</mi>
<mi>y</mi>
</mrow>
</msub>
<mo>=</mo>
<msub>
<mi>Y</mi>
<mi>k</mi>
</msub>
<mi>D</mi>
<mo>&UpArrow;</mo>
<mo>,</mo>
</mrow>
Ylow-frequency=(YkBkD ↓) D ↑ D ↑,
For a frame low-resolution image Yk, its corresponding high resolution graphics
Picture is X, and this mapping relations are represented by, Yk=(D ↓) BkWkX+nk, k=1,2 ..., K, wherein,Represent to original
Beginning image is up-sampled, Ylow-frequencyExpression low frequency component information, the D ↑ up-sampling factor, D ↓ it is the down-sampling factor, BkRepresent
Fuzzy matrix, WkRepresent geometry motion matrix, nkThe additive noise in image acquisition process is represented, K represents kth frame image, finally
Obtained high-frequency information components are Yk(high-frequency)。
3. the Infrared image reconstruction method that edge according to claim 1 is kept, it is characterised in that the step 2) in base
Estimate that the super-resolution rebuilding of operator is to decline optimization using gradient to be solved in L2 norms, high frequency letter is obtained when convergence
Cease reconstruction image.
4. the Infrared image reconstruction method that edge according to claim 3 is kept, it is characterised in that described, described to be based on L2
The super-resolution rebuilding of norm estimation operator is specifically included：The full variation canonical of image border can be kept using regularization term use
, radio-frequency component rebuild cost function be,
Bilateral filtering operator weight control coefrficient is represented,Horizontal direction transformation matrix is represented,Represent vertical direction transformation matrix, Xh
Represent the radio-frequency component of original high-resolution image.
5. the Infrared image reconstruction method that edge according to claim 4 is kept, it is characterised in that the step 3) in it is right
Original low-resolution image sequence information carries out estimating that the super-resolution algorithms reconstruction of operator uses gradient based on L1 norms
Descent method optimization is solved, and the high-definition picture X of original reconstruction is obtained when convergenceoriginal。
6. the Infrared image reconstruction method that edge according to claim 5 is kept, it is characterised in that the step 3) use
The full variation regular terms of image border can be kept, the cost function that original image is rebuild is,Represent
Image after being up-sampled to low-resolution image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710326271.XA CN107316274A (en) | 2017-05-10 | 2017-05-10 | A kind of Infrared image reconstruction method that edge is kept |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710326271.XA CN107316274A (en) | 2017-05-10 | 2017-05-10 | A kind of Infrared image reconstruction method that edge is kept |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107316274A true CN107316274A (en) | 2017-11-03 |
Family
ID=60184725
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710326271.XA Pending CN107316274A (en) | 2017-05-10 | 2017-05-10 | A kind of Infrared image reconstruction method that edge is kept |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107316274A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108090870A (en) * | 2017-12-13 | 2018-05-29 | 苏州长风航空电子有限公司 | A kind of infrared image super resolution ratio reconstruction method based on thaumatropy self similarity |
| WO2020124355A1 (en) * | 2018-12-18 | 2020-06-25 | 深圳市大疆创新科技有限公司 | Image processing method, image processing device, and unmanned aerial vehicle |
| WO2021083020A1 (en) * | 2019-11-01 | 2021-05-06 | RealMe重庆移动通信有限公司 | Image fusion method and apparatus, and storage medium and terminal device |
| CN114841901A (en) * | 2022-07-01 | 2022-08-02 | 北京大学深圳研究生院 | Image reconstruction method based on generalized depth expansion network |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101441765A (en) * | 2008-11-19 | 2009-05-27 | 西安电子科技大学 | Self-adapting regular super resolution image reconstruction method for maintaining edge clear |
| CN102360498A (en) * | 2011-10-27 | 2012-02-22 | 江苏省邮电规划设计院有限责任公司 | Reconstruction method for image super-resolution |
| US20130030289A1 (en) * | 2006-07-06 | 2013-01-31 | Quing Zhu | Method and apparatus for medical imaging using near-infrared optical tomography and flourescence tomography combined with ultrasound |
-
2017
- 2017-05-10 CN CN201710326271.XA patent/CN107316274A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130030289A1 (en) * | 2006-07-06 | 2013-01-31 | Quing Zhu | Method and apparatus for medical imaging using near-infrared optical tomography and flourescence tomography combined with ultrasound |
| CN101441765A (en) * | 2008-11-19 | 2009-05-27 | 西安电子科技大学 | Self-adapting regular super resolution image reconstruction method for maintaining edge clear |
| CN102360498A (en) * | 2011-10-27 | 2012-02-22 | 江苏省邮电规划设计院有限责任公司 | Reconstruction method for image super-resolution |
Non-Patent Citations (1)
| Title |
|---|
| 代少升等: "Edge preserving super-resolution infrared image reconstruction based on L1- and L2-norms", 《FRONTIERS OF OPTOELECTRONICS》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108090870A (en) * | 2017-12-13 | 2018-05-29 | 苏州长风航空电子有限公司 | A kind of infrared image super resolution ratio reconstruction method based on thaumatropy self similarity |
| WO2020124355A1 (en) * | 2018-12-18 | 2020-06-25 | 深圳市大疆创新科技有限公司 | Image processing method, image processing device, and unmanned aerial vehicle |
| WO2021083020A1 (en) * | 2019-11-01 | 2021-05-06 | RealMe重庆移动通信有限公司 | Image fusion method and apparatus, and storage medium and terminal device |
| CN112767290A (en) * | 2019-11-01 | 2021-05-07 | RealMe重庆移动通信有限公司 | Image fusion method, image fusion device, storage medium and terminal device |
| CN112767290B (en) * | 2019-11-01 | 2022-11-11 | RealMe重庆移动通信有限公司 | Image fusion method, image fusion device, storage medium and terminal device |
| CN114841901A (en) * | 2022-07-01 | 2022-08-02 | 北京大学深圳研究生院 | Image reconstruction method based on generalized depth expansion network |
| CN114841901B (en) * | 2022-07-01 | 2022-10-25 | 北京大学深圳研究生院 | Image reconstruction method based on generalized depth expansion network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110310227B (en) | Image super-resolution reconstruction method based on high-low frequency information decomposition | |
| CN107316274A (en) | A kind of Infrared image reconstruction method that edge is kept | |
| CN105046672B (en) | A kind of image super-resolution rebuilding method | |
| CN102663361B (en) | Face image reversible geometric normalization method facing overall characteristics analysis | |
| CN107292819A (en) | A kind of infrared image super resolution ratio reconstruction method protected based on edge details | |
| CN110634147A (en) | Image matting method based on bilateral boot up-sampling | |
| Li et al. | Sparse representation-based volumetric super-resolution algorithm for 3D CT images of reservoir rocks | |
| Guo et al. | Towards efficient motion-blurred public security video super-resolution based on back-projection networks | |
| Wen et al. | Residual network with detail perception loss for single image super-resolution | |
| Khattab et al. | A Hybrid Regularization-Based Multi-Frame Super-Resolution Using Bayesian Framework. | |
| Wu et al. | Multi-image hybrid super-resolution reconstruction via interpolation and multi-scale residual networks | |
| CN110211216B (en) | Three-dimensional image spatial domain fusion method based on volume rendering opacity weighting | |
| CN111899166A (en) | Medical hyperspectral microscopic image super-resolution reconstruction method based on deep learning | |
| CN106558021A (en) | Video enhancement method based on super-resolution technique | |
| Zhang | A selection of image processing techniques: from fundamentals to research front | |
| Fan et al. | Evaluation of and compensation for spatial noise of LCDs in medical applications | |
| Zhang et al. | Video superresolution reconstruction using iterative back projection with critical-point filters based image matching | |
| San-You et al. | Adaptive diagonal total-variation generative adversarial network for super-resolution imaging | |
| Li | Image super-resolution algorithm based on RRDB model | |
| Zhang et al. | Multi Morphological Sparse Regularized Image Super-Resolution Reconstruction Based on Machine Learning Algorithm | |
| CN106097248A (en) | A kind of compression sensing method based on high-definition picture knowledge priori and mixing visual system thereof | |
| CN110895790A (en) | Scene image super-resolution method based on posterior degradation information estimation | |
| Han et al. | Dual discriminators generative adversarial networks for unsupervised infrared super-resolution | |
| Wang et al. | Superresolution of single gaofen-4 visible-light and near-infrared (VNIR) image based on texture image extraction | |
| Mubarak et al. | Enhancement of Medical Images Using Super Resolution Convolutional 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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171103 |