CN112907461B - Defogging enhancement method for infrared foggy-day degraded image - Google Patents
Defogging enhancement method for infrared foggy-day degraded image Download PDFInfo
- Publication number
- CN112907461B CN112907461B CN202110100817.6A CN202110100817A CN112907461B CN 112907461 B CN112907461 B CN 112907461B CN 202110100817 A CN202110100817 A CN 202110100817A CN 112907461 B CN112907461 B CN 112907461B
- Authority
- CN
- China
- Prior art keywords
- image
- super pixel
- value
- super
- pixel
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000002834 transmittance Methods 0.000 claims abstract description 12
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 5
- 230000011218 segmentation Effects 0.000 claims description 11
- 238000004422 calculation algorithm Methods 0.000 claims description 7
- 238000010845 search algorithm Methods 0.000 claims description 4
- 230000002708 enhancing effect Effects 0.000 claims description 3
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 claims 1
- 238000003384 imaging method Methods 0.000 abstract description 7
- 230000000007 visual effect Effects 0.000 abstract description 4
- 238000011156 evaluation Methods 0.000 abstract 1
- 238000003709 image segmentation Methods 0.000 abstract 1
- 230000001788 irregular Effects 0.000 abstract 1
- 239000011159 matrix material Substances 0.000 abstract 1
- 238000006731 degradation reaction Methods 0.000 description 6
- 230000015556 catabolic process Effects 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 125000001475 halogen functional group Chemical group 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 230000000903 blocking effect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 239000002245 particle Substances 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000003331 infrared imaging Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005293 physical law Methods 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/73—Deblurring; Sharpening
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
- G06T2207/30192—Weather; Meteorology
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a defogging enhancement method for an infrared foggy weather degraded image. The invention realizes image defogging treatment based on irregular local area information and an atmospheric physical model. According to the complexity of the infrared image scene, the super pixel number and the boundary compactness of the image are preset, and the image segmentation is realized by using simple linear iterative clustering. And obtaining the global atmospheric light value A by using an improved automatic searching method for quadtree decomposition. In the obtained super pixel blocks, according to the value of the pixel matrix in each super pixel block, the visual visibility evaluation method and the upper and lower limit constraint relation of information truncation are analyzed to obtain the optimal transmittance t'. The haze-free image J can be inverted according to the obtained input image I, the atmospheric light value a, and the optimal transmittance t'. According to the method, the optimal parameters of the atmospheric physical model are calculated through the image local information, the method has adaptability to different fog concentrations, defogging inversion of the infrared image is realized, the structure is simple, the implementation is easy, and the imaging quality of the infrared foggy days is effectively improved.
Description
Technical field:
the invention belongs to the technical field of image processing, and mainly aims at a degradation image enhancement algorithm of a single-band imaging system in a haze day, which is particularly suitable for enhancing the local blurring of an image, which is caused by high overall brightness and low contrast of an image on an infrared remote sensing image in a haze day, so that the influence of haze on the imaging system is reduced, and the improvement of the image quality is realized.
The background technology is as follows:
in the imaging process of an outdoor long-distance optical system in a haze weather, an infrared imaging system is often influenced by scattering, absorption and reflection effects generated by suspended particles in the atmosphere, so that the phenomena of high overall brightness value, low contrast, reduced dynamic range, reduced definition, blurring and the like are caused by degradation of an observation image, and the influence degree of the degradation on the image is enhanced along with the increase of the distance between a target and observation equipment. The problem of image degradation results in the inability to obtain clear and stable high quality images or sequences, and in the inability of subsequent algorithms for identification, tracking, segmentation, etc. to work effectively. Meanwhile, along with the rapid development of automatic driving and auxiliary driving at present, new requirements on the imaging quality of haze days are also provided. Defogging techniques can currently be divided into two main categories: an enhancement method based on image processing and an enhancement method based on image restoration.
The defogging method based on image enhancement starts from the characteristic of low contrast of a foggy image, improves visual quality through various filtering and linear nonlinear stretching transformation of the image contrast, and the mainstream methods comprise histogram equalization, a filtering method, wavelet change, a Retinex algorithm, an atmospheric modulation transfer function method and the like. These methods have a good effect to some extent, but have problems. For targets with uneven illumination or complex distance change, halo phenomenon and blockiness are easy to cause, the method has poor universality, simultaneously has the problem of large calculation amount, and defogging images do not accord with actual physical laws to a certain extent only from the visual point of view.
Based on the image restoration method, the physical process under the image quality caused by haze weather is started, and the defogging purpose is achieved by inverting the physical degradation process. According to the basic research of the atmospheric particle optics, an atmospheric physical scattering model is obtained, and the haze-free image can be inverted by solving the optimal parameters of the atmospheric physical scattering model. Based on this, this type of approach aims at finding the model optimal parameters, so most need to provide additional prior information in addition to the image, such as scene depth information or additional means to obtain depth assistance information, fog concentration information or scene polarization information, etc. Due to the difficulty in acquiring additional information and the complexity of operation, defogging methods based only on image information are rapidly developed. The dark channel method is a statistical priori information proposed by He Kaiming, and is the most effective defogging priori information at present, but the dark channel method is extremely easy to generate halo and blocky effect due to unreasonable blocking, and a large amount of calculation is needed for eliminating the blocky effect. However, the dark channel method is obtained by counting a large amount of three-channel information of the color image, and is not suitable for the single-band infrared image, so that effective constraint conditions need to be searched to guide the physical restoration process of the infrared degraded image.
The invention comprises the following steps:
in order to overcome the defects in the prior art, the invention provides an infrared foggy weather degradation image defogging enhancement method. The method is based on an image restoration method, the artifact phenomenon is eliminated by replacing rectangular segmentation with super-pixel segmentation, and meanwhile, the physical restoration process is guided by upper and lower limit constraint, so that the calculated amount is small, and the imaging quality is obviously improved.
The above object of the present invention is achieved by the following technical solutions:
the defogging and enhancing method for the infrared foggy-day degraded image is characterized in that after an image is segmented by utilizing a super-pixel segmentation algorithm, an improved quadtree decomposition search algorithm and upper and lower limit constraints are combined to obtain optimal parameters of an atmospheric physical model inversion equation so as to reconstruct the image, and the method comprises the following steps:
(1) The simple linear iterative clustering segmentation algorithm needs to preset the number k of super pixel blocks and the boundary compactness m, wherein the number k of super pixels and the boundary compactness m are respectively set to 40-80 and 30-50;
(2) And (3) initializing cluster centers according to a formula (1) and the super pixel number k, calculating an interval S of the initialized super pixel cluster centers, and uniformly distributing the initialized cluster centers on the infrared image according to the interval S. In order to avoid that the clustering center falls on the image edge to enable the point to form a closed connected domain with surrounding areas, taking a gradient minimum value point in a 3x3 neighborhood range of an initial clustering center point as an initial clustering center;
where N is the number of total pixels of the image, k is the number of superpixels,
rounding down for reject fraction;
(3) Combining the infrared image brightness domain l and the spatial domain (x, y) into a three-dimensional spatial domain v= [ l, x, y] T D is defined as the distance between the pixel point and the cluster center in three-dimensional space. Calculating the distance D between the pixel point and the clustering center according to the formulas (2), (3) and (4), wherein the pixel point belongs to the minimum DAnd clustering the super pixel blocks in the center. The farther the cluster center is from the pixel point, the probability of the super pixel block belonging to the cluster center is rapidly reduced, and in order to increase the operation speed, the 2S multiplied by 2S range of the neighborhood of the pixel point is taken as the search domain. Updating the clustering center after all the image pixel points are clustered, taking the center point of each clustering block as a new clustering center, and obtaining a super-pixel map with accurate segmentation after four iterations;
wherein d is s For space dimension distance, i.e. coordinate difference, d l For the brightness dimension distance, i.e. the gray value difference, (x) i ,y i ) And (x) j ,y j ) Coordinates of the image pixel point and the clustering center point, l i And l j Respectively the gray values of the pixel points and the clustering center points;
(4) The atmospheric light value A is obtained by using an improved quadtree decomposition search algorithm, and the specific algorithm comprises the following steps of replacing a regular rectangular quartile method by a quartile method based on a super pixel block. Firstly, after the image super-pixel is segmented, each super-pixel block is numbered, so that each pixel point is marked with a corresponding super-pixel block number label. Then dividing the image into four equal parts, averaging the four parts, counting the label numbers of the contained pixels by the part with the largest average value, and forming a new sub-block by the pixels corresponding to the counted label numbers. The newly composed sub-block coordinate area is then divided into four equal parts. This process is repeated until only one tag number is contained in the block with the largest average value. And (5) taking an average value of the gray values of the pixel points of the super pixel block corresponding to the label number according to the formula (5) to obtain a global atmosphere light value A. In some cases, the iteration process falls into a local loop under a special condition that after a plurality of iterations, the number of super pixel blocks contained in the newly formed sub-blocks is less than or equal to four, so that when the new sub-blocks are divided into four equal parts, the label number appearing in the area with the largest statistical average value is equal to the label number of the new sub-blocks, so that the label number is consistent with the atomic block when the next-stage sub-blocks are constructed, and the local loop falls into the local loop, and the local loop can be taken as an iteration termination condition. At this time, since the number of super pixel blocks included in the sub-blocks is small, according to formula (5), calculating the average value of all the super pixel blocks in the sub-blocks, and selecting the maximum value as the global atmospheric light value A;
wherein I (x, y) represents a gray value at a coordinate point (x, y) in the super pixel block Ω, and n is a total number of pixels in the super pixel block Ω;
(5) Because of poor non-uniformity of the infrared image, in order to avoid that the maximum gray value point and the minimum gray value point of each super pixel block are just positioned at the blind pixels, the gray values of the pixel points of the super pixel blocks are ordered from small to large, and according to the formula (6) (7), the average value of the gray values of the front 5% and the rear 5% is calculated as the minimum maximum value of the super pixel blocks and is recorded as min_m and max_m:
wherein, Ω_min and Ω_max are respectively the sets of pixel points corresponding to the first 5% gray value and the last 5% gray value of the super pixel block Ω, and n is the number of total pixel points in the super pixel block;
(6) The Mean square Error MSE (Mean-square Error) is an effective parameter for measuring image quality and can be calculated by the formula (8)A higher value of MSE means that the image has a higher contrast, as calculated, to conform to the visual characteristics of the human eye. From equation (8), it can be deduced that the value of MSE is inversely related to the transmittance t, and a smaller t can obtain a higher value of MSE. For infrared images, a smaller t makes the gray value of the image pixel point more prone to generate data truncation phenomenon, thereby causing information loss, and defining the loss value caused by data truncation as C LOSS Namely, formula (9):
wherein the method comprises the steps of
Is the gray average value of the pixel points of the original image. To prevent information loss, C should be satisfied LOSS Less than or equal to zero. The value of the transmittance that does not cause data loss while making the MSE value larger is the optimal transmittance t'. Deriving the upper and lower limit constraints that can prevent information loss from equation (9) so that the value of the optimal transmittance of the super pixel block can be obtained from equation (10):
wherein ω is a constant for controlling the defogging degree, and is set to 0.95;
(7) The optimal atmospheric light value and the optimal transmissivity can be obtained through the steps (4) and (6), and the defogging reconstructed image J can be obtained through calculation according to an inversion formula (11) of an atmospheric physical model;
wherein J (x) and I (x) are pixel points of the reconstructed image and the original image respectively, and t' (x) is the optimal transmittance value of the super-pixel block corresponding to the point;
compared with the prior art, the invention has the beneficial effects that:
1) The defogging process is easy to realize, and the transmittance value can be automatically adjusted by self-adapting the distance degree of a target and the haze concentration without obtaining additional information beyond an infrared image and only needing single-frame information and multi-frame information;
2) Through super-pixel segmentation, halo phenomenon and blocking effect can be avoided, so that the imaging quality is more natural, and the human eye vision characteristic is more met.
3) The invention has less input parameters, only needs the input parameters in the segmentation stage, and the acquisition of the global atmosphere light value and the upper and lower limit constraint optimal transmissivity can be automatically carried out by an improved quadtree decomposition search algorithm, so that the defogging reconstruction process can be realized without human intervention.
Drawings
FIG. 1 is a block diagram of an implementation flow of the present invention;
fig. 2 is an original infrared image of a haze day as an input image in the present invention.
Fig. 3 is an output image of the present invention, i.e., defogging reconstructed image.
Detailed Description
The technical solutions in the embodiments of the present invention will be described in detail below with reference to the accompanying drawings in the embodiments of the present invention. Several parameters are involved which need to be adjusted for a specific processing environment to achieve good performance.
The test picture used in the invention is obtained by shooting a 640X 512 short wave infrared camera developed by Shanghai technology physical research institute of Chinese sciences.
Simulation environment: matlab2018b;
test image: short wave infrared image, size 640 x 512, scene is city building;
for the test image, the super pixel number k=60, the boundary compactness m=40, and the defogging degree control constant ω=0.95 in the defogging step (6) are set.
Claims (1)
1. The defogging and enhancing method for the infrared foggy-day degraded image is characterized by comprising the following steps of:
(1) Presetting the number k of super pixel blocks and the compactness m of the super pixel segmentation boundary in a simple linear iterative clustering super pixel segmentation algorithm, wherein the number k of the super pixel blocks is set to be 40-80, and the compactness m is set to be 30-50;
(2) Initializing a cluster center according to the number k of super pixel blocks, wherein the initial cluster center is uniformly distributed in an image by the size S of an interval grid, and in order to avoid that the initial cluster center is just positioned at the edge of the image, taking a pixel point with the minimum gradient value in the 3x3 neighborhood of the initial center point as the initial cluster center, wherein the size S of the grid is as follows:
wherein N is the number of total pixel points of the image, k is the number of super pixel blocks,
rounding down for reject fraction;
(3) Combining the one-dimensional brightness dimension l and the two-dimensional space dimension (x, y) of the infrared image into a three-dimensional space domain V= [ l, x, y ]] T Defining D as the distance between the pixel points and the clustering center in the three-dimensional space, classifying the pixel points into super pixel blocks for minimizing D, updating the clustering center after all the pixel points of the image are classified, and obtaining a super pixel map with accurate segmentation after four iterations, wherein in order to avoid overlarge calculated amount, the classifying search domain of the pixel points is limited in a range of 2S multiplied by 2S of a neighborhood around the pixel points:
wherein m is a compactness parameter, d s For space dimension distance, i.e. coordinate difference, d l For the brightness dimension distance, i.e. the gray value difference, (x) i ,y i ) And (x) j ,y j ) Coordinates of the image pixel point and the clustering center point, l i And l j Respectively the gray values of the pixel points and the clustering center points;
(4) Using an improved quad-tree decomposition search algorithm based on the super pixel block, obtaining a final super pixel block when the iteration termination condition is met, and calculating an average value re_m of the super pixel block to obtain an optimal global atmospheric light value A;
wherein I (x, y) represents a gray value at a coordinate point (x, y) in the super pixel block Ω, and n is a total number of pixels in the super pixel block Ω;
(5) The gray values corresponding to the pixel points in each super pixel block are arranged in a descending order, and the average value of the gray values of the first 5% and the last 5% in each block is calculated to be used as the minimum value and the maximum value of the super pixel block and respectively recorded as min_m and max_m:
wherein, Ω_min and Ω_max are respectively the sets of pixel points corresponding to the first 5% gray value and the last 5% gray value of the super pixel block Ω, and n is the number of total pixel points in the super pixel block;
(6) To obtain a higher mean square error value C MSE Namely, formula (8) and preventing information truncation from occurringC LOSS Namely, formula (9), the upper and lower limit constraints of the transmittance can be obtained; the optimal transmittance value t' of the super pixel block can be obtained according to an upper limit constraint formula (10) of the transmittance by utilizing the maximum and minimum average values max_m and min_m of the super pixel block:
wherein the method comprises the steps of
The gray average value of the pixel points of the original image is set to be 0.95, t is the transmissivity, omega is a constant for controlling defogging degree;
(7) Calculating an inversion formula (11) of the atmospheric physical model by using the obtained global atmospheric light value A and the optimal transmissivity t' of each super pixel block, and obtaining defogging reconstructed images:
wherein J (x) and I (x) are the gray values of the pixels of the reconstructed image and the original image, respectively, and t' (x) is the optimal transmittance value of the super-pixel block corresponding to the point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110100817.6A CN112907461B (en) | 2021-01-26 | 2021-01-26 | Defogging enhancement method for infrared foggy-day degraded image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110100817.6A CN112907461B (en) | 2021-01-26 | 2021-01-26 | Defogging enhancement method for infrared foggy-day degraded image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112907461A CN112907461A (en) | 2021-06-04 |
| CN112907461B true CN112907461B (en) | 2023-05-05 |
Family
ID=76119218
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110100817.6A Active CN112907461B (en) | 2021-01-26 | 2021-01-26 | Defogging enhancement method for infrared foggy-day degraded image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112907461B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114549336B (en) * | 2021-11-25 | 2024-05-03 | 湖南科技大学 | Unsupervised image defogging method |
| CN115063404B (en) * | 2022-07-27 | 2022-11-08 | 建首(山东)钢材加工有限公司 | Weathering resistant steel weld joint quality detection method based on X-ray flaw detection |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111325688A (en) * | 2020-02-18 | 2020-06-23 | 长安大学 | Unmanned aerial vehicle image defogging method fusing morphological clustering and optimizing atmospheric light |
| CN111667433A (en) * | 2020-06-09 | 2020-09-15 | 中国电子科技集团公司第五十四研究所 | Unmanned aerial vehicle image defogging method based on simple linear iterative clustering optimization |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104794688B (en) * | 2015-03-12 | 2018-04-03 | 北京航空航天大学 | Single image to the fog method and device based on depth information separation sky areas |
| CN108596849B (en) * | 2018-04-23 | 2021-11-23 | 南京邮电大学 | Single image defogging method based on sky region segmentation |
| CN110428371A (en) * | 2019-07-03 | 2019-11-08 | 深圳大学 | Image defogging method, system, storage medium and electronic equipment based on super-pixel segmentation |
| CN111598788B (en) * | 2020-04-08 | 2023-03-07 | 西安理工大学 | Single image defogging method based on quadtree decomposition and non-local prior |
| CN111899198A (en) * | 2020-08-06 | 2020-11-06 | 北京科技大学 | Defogging method and device for marine image |
-
2021
- 2021-01-26 CN CN202110100817.6A patent/CN112907461B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111325688A (en) * | 2020-02-18 | 2020-06-23 | 长安大学 | Unmanned aerial vehicle image defogging method fusing morphological clustering and optimizing atmospheric light |
| CN111667433A (en) * | 2020-06-09 | 2020-09-15 | 中国电子科技集团公司第五十四研究所 | Unmanned aerial vehicle image defogging method based on simple linear iterative clustering optimization |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112907461A (en) | 2021-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Fast image dehazing method based on linear transformation | |
| Zhu et al. | Haze removal method for natural restoration of images with sky | |
| CN106157267B (en) | Image defogging transmissivity optimization method based on dark channel prior | |
| CN111292258B (en) | Image defogging method based on dark channel prior and bright channel prior | |
| CN103020920B (en) | Method for enhancing low-illumination images | |
| Gao et al. | Sand-dust image restoration based on reversing the blue channel prior | |
| CN111667433B (en) | Unmanned aerial vehicle image defogging method based on simple linear iterative clustering optimization | |
| CN109685045B (en) | Moving target video tracking method and system | |
| CN111598791B (en) | Image defogging method based on improved dynamic atmospheric scattering coefficient function | |
| CN112907461B (en) | Defogging enhancement method for infrared foggy-day degraded image | |
| CN111145105B (en) | Image rapid defogging method and device, terminal and storage medium | |
| CN112950589A (en) | Dark channel prior defogging algorithm of multi-scale convolution neural network | |
| CN112419163B (en) | Single image weak supervision defogging method based on priori knowledge and deep learning | |
| Fu et al. | Scene-awareness based single image dehazing technique via automatic estimation of sky area | |
| Fu et al. | An anisotropic Gaussian filtering model for image de-hazing | |
| CN110349113B (en) | Adaptive image defogging method based on dark primary color priori improvement | |
| CN109345479B (en) | Real-time preprocessing method and storage medium for video monitoring data | |
| Zhen et al. | Single Image Defogging Algorithm based on Dark Channel Priority. | |
| Li et al. | DLT-Net: deep learning transmittance network for single image haze removal | |
| CN113822816A (en) | Haze removing method for single remote sensing image optimized by aerial fog scattering model | |
| CN117611501A (en) | Low-illumination image enhancement method, device, equipment and readable storage medium | |
| CN109360169B (en) | Signal processing method for removing rain and mist of single image | |
| CN115631108A (en) | RGBD-based image defogging method and related equipment | |
| CN112598777B (en) | Haze fusion method based on dark channel prior | |
| CN111932469A (en) | Significance weight quick exposure image fusion method, device, equipment and medium |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |