CN106296612A - Hierarchical monitoring video sharpening system and method for image quality evaluation and weather condition guidance - Google Patents
Hierarchical monitoring video sharpening system and method for image quality evaluation and weather condition guidance Download PDFInfo
- Publication number
- CN106296612A CN106296612A CN201610649883.8A CN201610649883A CN106296612A CN 106296612 A CN106296612 A CN 106296612A CN 201610649883 A CN201610649883 A CN 201610649883A CN 106296612 A CN106296612 A CN 106296612A
- Authority
- CN
- China
- Prior art keywords
- image
- module
- video
- mist elimination
- weather
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 63
- 238000013441 quality evaluation Methods 0.000 title claims abstract description 29
- 238000012544 monitoring process Methods 0.000 title claims abstract description 12
- 238000012545 processing Methods 0.000 claims abstract description 23
- 238000006243 chemical reaction Methods 0.000 claims abstract description 16
- 239000003595 mist Substances 0.000 claims description 127
- 238000003379 elimination reaction Methods 0.000 claims description 101
- 230000008030 elimination Effects 0.000 claims description 90
- 230000006835 compression Effects 0.000 claims description 20
- 238000007906 compression Methods 0.000 claims description 20
- 238000002835 absorbance Methods 0.000 claims description 18
- 230000001186 cumulative effect Effects 0.000 claims description 12
- 230000000007 visual effect Effects 0.000 claims description 12
- 238000001914 filtration Methods 0.000 claims description 10
- 238000001303 quality assessment method Methods 0.000 claims description 8
- 238000005457 optimization Methods 0.000 claims description 6
- 238000002834 transmittance Methods 0.000 claims description 6
- 238000001514 detection method Methods 0.000 claims description 4
- 230000011218 segmentation Effects 0.000 claims description 4
- 230000006978 adaptation Effects 0.000 claims description 3
- 230000005540 biological transmission Effects 0.000 claims description 3
- 230000000903 blocking effect Effects 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 3
- 238000011084 recovery Methods 0.000 claims description 3
- 238000005352 clarification Methods 0.000 abstract description 6
- 238000004458 analytical method Methods 0.000 abstract description 3
- 230000015556 catabolic process Effects 0.000 abstract description 3
- 238000006731 degradation reaction Methods 0.000 abstract description 3
- 238000000889 atomisation Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000003384 imaging method Methods 0.000 description 3
- 230000004087 circulation Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 230000002708 enhancing effect Effects 0.000 description 2
- 239000002245 particle Substances 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007850 degeneration Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 125000001475 halogen functional group Chemical group 0.000 description 1
- 238000003707 image sharpening Methods 0.000 description 1
- 230000001678 irradiating effect Effects 0.000 description 1
- 230000007257 malfunction Effects 0.000 description 1
- 230000011514 reflex Effects 0.000 description 1
- 238000009738 saturating Methods 0.000 description 1
- 230000001568 sexual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/73—Deblurring; Sharpening
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration using histogram techniques
-
- 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
- G06T2207/10016—Video; Image sequence
-
- 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
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a hierarchical monitoring video sharpening system and method for image quality evaluation and weather condition guidance, and belongs to the technical field of video processing and analysis. The video conversion system comprises a video acquisition module, a video conversion module, a video image processing module and a display module. The video acquisition module and the video conversion module acquire images and convert the images into digital image frames, the video image processing module integrates weather conditions and image quality evaluation, judges the defogging grade of the images and adopts corresponding algorithms to defogg the images of different grades. The invention has simple and intelligent structure and easy realization, can greatly improve the efficiency and quality of online video clarification processing of monitoring, can effectively avoid the limitation that one defogging algorithm cannot meet the degradation of different types of videos, and has wide application value in the fields of outdoor video monitoring and the like.
Description
Technical field
The present invention relates to Video processing and analysis technical field, more particularly, it relates to a kind of image quality evaluation and sky
The stagewise monitor video sharpening system and method that vaporous condition guides, the present invention is directed to different weather situation and according to image
Quality evaluation result, to visual classification classification, automatically selects corresponding algorithm for image clearness and carries out real-time video mist elimination.
Background technology
Due to reasons such as environmental pollution, regional climates, haze weather is the most increasingly common, this situation
City seems particularly evident.Under this kind of weather condition, the quality of monitor video can degradation.Dislike at mist, haze and rain etc.
The video photographed under bad weather conditions, is characterized in: contrast is low, narrow dynamic range, color are the distinctest and saturation is low, very
To producing color displacement.This degeneration brings many inconvenience to the information that people extract in image.Additionally, regard at computer
In feel field, many algorithms all based on picture rich in detail, such as feature extraction, target recognition, behavior analysis etc..And the greasy weather etc.
Under weather condition, the image of shooting causes great majority algorithm cisco unity malfunction based on computer vision.
In recent years, the scholar in the field such as Digital Image Processing, computer vision proposes a series of method for removing
The impact of weather in haze image, to reach the purpose of image sharpening.Fattal, He Kaiming, Tarel etc. propose respective mist elimination
Algorithm, has greatly promoted the development of single image mist elimination technology, and has made image mist elimination technology be applied to batch processing with in real time
Sexual system is possibly realized.
But present stage existing algorithm is some unsatisfactory place total during mist elimination, and the most triumphant bright algorithm is to dense
The process of mist part can manage it, but the enhancing of the contrast of image and minutia is not enough, and the speed of service is the most slow.
Chinese Patent Application No. 201510609664.2, filing date JIUYUE in 2015 23 days, invention and created name is: one
Plant digital picture defogging method based on histogram equalization；This application case discloses and a kind of utilizes traditional histogram equalization to calculate
Method carries out the scheme of digital picture mist elimination, and the reflection light of solution body surface is during arriving imaging device, due to air
The scattering of particle and decay；Natural light enters imaging device participate in the problems such as imaging because of atmospheric particles scattering.Process step
Suddenly it is: (1) obtains former atomization image；(2) it is HSI to former atomization image RGB model conversion, I component is analyzed image histogram；
(3) applicable size masterplate is set, I component is carried out histogram equalization enhancing, obtain the rectangular histogram of conversion atomization image.Tradition
Histogram equalization algorithm preferable at mist part effectiveness comparison, but when the concentration change of the degree of depth in figure or mist is bigger
Time, algorithm process effect is bad.
And for example Chinese Patent Application No. 201110282105.7, filing date JIUYUE in 2011 21 days, invention and created name
For: the image clarification method in foggy day of a kind of multi-scale Retinex model based on HIS space, the process step of this application case
For: one, gather source images, if source images is black white image, then it is converted to double type by byte type；If cromogram
Picture, then be converted to double type from the monochrome pixels value of tri-passages of R, G, B by byte type respectively by it.Two, by R, G, B tri-
Channel-shifted is to H, I, S；Three, the multiple dimensioned retinex (MSR) that the pixel being obtained step 2 improves respectively calculates
Method, obtains new image；Four, the image being obtained step 3 carries out linear contrast's broadening；Five, step 4 is obtained
H, I, S are transformed into R, G, B respectively；Six, R, G, the B being obtained step 5 synthesizes, the image after display sharpening.This Shen
Case broad image captured under the conditions of the greasy weather please be carried out sharpening process, the effective letter in greasy weather broad image can be recovered
Breath, but this application case uses the easy distortion of tone that multiple dimensioned Retinex algorithm processes, and has halo effect at edge portions.
Therefore, it is necessary to propose a kind of scheme that can carry out mist elimination for the image of different quality.
Summary of the invention
1. invention to solve the technical problem that
Quality for different weather situation video image is different, and general mist elimination algorithm cannot meet dissimilar video fall
The problems such as matter, the invention provides a kind of image quality evaluation and the stagewise monitor video sharpening system of weather conditions guiding
And method, the weather such as mist, rain can be descended the video photographed to carry out real-time sharpening and process by the present invention, simple to operate, easily
In realizing and to enable to the result that monitor video image processes the most more satisfactory every time.
2. technical scheme
For reaching above-mentioned purpose, the technical scheme that the present invention provides is:
The stagewise monitor video sharpening system that a kind of image quality evaluation of the present invention and weather conditions guide, including
Video acquisition module, video conversion module, Computer Vision module and display module；Wherein, video acquisition module is used for obtaining
Take the video image information of monitoring scene；Video conversion module receives the video signal that video acquisition module sends, by decoding
The video signal collected is converted into the form of digital image frames and flows to Computer Vision module by means；At video image
Reason module, by Implementation of Embedded System, starts image quality evaluation function and mist elimination function bad weather when automatically,
And image is compressed coding；Display module is for showing compressed encoded video.
Further, described Computer Vision module includes weather receiver module, system control module, figure picture element
Amount evaluation module, mist elimination module, image compression module, wherein:
Weather receiver module for receiving the weather condition on the same day according to short-range forecast, and tentatively judges mist elimination etc.
Level, concrete classification is as follows, 1 grade: mist；2 grades: middle mist, light rain, slight snow；3 grades: thick fog, rain；The weather feelings that picture quality is similar
Condition is classified as one-level；
System control module is connected with video acquisition module, comments for automatically starting picture quality bad weather when
Valency function and mist elimination function, control focal length and the orientation of CCTV camera simultaneously；
Image quality assessment module is for for the inaccurate situation of weather forecast information, by visual contrast VCM pair
The quality of image is estimated, if picture quality is preferable, it is determined that weather forecast is reported by mistake, and image is transmitted directly to compression of images
Module, is then forwarded to image compression module after otherwise carrying out mist elimination process；
Mist elimination module uses different mist elimination algorithms automatically according to mist elimination grade, and input picture is carried out mist elimination process；
Image compression module for becoming convenient transmission and the video codes of display to the digital image frames compressed encoding received
Stream.
The stagewise monitor video clarification method that a kind of image quality evaluation of the present invention and weather conditions guide, its step
Suddenly it is:
Step one, the video image information of video acquisition module acquisition monitoring scene, and through video conversion module by solving
Code means are converted into the form of digital image frames, flow to Computer Vision module；
Step 2, the weather receiver module of Computer Vision module receive the weather condition on the same day according to weather forecast；
Picture quality is estimated by image quality assessment module by visual contrast VCM, if picture quality is not up to mist elimination etc.
Level, it is determined that weather forecast is reported by mistake, and image is transmitted directly to image compression module；Otherwise it is sent to mist elimination module carry out at mist elimination
Reason；
Step 3, mist elimination module use different mist elimination algorithms automatically according to the mist elimination grade that step 2 is assessed, after mist elimination
Image be sent to image compression module；
Step 4, image compression module become video code flow to the digital image frames compressed encoding received, and flow to show
Show that module shows.
Further, by specifically comprising the following steps that picture quality is estimated by visual contrast VCM in step 2
(1) collection image being divided into some sub-blocks, sub-block size is 0.05 × min (H, W), the height of H representative image,
W representative image width；
(2) the local variance size of the most each sub-block is calculated；
(3) determining a threshold value by maximum variance between clusters, statistical variance accounts for total son more than the sub-block number of this threshold value
The ratio of block number, its formula is as follows:
VCM=100*Rv/Rt
Wherein, RvRepresent that variance exceedes the sub-block number of threshold value, RtRepresent the sub-block sum in single image.
Further, when step 2 setting VCM value as 0-10, it is 3 grades of thick fogs, rain image；When VCM value is 10-30,
It is mist in 2 grades, light rain, image in slight snow；When VCM value is 30-45, it it is 1 grade of mist image.
Further, in step 3 when mist elimination grade is 1 grade, use and limit the calculation of contrast self-adapting histogram equilibrium
Method carries out image procossing；When mist elimination grade is 2 grades, improvement MSRCR algorithm is used to carry out image procossing；When mist elimination grade is 3 grades,
Use and carry out image procossing based on edge optimization absorbance algorithm for estimating.
Further, in step 3, mist elimination grade is that image processing process when 1 grade is:
(1) first image is divided into the most nonoverlapping n sub-block；
(2) cumulative histogram of all sub-blocks in calculating input image:
In formula, M is the pixel number that sub-block comprises, and N is the gray level that sub-block is total, Hi,jN () is cumulative histogram, hi,j
K () is sub-block rectangular histogram；
(3) ask for cumulative histogram shear ultimate value:
Wherein, β is that rectangular histogram shears the limit, smaxFor cumulative histogram Hi,jN the greatest gradient of (), α is truncation function；
(4) the pixel rectangular histogram in sub-block is sheared and is redistributed:
Specifying that the number of pixels that each gray level comprises not can exceed that β, being sheared, if being sheared sum beyond part
For Nt, the number of pixels that average each gray level is assigned to is Aver=Nt/ N, redistributes as the following formula:
1)hi,j(k) > β, hnew=β
2)hi,j(k)+Aver > β, hnew=β
3)hi,j(k)+Aver < β, hnew=β+Aver
Wherein, hnewFor limiting contrast self-adapting histogram；
(5) to limiting contrast self-adapting histogram hnewCarry out histogram equalization process, the first each ash of statistic histogram
The number of times that degree level occurs, then add up normalized rectangular histogram, finally calculate new pixel value；
(6) image after equalization is carried out interpolation processing, it is thus achieved that final image.
Further, in step 3, mist elimination grade is that image processing process when 2 grades is:
(1) using the smothing filtering template R respectively to original color image, tri-passages of G, B carry out denoising；
(2) coloured image after denoising is carried out self adaptation overall situation brightness adjustment, adjusts formula as follows:
Wherein, S (x, y) is image R, the meansigma methods of G, B triple channel brightness value, S ' (x, y) is the image after brightness adjustment,
1/r=min (1,6Sal+ 2/3), SalFor image overall intensity meansigma methods of gray scale in log-domain；
(3) after processing, coloured image image pixel Value Types under R, tri-components of G, B is converted to double type, and
It is transformed into log-domain；
(4) choose basic, normal, high three different Gauss yardsticks and image is carried out convolution, obtain the irradiation under different scale
(x, y), (x y) is write as the Gauss convolution around function with corresponding passage light intensity, i.e. to component L will to irradiate component L
Log L (x, y)=log [Gk(x,y)*Si(x,y)]
(5) the reflecting component r of each passage is calculatedi(x, y), expression formula is as follows:
Wherein, Ci(x, y) is the color recovery factor, and β is gain coefficient, Si(x y) is i-th Color Channel light intensity, i table
Show port number, WkIt is weight coefficient, Gk(x, y) be the Gauss of different scale around function, α is non-linear controlled intensity；
(6) calculate average and the mean square deviation of each passage respectively, quantify according to formula, finally by R, G, B three-component
Merge, obtain final output image R (x, y)；Formula is as follows:
Mini=Meani-Dynamic*Vari
Maxi=meani+Dynamic*Vari
Ri(x, y)=(ri(x,y)-Mini)/Maxi-Mini*(255-0)
Wherein, Mini、Maxi、Meani、VariFor ri(x, y) minima of each passage, maximum, average and mean square deviation,
Dynamic is dynamic parameter.
Further, in step 3, mist elimination grade is that image processing process when 3 grades is:
(1) original image is carried out mini-value filtering and calculates air light value A；
(2) filtered minima image is carried out rim detection, extract the marginal zone that the depth of field in image changes greatly
Territory；
(3) marginal area is carried out Threshold segmentation, determine different depth of field regional boundary lines in image；
(4) the different depth of field regions distinguished are carried out absorbance estimation, concretely comprise the following steps: first judge that current pixel point is
No it is in marginal area, if current pixel point is around depth of field Sudden change region, then centered by just selecting current pixel point
The rectangular area of 15*15, for the pixel higher than threshold value in this region, selects the minima of the part that the depth of field is big as thoroughly
Radiance rate value, for the pixel less than threshold value, selects the minima of the part that the depth of field is little as transmittance values；If current pixel point
Not around depth of field Sudden change region, with regard to minima in use 15*15 block region as the transmittance values of current pixel point；
(5) Guided Filter is used to process to remove blocking effect to absorbance figure, it is thus achieved that final absorbance figure t
(x)；
(6) carry out restored image by absorbance scattergram t (x) and air light value A, restore formula as follows:
J (x)=[I (x)-A)]/t (x)+A
Wherein, J (x) is the picture rich in detail obtained, and I (x) is original mist elimination image.
Further, the calculating process of air light value A is: first, choose the brightest in the figure after mini-value filtering
0.1% pixel, then finds out the triple channel pixel value of input picture I corresponding to these pixels, chooses in these pixels the brightest
Pixel value as air light value A.
3 beneficial effects
For reaching above-mentioned purpose, the technical scheme that the present invention provides is:
(1) the stagewise monitor video sharpening system that a kind of image quality evaluation of the present invention and weather conditions guide,
Including video acquisition module, video conversion module, Computer Vision module and display module, automatic bad weather when
Start image quality evaluation function and mist elimination function, it is possible to the weather such as mist, rain descends the video photographed carry out real-time clear
Change processes, simple in construction and intelligence, it is easy to accomplish, hardware cost is low, and flexible configuration is convenient, at field tools such as life outdoor videos monitoring
Have a wide range of applications；
(2) the stagewise monitor video clarification method that a kind of image quality evaluation of the present invention and weather conditions guide,
By visual contrast VCM, picture quality is estimated, if picture quality is not up to mist elimination grade, it is determined that weather forecast
Wrong report, image is directly compressed coding output；Otherwise carry out mist elimination process, it is possible to avoid weather forecast to judge by accident, to need not
The image of mist elimination carries out mist elimination and makes image become worse；
(3) the stagewise monitor video clarification method that a kind of image quality evaluation of the present invention and weather conditions guide,
Automatically use different mist elimination algorithms according to mist elimination grade, when mist elimination grade is 1 grade, uses and limit contrast self-adapting histogram
Equalization algorithm；When mist elimination grade is 2 grades, uses and improve MSRCR algorithm；When mist elimination grade is 3 grades, use saturating based on edge optimization
Penetrate rate algorithm for estimating, it is possible to flexibly monitor video image is carried out according to weather condition mist elimination process, improve monitor video
Quality, effectively avoids a kind of mist elimination algorithm cannot meet the limitation that dissimilar video degrades simultaneously.
Accompanying drawing explanation
Fig. 1 is the structured flowchart of stagewise monitor video sharpening system in the present invention；
Fig. 2 is mist elimination image quality evaluation algorithm flow chart in the present invention；
Fig. 3 is that in the present invention, mist removes algorithm flow chart；
Fig. 4 is that in the present invention, middle mist removes algorithm flow chart；
Fig. 5 is that in the present invention, thick fog removes algorithm flow chart.
Detailed description of the invention
For further appreciating that present disclosure, the present invention is described in detail in conjunction with the accompanying drawings and embodiments.
Embodiment 1
The stagewise monitor video guided in conjunction with Fig. 1, a kind of image quality evaluation of the present embodiment and weather conditions is clear
Change system, including video acquisition module 1, video conversion module 2, Computer Vision module 3 and display module 4.Video acquisition
Module 1 is for obtaining the video image information of monitoring scene bad weather when, and converts mould by twisted-pair feeder with video
Block 2 connects, the video signal AV port output that video acquisition module 1 will get.
Video conversion module 2 receives the AV video signal that video acquisition module 1 sends, and will be collected by decoding means
Video signal is converted into form the input video processing unit of digital image frames (yuv format), is connect by special photographic head
Mouth is connected with Computer Vision module 3；It is real that this video conversion module 2 uses the low-power chip TVP5150AM1PBSR of TI to come
Now decode.
Computer Vision module 3, by Implementation of Embedded System, starts picture quality bad weather when automatically
Function of Evaluation and mist elimination function, by the image compression encoding after mist elimination.This Computer Vision module 3 includes weather receiver module
3-1, system control module 3-2, image quality assessment module 3-3, mist elimination module 3-4, image compression module 3-5.Wherein:
Described weather receiver module 3-1 for receiving the weather condition on the same day according to short-range forecast, and tentatively sentences
Disconnected mist elimination grade, concrete classification is as follows, 1 grade: mist；2 grades: middle mist, light rain, slight snow；3 grades: thick fog, rain；Picture quality is similar
Weather condition be classified as a class.
Described system control module 3-2 is connected with video acquisition module 1 by RS232 universal serial bus, at weather
Automatically start image quality evaluation function and mist elimination function when of bad, can also control simultaneously CCTV camera focal length and
The parameters such as orientation.
Described image quality assessment module 3-3, for the possible inaccurate situation of information for weather forecast, passes through
The quality of image is estimated by visual contrast (Visual Contrast Measure, VCM), if picture quality is preferable,
Then determining that weather forecast is reported by mistake, image is transmitted directly to compressed encoding module；Otherwise carry out mist elimination process, concrete steps such as Fig. 2
Shown in.
Described mist elimination module 3-4 uses different mist elimination algorithms automatically according to mist elimination grade, when mist elimination grade is 1 grade,
Use and limit contrast self-adapting histogram equilibrium algorithm；When mist elimination grade is 2 grades, uses and improve MSRCR algorithm；Mist elimination grade
When being 3 grades, using based on edge optimization absorbance algorithm for estimating, the image after mist elimination is sent to video image compression module 3-5.
Described image compression module 3-5 for becoming convenient transmission and display to the digital image frames compressed encoding received
Video code flow, it is simple to transmit and show.
Display module 4 is for showing compressed encoded video, and this display module 4 is by VGA line and Video processing list
Unit is connected.
The stagewise monitor video clarification method that a kind of image quality evaluation of the present embodiment and weather conditions guide, tool
Body process is as follows:
Step one, the video image information of video acquisition module 1 acquisition monitoring scene, and pass through through video conversion module 2
Decoding means are converted into yuv format, flow to Computer Vision module 3；
Step 2, the weather receiver module 3-1 of Computer Vision module 3 receive the weather feelings on the same day according to weather forecast
Condition；Picture quality is estimated by image quality assessment module 3-3 by visual contrast, if picture quality is not up to mist elimination
Grade, it is determined that weather forecast is reported by mistake, and image is transmitted directly to image compression module 3-5；Otherwise it is sent to mist elimination module 3-4 enter
Row mist elimination processes；
Step 3, mist elimination module 3-4 use different mist elimination algorithms, mist elimination automatically according to the mist elimination grade that step 2 is assessed
When grade is 1 grade, uses and limit contrast self-adapting histogram equilibrium algorithm；When mist elimination grade is 2 grades, uses and improve MSRCR
Algorithm；When mist elimination grade is 3 grades, using based on edge optimization absorbance algorithm for estimating, the image after mist elimination is sent to image pressure
Contracting module 3-5；
Step 4, image compression module 3-5 become video code flow to the digital image frames compressed encoding received, and flow to
Display module 4 shows.
Referring to Fig. 2, owing to having mist image relative to the contrast degradation without mist image under same scene, VCM can
To estimate the contrast of image, having mist scene image, its corresponding VCM is less；Clear without mist image under same scene
Visual effect is preferable, and its corresponding VCM value becomes big.When the present embodiment sets thick fog image (3 grades), VCM value is 0-10；In
Mist, during light rain (2 grades), VCM value is 10-30；During mist (1 grade), VCM value is 30-45.Image VCM value is used to judge to gather figure
Seem no to reach mist elimination standard, it is possible to avoid weather forecast to judge by accident, the image that need not mist elimination is carried out mist elimination image is become
Obtain worse.Detailed process is as follows:
(1) collection image being divided into some sub-blocks, the present embodiment neutron block size is that 0.05 × min (H, W), H represent
The height of image, W representative image width；
(2) the local variance size of the most each sub-block is calculated；
(3) statistical variance accounts for the ratio of total sub-block number more than the sub-block number of threshold value, and its formula is as follows:
VCM=100*Rv/Rt
Wherein, RvRepresent that variance exceedes the sub-block number of threshold value, RtRepresent the sub-block sum in single image.
Threshold value is determined by maximum variance between clusters (OSTU), determines that process is summarized as follows:
First initialize threshold value t, image is divided into A and B two class；Calculate between average and the class of A and B two class collection of pixels
Variance；By t from 0 to 255 circulations, when working as A, B inter-class variance maximum, corresponding variance D is threshold value.
(4) according to the image VCM value obtained, it is judged that gather image mist elimination to be carried out.
Referring to Fig. 3, in described mist elimination module 3-4, when picture quality is 1 grade, the present embodiment uses and limits contrast certainly
Mist image is processed by adaptive histogram equalization algorithm, and detailed process is as follows:
(1) first image is divided into the most nonoverlapping sub-block, the most each sub-block equal in magnitude or near
Patibhaga-nimitta etc.；
(2) cumulative histogram of all sub-blocks in calculating input image:If M is sub-block
The pixel number comprised, N is the gray level that sub-block is total, Hi,jN () is cumulative histogram, hi,jK () is sub-block rectangular histogram；
(3) ask for cumulative histogram shear ultimate value:
Wherein, β is that rectangular histogram shears the limit, smaxFor cumulative histogram Hi,jN the greatest gradient of (), its span is 1
～the integer between 4, α is truncation function, and its span is between 0～100.
(4) the pixel rectangular histogram in sub-block is sheared and is redistributed:
Specifying that the number of pixels that each gray level comprises not can exceed that β, being sheared, if being sheared sum beyond part
For Nt, the number of pixels that average each gray level is assigned to is Aver=Nt/ N, redistributes as the following formula:
1)hi,j(k) > β, hnew=β
2)hi,j(k)+Aver > β, hnew=β
3)hi,j(k)+Aver < β, hnew=β+Aver
Wherein: hnewFor limiting contrast self-adapting histogram；hi,jK () is the initial rectangular histogram of sub-block.
(5) carrying out histogram equalization process, detailed process is described as follows:
To hnewCarry out common histogram equalization to process, the number of times that the first each gray level of statistic histogram occurs, then tire out
Count normalized rectangular histogram, finally calculate new pixel value.
(6) image after equalization being carried out interpolation processing, it is thus achieved that final image, interpolation processing detailed process describes such as
Under:
To in sub-block pixel (x, y), calculates as follows:
L (i)=a [bl-(i)+(1-b)l+-(i)]+(1-a)[bl-+(i)+(1-b)l++(i)]
(x-,x-)、(x-,x+)、(x+,x+)、(x+,x-) it is respectively pixel (x, y) adjacent four module centers point coordinates；l-
(i)、l-+(i)、l+-(i)、l++It is respectively (x-,x-)、(x-,x+)、(x+,x+)、(x+,x-) gray value at place, l (i) be (x, y)
Gray value.
Referring to Fig. 4, in described mist elimination module 3-4, when picture quality is 2 grades, the present embodiment uses improvement band color extensive
Multiple multi-Scale Retinex Algorithm (MSRCR) centering mist, light rain image carries out mist elimination process, specifically comprises the following steps that
(1) using the smothing filtering template R respectively to original color image, tri-passages of G, B carry out denoising, smooth
Template is as follows:
(2) because the brightness making image overall after denoising is higher, the coloured image after denoising is carried out the self adaptation overall situation
Brightness adjustment, strengthens the proportion shared by relatively dark pixel in image, adjusts formula as follows:
Wherein, S (x, y) is image R, the meansigma methods of G, B triple channel brightness value, S ' (x, y) is the image after brightness adjustment,
1/r=min (1,6Sal+ 2/3), SalFor image the overall intensity meansigma methods of gray scale, S in log-domainalIt is worth the least, in image relatively
The level of stretch of dark-part is the biggest, and the proportion shared by relatively dark pixel is the highest.
(3) R of the coloured image after processing, the image pixel Value Types under tri-components of G, B is converted to double type,
And it is transformed into log-domain, so so that the plus and minus calculation that is transformed in log-domain of convolution algorithm in real number field；
(4) basic, normal, high three different Gauss yardsticks (value that the present embodiment is chosen is respectively 15,80,250) are chosen right
Image carries out convolution, obtain irradiation component L under different scale (x, y), simultaneously can (x y) be write as Gauss ring by irradiating component L
Around the convolution of function with corresponding passage light intensity, i.e.
Log L (x, y)=log [Gk(x,y)*Si(x,y)]
(5) the reflecting component r of each passage is calculated according to formulai(x, y), expression formula is as follows:
Wherein Ci(x, y) is the color recovery factor, and β is gain coefficient, Si(x is y) that i-th Color Channel inputs light intensity, i
Represent port number, WkIt is weight coefficient, Gk(x, y) be the Gauss of different scale around function, α is non-linear controlled intensity, and α takes
Value 123, when β takes 46, effect is best
(6) calculate average and the mean square deviation of each passage respectively, quantify according to formula, finally by R, G, B three-component
Merge, obtain final output image R (x, y).Formula is:
Mini=Meani-Dynamic*Vari
Maxi=Meani+Dynamic*Vari
Ri(x, y)=(ri(x,y)-Mini)/Maxi-Mini*(255-0)
Wherein, Mini、Maxi、Meani、VariFor ri(x, y) minima of each passage, maximum, average and mean square deviation,
Dynamic is dynamic parameter.
In conjunction with Fig. 5, in described mist elimination module 3-4, when picture quality is 3 grades, uses and estimate based on edge optimization absorbance
Thick fog image is processed by calculating method, specifically comprises the following steps that
(1) mini-value filtering template is used original image to carry out mini-value filtering and calculates air light value A, template size
For 15*15, the calculation procedure of air light value A is as follows: first, chooses the brightest 0.1% picture in the figure after mini-value filtering
Element, then finds out the triple channel pixel value of input picture I corresponding to these pixels, chooses pixel value the brightest in these pixels and makees
For air light value A.
(2) filtered minima figure carrying out rim detection, use Laplace operator to carry out rim detection, template is such as
Under:
(3) marginal area being carried out Threshold segmentation, determine different depth of field regional boundary lines in image, threshold value determines face one, front
Sample uses Ostu algorithm segmentation threshold.
First initialize threshold value t, image is divided into A and B two class；Calculate between average and the class of A and B two class collection of pixels
Variance；By t from 0 to 255 circulations, when working as A, B inter-class variance maximum, corresponding t is threshold value.
(4) the different depth of field regions distinguished are carried out absorbance estimation, concretely comprise the following steps: first judge that current pixel point is
No it is in marginal area, if current pixel point is around depth of field Sudden change region, then centered by just selecting current pixel point
The rectangular area of 15*15, for the pixel higher than threshold value in this region, selects the minima of the part that the depth of field is big as thoroughly
Radiance rate value, for the pixel less than threshold value, selects the minima of the part that the depth of field is little as transmittance values；If current pixel point
Not around depth of field Sudden change region, with regard to minima in use 15*15 block region as the transmittance values of current pixel point；
(5) Guided Filter (Steerable filter device) is used to process to remove blocking effect to absorbance figure, it is thus achieved that
Whole absorbance figure t (x).
(6) by absorbance scattergram and air light value reflex original image, formula is restored as follows:
J (x)=[I (x)-A)]/t (x)+A
Wherein, J (x) is the picture rich in detail obtained, and I (x) is original mist elimination image, and A is air light value.
The stagewise monitor video sharpening side that a kind of image quality evaluation of the present embodiment proposition and weather conditions guide
Method, it is possible to monitor video image is carried out mist elimination process according to weather condition, it is possible to improve the quality of monitor video greatly, with
Time can effectively avoid general mist elimination algorithm cannot meet the limitation that dissimilar video degrades.The present embodiment propose be
System simple in construction and intelligence, it is easy to accomplish, hardware cost is low, and flexible configuration is convenient, has extensively in fields such as life outdoor videos monitoring
General using value.
Schematically being described the present invention and embodiment thereof above, this description does not has restricted, institute in accompanying drawing
Show is also one of embodiments of the present invention, and actual structure is not limited thereto.So, if the common skill of this area
Art personnel enlightened by it, in the case of without departing from the invention objective, designs and this technical scheme without creative
Similar frame mode and embodiment, all should belong to protection scope of the present invention.
Claims (10)
1. the stagewise monitor video sharpening system that an image quality evaluation and weather conditions guide, it is characterised in that: bag
Include video acquisition module (1), video conversion module (2), Computer Vision module (3) and display module (4)；Wherein, video
Acquisition module (1) is for obtaining the video image information of monitoring scene；Video conversion module (2) receives video acquisition module (1)
The video signal sent, flows to video by decoding means by the form that the video signal collected is converted into digital image frames
Image processing module (3)；Computer Vision module (3) passes through Implementation of Embedded System, automatically opens bad weather when
Motion video quality evaluation function and mist elimination function, and image is compressed coding；Display module (4) is for regarding compressed encoding
Frequency shows.
The stagewise monitor video sharpening that a kind of image quality evaluation the most according to claim 1 and weather conditions guide
System, it is characterised in that: described Computer Vision module (3) includes weather receiver module (3-1), system control module
(3-2), image quality assessment module (3-3), mist elimination module (3-4), image compression module (3-5), wherein:
Weather receiver module (3-1) for receiving the weather condition on the same day according to short-range forecast, and tentatively judges mist elimination etc.
Level, concrete classification is as follows, 1 grade: mist；2 grades: middle mist, light rain, slight snow；3 grades: thick fog, rain；The weather feelings that picture quality is similar
Condition is classified as one-level；
System control module (3-2) is connected with video acquisition module (1), for automatically starting figure picture element bad weather when
Amount Function of Evaluation and mist elimination function, control focal length and the orientation of CCTV camera simultaneously；
Image quality assessment module (3-3) is for for the inaccurate situation of weather forecast information, by visual contrast VCM pair
The quality of image is estimated, if picture quality is preferable, it is determined that weather forecast is reported by mistake, and image is transmitted directly to compression of images
Module (3-5), is then forwarded to image compression module (3-5) after otherwise carrying out mist elimination process；
Mist elimination module (3-4) uses different mist elimination algorithms automatically according to mist elimination grade, and input picture is carried out mist elimination process；
Image compression module (3-5) for becoming convenient transmission and the video codes of display to the digital image frames compressed encoding received
Stream.
3. the stagewise monitor video sharpening system that a kind utilizes described in claim 2 carries out the side of video sharpening process
Method, the steps include:
Step one, the video image information of video acquisition module (1) acquisition monitoring scene, and pass through through video conversion module (2)
Decoding means are converted into the form of digital image frames, flow to Computer Vision module (3)；
Step 2, the weather receiver module (3-1) of Computer Vision module (3) receive the weather feelings on the same day according to weather forecast
Condition；Picture quality is estimated by image quality assessment module (3-3) by visual contrast VCM, if picture quality does not reaches
To mist elimination grade, it is determined that weather forecast is reported by mistake, and image is transmitted directly to image compression module (3-5)；Otherwise it is sent to mist elimination
Module (3-4) carries out mist elimination process；
Step 3, mist elimination module (3-4) use different mist elimination algorithms automatically according to the mist elimination grade that step 2 is assessed, after mist elimination
Image be sent to image compression module (3-5)；
Step 4, image compression module (3-5) become video code flow to the digital image frames compressed encoding received, and flow to show
Show that module (4) shows.
The stagewise monitor video sharpening that a kind of image quality evaluation the most according to claim 3 and weather conditions guide
Method, it is characterised in that: by specifically comprising the following steps that picture quality is estimated by visual contrast VCM in step 2
(1) collection image being divided into some sub-blocks, sub-block size is 0.05 × min (H, W), the height of H representative image, W generation
Table picture traverse；
(2) the local variance size of the most each sub-block is calculated；
(3) determining a threshold value by maximum variance between clusters, statistical variance accounts for total sub-block more than the sub-block number of this threshold value
The ratio of number, its formula is as follows:
VCM=100*Rv/Rt
Wherein, RvRepresent that variance exceedes the sub-block number of threshold value, RtRepresent the sub-block sum in single image.
The stagewise monitor video sharpening that a kind of image quality evaluation the most according to claim 4 and weather conditions guide
Method, it is characterised in that: when step 2 setting VCM value as 0-10, it is 3 grades of thick fogs, rain image；When VCM value is 10-30, it is 2
Mist, light rain, image in slight snow in Ji；When VCM value is 30-45, it it is 1 grade of mist image.
The stagewise monitor video sharpening that a kind of image quality evaluation the most according to claim 5 and weather conditions guide
Method, it is characterised in that: in step 3 when mist elimination grade is 1 grade, use and limit contrast self-adapting histogram equilibrium algorithm
Carry out image procossing；When mist elimination grade is 2 grades, improvement MSRCR algorithm is used to carry out image procossing；When mist elimination grade is 3 grades, adopt
Image procossing is carried out with based on edge optimization absorbance algorithm for estimating.
7. the stagewise monitor video guided according to a kind of image quality evaluation described in claim 5 or 6 and weather conditions is clear
Clearization method, it is characterised in that: in step 3, mist elimination grade is that image processing process when 1 grade is:
(1) first image is divided into the most nonoverlapping n sub-block；
(2) cumulative histogram of all sub-blocks in calculating input image:
In formula, M is the pixel number that sub-block comprises, and N is the gray level that sub-block is total, Hi,jN () is cumulative histogram, hi,j(k)
For sub-block rectangular histogram；
(3) ask for cumulative histogram shear ultimate value:
Wherein, β is that rectangular histogram shears the limit, smaxFor cumulative histogram Hi,jN the greatest gradient of (), α is truncation function；
(4) the pixel rectangular histogram in sub-block is sheared and is redistributed:
Specify that the number of pixels that each gray level comprises not can exceed that β, be sheared beyond part, if being sheared sum is Nt,
The number of pixels that average each gray level is assigned to is Aver=Nt/ N, redistributes as the following formula:
1)hi,j(k) > β, hnew=β
2)hi,j(k)+Aver > β, hnew=β
3)hI, j(k)+Aver < β, hnew=β+Aver
Wherein, hnewFor limiting contrast self-adapting histogram；
(5) to limiting contrast self-adapting histogram hnewCarry out histogram equalization process, the first each gray level of statistic histogram
The number of times occurred, then add up normalized rectangular histogram, finally calculate new pixel value；
(6) image after equalization is carried out interpolation processing, it is thus achieved that final image.
8. the stagewise monitor video guided according to a kind of image quality evaluation described in claim 5 or 6 and weather conditions is clear
Clearization method, it is characterised in that: in step 3, mist elimination grade is that image processing process when 2 grades is:
(1) using the smothing filtering template R respectively to original color image, tri-passages of G, B carry out denoising；
(2) coloured image after denoising is carried out self adaptation overall situation brightness adjustment, adjusts formula as follows:
Wherein, (x, y) is image R to S, the meansigma methods of G, B triple channel brightness value, and (x y) is the image after brightness adjustment, 1/r to S '
=min (1,6Sal+ 2/3), SalFor image overall intensity meansigma methods of gray scale in log-domain；
(3) after processing, coloured image image pixel Value Types under R, tri-components of G, B is converted to double type, and changes
To log-domain；
(4) choose basic, normal, high three different Gauss yardsticks and image is carried out convolution, obtain the irradiation component L under different scale
(x, y), (x y) is write as the Gauss convolution around function with corresponding passage light intensity, i.e. will to irradiate component L
Log L (x, y)=log [Gk(x,y)*Si(x,y)]
(5) the reflecting component r of each passage is calculatedi(x, y), expression formula is as follows:
Wherein, Ci(x, y) is the color recovery factor, and β is gain coefficient, Si(x, y) is i-th Color Channel light intensity, and i represents logical
Number of channels, WkIt is weight coefficient, Gk(x, y) be the Gauss of different scale around function, α is non-linear controlled intensity；
(6) calculate gray value average and the mean square deviation of each passage respectively, quantify according to formula, finally by R, G, B tri-points
Amount merges, obtain final output image R (x, y)；Formula is as follows:
Mini=Meani-Dynamic*Vari
Maxi=Meani+Dynamic*Vari
Ri(x, y)=(ri(x,y)-Mini)/Maxi-Mini*(255-0)
Wherein, Mini、Maxi、Meani、VariFor ri(x, y) minima of each passage, maximum, average and mean square deviation,
Dynamic is dynamic parameter.
9. the stagewise monitor video guided according to a kind of image quality evaluation described in claim 5 or 6 and weather conditions is clear
Clearization method, it is characterised in that: in step 3, mist elimination grade is that image processing process when 3 grades is:
(1) original image is carried out mini-value filtering and calculates air light value A；
(2) filtered minima image is carried out rim detection, extract the marginal area that the depth of field in image changes greatly；
(3) marginal area is carried out Threshold segmentation, determine different depth of field regional boundary lines in image；
(4) the different depth of field regions distinguished are carried out absorbance estimation, concretely comprise the following steps: first judge whether current pixel point is located
In marginal area, if current pixel point is around depth of field Sudden change region, then just select the 15*15 centered by current pixel point
Rectangular area, for the pixel higher than threshold value in this region, select the minima of the big part of the depth of field as absorbance
Value, for the pixel less than threshold value, selects the minima of the part that the depth of field is little as transmittance values；If current pixel point does not exists
Around depth of field Sudden change region, with regard to minima in use 15*15 block region as the transmittance values of current pixel point；
(5) Guided Filter is used to process to remove blocking effect to absorbance figure, it is thus achieved that final absorbance figure t (x)；
(6) carry out restored image by absorbance scattergram t (x) and air light value A, restore formula as follows:
J (x)=[I (x)-A)]/t (x)+A
Wherein, J (x) is the picture rich in detail obtained, and I (x) is original mist elimination image.
The stagewise monitor video that a kind of image quality evaluation the most according to claim 9 and weather conditions guide is clear
Change method, it is characterised in that: the calculating process of air light value A is: first, choose the brightest in the figure after mini-value filtering
0.1% pixel, then finds out the triple channel pixel value of input picture I corresponding to these pixels, chooses in these pixels the brightest
Pixel value as air light value A.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610649883.8A CN106296612B (en) | 2016-08-09 | 2016-08-09 | Hierarchical monitoring video sharpening system and method for image quality evaluation and weather condition guidance |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610649883.8A CN106296612B (en) | 2016-08-09 | 2016-08-09 | Hierarchical monitoring video sharpening system and method for image quality evaluation and weather condition guidance |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106296612A true CN106296612A (en) | 2017-01-04 |
| CN106296612B CN106296612B (en) | 2019-07-02 |
Family
ID=57667565
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610649883.8A Expired - Fee Related CN106296612B (en) | 2016-08-09 | 2016-08-09 | Hierarchical monitoring video sharpening system and method for image quality evaluation and weather condition guidance |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106296612B (en) |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106846274A (en) * | 2017-01-23 | 2017-06-13 | 重庆师范大学 | Haze method is gone with reference to the MSRCR traffic images of HE and guiding filtering |
| CN107657166A (en) * | 2017-09-27 | 2018-02-02 | 广东欧珀移动通信有限公司 | Solve lock control method and Related product |
| CN107767353A (en) * | 2017-12-04 | 2018-03-06 | 河南工业大学 | A kind of adapting to image defogging method based on definition evaluation |
| CN107845078A (en) * | 2017-11-07 | 2018-03-27 | 北京航空航天大学 | A kind of unmanned plane image multithreading clarification method of metadata auxiliary |
| CN107943083A (en) * | 2017-12-11 | 2018-04-20 | 梁金凤 | A kind of flight system of precise control |
| CN108294937A (en) * | 2017-11-28 | 2018-07-20 | 王洪燕 | The anti-injury cleaning platform of nasal cavity |
| CN108734189A (en) * | 2017-04-20 | 2018-11-02 | 天津工业大学 | Vehicle License Plate Recognition System based on atmospherical scattering model and deep learning under thick fog weather |
| CN108986046A (en) * | 2018-07-09 | 2018-12-11 | 西安理工大学 | A kind of traveling monitoring image defogging method |
| CN108986049A (en) * | 2018-07-20 | 2018-12-11 | 百度在线网络技术（北京）有限公司 | Method and apparatus for handling image |
| CN109584186A (en) * | 2018-12-25 | 2019-04-05 | 西北工业大学 | A kind of unmanned aerial vehicle onboard image defogging method and device |
| CN109889817A (en) * | 2019-02-26 | 2019-06-14 | 北京华夏电通科技有限公司 | Video quality detection system and method applied to court |
| CN109886920A (en) * | 2019-01-16 | 2019-06-14 | 安徽谛听信息科技有限公司 | A kind of greasy weather stage division, greasy weather hierarchy system |
| CN110033419A (en) * | 2019-04-17 | 2019-07-19 | 山东超越数控电子股份有限公司 | A kind of processing method being adapted to warship basic image defogging |
| CN110072107A (en) * | 2019-04-25 | 2019-07-30 | 南京理工大学 | A kind of haze video-frequency compression method shared based on estimation |
| CN110097523A (en) * | 2019-05-30 | 2019-08-06 | 大连海事大学 | A kind of video image fog concentration classification and adaptive defogging method |
| CN110136210A (en) * | 2018-02-07 | 2019-08-16 | 中国石油化工股份有限公司 | A kind of intelligent early-warning system of monitoring device leakage |
| CN110136089A (en) * | 2019-05-23 | 2019-08-16 | 大连理工大学 | A kind of human embryos cardiac ultrasound images Enhancement Method |
| CN110188680A (en) * | 2019-05-29 | 2019-08-30 | 南京林业大学 | Tea tree tender shoots intelligent identification Method based on factor iteration |
| CN111353332A (en) * | 2018-12-21 | 2020-06-30 | 中国电信股份有限公司 | Fingerprint image processing method and device and computer readable storage medium |
| CN111369634A (en) * | 2020-03-26 | 2020-07-03 | 苏州瑞立思科技有限公司 | Image compression method and device based on weather conditions |
| CN111461088A (en) * | 2020-06-17 | 2020-07-28 | 长沙超创电子科技有限公司 | Rail transit obstacle avoidance system based on image processing and target recognition |
| CN112561906A (en) * | 2020-12-24 | 2021-03-26 | 百果园技术(新加坡)有限公司 | Image processing method, device, equipment and medium |
| CN112668613A (en) * | 2020-12-07 | 2021-04-16 | 中国西安卫星测控中心 | Satellite infrared imaging effect prediction method based on weather forecast and machine learning |
| CN113014773A (en) * | 2021-03-02 | 2021-06-22 | 山东鲁能软件技术有限公司智能电气分公司 | Overhead line video visual monitoring system and method |
| CN113628133A (en) * | 2021-07-28 | 2021-11-09 | 武汉三江中电科技有限责任公司 | Rain and fog removing method and device based on video image |
| CN114494049A (en) * | 2022-01-12 | 2022-05-13 | 宁夏回族自治区大气探测技术保障中心 | Weather radar running state image recognition algorithm |
| CN116630447A (en) * | 2023-07-24 | 2023-08-22 | 成都海风锐智科技有限责任公司 | Weather prediction method based on image processing |
| CN117423113A (en) * | 2023-12-18 | 2024-01-19 | 青岛华正信息技术股份有限公司 | Adaptive denoising method for archive OCR (optical character recognition) image |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103164845A (en) * | 2011-12-16 | 2013-06-19 | 中国科学院沈阳自动化研究所 | Real-time image mist elimination device and method |
| US20140140619A1 (en) * | 2011-08-03 | 2014-05-22 | Sudipta Mukhopadhyay | Method and System for Removal of Fog, Mist, or Haze from Images and Videos |
| CN104506755A (en) * | 2015-01-13 | 2015-04-08 | 武汉烽火众智数字技术有限责任公司 | Method for real-time automatic defogging of high-definition videos based on FPGA |
| CN105678735A (en) * | 2015-10-13 | 2016-06-15 | 中国人民解放军陆军军官学院 | Target salience detection method for fog images |
-
2016
- 2016-08-09 CN CN201610649883.8A patent/CN106296612B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140140619A1 (en) * | 2011-08-03 | 2014-05-22 | Sudipta Mukhopadhyay | Method and System for Removal of Fog, Mist, or Haze from Images and Videos |
| CN103164845A (en) * | 2011-12-16 | 2013-06-19 | 中国科学院沈阳自动化研究所 | Real-time image mist elimination device and method |
| CN104506755A (en) * | 2015-01-13 | 2015-04-08 | 武汉烽火众智数字技术有限责任公司 | Method for real-time automatic defogging of high-definition videos based on FPGA |
| CN105678735A (en) * | 2015-10-13 | 2016-06-15 | 中国人民解放军陆军军官学院 | Target salience detection method for fog images |
Non-Patent Citations (2)
| Title |
|---|
| SRINIVASA G.NARASIMHAN等: "Removing weather effects from monochrome images", 《PROCEEDINGS OF THE 2001 IEEE COMPUTER SOCIETY CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION》 * |
| 宋洪军等: "基于交通视频的雾天检测与去雾方法研究", 《控制工程》 * |
Cited By (37)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106846274A (en) * | 2017-01-23 | 2017-06-13 | 重庆师范大学 | Haze method is gone with reference to the MSRCR traffic images of HE and guiding filtering |
| CN108734189A (en) * | 2017-04-20 | 2018-11-02 | 天津工业大学 | Vehicle License Plate Recognition System based on atmospherical scattering model and deep learning under thick fog weather |
| CN107657166A (en) * | 2017-09-27 | 2018-02-02 | 广东欧珀移动通信有限公司 | Solve lock control method and Related product |
| CN107845078A (en) * | 2017-11-07 | 2018-03-27 | 北京航空航天大学 | A kind of unmanned plane image multithreading clarification method of metadata auxiliary |
| CN107845078B (en) * | 2017-11-07 | 2020-04-14 | 北京航空航天大学 | Unmanned aerial vehicle image multithreading sharpening method assisted by metadata |
| CN108294937A (en) * | 2017-11-28 | 2018-07-20 | 王洪燕 | The anti-injury cleaning platform of nasal cavity |
| CN107767353A (en) * | 2017-12-04 | 2018-03-06 | 河南工业大学 | A kind of adapting to image defogging method based on definition evaluation |
| CN107943083A (en) * | 2017-12-11 | 2018-04-20 | 梁金凤 | A kind of flight system of precise control |
| CN110136210A (en) * | 2018-02-07 | 2019-08-16 | 中国石油化工股份有限公司 | A kind of intelligent early-warning system of monitoring device leakage |
| CN110136210B (en) * | 2018-02-07 | 2023-06-09 | 中国石油化工股份有限公司 | Intelligent early warning system for leakage of monitoring device |
| CN108986046B (en) * | 2018-07-09 | 2021-12-21 | 西安理工大学 | Driving monitoring image defogging method |
| CN108986046A (en) * | 2018-07-09 | 2018-12-11 | 西安理工大学 | A kind of traveling monitoring image defogging method |
| CN108986049A (en) * | 2018-07-20 | 2018-12-11 | 百度在线网络技术（北京）有限公司 | Method and apparatus for handling image |
| CN111353332A (en) * | 2018-12-21 | 2020-06-30 | 中国电信股份有限公司 | Fingerprint image processing method and device and computer readable storage medium |
| CN111353332B (en) * | 2018-12-21 | 2023-06-02 | 中国电信股份有限公司 | Fingerprint image processing method, fingerprint image processing device and computer readable storage medium |
| CN109584186A (en) * | 2018-12-25 | 2019-04-05 | 西北工业大学 | A kind of unmanned aerial vehicle onboard image defogging method and device |
| CN109886920A (en) * | 2019-01-16 | 2019-06-14 | 安徽谛听信息科技有限公司 | A kind of greasy weather stage division, greasy weather hierarchy system |
| CN109889817A (en) * | 2019-02-26 | 2019-06-14 | 北京华夏电通科技有限公司 | Video quality detection system and method applied to court |
| CN110033419A (en) * | 2019-04-17 | 2019-07-19 | 山东超越数控电子股份有限公司 | A kind of processing method being adapted to warship basic image defogging |
| CN110072107A (en) * | 2019-04-25 | 2019-07-30 | 南京理工大学 | A kind of haze video-frequency compression method shared based on estimation |
| CN110136089A (en) * | 2019-05-23 | 2019-08-16 | 大连理工大学 | A kind of human embryos cardiac ultrasound images Enhancement Method |
| CN110188680B (en) * | 2019-05-29 | 2021-08-24 | 南京林业大学 | Tea tree tender shoot intelligent identification method based on factor iteration |
| CN110188680A (en) * | 2019-05-29 | 2019-08-30 | 南京林业大学 | Tea tree tender shoots intelligent identification Method based on factor iteration |
| CN110097523B (en) * | 2019-05-30 | 2022-10-18 | 大连海事大学 | Video image fog concentration classification and self-adaptive defogging method |
| CN110097523A (en) * | 2019-05-30 | 2019-08-06 | 大连海事大学 | A kind of video image fog concentration classification and adaptive defogging method |
| CN111369634A (en) * | 2020-03-26 | 2020-07-03 | 苏州瑞立思科技有限公司 | Image compression method and device based on weather conditions |
| CN111461088A (en) * | 2020-06-17 | 2020-07-28 | 长沙超创电子科技有限公司 | Rail transit obstacle avoidance system based on image processing and target recognition |
| CN111461088B (en) * | 2020-06-17 | 2020-09-08 | 长沙超创电子科技有限公司 | Rail transit obstacle avoidance system based on image processing and target recognition |
| CN112668613A (en) * | 2020-12-07 | 2021-04-16 | 中国西安卫星测控中心 | Satellite infrared imaging effect prediction method based on weather forecast and machine learning |
| CN112561906A (en) * | 2020-12-24 | 2021-03-26 | 百果园技术(新加坡)有限公司 | Image processing method, device, equipment and medium |
| CN113014773A (en) * | 2021-03-02 | 2021-06-22 | 山东鲁能软件技术有限公司智能电气分公司 | Overhead line video visual monitoring system and method |
| CN113628133A (en) * | 2021-07-28 | 2021-11-09 | 武汉三江中电科技有限责任公司 | Rain and fog removing method and device based on video image |
| CN114494049A (en) * | 2022-01-12 | 2022-05-13 | 宁夏回族自治区大气探测技术保障中心 | Weather radar running state image recognition algorithm |
| CN116630447A (en) * | 2023-07-24 | 2023-08-22 | 成都海风锐智科技有限责任公司 | Weather prediction method based on image processing |
| CN116630447B (en) * | 2023-07-24 | 2023-10-20 | 成都海风锐智科技有限责任公司 | Weather prediction method based on image processing |
| CN117423113A (en) * | 2023-12-18 | 2024-01-19 | 青岛华正信息技术股份有限公司 | Adaptive denoising method for archive OCR (optical character recognition) image |
| CN117423113B (en) * | 2023-12-18 | 2024-03-05 | 青岛华正信息技术股份有限公司 | Adaptive denoising method for archive OCR (optical character recognition) image |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106296612B (en) | 2019-07-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106296612A (en) | Hierarchical monitoring video sharpening system and method for image quality evaluation and weather condition guidance | |
| CN106530246A (en) | Image dehazing method and system based on dark channel and non-local prior | |
| CN110378849B (en) | Image defogging and rain removing method based on depth residual error network | |
| CN108389175A (en) | Merge the image defogging method of variogram and color decaying priori | |
| CN110148093B (en) | Image defogging improvement method based on dark channel prior | |
| CN108665417B (en) | License plate image deblurring method, device and system | |
| CN108765336A (en) | Image defogging method based on dark bright primary colors priori with auto-adaptive parameter optimization | |
| CN109993804A (en) | A kind of road scene defogging method generating confrontation network based on condition | |
| CN105976337B (en) | A kind of image defogging method based on intermediate value guiding filtering | |
| CN103996178A (en) | Sand and dust weather color image enhancing method | |
| CN109255759A (en) | Image defogging method based on sky segmentation and transmissivity adaptive correction | |
| CN106875351A (en) | A kind of defogging method towards large area sky areas image | |
| CN111709888B (en) | Aerial image defogging method based on improved generation countermeasure network | |
| CN111598791B (en) | Image defogging method based on improved dynamic atmospheric scattering coefficient function | |
| CN104253930A (en) | Real-time video defogging method | |
| CN106296618A (en) | A kind of color image defogging method based on Gaussian function weighted histogram regulation | |
| CN111325688B (en) | Unmanned aerial vehicle image defogging method for optimizing atmosphere light by fusion morphology clustering | |
| CN107316278A (en) | A kind of underwater picture clearness processing method | |
| CN103049765A (en) | Method for judging crowd density and number of people based on fish eye camera | |
| CN109118450B (en) | Low-quality image enhancement method under sand weather condition | |
| CN102855617A (en) | Method and system for processing adaptive images | |
| CN109816610A (en) | A kind of image defogging system | |
| CN104717400A (en) | Real-time defogging method of monitoring video | |
| CN109359593A (en) | A kind of sleet environment fuzzy pictures monitoring and pre-alarming method based on image local grid | |
| CN104253994B (en) | A kind of night monitoring video real time enhancing method merged based on sparse coding |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20170612 Address after: 210168 Jiangsu, Nanjing, Pu Pu Road, Pukou District No. 30 Applicant after: Nanjing Tech University Applicant after: Taizhou Yuandong hi tech Automation Engineering Co.,Ltd. Address before: 210168 Jiangsu, Nanjing, Pu Pu Road, Pukou District No. 30 Applicant before: Nanjing Tech University |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190702 |