CN110503634A - Visibility measurement method based on image automatic identification - Google Patents
Visibility measurement method based on image automatic identification Download PDFInfo
- Publication number
- CN110503634A CN110503634A CN201910690633.2A CN201910690633A CN110503634A CN 110503634 A CN110503634 A CN 110503634A CN 201910690633 A CN201910690633 A CN 201910690633A CN 110503634 A CN110503634 A CN 110503634A
- Authority
- CN
- China
- Prior art keywords
- image
- visibility
- formula
- mass center
- puppet
- 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
- 238000000691 measurement method Methods 0.000 title claims abstract description 19
- 238000000034 method Methods 0.000 claims abstract description 40
- 238000012360 testing method Methods 0.000 claims abstract description 27
- 230000011218 segmentation Effects 0.000 claims abstract description 25
- 230000000694 effects Effects 0.000 claims abstract description 24
- 239000000523 sample Substances 0.000 claims abstract description 10
- 229910003460 diamond Inorganic materials 0.000 claims abstract description 7
- 239000010432 diamond Substances 0.000 claims abstract description 7
- 230000008033 biological extinction Effects 0.000 claims description 9
- 238000012935 Averaging Methods 0.000 claims description 6
- 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 description 3
- 230000010339 dilation Effects 0.000 claims description 3
- 235000019988 mead Nutrition 0.000 claims description 3
- 238000013519 translation Methods 0.000 claims description 3
- 230000001143 conditioned effect Effects 0.000 claims description 2
- 230000011514 reflex Effects 0.000 claims description 2
- 238000012417 linear regression Methods 0.000 description 8
- 238000004458 analytical method Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 2
- 238000004064 recycling Methods 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002902 bimodal effect Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000002356 laser light scattering Methods 0.000 description 1
- 239000003550 marker Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G06T5/73—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/155—Segmentation; Edge detection involving morphological operators
Abstract
The present invention discloses the visibility measurement method based on image automatic identification, acquires original image and test image containing target；Maximum kind variance method is respectively adopted, Threshold segmentation is carried out to original image, test image；The test image after Threshold segmentation is continuously expanded using diamond shape probe, obtains expanding image；It is counter to expanding image to be coloured, using centroid principle, " puppet " the mass center calibration of more connected regions after searching continuous expansion；By " puppet " mass center as the quasi- center for selecting multiple target region, multiple target region to be selected is marked；In the dark effect picture of original image after singulation, find the maximum central point of " puppet " mass center light intensity in corresponding label multiple target region to be selected, the final goal region that region where the central point is calculated as visibility, selection suitably remove fog coefficient ω, carry out the calculating of visibility.The image for the different scenes that can be shot to picture pick-up device, selection is most suitable to go fog coefficient, calculates visibility.
Description
Technical field
The invention belongs to technical field of image processing, and in particular to the visibility measurement method based on image automatic identification.
Background technique
With the continuous propulsion of reform and opening-up process, economy rapid development and urban population increase, and consequent is energy
The consumption in source rapidly increases, and atmosphere pollution is caused to aggravate, and visibility level decline is brought to daily life etc.
Serious influence, especially aerospace, in terms of it is very harmful.Though in nationwide by every behave into
Row control, but difficulty is larger, still can see in recent years, various regions haze weather takes place frequently.
There are three types of common visibility detecting methods: artificial ocular estimate, device survey method and digital camera method.Ocular estimate error
Greatly, it is affected by observer personal considerations；Device survey method is expensive, and site requirements is high, cannot be widely used；Digital camera method
Influence practical, at low cost, that adverse weather can be overcome, therefore studied extensively by domestic and international experts and scholars.
Summary of the invention
The object of the present invention is to provide the visibility measurement methods based on image automatic identification, can shoot to picture pick-up device
Different scenes image, choose it is most suitable goes fog coefficient, calculate visibility.
The technical solution adopted by the present invention is that the visibility measurement method based on image automatic identification, specifically according to following
Step is implemented:
Step 1, original image of the acquisition containing target, same date, identical place, same shooting angle do not acquire not on the same day
The target image of gas situation is as test image；
Maximum kind variance method is respectively adopted to original image, test image progress Threshold segmentation in step 2；
Step 3 continuously expands the test image after Threshold segmentation using diamond shape probe, obtains expanding image；
Step 4, it is counter to expanding image coloured, using centroid principle, more connected regions after searching continuous expansion
The calibration of " puppet " mass center；
Step 5, the center by " puppet " mass center as quasi- selection multiple target region, mark multiple target region to be selected；
Step 6, original image after singulation dark effect picture in, find corresponding label multiple target area to be selected
" puppet " mass center light intensity I in domainc(y) maximum central point, the final goal that the region where the central point is calculated as visibility
Region, selection suitably remove fog coefficient ω, carry out the calculating of visibility.
The features of the present invention also characterized in that:
Step 2 detailed process are as follows: the Threshold segmentation of original image, test image is carried out using maximum kind variance method；
It is g (x, y) after Threshold segmentation if the pixel of original image or test image is f (x, y), then:
If the size of original image or test image is M × N, segmentation threshold is denoted as T.
The calculating process of segmentation threshold are as follows: target image vegetarian refreshments is in the ratio of entire imageImage averaging gray scale is
μ0；Background pixel point is in the ratio of entire imageImage averaging gray scale is μ1；The gray value of pixel is less than threshold value in image
The number of pixels of T is denoted as N0, gray scale is denoted as N greater than the number of pixels of threshold value T1；The overall average gray scale of image is denoted as μ, inter-class variance
It is denoted as g；Following formula then can be obtained:
The threshold value T for making the maximum threshold value of inter-class variance g, as acquiring is obtained using traversal.
Step 3 detailed process are as follows:
" probe ", which is done, using structural element collects target image information；The condition about origin is first carried out to structural element S
Reflection obtains SV, then in test image after segmentation by SVTranslate x, those translation after at least one non-zero of test image X
When element intersects, corresponding origin position set is dilation operation result；
Then expansion results are write as formula (3):
Structural element S is diamond shape probe；
Structural element S continuously expands repeatedly, obtains expanding image.
Step 4 detailed process are as follows:
Step 4.1 searches the pixel of the image after continuous expansion one by one, and the part all the points that pixel is 0 are assigned
Value is 1, and all the points that pixel is 1 are assigned a value of 0, obtains anti-rendered image；
Step 4.2 utilizes bwlabel function to anti-rendered image, finds all connected domains separated on anti-rendered image,
The principle for recycling mass center to search searches " puppet " mass center, and mass center searches principle such as formula (5):
Wherein, M10It is the summation of all the points x coordinate on object, M01It is the summation of all the points y-coordinate on object, M00It is object
The area of body；
Indicate the position of " puppet " mass center.
Step 5 detailed process are as follows: by " puppet " mass center as the quasi- center for selecting multiple target region, utilize Rectangle square
Shape function marks multiple target region to be selected, the rectangular area that setting area size is 6 × 6；
‘Position’[x,y,w,h] (6)
In formula (6), x, y are lower-left angular coordinate, and w, h respectively represent wide and high.
Step 6 detailed process are as follows: Cauchy's Mead daytime visibility measuring principle, such as formula (7):
Wherein VmetIndicate that atmospheric visibility, β indicate that extinction coefficient, ε indicate to select 0.05 than sense threshold value；
Then formula (7) is rewritten as formula (8):
The dark effect picture of original image after being divided by dark principle, wherein dark channel prior principle, such as
Formula (9):
WhereinIndicate transmissivity, Ic(y) haze image luminous intensity, A are indicatedcIndicate global atmosphere luminous intensity, ω is indicated
Fog coefficient is gone, value is 0 < ω < 1；
Meet Lambert-Beer's law between transmissive portions light splitting and incident intensity；
Meet the relationship such as formula (10) between atmospheric extinction coefficient and transmissivity:
T (x)=e-βd(x) (10)
Wherein, β is extinction coefficient, and d is object at a distance from receiving point, can be calculated by marking on map；
According to formula (9), I is chosenc(y) the final goal area that the region where maximum central point is calculated as visibility
Domain, selection suitably remove fog coefficient ω, go the presence of fog coefficient and visibility to choose rule, convolution (8), (9), (10) can
It is finally inversed by atmospheric visibility.
The presence of fog coefficient and visibility is gone to choose rule as shown in table 1:
Table 1
The beneficial effects of the present invention are:
The present invention is based on the visibility measurement methods of image automatic identification, can be to the different scenes that picture pick-up device is shot
Image, selection is most suitable to go fog coefficient, carries out the automatic selection and visibility measuring and calculating of target area；Measurement method of the invention
Also have the characteristics that it is practical, at low cost, influenced by Changes in weather it is small.
Detailed description of the invention
Fig. 1 is the grayscale image for shooting original image；
Fig. 2 is the effect picture of different threshold segmentation methods；
Wherein, (a) is Local threshold segmentation method；It (b) is iteration global threshold split plot design；(c) complete for maximum kind variance method
Office's thresholding method；
Fig. 3 is expansion process schematic diagram；
Wherein, (a) image；(b) structural element；(c) conditioned reflex；(d) expansion results；
Fig. 4 is expansion process figure；
Wherein, (a) is 1 expansion effect figure；It (b) is 2 expansion effect figures；It (c) is 3 expansion effect figures；It (d) is 4
Secondary expansion effect figure；It (e) is 5 expansion effect figures；It (f) is 6 expansion effect figures；It (g) is 7 expansion effect figures；It (h) is 8
Secondary expansion effect figure；
Fig. 5 is to negate effect picture after repeatedly expanding；
Wherein, (a) is multiple expansion plans, (b) is anti-color-patch map；
Fig. 6 is that more connected regions " puppet " mass center demarcates effect picture；
Fig. 7 is the quasi- selected target area schematic using 6 × 6 size of rectangular function label；
Fig. 8 is original shooting image target area marker effect picture；
Fig. 9 is the dark effect picture of original image；
Figure 10 is several hundred days visibility comparison diagrams；
Figure 11 is that visibility measuring and calculating in several hundred days is recorded a demerit and laser light scattering instrument (CJY-1G) true value overall linear regression analysis
Figure；
Figure 12 is the linear regression analysis figure for calculating visibility V < 2000m；
Figure 13 is measuring and calculating 2000≤V of visibility≤10000m linear regression analysis figure；
Figure 14 is the linear regression analysis figure for calculating visibility V > 10000m.
Specific embodiment
The following describes the present invention in detail with reference to the accompanying drawings and specific embodiments.
The present invention is based on the visibility measurement methods of image automatic identification, are specifically implemented according to the following steps:
Step 1, original image of the acquisition containing target, same date, identical place, same shooting angle do not acquire not on the same day
The target image of gas situation is as test image；
Maximum kind variance method is respectively adopted to original image, test image progress Threshold segmentation in step 2；
Detailed process are as follows: the Threshold segmentation of original image, test image is carried out using maximum kind variance method；
It is g (x, y) after Threshold segmentation if the pixel of original image or test image is f (x, y), then:
If the size of original image or test image is M × N, segmentation threshold is denoted as T.
The calculating process of segmentation threshold are as follows: target image vegetarian refreshments is in the ratio of entire imageImage averaging gray scale is
μ0；Background pixel point is in the ratio of entire imageImage averaging gray scale is μ1；The gray value of pixel is less than threshold value in image
The number of pixels of T is denoted as N0, gray scale is denoted as N greater than the number of pixels of threshold value T1；The overall average gray scale of image is denoted as μ, inter-class variance
It is denoted as g；Following formula then can be obtained:
The threshold value T for making the maximum threshold value of inter-class variance g, as acquiring is obtained using traversal.
We carry out gray proces to the original image of shooting and obtain result such as Fig. 1, recycle three kinds of different threshold value sides
Method processing, obtains different treatment effects such as Fig. 2.Finally there are obvious bimodal image background and target to histogram because of Da-Jin algorithm
Separating effect is preferable, and mistake divides rate minimum, therefore selects OTSU Da-Jin algorithm.
Step 3 continuously expands the test image after Threshold segmentation using diamond shape probe, obtains expanding image；
Detailed process are as follows:
" probe ", which is done, using structural element collects target image information；The condition about origin is first carried out to structural element S
Reflection obtains SV, then in test image after segmentation by SVTranslate x, those translation after at least one non-zero of test image X
When element intersects, corresponding origin position set is dilation operation result；The density bullet of (a) represents target figure in Fig. 3
As X, (b) representative structure element S, figure (c) is shown in its reflection, and expansion results are figure (d).
Then expansion results are write as formula (3):
Structural element S is diamond shape probe；
Structural element S continuously expands repeatedly, obtains expanding image.Effect is as shown in Figure 4.
Step 4, it is counter to expanding image coloured, using centroid principle, more connected regions after searching continuous expansion
The calibration of " puppet " mass center；
Detailed process are as follows:
Step 4.1 searches the pixel of the image after continuous expansion one by one, and the part all the points that pixel is 0 are assigned
Value is 1, and all the points that pixel is 1 are assigned a value of 0, obtains anti-rendered image, as shown in Figure 5.
Step 4.2 utilizes bwlabel function to anti-rendered image, finds all connected domains separated on anti-rendered image,
The principle for recycling mass center to search searches " puppet " mass center, and mass center searches principle such as formula (5):
Wherein, M10It is the summation of all the points x coordinate on object, M01It is the summation of all the points y-coordinate on object, M00It is object
The area of body；According to (6) formula, all connected domains " puppet " mass center is found, is marked with " * ", effect such as Fig. 6.
Indicate the position of " puppet " mass center.
Step 5, the center by " puppet " mass center as quasi- selection multiple target region, mark multiple target region to be selected；
Detailed process are as follows: by " puppet " mass center as the quasi- center for selecting multiple target region, utilize Rectangle rectangular function
Mark multiple target region to be selected, the rectangular area that setting area size is 6 × 6；Marking serial numbers are easily distinguishable with that can select
Color regular font indicates, selects black, signal such as Fig. 7 herein.
‘Position’[x,y,w,h] (6)
In formula (6), x, y are lower-left angular coordinate, and w, h respectively represent wide and high.Final label result is reflected in the original of shooting
Such as Fig. 8 on beginning image.
Step 6, original image after singulation dark effect picture in, find corresponding label multiple target area to be selected
" puppet " mass center light intensity I in domainc(y) maximum central point, the final goal that the region where the central point is calculated as visibility
Region, selection suitably remove fog coefficient ω, carry out the calculating of visibility；
Detailed process are as follows: Cauchy's Mead daytime visibility measuring principle, such as formula (7):
Wherein VmetIndicate that atmospheric visibility, β indicate that extinction coefficient, ε indicate to select 0.05 than sense threshold value；
Then formula (7) is rewritten as formula (8):
The dark effect picture of original image is as shown in Figure 9 after being divided by dark principle, wherein dark is first
Principle is tested, such as formula (9):
WhereinIndicate transmissivity, Ic(y) haze image luminous intensity, A are indicatedcIndicate global atmosphere luminous intensity, ω is indicated
Fog coefficient is gone, value is 0 < ω < 1；
When light is propagated in the medium, a part can be converted into thermal energy release by Absorption of Medium, and another part is deposited in medium
Particle scattering, deviate the original direction of propagation, lose information, remaining part just passes through Jie by the original direction of propagation
Matter, into video camera.Therefore, meet Lambert-Beer's law between transmissive portions light splitting and incident intensity.
Meet the relationship such as formula (10) between atmospheric extinction coefficient and transmissivity:
T (x)=e-βd(x) (10)
Wherein, β is extinction coefficient, and d is object at a distance from receiving point, can be calculated by marking on map；
According to formula (9), I is chosenc(y) the final goal area that the region where maximum central point is calculated as visibility
Domain, selection suitably remove fog coefficient ω, go the presence of fog coefficient and visibility to choose rule, convolution (8), (9), (10) can
It is finally inversed by atmospheric visibility.
The presence of fog coefficient and visibility is gone to choose rule as shown in table 1:
Table 1
Embodiment
Several hundred days images for calculating shooting, using six building, certain teaching as object, to visibility meter pair before the CJY-1G of roof
Than verification.Obtain comparison diagram such as Figure 10.From the point of view of the overall trend of figure, the visibility value and instrument value basic trend one that calculate
It causes, thinks that the goodness of fit is higher.But at the 45th, 54,55,56,57,84 day, it is seen that two curves have apparent differentiation.This 6 days because of energy
View degree instrument failure, obtained data are 0, and therefore, this 6 days data can not verify, and need to reject.Finally, by measuring and calculating, have 8 days
Data error be more than allow worst error, therefore the accuracy rate of context of methods be 92.30%.
Measuring and calculating value and CJY-1G true value are subjected to whole and piecewise linear regression analysis, establish visibility measurement model, X
Axis indicates that measuring and calculating value, Y-axis indicate CJY-1G true value.It is obtained by fitting such as Figure 11, obtaining unary linear regression equation is formula
(12):
Y=1.11473x+431.48036 (12)
Its fitting coefficient R2=0.96296.It is 80.77% that model accuracy rate is obtained after measuring and calculating comparison.
Its reason is analyzed, is the scatter plot as shown in Figure 11, in low visibility and number of days phase when compared with high-visibility
To less, therefore this partial data may being affected for linear regression fit.Therefore, we attempt the several hundred of measuring and calculating
Its visibility carries out regression analysis stage by stage.If Figure 12, Figure 13, Figure 14 are linear visibility linear regression stage by stage.By quasi-
It is formula (13) that conjunction, which obtains regression model:
It is compared through measuring and calculating, obtaining segmented model accuracy rate is 91.35%.
By the above-mentioned means, can be clapped picture pick-up device the present invention is based on the visibility measurement method of image automatic identification
The image for the different scenes taken the photograph, selection is most suitable to go fog coefficient, carries out the automatic selection and visibility measuring and calculating of target area；This
The measurement method of invention also have the characteristics that it is practical, at low cost, influenced by Changes in weather it is small；Experiment proves that accuracy
It is higher.
Claims (8)
1. the visibility measurement method based on image automatic identification, which is characterized in that be specifically implemented according to the following steps:
Step 1, original image of the acquisition containing target, same date, identical place, same shooting angle do not acquire different weather feelings
The target image of condition is as test image；
Maximum kind variance method is respectively adopted to original image, test image progress Threshold segmentation in step 2；
Step 3 continuously expands the test image after Threshold segmentation using diamond shape probe, obtains expanding image；
Step 4, it is counter to expanding image coloured, utilize centroid principle, " puppet " of more connected regions after searching continuous expansion
Mass center calibration；
Step 5, the center by " puppet " mass center as quasi- selection multiple target region, mark multiple target region to be selected；
Step 6, original image after singulation dark effect picture in, find in corresponding label multiple target region to be selected
" puppet " mass center light intensity Ic(y) maximum central point, the final goal region that the region where the central point is calculated as visibility,
Selection suitably removes fog coefficient ω, carries out the calculating of visibility.
2. according to claim 1 based on the visibility measurement method of image automatic identification, which is characterized in that step 2 is specific
Process are as follows: the Threshold segmentation of original image, test image is carried out using maximum kind variance method；
It is g (x, y) after Threshold segmentation if the pixel of original image or test image is f (x, y), then:
If the size of original image or test image is M × N, segmentation threshold is denoted as T.
3. according to claim 2 based on the visibility measurement method of image automatic identification, which is characterized in that the segmentation threshold
The calculating process of value are as follows: target image vegetarian refreshments is in the ratio of entire imageImage averaging gray scale is μ0；Background pixel point exists
The ratio of entire image isImage averaging gray scale is μ1；Number of pixels of the gray value of pixel less than threshold value T is denoted as in image
N0, gray scale is denoted as N greater than the number of pixels of threshold value T1；The overall average gray scale of image is denoted as μ, and inter-class variance is denoted as g；Then it can be obtained
Following formula:
The threshold value T for making the maximum threshold value of inter-class variance g, as acquiring is obtained using traversal.
4. according to claim 1 based on the visibility measurement method of image automatic identification, which is characterized in that step 3 is specific
Process are as follows:
" probe ", which is done, using structural element collects target image information；The conditioned reflex about origin is first carried out to structural element S
Obtain SV, then in test image after segmentation by SVTranslate x, those translation after at least one nonzero element of test image X
When intersection, corresponding origin position set is dilation operation result；
Then expansion results are write as formula (3):
Structural element S is diamond shape probe；
Structural element S continuously expands repeatedly, obtains expanding image.
5. according to claim 1 based on the visibility measurement method of image automatic identification, which is characterized in that step 4 is specific
Process are as follows:
Step 4.1 searches the pixel of the image after continuous expansion one by one, and the part all the points that pixel is 0 are assigned a value of
1, all the points that pixel is 1 are assigned a value of 0, obtain anti-rendered image；
Step 4.2 utilizes bwlabel function to anti-rendered image, finds all connected domains separated on anti-rendered image, then benefit
" puppet " mass center is searched with the principle that mass center is searched, mass center searches principle such as formula (5):
Wherein, M10It is the summation of all the points x coordinate on object, M01It is the summation of all the points y-coordinate on object, M00It is object
Area；
Indicate the position of " puppet " mass center.
6. according to claim 1 based on the visibility measurement method of image automatic identification, which is characterized in that step 5 is specific
Process are as follows: by " puppet " mass center as the quasi- center for selecting multiple target region, mark more mesh to be selected using Rectangle rectangular function
Mark region, the rectangular area that setting area size is 6 × 6；
‘Position’[x,y,w,h] (6)
In formula (6), x, y are lower-left angular coordinate, and w, h respectively represent wide and high.
7. according to claim 1 based on the visibility measurement method of image automatic identification, which is characterized in that step 6 is specific
Process are as follows: Cauchy's Mead daytime visibility measuring principle, such as formula (7):
Wherein VmetIndicate that atmospheric visibility, β indicate that extinction coefficient, ε indicate to select 0.05 than sense threshold value；
Then formula (7) is rewritten as formula (8):
The dark effect picture of original image after being divided by dark principle, wherein dark channel prior principle, such as formula
(9):
WhereinIndicate transmissivity, Ic(y) haze image luminous intensity, A are indicatedcIndicate global atmosphere luminous intensity, ω indicates defogging
Coefficient, value are 0 < ω < 1；
Meet Lambert-Beer's law between transmissive portions light splitting and incident intensity；
Meet the relationship such as formula (10) between atmospheric extinction coefficient and transmissivity:
T (x)=e-βd(x) (10)
Wherein, β is extinction coefficient, and d is object at a distance from receiving point, can be calculated by marking on map；
According to formula (9), I is chosenc(y) the final goal region that the region where maximum central point is calculated as visibility is chosen
Fog coefficient ω suitably is removed, the presence of fog coefficient and visibility is gone to choose rule, convolution (8), (9), (10) can be finally inversed by big
Gas visibility.
8. according to claim 1 based on the visibility measurement method of image automatic identification, which is characterized in that the defogging system
Several and visibility presence chooses rule as shown in table 1:
Table 1
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910690633.2A CN110503634B (en) | 2019-07-29 | 2019-07-29 | Visibility measuring method based on automatic image identification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910690633.2A CN110503634B (en) | 2019-07-29 | 2019-07-29 | Visibility measuring method based on automatic image identification |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110503634A true CN110503634A (en) | 2019-11-26 |
| CN110503634B CN110503634B (en) | 2022-07-15 |
Family
ID=68587662
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910690633.2A Active CN110503634B (en) | 2019-07-29 | 2019-07-29 | Visibility measuring method based on automatic image identification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110503634B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111192275A (en) * | 2019-12-30 | 2020-05-22 | 西安金路交通工程科技发展有限责任公司 | Highway fog visibility identification method based on dark channel prior theory |
| CN114280056A (en) * | 2021-12-20 | 2022-04-05 | 北京普测时空科技有限公司 | Visibility measurement system |
| CN114720425A (en) * | 2022-04-24 | 2022-07-08 | 安徽气象信息有限公司 | Visibility monitoring system and method based on image recognition |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101901477A (en) * | 2010-07-27 | 2010-12-01 | 中国农业大学 | Method and system for extracting field image edges of plant leaves |
| CN105931220A (en) * | 2016-04-13 | 2016-09-07 | 南京邮电大学 | Dark channel experience and minimal image entropy based traffic smog visibility detection method |
| CN107784669A (en) * | 2017-10-27 | 2018-03-09 | 东南大学 | A kind of method that hot spot extraction and its barycenter determine |
-
2019
- 2019-07-29 CN CN201910690633.2A patent/CN110503634B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101901477A (en) * | 2010-07-27 | 2010-12-01 | 中国农业大学 | Method and system for extracting field image edges of plant leaves |
| CN105931220A (en) * | 2016-04-13 | 2016-09-07 | 南京邮电大学 | Dark channel experience and minimal image entropy based traffic smog visibility detection method |
| CN107784669A (en) * | 2017-10-27 | 2018-03-09 | 东南大学 | A kind of method that hot spot extraction and its barycenter determine |
Non-Patent Citations (4)
| Title |
|---|
| HE KAIMING，ET AL: "《Single Image Haze Removal Using Dark Channel Prior》", 《IEEE TRANSACTION ON PATTERN ANALYSIS MACHINE INTELLIGENCE》 * |
| PAN_JINQUAN: "《Matlab形态学图像处理：二值图像分割 标记连通区域和重心位置 删除连通区域》", 《HTTP://BLOG.CSDN.NET/GUYUEALIAN/ARTICLE/DETAILS/71440949》 * |
| 赵瑞: "《基于机器视觉的大气能见度测量方法研究》", 《中国优秀硕士学位论文全文数据库 基础科学辑》 * |
| 赵瑞等: "《基于暗原色先验理论的大气能见度测量》", 《激光与光电子学进展》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111192275A (en) * | 2019-12-30 | 2020-05-22 | 西安金路交通工程科技发展有限责任公司 | Highway fog visibility identification method based on dark channel prior theory |
| CN114280056A (en) * | 2021-12-20 | 2022-04-05 | 北京普测时空科技有限公司 | Visibility measurement system |
| CN114720425A (en) * | 2022-04-24 | 2022-07-08 | 安徽气象信息有限公司 | Visibility monitoring system and method based on image recognition |
| CN114720425B (en) * | 2022-04-24 | 2023-02-21 | 安徽气象信息有限公司 | Visibility monitoring system and method based on image recognition |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110503634B (en) | 2022-07-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | CCTSDB 2021: a more comprehensive traffic sign detection benchmark | |
| Heinzler et al. | Cnn-based lidar point cloud de-noising in adverse weather | |
| CN109816024A (en) | A kind of real-time automobile logo detection method based on multi-scale feature fusion and DCNN | |
| CN107516077B (en) | Traffic sign information extraction method based on fusion of laser point cloud and image data | |
| CN105373794B (en) | A kind of licence plate recognition method | |
| CN110503634A (en) | Visibility measurement method based on image automatic identification | |
| CN108596165B (en) | Road traffic marking detection method and system based on unmanned plane low latitude Aerial Images | |
| CN105608417B (en) | Traffic lights detection method and device | |
| CN106679633B (en) | A kind of vehicle-mounted distance-finding system base and method | |
| CN114677554A (en) | Statistical filtering infrared small target detection tracking method based on YOLOv5 and Deepsort | |
| CN107729843B (en) | Low-floor tramcar pedestrian identification method based on radar and visual information fusion | |
| CN111582339B (en) | Vehicle detection and recognition method based on deep learning | |
| CN111340855A (en) | Road moving target detection method based on track prediction | |
| CN108764338B (en) | Pedestrian tracking method applied to video analysis | |
| CN111898627B (en) | SVM cloud microparticle optimization classification recognition method based on PCA | |
| CN103049909B (en) | A kind of be focus with car plate exposure method | |
| CN115546170B (en) | Fan blade defect positioning method and system based on laser ranging | |
| CN106407951A (en) | Monocular vision-based nighttime front vehicle detection method | |
| CN112488083A (en) | Traffic signal lamp identification method, device and medium for extracting key points based on heatmap | |
| CN111383286B (en) | Positioning method, positioning device, electronic equipment and readable storage medium | |
| CN114998308A (en) | Defect detection method and system based on photometric stereo | |
| CN113312987B (en) | Recognition method based on unmanned aerial vehicle road surface crack image | |
| CN114565848A (en) | Liquid medicine level detection method and system in complex scene | |
| CN116228756B (en) | Method and system for detecting bad points of camera in automatic driving | |
| CN111353481A (en) | Road obstacle identification method based on laser point cloud and video image |
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 |