CN107301623B - Traffic image defogging method and system based on dark channel and image segmentation - Google Patents
Traffic image defogging method and system based on dark channel and image segmentation Download PDFInfo
- Publication number
- CN107301623B CN107301623B CN201710330823.4A CN201710330823A CN107301623B CN 107301623 B CN107301623 B CN 107301623B CN 201710330823 A CN201710330823 A CN 201710330823A CN 107301623 B CN107301623 B CN 107301623B
- Authority
- CN
- China
- Prior art keywords
- image
- light transmittance
- atmospheric light
- traffic image
- foggy
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 68
- 238000003709 image segmentation Methods 0.000 title claims abstract description 33
- 238000002834 transmittance Methods 0.000 claims abstract description 191
- 238000012937 correction Methods 0.000 claims abstract description 82
- 230000015556 catabolic process Effects 0.000 claims abstract description 33
- 238000006731 degradation reaction Methods 0.000 claims abstract description 33
- 238000012545 processing Methods 0.000 claims abstract description 19
- 239000003595 mist Substances 0.000 claims abstract description 8
- 230000003044 adaptive effect Effects 0.000 claims description 50
- 238000004364 calculation method Methods 0.000 claims description 32
- 230000011218 segmentation Effects 0.000 claims description 27
- 238000001914 filtration Methods 0.000 claims description 18
- 230000000877 morphologic effect Effects 0.000 claims description 12
- 238000005457 optimization Methods 0.000 claims description 8
- 230000005540 biological transmission Effects 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 241000283070 Equus zebra Species 0.000 description 1
- 241000272168 Laridae Species 0.000 description 1
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 125000001475 halogen functional group Chemical group 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000001575 pathological effect Effects 0.000 description 1
- 230000000704 physical effect Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G06T5/73—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- 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
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20172—Image enhancement details
- G06T2207/20192—Edge enhancement; Edge preservation
Abstract
The invention discloses a traffic image defogging method and system based on a dark channel and image segmentation, wherein the method comprises the following steps: segmenting the foggy traffic image to obtain a close-range area and a sky area of the foggy traffic image; taking the average intensity value of the mist traffic image sky area as an atmospheric light coefficient, and calculating the scene atmospheric light transmittance by combining a dark channel map of the mist traffic image; and carrying out self-adaptive correction on the atmospheric light transmittance of the sky area in the scene atmospheric light transmittance, and combining the atmospheric light coefficient and the degradation model of the foggy traffic image to obtain the defogged traffic image. The invention comprehensively adopts the methods of image segmentation and dark channel to carry out the defogging of the traffic image, and avoids the situation that the transmissivity of the sky area is low or negative after the defogging through self-adaptive correction, thereby being more effective and better in universality; the average intensity value of the foggy traffic image sky area is taken as the atmospheric light coefficient, so that the method is more reasonable. The invention can be widely applied to the field of image processing.
Description
Technical Field
The invention relates to the field of image processing, in particular to a traffic image defogging method and system based on a dark channel and image segmentation.
Background
In recent years, with the advancement of urbanization and rapid development of industrialization in various parts of China, environmental problems become more serious, and meanwhile, haze weather also appears more and more frequently, so that the accuracy of image information obtained by a monitoring system is seriously influenced. Taking a highway monitoring system as an example, due to the fact that the visibility of a road is greatly reduced due to heavy fog, road condition information obtained by a driver through vision is often inaccurate, correct judgment of the driver on the environment is affected, and traffic accidents are easily caused. Meanwhile, degraded images obtained in foggy weather also cause great difficulty in monitoring the condition of a traffic road, so that traffic departments cannot obtain useful traffic information through cameras, and huge resistance is brought to traffic guidance. Therefore, the realization of high-quality image defogging has very important practical significance for image processing and computer vision application.
Image defogging algorithms mainly fall into two categories: one is a defogging algorithm based on an atmospheric scattering model, the atmospheric scattering rule is analyzed from the physical cause of fog, a corresponding degradation model is established, and the image scene is restored by using priori knowledge; the other type is a defogging algorithm based on image enhancement, which highlights the characteristics and valuable information of the scenery in the image by enhancing the contrast of the degraded image, thereby improving the quality of the image and achieving the aim of sharpening. The defogging algorithm based on image enhancement is mature and easy to realize, but the physical property of image imaging is not considered, so that a clear defogged image cannot be recovered when the haze concentration is high; the defogging algorithm based on the atmospheric scattering model adopts an image degradation model established by the atmospheric scattering principle, fully considers the reason of image degradation and the relation between the image degradation and the atmospheric scattering, and can obtain better image defogging effect.
At present, in the field of image defogging, a widely used method is a single image defogging algorithm based on an atmospheric scattering model, and the algorithm adopts prior information contained in a single image or provides some reasonable assumptions to achieve defogging of the image. The defogging algorithm based on dark channel prior is taken as a typical algorithm of the defogging algorithm of the single image based on the atmospheric scattering model, and the defogging of the image is realized by combining a degradation model of foggy day imaging and a dark channel principle. However, the existing defogging algorithm based on dark channel prior has the following defects:
1) the method is established on the basis of a dark primary color prior theory, and for images of bright areas such as sky, white cloud and the like without dark primary colors, the dark primary color prior theory is easy to cause that the calculated transmissivity of the sky area is low (because the dark primary color prior theory approximates the depth of field of the sky area to be infinite), even negative (because the estimated value of the dark primary color prior theory on the atmospheric light coefficient is smaller than the true value), the effective defogging of the images is difficult to realize, the method is not suitable for processing the images of large-area sky areas such as traffic images, and the universality is poor.
2) Usually, the first 0.1% of pixels are extracted from the dark channel image of the foggy image according to the brightness, and then the maximum value in the original image corresponding to the pixels is taken as the atmospheric light coefficient, so that the atmospheric light coefficients of all channels are close to the maximum pixel value of 255, and the local color spot and color cast phenomenon of the foggy image is caused, which is not reasonable enough.
Disclosure of Invention
To solve the above technical problems, the present invention aims to: the traffic image defogging method based on the dark channel and the image segmentation is effective, good in universality and reasonable.
Another object of the present invention is to: the traffic image defogging system based on the dark channel and the image segmentation is effective, good in universality and reasonable.
The technical scheme adopted by the invention is as follows:
a traffic image defogging method based on dark channels and image segmentation comprises the following steps:
segmenting the foggy traffic image to obtain a close-range area and a sky area of the foggy traffic image;
taking the average intensity value of the mist traffic image sky area as an atmospheric light coefficient, and calculating the scene atmospheric light transmittance by combining a dark channel map of the mist traffic image;
and carrying out self-adaptive correction on the atmospheric light transmittance of the sky area in the scene atmospheric light transmittance, and combining the atmospheric light coefficient and the degradation model of the foggy traffic image to obtain the defogged traffic image.
Further, the step of segmenting the foggy traffic image to obtain a close-range region and a sky region of the foggy traffic image includes:
taking the sky area as a background, taking an area outside the sky area in the foggy traffic image as a close-range area, and respectively calculating the probability omega of the close-range area by assuming that t is any segmentation threshold of the foggy traffic imagetProbability of sky region ωB；
Calculating the average gray value mu of the close shot regiontAnd average gray value mu of sky areaB；
Calculating the total average gray value mu of the whole foggy traffic imagerAnd between-class variance σ2Wherein, murAnd σ2The calculation formulas of (A) and (B) are respectively as follows: mu.sr＝ωt×μt+ωB×μB，σ2＝ωt×(μr-μt)2+ωB×(μr-μB)2；
Solving the variance sigma between classes by adopting a traversal method2Taking a T value corresponding to the maximum value as an optimal segmentation threshold T of the sky region and the close-range region;
segmenting the foggy traffic image according to the optimal segmentation threshold T to segment a candidate close-range area and a candidate sky area, wherein the gray value of the candidate close-range area is 0, and the gray value of the candidate sky area is 1;
and performing morphological processing on the segmented image to obtain a final close-range area and a final sky area of the foggy traffic image.
Further, the step of taking an average intensity value of the sky area of the foggy traffic image as an atmospheric light coefficient and calculating the atmospheric light transmittance of the scene by combining a dark channel map of the foggy traffic image includes:
calculating an atmospheric light coefficient of a foggy traffic image sky area, wherein an expression of the atmospheric light coefficient A of the foggy traffic image sky area is as follows:
wherein, Ig(v) Expressing a gray scale map of the foggy image, psi (v) expressing a sky area of the foggy traffic image, and mean used for solving an average value of all pixel points;
calculating a dark channel of the foggy traffic image based on a dark channel prior theory, wherein the dark channel prior theory enables the dark channel J of any one image J to be connecteddark(x) Is defined as:
where c is one of the color channels in { r, g, b } of image J, JcRepresents the c component of the color channel of J, omega (x) represents a square area with x as the center, y is any pixel point in the area omega (x), Jc(y) is image JcThe value of the pixel at the pixel point y,for finding the minimum value of the three channels of r, g and b,a minimum value filter;
calculating the scene atmosphere light transmittance of the foggy traffic image, wherein the calculation formula of the scene atmosphere light transmittance t (x) of the foggy traffic image is as follows:
wherein, Ic(y) is the c-channel component of the foggy traffic image I (y), AcIs the c-channel component of A, ω is a constant characterizing the degree of dehazing, ω ∈ (0, 1)]。
Further, the step of adaptively correcting the atmospheric light transmittance of the sky area in the scene atmospheric light transmittance and obtaining the defogged traffic image by combining the atmospheric light coefficient and the degradation model of the foggy traffic image includes:
carrying out self-adaptive correction on the atmospheric light transmittance of the sky area in the scene atmospheric light transmittance to obtain the scene atmospheric light transmittance after the self-adaptive correction;
optimizing the scene atmospheric light transmittance after the self-adaptive correction by adopting a guided filtering method to obtain the optimized scene atmospheric light transmittance;
and obtaining the defogged traffic image by combining the degradation model of the fog traffic image according to the atmospheric light coefficient and the optimized scene atmospheric light transmittance.
Further, the step of adaptively correcting the atmospheric light transmittance of the sky area in the scene atmospheric light transmittance to obtain an adaptively corrected scene atmospheric light transmittance includes:
obtaining atmospheric light transmittance t (v) of a sky area from the scene atmospheric light transmittance t (x);
the method comprises the following steps of carrying out adaptive correction on atmospheric light transmittance t (v) of a sky area to obtain adaptive-corrected atmospheric light transmittance t '(v) of the sky area, wherein the transmittance t' (v) is expressed as: t' (v) ═ min (| λ × t (v) |, μ), where v ∈ Ψ (x), λ is a constant for maintaining continuity of transmittance in the sky region, and μ is a threshold for transmittance correction;
and obtaining the scene atmosphere light transmittance t '(x) after the adaptive correction according to the sky area atmosphere light transmittance t' (v) after the adaptive correction.
Further, the step of obtaining the traffic image after defogging according to the atmospheric light coefficient and the optimized scene atmospheric light transmittance in combination with the degradation model of the traffic image after defogging specifically comprises:
obtaining a defogged traffic image by combining a degradation model of a fog traffic image I (x) according to an atmospheric light coefficient A and an optimized scene atmospheric light transmittance t' (x), wherein the expression of the defogged traffic image J (x) is as follows:
the other technical scheme adopted by the invention is as follows:
a traffic image defogging system based on dark channels and image segmentation, comprising:
the image segmentation module is used for segmenting the foggy traffic image to obtain a close-range area and a sky area of the foggy traffic image;
the scene atmosphere light transmittance calculation module is used for taking the average intensity value of the sky area of the foggy traffic image as an atmosphere light coefficient and calculating the scene atmosphere light transmittance by combining a dark channel map of the foggy traffic image;
and the self-adaptive correction and defogging module is used for carrying out self-adaptive correction on the atmospheric light transmittance of the sky area in the scene atmospheric light transmittance and obtaining the defogged traffic image by combining the atmospheric light coefficient and the degradation model of the foggy traffic image.
Further, the image segmentation module comprises:
a probability calculation unit for taking the sky area as a background, taking the area outside the sky area in the foggy traffic image as a close-range area,supposing that t is any segmentation threshold of the foggy traffic image, respectively calculating the probability omega of the close-range regiontProbability of sky region ωB；
A mean gray value calculation unit for calculating mean gray value μ t of the close-range region and mean gray value μ of the sky regionB；
A total average gray value and between-class variance calculating unit for calculating the total average gray value μ of the whole foggy traffic imagerAnd between-class variance σ2Wherein, murAnd σ2The calculation formulas of (A) and (B) are respectively as follows: mu.sr＝ωt×μt+ωB×μB，σ2＝ωt×(μr-μt)2+ωB×(μr-μB)2；
An optimal segmentation threshold calculation unit for calculating the inter-class variance σ by traversal2Taking a T value corresponding to the maximum value as an optimal segmentation threshold T of the sky region and the close-range region;
the segmentation unit is used for segmenting the foggy traffic image according to the optimal segmentation threshold T to segment a candidate close-range area and a candidate sky area, wherein the gray value of the candidate close-range area is 0, and the gray value of the candidate sky area is 1;
and the morphological processing unit is used for performing morphological processing on the segmented image to obtain a final close-range area and a final sky area of the foggy traffic image.
Further, the adaptive modification and defogging module comprises:
the adaptive correction unit is used for carrying out adaptive correction on the atmospheric light transmittance of the sky area in the scene atmospheric light transmittance to obtain the scene atmospheric light transmittance after the adaptive correction;
the optimization unit is used for optimizing the scene atmospheric light transmittance after the adaptive correction by adopting a guided filtering method to obtain the optimized scene atmospheric light transmittance;
and the defogging unit is used for obtaining a defogged traffic image by combining the degradation model of the defogged traffic image according to the atmospheric light coefficient and the optimized scene atmospheric light transmittance.
Further, the adaptive correction unit includes:
a first acquisition subunit configured to acquire an atmospheric light transmittance t (v) of a sky region from a scene atmospheric light transmittance t (x);
a sky region adaptive correction subunit, which performs adaptive correction on the atmospheric light transmittance t (v) of the sky region to obtain an adaptive corrected sky region atmospheric light transmittance t '(v), wherein the transmittance t' (v) is expressed by: t' (v) ═ min (| λ × t (v) |, μ), where v ∈ Ψ (x), λ is a constant for maintaining continuity of transmittance in the sky region, and μ is a threshold for transmittance correction;
and the second acquisition subunit is used for obtaining the scene atmosphere light transmittance t '(x) after the adaptive correction according to the sky area atmosphere light transmittance t' (v) after the adaptive correction.
The method of the invention has the beneficial effects that: comprises segmenting the foggy traffic image, taking the average intensity value of the foggy traffic image sky area as the atmospheric light coefficient, and calculating the scene atmosphere light transmittance by combining the dark channel map of the fog traffic image and carrying out adaptive correction on the atmosphere light transmittance of the sky area in the scene atmosphere light transmittance, and the step of obtaining the defogged traffic image by combining the atmospheric light coefficient and the degradation model of the foggy traffic image comprehensively adopts the methods of image segmentation and dark channel to defogg the traffic image, and adds the step of self-adaptive correction to the atmospheric light transmissivity of the sky area in the scene atmospheric light transmissivity, the situation that the transmissivity of the sky area is low or negative after defogging is avoided through self-adaptive correction, the method is suitable for processing images containing large-area sky areas such as traffic images, and is more effective and better in universality; the average intensity value of the foggy traffic image sky area is taken as the atmospheric light coefficient, so that the situation that the atmospheric light coefficients of all channels are close to the maximum pixel value of 255 is avoided, and the method is more reasonable. Furthermore, an improved Dajin algorithm is adopted when the foggy traffic image is segmented, so that the method has the advantages of simple calculation and good real-time performance of the traditional Dajin algorithm, and avoids the situation that a close-range area is wrongly divided into sky areas through morphological processing, so that the precision is higher. And further, the method comprises the step of optimizing the scene atmospheric light transmittance after the self-adaptive correction by adopting a guided filtering method, and the optimization method of the guided filtering method is adopted, so that the speed of optimizing the atmospheric light transmittance is improved, and the problems of halation and color distortion after the image is defogged are effectively solved.
The system of the invention has the advantages that: the system comprises an image segmentation module, a scene atmospheric light transmittance calculation module and a self-adaptive correction and defogging module, wherein the traffic image defogging is carried out by comprehensively adopting an image segmentation and dark channel method, the self-adaptive correction and defogging module is additionally provided with an operation of carrying out self-adaptive correction on the atmospheric light transmittance of a sky area in the scene atmospheric light transmittance, and the situation that the transmittance of the sky area is low or negative after defogging is avoided through the self-adaptive correction; the average intensity value of the fog traffic image sky area is taken as the atmospheric light coefficient in the scene atmospheric light transmittance calculation module, so that the situation that the atmospheric light coefficients of all channels are all close to the maximum pixel value of 255 is avoided, and the method is more reasonable. Furthermore, an improved Dajin algorithm is adopted when the foggy traffic image is segmented in the image segmentation module, so that the method has the advantages of simple calculation and good real-time performance of the traditional Dajin algorithm, avoids the situation that a close-range area is wrongly divided into sky areas through morphological processing, and is higher in precision. Furthermore, the adaptive correction and defogging module comprises an optimization unit for optimizing the scene atmospheric light transmittance after adaptive correction by adopting a guided filtering method, and the speed of optimizing the atmospheric light transmittance is improved and the problems of halation and color distortion after image defogging are effectively solved by adopting the guided filtering optimization method.
Drawings
FIG. 1 is an overall flowchart of a traffic image defogging method based on dark channels and image segmentation according to the present invention;
FIG. 2 is a block diagram of an image defogging algorithm according to an embodiment of the invention;
FIG. 3 is a comparison graph of the segmentation effect of the foggy traffic image sky area by using the Otsu algorithm and the improved Otsu algorithm, respectively, according to an embodiment of the present invention;
FIG. 4 is a schematic diagram illustrating a process of adaptively modifying and optimizing a transmission map of a foggy traffic image scene according to an embodiment of the present invention;
fig. 5 is a schematic view of traffic images before and after defogging according to an embodiment of the invention.
Detailed Description
Referring to fig. 1, a traffic image defogging method based on a dark channel and image segmentation comprises the following steps:
segmenting the foggy traffic image to obtain a close-range area and a sky area of the foggy traffic image;
taking the average intensity value of the mist traffic image sky area as an atmospheric light coefficient, and calculating the scene atmospheric light transmittance by combining a dark channel map of the mist traffic image;
and carrying out self-adaptive correction on the atmospheric light transmittance of the sky area in the scene atmospheric light transmittance, and combining the atmospheric light coefficient and the degradation model of the foggy traffic image to obtain the defogged traffic image.
Further, as a preferred embodiment, the step of segmenting the foggy traffic image to obtain a close-range area and a sky area of the foggy traffic image includes:
taking the sky area as a background, taking an area outside the sky area in the foggy traffic image as a close-range area, and respectively calculating the probability omega of the close-range area by assuming that t is any segmentation threshold of the foggy traffic imagetProbability of sky region ωB；
Calculating the average gray value mu of the close shot regiontAnd average gray value mu of sky areaB；
Calculating the total average gray value mu of the whole foggy traffic imagerAnd between-class variance σ2Wherein, murAnd σ2The calculation formulas of (A) and (B) are respectively as follows: mu.sr＝ωt×μt+ωB×μB，σ2＝ωt×(μr-μt)2+ωB×(μr-μB)2；
Solving the variance sigma between classes by adopting a traversal method2Taking a T value corresponding to the maximum value as an optimal segmentation threshold T of the sky region and the close-range region;
segmenting the foggy traffic image according to the optimal segmentation threshold T to segment a candidate close-range area and a candidate sky area, wherein the gray value of the candidate close-range area is 0, and the gray value of the candidate sky area is 1;
and performing morphological processing on the segmented image to obtain a final close-range area and a final sky area of the foggy traffic image.
Further, as a preferred embodiment, the step of taking an average intensity value of the sky area of the fog traffic image as an atmospheric light coefficient and calculating the atmospheric light transmittance of the scene by combining the dark channel map of the fog traffic image includes:
calculating an atmospheric light coefficient of a foggy traffic image sky area, wherein an expression of the atmospheric light coefficient A of the foggy traffic image sky area is as follows:
wherein, Ig(v) Expressing a gray scale map of the foggy image, psi (v) expressing a sky area of the foggy traffic image, and mean used for solving an average value of all pixel points;
calculating a dark channel of the foggy traffic image based on a dark channel prior theory, wherein the dark channel prior theory enables the dark channel J of any one image J to be connecteddark(x) Is defined as:
where c is one of the color channels in { r, g, b } of image J, JcRepresents the c component of the color channel of J, omega (x) represents a square area with x as the center, y is any pixel point in the area omega (x), Jc(y) is image JcThe value of the pixel at the pixel point y,
for finding the minimum value of the three channels of r, g and b,
a minimum value filter;
calculating the scene atmosphere light transmittance of the foggy traffic image, wherein the calculation formula of the scene atmosphere light transmittance t (x) of the foggy traffic image is as follows:
wherein, Ic(y) is the c-channel component of the foggy traffic image I (y), AcIs the c-channel component of A, ω is a constant characterizing the degree of dehazing, ω ∈ (0, 1)]。
Further, as a preferred embodiment, the step of adaptively correcting the atmospheric light transmittance of the sky area in the atmospheric light transmittance of the scene and obtaining the defogged traffic image by combining the atmospheric light coefficient and the degradation model of the foggy traffic image includes:
carrying out self-adaptive correction on the atmospheric light transmittance of the sky area in the scene atmospheric light transmittance to obtain the scene atmospheric light transmittance after the self-adaptive correction;
optimizing the scene atmospheric light transmittance after the self-adaptive correction by adopting a guided filtering method to obtain the optimized scene atmospheric light transmittance;
and obtaining the defogged traffic image by combining the degradation model of the fog traffic image according to the atmospheric light coefficient and the optimized scene atmospheric light transmittance.
In a further preferred embodiment, the step of adaptively correcting the atmospheric light transmittance of the sky region in the scene atmospheric light transmittance to obtain an adaptively corrected scene atmospheric light transmittance includes:
obtaining atmospheric light transmittance t (v) of a sky area from the scene atmospheric light transmittance t (x);
the method comprises the following steps of carrying out adaptive correction on atmospheric light transmittance t (v) of a sky area to obtain adaptive-corrected atmospheric light transmittance t '(v) of the sky area, wherein the transmittance t' (v) is expressed as: t' (v) ═ min (| λ × t (v) |, μ), where v ∈ Ψ (x), λ is a constant for maintaining continuity of transmittance in the sky region, and μ is a threshold for transmittance correction;
and obtaining the scene atmosphere light transmittance t '(x) after the adaptive correction according to the sky area atmosphere light transmittance t' (v) after the adaptive correction.
Further as a preferred embodiment, the step of obtaining the traffic image after defogging according to the atmospheric light coefficient and the optimized scene atmospheric light transmittance in combination with the degradation model of the traffic image after defogging specifically includes:
obtaining a defogged traffic image by combining a degradation model of a fog traffic image I (x) according to an atmospheric light coefficient A and an optimized scene atmospheric light transmittance t' (x), wherein the expression of the defogged traffic image J (x) is as follows:
referring to fig. 1, a traffic image defogging system based on a dark channel and image segmentation comprises:
the image segmentation module is used for segmenting the foggy traffic image to obtain a close-range area and a sky area of the foggy traffic image;
the scene atmosphere light transmittance calculation module is used for taking the average intensity value of the sky area of the foggy traffic image as an atmosphere light coefficient and calculating the scene atmosphere light transmittance by combining a dark channel map of the foggy traffic image;
and the self-adaptive correction and defogging module is used for carrying out self-adaptive correction on the atmospheric light transmittance of the sky area in the scene atmospheric light transmittance and obtaining the defogged traffic image by combining the atmospheric light coefficient and the degradation model of the foggy traffic image.
Further as a preferred embodiment, the image segmentation module comprises:
a probability calculation unit for using the sky area as background, using the area outside the sky area in the foggy traffic image as near-scene area, and simulatingLet t be any segmentation threshold of the foggy traffic image, and respectively calculate the probability omega of the close-range regiontProbability of sky region ωB；
A mean gray value calculation unit for calculating mean gray value μ of the close-range regiontAnd average gray value mu of sky areaB；
A total average gray value and between-class variance calculating unit for calculating the total average gray value μ of the whole foggy traffic imagerAnd between-class variance σ2Wherein, murAnd σ2The calculation formulas of (A) and (B) are respectively as follows: mu.sr＝ωt×μt+ωB×μB，σ2＝ωt×(μr-μt)2+ωB×(μr-μB)2；
An optimal segmentation threshold calculation unit for calculating the inter-class variance σ by traversal2Taking a T value corresponding to the maximum value as an optimal segmentation threshold T of the sky region and the close-range region;
the segmentation unit is used for segmenting the foggy traffic image according to the optimal segmentation threshold T to segment a candidate close-range area and a candidate sky area, wherein the gray value of the candidate close-range area is 0, and the gray value of the candidate sky area is 1;
and the morphological processing unit is used for performing morphological processing on the segmented image to obtain a final close-range area and a final sky area of the foggy traffic image.
Further preferably, the adaptive modification and defogging module includes:
the adaptive correction unit is used for carrying out adaptive correction on the atmospheric light transmittance of the sky area in the scene atmospheric light transmittance to obtain the scene atmospheric light transmittance after the adaptive correction;
the optimization unit is used for optimizing the scene atmospheric light transmittance after the adaptive correction by adopting a guided filtering method to obtain the optimized scene atmospheric light transmittance;
and the defogging unit is used for obtaining a defogged traffic image by combining the degradation model of the defogged traffic image according to the atmospheric light coefficient and the optimized scene atmospheric light transmittance.
Further preferably, the adaptive correction unit includes:
a first acquisition subunit configured to acquire an atmospheric light transmittance t (v) of a sky region from a scene atmospheric light transmittance t (x);
a sky region adaptive correction subunit, which performs adaptive correction on the atmospheric light transmittance t (v) of the sky region to obtain an adaptive corrected sky region atmospheric light transmittance t '(v), wherein the transmittance t' (v) is expressed by: t' (v) ═ min (| λ × t (v) |, μ), where v ∈ Ψ (x), λ is a constant for maintaining continuity of transmittance in the sky region, and μ is a threshold for transmittance correction;
and the second acquisition subunit is used for obtaining the scene atmosphere light transmittance t '(x) after the adaptive correction according to the sky area atmosphere light transmittance t' (v) after the adaptive correction.
The invention will be further explained and explained with reference to the drawings and the embodiments in the description.
Example one
The invention provides a novel traffic image defogging method based on a dark channel and image segmentation, and aims to solve the problems that effective defogging of an image is difficult to realize by an existing image defogging algorithm, the image including a large-area sky area is not suitable for processing a traffic image and the like, and the image is not reasonable enough. As shown in fig. 2, the method mainly comprises the following steps: firstly, segmenting a traffic image through an improved Dajin algorithm to obtain a close-range region and a sky region of the traffic image; then, taking the average intensity value of the sky area as an atmospheric light coefficient, and calculating the atmospheric light transmittance of the scene by combining a dark channel map of the fog traffic image; then, carrying out self-adaptive correction on the atmospheric light transmittance of the sky area, and optimizing the scene atmospheric light transmittance and the atmospheric light transmittance of the sky area by adopting guide filtering; and finally, defogging of the image is realized by combining the degradation model of the foggy traffic image.
The foggy day image degradation model widely used at present is formed by deformation of an atmospheric scattering model, and the specific expression of the atmospheric scattering model is as follows:
I(x)＝J(x)t(x)+A(1-t(x))
in the above formula, i (x) represents the observed image (i.e., the image to be defogged), j (x) is the haze-free image to be restored (i.e., the image after defogging), a is the atmospheric light coefficient, t (x) is the atmospheric light transmittance, and t (x) is the medium transmission function describing the percentage of atmospheric light emitted from the object or scene that reaches the camera. The goal of defogging is to estimate the atmospheric optical coefficient A and transmittance t (x) and recover J (x) from I (x). The difficulty with image defogging based on the physical model is that if a fogged image is input, image defogging is a pathological problem that lacks a constraint term. Therefore, the defogging method of the invention is generally divided into the following four steps:
and (I) segmenting the foggy traffic image through an improved Otsu algorithm to obtain a close-range area and a sky area of the foggy traffic image.
Image segmentation is a classical problem in the field of digital images. In recent years, the segmentation algorithm adopting the neural network or the mean shift can well segment the image, but the calculation amount is large, the consumed time is long, the convergence speed is low, and the method cannot be well applied to the traffic field with high real-time requirement. Compared with these algorithms, the Otsu algorithm (OTSU) has the obvious advantage of dividing the image into two parts, namely a background part and a foreground part according to the gray-scale characteristics of the image. The Dajin algorithm has obvious advantages in traffic scenes with high real-time requirements because of simple calculation and no influence of image brightness and contrast.
In a traffic scene, under the influence of fog, zebra crossings, car bodies, white buildings and the like have brightness close to the sky, and if a traffic image is segmented by adopting a traditional Dajin algorithm, the above areas are also wrongly divided into sky areas. For this purpose, the improved Dajin algorithm is adopted for image segmentation.
In the improved Dajin algorithm, the background is a sky area, and areas outside the sky area are close-range areas, and the specific algorithm steps are as follows:
1) let t be any of the imagesSegmenting threshold values, and respectively calculating the probability (i.e. the ratio of the number of pixels belonging to the close shot to the whole image) omega of the close shot regiontAnd the probability of sky region (i.e. the proportion of the number of pixels belonging to the background to the whole image)' omegaB；
2) Calculating the average gray value mu of the close shot regiontAnd average gray value mu of sky areaB；
3) Calculating the total average gray value mu of the whole imagerAnd between-class variance σ2；
4) Solving T corresponding to the maximum inter-class variance value as the optimal segmentation threshold T of the sky region and the close-range region;
5) segmenting the image according to a threshold value T, setting a near area as 0 and setting a sky area as 1;
6) and performing morphological processing on the segmented image to obtain a final close-range region and a final sky region of the image.
The result of using the improved Dajin algorithm to segment the foggy traffic image is shown in fig. 3, where (a) in fig. 3 is the foggy traffic image, and (b) in fig. 3 is the image obtained by using the traditional Dajin algorithm to segment the foggy traffic image; fig. 3 (c) is an image obtained by segmenting the foggy traffic image by using the improved algorithm of the present invention. As can be seen from fig. 3, the improved salid algorithm further morphologically processes the regions segmented by the traditional salid algorithm, thereby avoiding the occurrence of the situation that the close-range region is mistakenly segmented into the sky region, and achieving higher precision.
And (II) after the fog traffic image is segmented, taking the average intensity value of the sky area as an atmospheric light coefficient A, and calculating the atmospheric light transmittance of the scene by combining a dark channel map of the fog traffic image.
This process can be further subdivided into the following steps:
(1) calculation of atmospheric light coefficient A
In the image defogging process, solving the value of the atmospheric light coefficient A in the scene is a key step, how to more accurately estimate A is also a difficulty of image defogging, and whether the estimation accuracy influences the final image restoration effect to a great extent. In the existing image defogging algorithm, the first 0.1% of pixels are generally extracted from a dark channel image of a foggy image according to the brightness, and then the maximum value in an original image corresponding to the pixels is taken as an atmospheric light coefficient, so that the atmospheric light coefficients of all channels are close to the maximum pixel value of 255, and local color stains and color cast phenomena occur in the defogged image.
In order to better estimate the value of the atmospheric light A, the invention adopts an averaging strategy, and takes the average intensity value of the sky area segmented by using an improved Otsu algorithm as the estimated value of A, namely:
in the above formula, Ig(v) A gray scale map representing a foggy traffic image, Ψ (x) representing the sky region, and mean operation to sum and average all pixel points.
(2) Dark channel map for obtaining fog traffic image based on dark channel prior theory
The dark channel prior theory is provided by a hong Kong university Thanksgow gull team performing statistical observation on a large number of outdoor fog-free images: in most local areas not including sky areas, at least one cold color channel has some pixel points with low brightness, even approaching zero, which is equivalent to that the minimum brightness value in the area approaches zero. Thus, for any one image J, its dark channel is defined:
in the above formula, JcRepresents a certain color channel component of J, and Ω (x) represents a square area centered on x. For image JcPerforming minimum operation twice to obtain a dark channel output,
is used to process each of the pixels of the image,
is a minimum filter. These two minimum operations may be interchanged in the calculation. Therefore, based on the concept of dark channel, when J is an outdoor fog-free image, the sky area is removed, and the dark primary color J of JdarkIs very low and approaches 0, i.e. there is:
Jdark→0
(3) calculation of scene atmospheric light transmittance
Atmospheric light transmittance t (x) is a medium transmission function that describes the percentage of light emitted from an object or scene that reaches the camera. By combining the obtained atmospheric light coefficient A and a dark channel prior theory and a degradation model of a foggy day image, the scene atmospheric light transmittance t (x) is obtained as follows:
in order to make the defogged image look more natural, a certain degree of fog must be retained during defogging, and therefore, the value of ω may be set to 0.95.
And (III) after the atmospheric light transmittance of the scene is calculated, firstly, carrying out self-adaptive correction on the atmospheric light transmittance of the sky area, and then, optimizing the atmospheric light transmittance of the scene and the sky atmospheric light transmittance by adopting a guide filtering method.
This process can be further subdivided into:
(1) adaptive correction of transmissivity in sky regions
The transmittance of the sky area calculated by using the dark channel prior theory is often low and even negative. The low transmittance is mainly caused by approximating the depth of field of the sky to be infinite, and the negative transmittance is caused by the fact that the atmospheric light coefficient a is estimated to be smaller than the true value. In order to better restore the image and avoid the phenomenon that the defogged image is excessively enhanced or generates a negative value of the transmittance, aiming at the transmission image of the sky area, the invention adopts a self-adaptive processing mode to correct the atmospheric light transmittance t (v) of the sky area to obtain the self-adaptive corrected atmospheric light transmittance t' (v) of the sky area:
t′(v)＝min(|λ×t(v)|,μ)，v∈Ψ(x)
in the above formula, λ represents a constant, and λ is used to maintain the continuity of the transmittance of the sky region; mu represents the threshold of transmittance correction, and mu is used for preventing the transmittance from being too large to cause overexposure of the image. Through a series of experiments, it is found that when the value of lambda is set to 10 and the value of mu is set to 0.5, a good defogging effect can be obtained.
(2) The transmittance is optimized using guided filtering.
The transmissivity of a close-range area and a sky area is rough, a block phenomenon exists, and the phenomenon can cause a halo phenomenon of a defogged image or a phenomenon that fog winds around the edge of an object, so that the traditional image defogging method generally adopts a soft matting method to optimize the transmissivity. However, the soft matting method has a large calculation amount and long consumption time, and the guided filtering can smooth the image details and maintain the edge information of the image, and has the greatest advantage of high calculation speed, so that the method optimizes the scene atmosphere light transmittance t' (x) after the adaptive correction by using the guided filtering method to obtain the optimized scene atmosphere light transmittance t ″ (x). When the guiding filtering is carried out, the guiding graph adopts a minimum value channel of the foggy traffic image, and is marked as I.
The atmospheric light transmittance of the sky region is corrected, and the atmospheric light transmittance optimized as a whole by using the guide filter method is shown in fig. 4. Fig. 4 (a) is the scene atmosphere light transmittance calculated in step (ii), and fig. 4 (b) is the scene atmosphere light transmittance after adaptive correction; in fig. 4, (c) is the scene atmospheric light transmittance after the directional filtering optimization.
As can be seen from fig. 4, after the scene transmission diagram is corrected and optimized, the problem of too low transmittance in the sky region does not exist, and the transmission region in the close-range region becomes finer and more natural in visual effect.
And (IV) defogging of the traffic image is realized by combining the degradation model of the foggy traffic image.
After obtaining the atmospheric light coefficient A and the optimized scene atmospheric light transmittance t' (x), according to the degradation model of the foggy traffic image, the final traffic image restoration formula is as follows:
fig. 5 shows a comparison of the effects before and after defogging by the image defogging method of the present invention, where (a) in fig. 5 and (b) in fig. 5 are traffic image diagrams before and after defogging, respectively.
The invention relates to a traffic image defogging method and system based on a dark channel and image segmentation.A more accurate sky area is segmented by combining an improved Otsu algorithm, and a more reasonable atmospheric light coefficient is further calculated by taking an intensity mean value; by carrying out self-adaptive correction on the atmospheric light transmittance of the sky area, the phenomenon that the defogged sky area is excessively enhanced or a transmittance negative value is generated is avoided; when the atmospheric light transmissivity of a close-range region and a sky region is optimized, a guide filtering method is adopted, so that the speed of transmissivity optimization is improved, and the problems of halation, color distortion and the like after image defogging are effectively solved. The traffic image defogging method and system based on the dark channel and the image segmentation can provide an effective theoretical basis and technical support for traffic supervision.
While the preferred embodiments of the present invention have been illustrated and described, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (6)
1. A traffic image defogging method based on a dark channel and image segmentation is characterized by comprising the following steps: the method comprises the following steps:
segmenting the foggy traffic image to obtain a close-range area and a sky area of the foggy traffic image;
taking the average intensity value of the mist traffic image sky area as an atmospheric light coefficient, and calculating the scene atmospheric light transmittance by combining a dark channel map of the mist traffic image;
carrying out self-adaptive correction on the atmospheric light transmittance of a sky area in the scene atmospheric light transmittance, and combining an atmospheric light coefficient and a degradation model of a foggy traffic image to obtain a defogged traffic image;
the step of performing adaptive correction on the atmospheric light transmittance of the sky area in the scene atmospheric light transmittance and obtaining the defogged traffic image by combining the atmospheric light coefficient and the degradation model of the foggy traffic image comprises the following steps of:
carrying out self-adaptive correction on the atmospheric light transmittance of the sky area in the scene atmospheric light transmittance to obtain the scene atmospheric light transmittance after the self-adaptive correction;
optimizing the scene atmospheric light transmittance after the self-adaptive correction by adopting a guided filtering method to obtain the optimized scene atmospheric light transmittance;
obtaining a defogged traffic image by combining a degradation model of the defogged traffic image according to the atmospheric light coefficient and the optimized scene atmospheric light transmittance;
the step of adaptively correcting the atmospheric light transmittance of the sky region in the scene atmospheric light transmittance to obtain a scene atmospheric light transmittance after adaptive correction includes:
obtaining atmospheric light transmittance t (v) of a sky area from the scene atmospheric light transmittance t (x);
the method comprises the following steps of carrying out adaptive correction on atmospheric light transmittance t (v) of a sky area to obtain adaptive-corrected atmospheric light transmittance t '(v) of the sky area, wherein the transmittance t' (v) is expressed as: t' (v) ═ min (| λ × t (v) |, μ), where v ∈ Ψ (x), Ψ (x) represents a sky region, λ is a constant for maintaining continuity of transmittance of the sky region, and μ is a threshold of transmittance correction;
and obtaining the scene atmosphere light transmittance t '(x) after the adaptive correction according to the sky area atmosphere light transmittance t' (v) after the adaptive correction.
2. The traffic image defogging method based on the dark channel and the image segmentation as claimed in claim 1, wherein: the step of segmenting the foggy traffic image to obtain a close-range region and a sky region of the foggy traffic image includes:
taking the sky area as a background, taking an area outside the sky area in the foggy traffic image as a close-range area, and respectively calculating the probability omega of the close-range area by assuming that t is any segmentation threshold of the foggy traffic imagetProbability of sky region ωB；
Calculating the average gray value mu of the close shot regiontAnd average gray value mu of sky areaB；
Calculating the total average gray value mu of the whole foggy traffic imagerAnd between-class variance σ2Wherein, murAnd σ2The calculation formulas of (A) and (B) are respectively as follows: mu.sr＝ωt×μt+ωB×μB，σ2＝ωt×(μr-μt)2+ωB×(μr-μB)2；
Solving the variance sigma between classes by adopting a traversal method2Taking a T value corresponding to the maximum value as an optimal segmentation threshold T of the sky region and the close-range region;
segmenting the foggy traffic image according to the optimal segmentation threshold T to segment a candidate close-range area and a candidate sky area, wherein the gray value of the candidate close-range area is 0, and the gray value of the candidate sky area is 1;
and performing morphological processing on the segmented image to obtain a final close-range area and a final sky area of the foggy traffic image.
3. The traffic image defogging method based on the dark channel and the image segmentation as claimed in claim 1, wherein: the method comprises the following steps of taking the average intensity value of the sky area of the foggy traffic image as an atmospheric light coefficient, and calculating the atmospheric light transmittance of a scene by combining a dark channel map of the foggy traffic image, wherein the steps comprise:
calculating an atmospheric light coefficient of a foggy traffic image sky area, wherein an expression of the atmospheric light coefficient A of the foggy traffic image sky area is as follows:wherein, Ig(v) Expressing a gray scale map of the foggy image, psi (v) expressing a sky area of the foggy traffic image, and mean used for solving an average value of all pixel points;
calculating a dark channel of the foggy traffic image based on a dark channel prior theory, wherein the dark channel prior theory enables the dark channel J of any one image J to be connecteddark(x) Is defined as:
where c is one of the color channels in { r, g, b } of image J, JcRepresents the c component of the color channel of J, omega (x) represents a square area with x as the center, y is any pixel point in the area omega (x), Jc(y) is image JcThe value of the pixel at the pixel point y,
for finding the minimum value of the three channels of r, g and b,
a minimum value filter;
calculating the scene atmosphere light transmittance of the foggy traffic image, wherein the calculation formula of the scene atmosphere light transmittance t (x) of the foggy traffic image is as follows:
wherein, Ic(y) is the c-channel component of the foggy traffic image I (y), AcIs the c-channel component of A, ω is a constant characterizing the degree of dehazing, ω ∈ (0, 1)]。
4. The traffic image defogging method based on the dark channel and the image segmentation as claimed in claim 1, wherein: the step of obtaining the defogged traffic image according to the atmospheric light coefficient and the optimized scene atmospheric light transmittance by combining with the degradation model of the fog traffic image is specifically as follows:
obtaining a defogged traffic image by combining a degradation model of a fog traffic image I (x) according to an atmospheric light coefficient A and an optimized scene atmospheric light transmittance t' (x), wherein the expression of the defogged traffic image J (x) is as follows:
5. a traffic image defogging system based on a dark channel and image segmentation is characterized in that: the method comprises the following steps:
the image segmentation module is used for segmenting the foggy traffic image to obtain a close-range area and a sky area of the foggy traffic image;
the scene atmosphere light transmittance calculation module is used for taking the average intensity value of the sky area of the foggy traffic image as an atmosphere light coefficient and calculating the scene atmosphere light transmittance by combining a dark channel map of the foggy traffic image;
the self-adaptive correction and defogging module is used for carrying out self-adaptive correction on the atmospheric light transmittance of the sky area in the scene atmospheric light transmittance and obtaining a defogged traffic image by combining the atmospheric light coefficient and the degradation model of the foggy traffic image;
the adaptive correction and defogging module comprises:
the adaptive correction unit is used for carrying out adaptive correction on the atmospheric light transmittance of the sky area in the scene atmospheric light transmittance to obtain the scene atmospheric light transmittance after the adaptive correction;
the optimization unit is used for optimizing the scene atmospheric light transmittance after the adaptive correction by adopting a guided filtering method to obtain the optimized scene atmospheric light transmittance;
the defogging unit is used for obtaining a defogged traffic image by combining a degradation model of the defogged traffic image according to the atmospheric light coefficient and the optimized scene atmospheric light transmittance;
the adaptive correction unit includes:
a first acquisition subunit configured to acquire an atmospheric light transmittance t (v) of a sky region from a scene atmospheric light transmittance t (x);
a sky region adaptive correction subunit, which performs adaptive correction on the atmospheric light transmittance t (v) of the sky region to obtain an adaptive corrected sky region atmospheric light transmittance t '(v), wherein the transmittance t' (v) is expressed by: t' (v) ═ min (| λ × t (v) |, μ), where v ∈ Ψ (x), Ψ (x) represents a sky region, λ is a constant for maintaining continuity of transmittance of the sky region, and μ is a threshold of transmittance correction;
and the second acquisition subunit is used for obtaining the scene atmosphere light transmittance t '(x) after the adaptive correction according to the sky area atmosphere light transmittance t' (v) after the adaptive correction.
6. The traffic image defogging system based on the dark channel and the image segmentation as claimed in claim 5, wherein: the image segmentation module comprises:
a probability calculation unit for calculating the probability omega of the close-range region respectively by using the sky region as the background, using the region outside the sky region in the foggy traffic image as the close-range region, and assuming t as any segmentation threshold of the foggy traffic imagetProbability of sky region ωB；
A mean gray value calculation unit for calculating mean gray value μ of the close-range regiontAnd average gray value mu of sky areaB；
A total average gray value and between-class variance calculating unit for calculating the total average gray value μ of the whole foggy traffic imagerAnd between-class variance σ2Wherein, murAnd σ2The calculation formulas of (A) and (B) are respectively as follows: mu.sr＝ωt×μt+ωB×μB，σ2＝ωt×(μr-μt)2+ωB×(μr-μB)2；
An optimal segmentation threshold calculation unit for calculating the inter-class variance σ by traversal2Taking a T value corresponding to the maximum value as an optimal segmentation threshold T of the sky region and the close-range region;
the segmentation unit is used for segmenting the foggy traffic image according to the optimal segmentation threshold T to segment a candidate close-range area and a candidate sky area, wherein the gray value of the candidate close-range area is 0, and the gray value of the candidate sky area is 1;
and the morphological processing unit is used for performing morphological processing on the segmented image to obtain a final close-range area and a final sky area of the foggy traffic image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710330823.4A CN107301623B (en) | 2017-05-11 | 2017-05-11 | Traffic image defogging method and system based on dark channel and image segmentation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710330823.4A CN107301623B (en) | 2017-05-11 | 2017-05-11 | Traffic image defogging method and system based on dark channel and image segmentation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107301623A CN107301623A (en) | 2017-10-27 |
| CN107301623B true CN107301623B (en) | 2020-02-14 |
Family
ID=60137852
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710330823.4A Active CN107301623B (en) | 2017-05-11 | 2017-05-11 | Traffic image defogging method and system based on dark channel and image segmentation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107301623B (en) |
Families Citing this family (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108335307A (en) * | 2018-04-19 | 2018-07-27 | 云南佳叶现代农业发展有限公司 | Adaptive tobacco leaf picture segmentation method and system based on dark primary |
| CN108876743B (en) * | 2018-06-26 | 2020-12-29 | 中山大学 | Image rapid defogging method, system, terminal and storage medium |
| CN109118546A (en) * | 2018-08-30 | 2019-01-01 | 东南大学 | A kind of depth of field hierarchical estimation method based on single-frame images |
| CN111091501A (en) * | 2018-10-24 | 2020-05-01 | 天津工业大学 | Parameter estimation method of atmosphere scattering defogging model |
| CN110136079A (en) * | 2019-05-05 | 2019-08-16 | 长安大学 | Image defogging method based on scene depth segmentation |
| CN110458029B (en) * | 2019-07-15 | 2022-12-20 | 佛山科学技术学院 | Vehicle detection method and device in foggy environment |
| CN111325688B (en) * | 2020-02-18 | 2023-05-05 | 西安汇智信息科技有限公司 | Unmanned aerial vehicle image defogging method for optimizing atmosphere light by fusion morphology clustering |
| CN111899198A (en) * | 2020-08-06 | 2020-11-06 | 北京科技大学 | Defogging method and device for marine image |
| CN112200746B (en) * | 2020-10-16 | 2024-03-08 | 中南大学 | Defogging method and equipment for foggy-day traffic scene image |
| CN112465720B (en) * | 2020-11-27 | 2024-02-23 | 南京邮电大学 | Image defogging method and device based on image sky segmentation and storage medium |
| CN112816483A (en) * | 2021-03-05 | 2021-05-18 | 北京文安智能技术股份有限公司 | Group fog recognition early warning method and system based on fog value analysis and electronic equipment |
| CN113379619B (en) * | 2021-05-12 | 2022-02-01 | 电子科技大学 | Integrated processing method for defogging imaging, visibility extraction and depth of field estimation |
| CN113205469A (en) * | 2021-06-04 | 2021-08-03 | 中国人民解放军国防科技大学 | Single image defogging method based on improved dark channel |
| CN115496685A (en) * | 2022-09-21 | 2022-12-20 | 重庆地质矿产研究院 | Rapid cloud thinning method for high-resolution second-order satellite remote sensing image |
| CN115937144B (en) * | 2022-12-08 | 2023-08-25 | 郑州大学 | Image processing method and system in thoracoscopy |
| CN116630349B (en) * | 2023-07-25 | 2023-10-20 | 山东爱福地生物股份有限公司 | Straw returning area rapid segmentation method based on high-resolution remote sensing image |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103077500B (en) * | 2012-12-30 | 2016-03-30 | 贺江涛 | The defogging method capable of view data and device |
| CN105654440B (en) * | 2015-12-30 | 2018-07-27 | 首都师范大学 | Quick single image defogging algorithm based on regression model and system |
| CN106339988B (en) * | 2016-01-07 | 2019-06-14 | 深圳开阳电子股份有限公司 | A kind of image defogging method and device |
| CN106251301B (en) * | 2016-07-26 | 2019-10-15 | 北京工业大学 | A kind of single image to the fog method based on dark primary priori |
| CN106530246B (en) * | 2016-10-28 | 2019-04-16 | 大连理工大学 | Image defogging method and system based on dark Yu non local priori |
| CN106548463B (en) * | 2016-10-28 | 2019-05-10 | 大连理工大学 | Sea fog image automatic defogging method and system based on dark and Retinex |
-
2017
- 2017-05-11 CN CN201710330823.4A patent/CN107301623B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN107301623A (en) | 2017-10-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107301623B (en) | Traffic image defogging method and system based on dark channel and image segmentation | |
| CN106157267B (en) | Image defogging transmissivity optimization method based on dark channel prior | |
| CN102750674B (en) | Video image defogging method based on self-adapting allowance | |
| Gao et al. | Sand-dust image restoration based on reversing the blue channel prior | |
| CN107451966B (en) | Real-time video defogging method implemented by guiding filtering through gray level image | |
| CN106548461B (en) | Image defogging method | |
| CN109087254B (en) | Unmanned aerial vehicle aerial image haze sky and white area self-adaptive processing method | |
| CN110570360B (en) | Retinex-based robust and comprehensive low-quality illumination image enhancement method | |
| CN108133462B (en) | Single image restoration method based on gradient field region segmentation | |
| CN104809709A (en) | Single-image self-adaptation defogging method based on domain transformation and weighted quadtree decomposition | |
| CN106780380A (en) | A kind of image defogging method and system | |
| CN108182671B (en) | Single image defogging method based on sky area identification | |
| CN107067375A (en) | A kind of image defogging method based on dark channel prior and marginal information | |
| CN112053298B (en) | Image defogging method | |
| CN105023246B (en) | A kind of image enchancing method based on contrast and structural similarity | |
| CN107085830B (en) | Single image defogging method based on propagation filtering | |
| CN108629750A (en) | A kind of night defogging method, terminal device and storage medium | |
| CN106504216B (en) | Single image to the fog method based on Variation Model | |
| CN110458029B (en) | Vehicle detection method and device in foggy environment | |
| CN111325688A (en) | Unmanned aerial vehicle image defogging method fusing morphological clustering and optimizing atmospheric light | |
| CN114693548A (en) | Dark channel defogging method based on bright area detection | |
| CN107977941B (en) | Image defogging method for color fidelity and contrast enhancement of bright area | |
| CN112465720A (en) | Image defogging method and device based on image sky segmentation and storage medium | |
| CN109949239B (en) | Self-adaptive sharpening method suitable for multi-concentration multi-scene haze image | |
| CN113298730B (en) | Defogging restoration method based on image decomposition |
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 |