CN110163807B - Low-illumination image enhancement method based on expected bright channel - Google Patents
Low-illumination image enhancement method based on expected bright channel Download PDFInfo
- Publication number
- CN110163807B CN110163807B CN201910211765.2A CN201910211765A CN110163807B CN 110163807 B CN110163807 B CN 110163807B CN 201910211765 A CN201910211765 A CN 201910211765A CN 110163807 B CN110163807 B CN 110163807B
- Authority
- CN
- China
- Prior art keywords
- image
- low
- channel
- illumination
- max
- 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
- 238000005286 illumination Methods 0.000 title claims abstract description 53
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000002834 transmittance Methods 0.000 claims abstract description 31
- 238000009499 grossing Methods 0.000 claims description 16
- 238000005516 engineering process Methods 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 4
- 239000000126 substance Substances 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 abstract description 18
- 230000000694 effects Effects 0.000 abstract description 3
- 230000002708 enhancing effect Effects 0.000 abstract 1
- 238000003384 imaging method Methods 0.000 abstract 1
- 230000003287 optical effect Effects 0.000 description 4
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012634 optical imaging Methods 0.000 description 2
- 238000002310 reflectometry Methods 0.000 description 2
- LFQSCWFLJHTTHZ-UHFFFAOYSA-N Ethanol Chemical compound CCO LFQSCWFLJHTTHZ-UHFFFAOYSA-N 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000000969 carrier Substances 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 125000001475 halogen functional group Chemical group 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration by the use of histogram techniques
-
- G06T5/70—
-
- 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
- G06T2207/10024—Color image
Abstract
A low-illumination image enhancement method based on an expected bright channel relates to a low-illumination image enhancement method and belongs to the field of digital image processing. In order to solve the problem that parameters need to be adjusted manually when a transmittance image is obtained in the existing low-illumination image enhancement algorithm based on bright channel prior, the low-illumination image enhancement algorithm based on the expected bright channel is provided. The method firstly collects HDR images and carries out statistics on the HDR images to obtain a distribution histogram of a maximum value channel image of the HDR images as an expected bright channel histogram of an enhanced image. Next, the histogram specification processing is performed on the maximum value channel image of the low-illuminance image using the desired bright channel histogram, and a desired bright channel image is obtained. The transmittance image is then solved using the desired bright channel image and the atmospheric imaging equation. And finally, enhancing the low-illumination image by combining an atmospheric scattering model. The invention has better enhancement effect on low-illumination images with different brightness, and can obviously enhance the brightness and contrast of the images so as to make the image details clearer.
Description
Technical Field
The invention relates to the field of digital image processing, in particular to a low-illumination image enhancement method.
Background
Various optical imaging devices are increasingly applied to the civil and military fields at present, but are affected by the conditions of low sensitivity of a detector, insufficient ambient illumination conditions and the like, and images formed by an optical system have various degradation problems, such as reduced image contrast, insufficient brightness and the like, so that the analysis and identification of targets in the images are affected. At the same time, the performance of the optical system is improved by hardware at a great cost due to the limitation of optical manufacturing technology. Therefore, the optical image and video enhancement algorithm under the low illumination condition is deeply researched, so that various optical imaging systems can adaptively adjust the brightness and contrast of the image along with the change of the illumination condition, the application range of the equipment can be expanded, the manufacturing cost of the equipment can be greatly saved, the volume of the equipment is reduced, and the theoretical and practical application values are extremely high.
Currently, the commonly used low-illumination image enhancement methods mainly include a histogram equalization method, a Retinex method, an enhancement method based on a bright channel prior principle, a High Dynamic Range (HDR) method, and the like. The histogram equalization algorithm is small in calculation amount, has a good enhancement effect on images with poor details, is the most widely applied algorithm at present, and has the problems of detail loss, color distortion, severe image brightness change and the like. The Retinex algorithm considers that the observed image is the product of incident illumination and the reflectivity of the surface of the object, and the reflectivity of the surface of the object is not influenced by the illumination and can reflect the real information of the object. To estimate the incident illumination, the Retinex algorithm generally assumes that the illumination is uniformly distributed in space, and then obtains the illumination component through low-pass filtering. However, this assumption does not necessarily hold in practice, as the light source and the strong reflection source in the image will appear halo after enhancement. And because the gradient change information of illumination is not taken into consideration, the edge information of the image can be blurred to a certain degree in the enhancement process. In addition, the Retinex algorithm still has no good solution to the color preservation problem of color images. The HDR algorithm synthesizes images with different exposure degrees in the same scene, has a good application effect in a scene with a large dynamic range, and can give consideration to both a high-brightness region and a low-brightness region in the scene. But HDR images require multiple frames of images to be combined, thus requiring a long capture time and requiring a relatively static hold between the scene and the camera during capture, which makes the algorithm difficult to apply to moving carriers or dynamic object scenes. In addition, when the ambient light intensity is low, it is difficult to distinguish target information and background in a dark place even if the exposure time is increased, and thus an ideal enhanced image cannot be obtained. Inspired by defogging by using a dark primary color prior principle, a bright channel prior is proposed in a paper 'An effective and integrated adaptive enhancement in a charting lighting condition' and is applied to image enhancement, but a hyper-parameter exists in the algorithm, and manual parameter adjustment is needed according to a specific image so as to ensure the quality of the enhanced image, so that the application range of the algorithm is greatly limited. To solve this problem, the following scholars propose some improved methods such as automatically selecting enhancement parameters and modifying the transmittance formula to stretch the image dynamic range, etc. However, these improvements are essentially enhanced by compressing the dynamic range of the bright area and expanding the dynamic range of the dark area, and the purpose of simultaneously retaining the information of the bright area and the dark area cannot be achieved, so that the problem of over-enhancement or under-enhancement of the bright area or the dark area of the image may occur.
The HDR image is subjected to statistical analysis, and a bright channel prior principle is combined, a low-illumination image enhancement algorithm based on an expected bright channel is provided, a maximum channel of the enhanced image is taken as the expected bright channel, a histogram obtained by HDR image statistics is taken as prior information, and the expected bright channel image is assumed to accord with a prior histogram. An expected bright channel image is obtained through histogram specification, and then the transmittance image and the low-illumination image are obtained by combining an atmospheric scattering model, so that the problems that manual parameter adjustment needs to be carried out on images with different brightness and over-enhancement or under-enhancement caused by parameter adjustment are solved.
Disclosure of Invention
The method aims to solve the problems that the prior image enhancement algorithm based on the bright channel prior needs artificial parameter adjustment and over-enhancement or under-enhancement caused by parameter adjustment.
A low-illumination image enhancement method based on a desired bright channel comprises the following steps:
collecting HDR images, establishing an HDR image database, counting all the HDR images in the database, and acquiring a distribution histogram H of a maximum value channel image;
step two, inputting the low-illumination image L, calculating the maximum value of the three channels of the low-illumination image RGB, and obtaining a maximum value channel image L max ；
Step threeAnd a maximum value channel image L of the low-illumination image L by using the distribution histogram H max Performing histogram specification processing to obtain an expected bright channel image J expect_max ；
Step four, the expected bright channel image J is obtained expect_max And a maximum value channel image L max Substituting the transmittance image T into an atmospheric scattering model to obtain a transmittance image T;
step five, smoothing the transmissivity image T by utilizing guide filtering to obtain a transmissivity smooth image T guidefilt ；
Step six, smoothing the image T by utilizing the transmissivity guidefilt Combining an atmospheric scattering model, and performing enhancement operation on the RGB three channels of the low-illumination image L to obtain an enhanced image J;
preferably, the specific process of the step one comprises the following steps:
searching related images in Google by using keywords 'HDR' and 'High dynamic image', downloading the images by using a web crawler technology, and generating an HDR image database;
and step two, acquiring maximum value channel images in three color channels of RGB of all images in the HDR image database, respectively calculating distribution histograms of the maximum value channel images, and adding and normalizing all the distribution histograms to obtain a distribution histogram H of the maximum value channel image in the HDR image database.
Preferably, the maximum value channel image L is obtained in the second step max The formula is obtained as follows:
wherein L is max (i, j) is the maximum value channel image L max Row i and column j; max represents a max operation; c, taking R, G, B, corresponding to three color channels of red, green and blue, L in RGB color space c (i, j) is the ith row and the jth column of the channel of the low-illumination image L in the RGB color space.
Preferably, step four will expect a bright throughRoad image J expect_max And a maximum value channel image L max Substituting an atmospheric scattering model to obtain a transmittance image T, wherein the specific formula is as follows:
wherein T is a transmittance image; l is a radical of an alcohol max The maximum value channel image of the low illumination image is obtained; j. the design is a square expect_max The maximum channel image expected for the enhanced image, namely the expected bright channel image; and A is an estimate of atmospheric light intensity.
Preferably, the specific process of step five includes the following steps:
an image block with M × M centered on the ith row and jth column elements in the transmittance image T is referred to as an image block W '(i, j), and the transmittance smoothed image T is referred to as an image block W' (i, j) guidefilt The image block with the same size and position as the image block W' (i, j) is marked as the image block W (i, j); m is the number of pixels, and the value of M is 11;
smoothing image T at transmittance guidefilt In, pixel point T guidefilt (i, j) as an approximation of all pixels in an image block W (i, j) having a center size of M × M:
T guidefilt (m′,n′)＝k (i,j) T(m′,n′)+l (i,j)
where m ', n ' represents the position of the pixel, (m ', n '). Epsilon.W (i, j), and (m ', n '). Epsilon.W ' (i, j); t is a unit of guidefilt (m ', n') is a transmittance smoothed image T guidefilt The m 'th row and the n' th column of pixel points; t (m ', n') is a pixel point of the m 'th row and the n' th column of the transmissivity image T; k is a radical of (i,j) ，l (i,j) Are approximate parameters;
l(i,j)＝u (i,j) (1-k (i,j) )
wherein the content of the first and second substances,
is the variance of all elements in the image block W' (i, j) in T; u. of (i,j) Is the average value of all elements in the image block W' (i, j) in T; count (W ') is the number of all pixels within the image block W' (i, j); epsilon is a smoothing parameter;
preferably, the specific process of step six includes the following steps:
smoothing image T with transmissivity guidefilt Combining with an atmospheric scattering model, performing enhancement operation on three channels of RGB of the low-illumination image L to obtain an enhanced image J, wherein a specific formula is as follows:
wherein, L is an input low-illumination image; j is the enhanced image; t is a unit of guidefilt Is a transmittance smoothed image; and A is an estimate of atmospheric light intensity.
The invention has the following beneficial effects:
the invention provides an image enhancement algorithm based on an expected bright channel on the basis of an image enhancement algorithm based on bright channel prior. The expected bright channel is obtained by a method of performing specified processing on the maximum channel image of the low-illumination image, so that the transmissivity image is calculated and the low-illumination image is enhanced, and the problem that parameter adjustment needs to be performed on images with different brightness in an original bright channel algorithm is solved. The invention does not need to adjust manual parameters, can obviously improve the brightness and contrast of images with different brightness, and has certain real-time property.
Drawings
FIG. 1 is a flow chart of an image enhancement algorithm of the present invention;
FIG. 2 is a distribution histogram of a maximum channel image of a HDR data set collected by the present invention;
fig. 3 is an original image before image enhancement in the embodiment, and fig. 4 is an image after image enhancement by applying the invention in the embodiment.
Detailed Description
The first embodiment is as follows: the present embodiment is described in connection with figure 1,
collecting HDR images, establishing an HDR image database, counting all the HDR images in the database, and acquiring a distribution histogram H of a maximum value channel image;
step two, inputting the low-illumination image L, calculating the maximum value of the three channels of the low-illumination image RGB, and obtaining a maximum value channel image L max ；
Thirdly, utilizing the distribution histogram H to carry out maximum value channel image L on the low-illumination image L max Performing histogram specification processing to obtain an expected bright channel image J expect_max ；
Step four, the expected bright channel image J is obtained expect_max And a maximum value channel image L max Substituting the transmittance image T into an atmospheric scattering model to obtain a transmittance image T;
step five, smoothing the transmissivity image T by utilizing guide filtering to obtain a transmissivity smooth image T guidefilt ；
Step six, smoothing the image T by utilizing the transmissivity guidefilt Combining an atmospheric scattering model, and performing enhancement operation on the RGB three channels of the low-illumination image L to obtain an enhanced image J;
the second embodiment is as follows:
the specific process of the first step of the embodiment comprises the following steps:
searching related images in Google by using keywords 'HDR' and 'High dynamic image', downloading the images by using a web crawler technology, and generating an HDR image database;
and step two, acquiring maximum value channel images in three color channels of all RGB images in the HDR image database, respectively calculating distribution histograms of the maximum value channel images, and adding and normalizing all the distribution histograms to obtain a distribution histogram H of the maximum value channel image in the HDR image database.
Other steps and parameters are the same as in the first embodiment.
The third concrete implementation mode:
obtaining the maximum value channel image L in the second step of the present embodiment max The acquisition formula is as follows:
wherein L is max (i, j) is the maximum value channel image L max Row i and column j; max represents a max operation; c, taking R, G, B as the color space corresponding to three color channels of red, green and blue, L c (i, j) is the ith row and the jth column of the channel of the low-illumination image L in the RGB color space.
Other steps and parameters are the same as in the first or second embodiment.
The fourth concrete implementation mode:
the expected bright channel image J in step four of the present embodiment expect_max And a maximum value channel image L max Substituting the atmospheric scattering model to obtain a transmittance image T, wherein the specific formula is as follows:
wherein T is a transmittance image; l is max The maximum value channel image of the low illumination image is obtained; j. the design is a square expect_max A maximum channel image expected for the enhanced image, namely an expected bright channel image; and A is an estimate of atmospheric light intensity.
Other steps and parameters are the same as in one of the first to third embodiments.
The fifth concrete implementation mode:
the concrete process of the fifth step comprises the following steps:
an image block with the size of M x M with the i-th row and j-th column elements as the center in the transmittance image T is recorded as an image block W' (i, j), and the transmittance smoothed image T is recorded guidefilt The image block with the same size and the same position as the image block W' (i, j) is marked as an image block W (i, j), and the size of M is 11;
smoothing image T at transmittance guidefilt In, pixel point T guidefilt (i, j) as an approximation of all pixels in an image block W (i, j) having a center size of M × M:
T guidefilt (m′,n′)＝k (i,j) T(m′,n′)+l (i,j)
where m ', n ' represents the position of the pixel, (m ', n ') ∈ W (i, j), and (m ', n ') ∈ W ' (i, j); t is guidefilt (m ', n') is a transmittance smoothed image T guidefilt The m 'th row and the n' th column of pixel points; t (m ', n') is a pixel point of the m 'th row and the n' th column of the transmissivity image T; k is a radical of (i,j) ，l (i,j) Are approximate parameters;
l(i,j)＝u (i,j) (1-k (i,j) )
wherein the content of the first and second substances,
is the variance of all elements in the image block W' (i, j) in T; u. of (i,j) Is the mean value of all elements in the image block W' (i, j) in T; count (W ') is the number of all pixels within the image block W' (i, j); epsilon is a smoothing parameter; epsilon is a smoothing parameter, epsilon is a small value, and k is prevented (i,j) If the epsilon is too large, the epsilon is 0.01;
other steps and parameters are the same as in one of the first to fourth embodiments.
The sixth specific implementation mode:
the concrete process of the step six comprises the following steps:
smoothing image T with transmissivity guidefilt Combining with an atmospheric scattering model, performing enhancement operation on three channels of RGB of the low-illumination image L to obtain an enhanced image J, wherein a specific formula is as follows:
whereinL is an input low-illumination image; j is the enhanced image; t is guidefilt Is a transmittance smoothed image; a is an estimate of atmospheric light intensity, and A is taken
Other steps and parameters are the same as in one of the first to fifth embodiments.
According to the invention, manual parameter adjustment is not needed, the low-illumination images in different brightness ranges can be enhanced, the brightness and the contrast of the low-illumination images are obviously improved, and certain real-time performance is achieved.
Claims (6)
1. A low-illumination image enhancement method based on a desired bright channel is characterized by comprising the following steps:
collecting HDR images, establishing an HDR image database, counting all the HDR images in the database, and acquiring a distribution histogram H of a maximum value channel image;
step two, inputting the low-illumination image L, calculating the maximum value of the three channels of the low-illumination image RGB, and obtaining a maximum value channel image L max ；
Thirdly, utilizing the distribution histogram H to carry out maximum value channel image L on the low-illumination image L max Performing histogram specification processing to obtain an expected bright channel image J expect_max ；
Step four, the expected bright channel image J is obtained expect_max And a maximum value channel image L max Substituting the transmittance image T into an atmospheric scattering model to obtain a transmittance image T;
step five, smoothing the transmissivity image T by utilizing guide filtering to obtain a transmissivity smooth image T guidefilt ；
Step six, smoothing the image T by utilizing the transmissivity guidefilt And (4) combining an atmospheric scattering model, and performing enhancement operation on the RGB three channels of the low-illumination image L to obtain an enhanced image J.
2. The method according to claim 1, wherein the specific process of step one comprises the following steps:
searching related images in Google by using keywords 'HDR' and 'High dynamic image', downloading the images by using a web crawler technology, and generating an HDR image database;
and step two, acquiring maximum value channel images in three color channels of RGB of all images in the HDR image database, respectively calculating distribution histograms of the maximum value channel images, and adding and normalizing all the distribution histograms to obtain a distribution histogram H of the maximum value channel image in the HDR image database.
3. A low-illumination image enhancement method based on expected bright channel as claimed in claim 1, wherein the maximum channel image L in step two max The acquisition formula is as follows:
wherein L is max (i, j) is the maximum value channel image L max Row i and column j; max represents a max operation; c, taking R, G, B as the color space corresponding to three color channels of red, green and blue, L c (i, j) is the ith row and the jth column of the channel of the low-illumination image L in the RGB color space.
4. The method according to claim 1, wherein said step four of obtaining a desired bright channel image J expect_max And a maximum value channel image L max Substituting the atmospheric scattering model to obtain a transmittance image T, wherein the specific formula is as follows:
wherein T is a transmittance image; l is max Is the maximum channel of the low-light imageAn image; j. the design is a square expect_max A maximum channel image expected for the enhanced image, namely an expected bright channel image; and A is an estimate of atmospheric light intensity.
5. The low-illumination image enhancement method based on the expected bright channel as claimed in claim 1, wherein the specific process of step five comprises the following steps:
an image block with the size of M x M with the i-th row and j-th column elements as the center in the transmittance image T is recorded as an image block W' (i, j), and the transmittance smoothed image T is recorded guidefilt The image block with the same size and position as the image block W' (i, j) is marked as the image block W (i, j); m is the number of pixels, and the value of M is 31;
smoothing image T at transmittance guidefilt In, pixel point T guidefilt (i, j) as an approximation of all pixels in an image block W (i, j) having a center size of M × M:
T guidefilt (m′,n′)＝k (i,j) T(m′,n′)+l (i,j)
where m ', n ' represents the position of the pixel, (m ', n '). Epsilon.W (i, j), and (m ', n '). Epsilon.W ' (i, j); t is a unit of guidefilt (m ', n') is a transmittance smoothed image T guidefilt The m 'th row and the n' th column of pixel points; t (m ', n') is a pixel point of the m 'th row and the n' th column of the transmissivity image T; k is a radical of (i,j) ，l (i,j) Are approximate parameters;
l(i,j)＝u (i,j) (1-k (i,j) )
wherein the content of the first and second substances,
is the variance of all elements in the image block W' (i, j) in T; u. of (i,j) Is the mean value of all elements in the image block W' (i, j) in T; count (W ') is the number of all pixels within the image block W' (i, j); ε is the smoothing parameter.
6. The low-illumination image enhancement method based on the expected bright channel as claimed in claim 1, wherein the specific process of step six comprises the following steps:
smoothing image T with transmissivity guidefilt Combining with an atmospheric scattering model, performing enhancement operation on three channels of RGB of the low-illumination image L to obtain an enhanced image J, wherein a specific formula is as follows:
wherein, L is an input low-illumination image; j is the enhanced image; t is guidefilt Is a transmittance smoothed image; and A is an estimate of atmospheric light intensity.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910211765.2A CN110163807B (en) | 2019-03-20 | 2019-03-20 | Low-illumination image enhancement method based on expected bright channel |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910211765.2A CN110163807B (en) | 2019-03-20 | 2019-03-20 | Low-illumination image enhancement method based on expected bright channel |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110163807A CN110163807A (en) | 2019-08-23 |
| CN110163807B true CN110163807B (en) | 2023-04-07 |
Family
ID=67638825
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910211765.2A Active CN110163807B (en) | 2019-03-20 | 2019-03-20 | Low-illumination image enhancement method based on expected bright channel |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110163807B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111192210B (en) * | 2019-12-23 | 2023-05-26 | 杭州当虹科技股份有限公司 | Self-adaptive enhanced video defogging method |
| CN111402145B (en) * | 2020-02-17 | 2022-06-07 | 哈尔滨工业大学 | Self-supervision low-illumination image enhancement method based on deep learning |
| CN113284054A (en) * | 2020-02-19 | 2021-08-20 | 华为技术有限公司 | Image enhancement method and image enhancement device |
| CN111899201B (en) * | 2020-08-17 | 2022-06-07 | 哈尔滨工业大学 | Low-illumination image enhancement method based on conditional re-enhancement network |
| CN113256522A (en) * | 2021-05-21 | 2021-08-13 | 上海国茂数字技术有限公司 | Method and device for enhancing low-illumination image and readable storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104809700A (en) * | 2015-04-16 | 2015-07-29 | 北京工业大学 | Low-light video real-time enhancement method based on bright channel |
| CN106846263A (en) * | 2016-12-28 | 2017-06-13 | 中国科学院长春光学精密机械与物理研究所 | The image defogging method being immunized based on fusion passage and to sky |
| CN106886985A (en) * | 2017-04-25 | 2017-06-23 | 哈尔滨工业大学 | A kind of self adaptation enhancement method of low-illumination image for reducing colour cast |
| CN107437240A (en) * | 2017-07-14 | 2017-12-05 | 浙江大学 | A kind of enhancement method of low-illumination image for suppressing highlight regions |
| CN109064426A (en) * | 2018-07-26 | 2018-12-21 | 电子科技大学 | A kind of method and device for inhibiting dazzle in low-light (level) image and enhancing image |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10477128B2 (en) * | 2017-01-06 | 2019-11-12 | Nikon Corporation | Neighborhood haze density estimation for single-image dehaze |
-
2019
- 2019-03-20 CN CN201910211765.2A patent/CN110163807B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104809700A (en) * | 2015-04-16 | 2015-07-29 | 北京工业大学 | Low-light video real-time enhancement method based on bright channel |
| CN106846263A (en) * | 2016-12-28 | 2017-06-13 | 中国科学院长春光学精密机械与物理研究所 | The image defogging method being immunized based on fusion passage and to sky |
| CN106886985A (en) * | 2017-04-25 | 2017-06-23 | 哈尔滨工业大学 | A kind of self adaptation enhancement method of low-illumination image for reducing colour cast |
| CN107437240A (en) * | 2017-07-14 | 2017-12-05 | 浙江大学 | A kind of enhancement method of low-illumination image for suppressing highlight regions |
| CN109064426A (en) * | 2018-07-26 | 2018-12-21 | 电子科技大学 | A kind of method and device for inhibiting dazzle in low-light (level) image and enhancing image |
Non-Patent Citations (5)
| Title |
|---|
| Guided Image Filtering;K. He等;《IEEE Transactions on Pattern Analysis and Machine Intelligence》;20121231;第35卷(第6期);第1397-1409页 * |
| Restoration and enhancement of underwater images based on bright channel prior;Gao Y等;《 Mathematical Problems in Engineering》;20161122;第1-15页 * |
| 一种改进的色彩保持低照度图像增强方法;遆晓光等;《哈尔滨工业大学学报》;20141231;第46卷(第3期);第1-7页 * |
| 亮通道先验Retinex对低照度图像的光照补偿;李赓飞等;《光学精密工程》;20181231;第26卷(第5期);第1191-1200页 * |
| 复杂环境下的机器人视觉图像增强算法的研究;马金荣;《中国优秀硕士学位论文全文数据库 信息科技辑》;20180515(第05期);第I138-389页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110163807A (en) | 2019-08-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110163807B (en) | Low-illumination image enhancement method based on expected bright channel | |
| US11127122B2 (en) | Image enhancement method and system | |
| CN105761227B (en) | Underwater picture Enhancement Method based on dark channel prior and white balance | |
| Tao et al. | Low-light image enhancement using CNN and bright channel prior | |
| CN106886985B (en) | A kind of adaptive enhancement method of low-illumination image reducing colour cast | |
| CN107292830B (en) | Low-illumination image enhancement and evaluation method | |
| CN110599433B (en) | Double-exposure image fusion method based on dynamic scene | |
| CN104240194B (en) | A kind of enhancement algorithm for low-illumination image based on parabolic function | |
| Wang et al. | A fast single-image dehazing method based on a physical model and gray projection | |
| CN106897981A (en) | A kind of enhancement method of low-illumination image based on guiding filtering | |
| Zheng et al. | Infrared traffic image enhancement algorithm based on dark channel prior and gamma correction | |
| Wang et al. | Variational single nighttime image haze removal with a gray haze-line prior | |
| CN109639994B (en) | Dynamic adjusting method for exposure time of embedded vehicle-mounted camera | |
| CN108133462B (en) | Single image restoration method based on gradient field region segmentation | |
| Feng et al. | Low-light image enhancement based on multi-illumination estimation | |
| CN112419181A (en) | Wide dynamic infrared image detail enhancement method | |
| CN110288539A (en) | A kind of mobile clear method of underwater picture with dark channel prior in color combining space | |
| CN115578297A (en) | Generalized attenuation image enhancement method for self-adaptive color compensation and detail optimization | |
| CN105989583B (en) | A kind of image defogging method | |
| Zhang et al. | Image dehazing based on dark channel prior and brightness enhancement for agricultural remote sensing images from consumer-grade cameras | |
| Wei et al. | An image fusion dehazing algorithm based on dark channel prior and retinex | |
| Fu et al. | An anisotropic Gaussian filtering model for image de-hazing | |
| CN112330613B (en) | Evaluation method and system for cytopathology digital image quality | |
| Han et al. | Automatic illumination and color compensation using mean shift and sigma filter | |
| CN110992287B (en) | Method for clarifying non-uniform illumination video |
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 |