CN107610072B - Adaptive noise reduction method for low-light-level video image based on gradient guided filtering - Google Patents
Adaptive noise reduction method for low-light-level video image based on gradient guided filtering Download PDFInfo
- Publication number
- CN107610072B CN107610072B CN201710932558.7A CN201710932558A CN107610072B CN 107610072 B CN107610072 B CN 107610072B CN 201710932558 A CN201710932558 A CN 201710932558A CN 107610072 B CN107610072 B CN 107610072B
- Authority
- CN
- China
- Prior art keywords
- image
- low
- filtering
- light
- video image
- 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
Images
Landscapes
- Picture Signal Circuits (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a self-adaptive noise reduction method for a low-light-level video image based on gradient guided filtering, and belongs to the technical field of image processing. The implementation method of the invention comprises the following steps: step 1, detecting a motion area in a low-light-level video image; step 2, estimating the noise intensity of the low-light-level video image; step 3, performing adaptive gradient guided filtering and iterative guided filtering processing on the low-light-level video image so as to obtain a final filtered output image; the invention aims to provide a self-adaptive noise reduction method of a low-light-level video image based on gradient guide filtering, which reduces image noise, improves the signal-to-noise ratio of the image and improves the image quality on the premise of reducing human intervention as much as possible. The invention has good practical application prospect in the field of low-light night vision imaging.
Description
Technical Field
The invention relates to a self-adaptive noise reduction method for a low-light-level video image based on gradient guided filtering, and belongs to the technical field of image processing.
Background
Low-light-level night vision imaging is one of the common important means in the night vision field, inherent random noise of a low-light-level night vision image is a key factor influencing imaging quality, and low-light-level night vision image filtering is also a subject of long-term research at home and abroad.
The main night vision low-light imaging devices at present are: such as tube-coupled CCD (ICCD), electron multiplying CCD (EMCCD), electron bombarded CCD (EBCCD), and high-sensitivity CMOS. Because the intensity of visible light at night is relatively weak, a high-sensitivity photosensitive device is needed, and meanwhile, in order to obtain scene detail information as much as possible, photoelectric conversion signals are amplified and multiplied, noise is enhanced in the process, and therefore, the noise of the low-light-level imaging device is obvious. Human eyes are very sensitive to noise, and a large amount of noise can seriously reduce the imaging quality and influence the normal observation of people, so that how to effectively reduce the noise in night vision low-light-level imaging is always a key technology for the research of night vision imaging technology.
The typical simple methods for image noise reduction mainly include: mean filtering, median filtering, bilateral filtering and the like, although the method is simple and easy to implement, the effect is often poor; typical complex methods are mainly: the method comprises a non-local area average (NLMS) method, a Bayesian least square-Gaussian scale mixing method (BLS-GSM), a three-dimensional block matching (BM3D) method and the like, and although the method has a good effect, the algorithm is complex, the calculation amount is large, and the real-time application is difficult.
In summary, it is still a challenging problem to effectively reduce the noise in low-light-level video images and to operate quickly and in real time.
Disclosure of Invention
The invention aims to provide a self-adaptive noise reduction method of a low-light-level video image based on gradient guide filtering, which reduces image noise, improves the signal-to-noise ratio of the image and improves the image quality on the premise of reducing human intervention as much as possible.
The purpose of the invention is realized by the following technical scheme:
the invention discloses a self-adaptive noise reduction method of a low-light-level video image based on gradient guide filtering, which comprises the following steps:
step 1, detecting a motion area in a low-light-level video image.
Step 1.1, respectively conducting guiding filtering on the current frame low-light-level video image and the previous frame low-light-level video image to obtain the low-frequency information of the video image.
The step 1.1 is realized by the following specific method: respectively conducting guide filtering on the current x-th frame low-light level video image g (x) and the previous frame low-light level video image g (x-1), taking an input image to be filtered as a guide image, and respectively obtaining low-frequency information g of the current frame and the previous frame video image after filteringL(x) And gL(x-1)。
And step 1.2, subtracting the low-frequency information of two adjacent frames of video images, and then taking an absolute value to obtain a low-frequency absolute difference image.
The step 1.2 is realized by the following specific method: obtaining low-frequency image information g of current frameL(x) And the last frame of low-frequency image information gL(x-1) and taking the absolute value to obtain a low-frequency absolute difference image DL(x) I.e. by
DL(x)＝|gL(x)-gL(x-1)| (1)
And step 1.3, firstly, obtaining the average value of the low-frequency absolute difference image, and then obtaining the threshold value of motion judgment according to the average value.
The step 1.3 is realized by the following specific method: firstly, obtaining low-frequency absolute difference image DL(x) Mean value of (2)LThe threshold value MGate for judging the movement is obtained by the following formula
MGate＝α·mDL(2)
In the formula, α is a constant, and the α value is adjusted empirically as appropriate for different motion situations.
And 1.4, comparing the pixels of the low-frequency absolute difference image with a motion judgment threshold respectively to obtain a motion area of the current frame image.
The step 1.4 is realized by the following specific method: low frequency absolute difference image DL(x) Respectively judging each pixel value and the motion threshold value MGate to obtain a motion region Motionmap (x) of the current frame image, namely
And 2, estimating the noise intensity of the low-light-level video image.
And 2.1, combining the motion areas of the current frame image to respectively obtain the static areas of the current frame image and the previous frame image.
The step 2.1 is realized by the following specific method: combining motion areas Motionmap (x) of the current x frame video image to respectively obtain static areas g of the current x frame image and the previous frame imageS(x) And gS(x-1), i.e.
And 2.2, performing difference between the static area of the current frame image and the static area of the previous frame image and taking an absolute value to obtain an absolute difference image of the static areas of adjacent frames.
The step 2.2 is realized by the following specific method: the current x-th frame still area image gS(x) And the previous frame still region image gS(x-1) making difference and taking absolute value to obtain absolute difference image D of adjacent frame static areaS(x) I.e. by
DS(x)＝|gS(x)-gS(x-1)| (5)
And 2.3, solving an effective average value of absolute difference images of static areas of adjacent frames as an estimated value of the noise intensity.
The step 2.3 is realized by the following specific method: obtaining absolute difference image D of static area of adjacent frameS(x) As the noise intensity estimation value NoiseV (x) of the xth frame video image, i.e. the effective mean value of
Where sum () is the sum of all pixels in the image matrix, and M and N are the number of columns and rows, respectively, of the video image.
And 3, performing adaptive gradient guided filtering and iterative guided filtering on the low-light-level video image so as to obtain a final filtered output image.
And 3.1, combining the motion area of the current frame image, and utilizing the weighted average of adjacent frames to obtain an initial guide image.
The step 3.1 is realized by the following specific method: combining the motion region Motionmap (x) of the current frame image, and utilizing the weighted average of the information of two adjacent frames to obtain the initial guide image I (x), namely
I(x)＝MotionMap(x)·g(x)+(1-MotionMap(x))·(m·g(x)+(1-m)·g(x-1)) (7)
In the formula, m is a constant, and 0< m <1, and m is set as appropriate empirically in various cases.
And 3.2, performing gradient guide filtering on the low-light-level video image according to the initial guide image, and setting filtering parameters by combining the noise intensity estimation of the current frame video image to obtain an initial filtering result image.
The step 3.2 is realized by the following specific method: according to the initial guide image I (x), gradient guide filtering is firstly carried out, meanwhile, filtering parameters are set according to the noise intensity estimated value NoiseV (x) of the current frame video image, and an initial filtering result image F is obtained0(x)。
In which GDGF () is the image gradient guided filter function, r0To be the size of the filtering window,0is a filter smoothing coefficient, wherein the filter parameter is related to the noise strength estimate noisev (x).
0＝k0·(NoiseV(x)/Lrange)2(9)
In the formula, k0The value is constant, and is adjusted according to the filter strength requirement, and Lrange is the dynamic range of the input image brightness.
And 3.3, performing iteration guide filtering on the initial filtering result image for multiple times, setting filtering parameters according to the noise intensity estimation value NoiseV (x) of the current frame video image, and obtaining a final filtered output image.
Step 3.3 the method is implemented specifically as follows: for the initial filtering result image F0(x) Performing iterative guided filtering for n times, wherein a guide image and an input image of each filtering are output images of the previous filtering, simultaneously setting filtering parameters according to a noise intensity estimated value NoiseV (x) of a video image of the current frame, and the filtering result of the nth time is an output image F after final filteringn(x) For the jth filtering, the following expression is given
Where GF () is the image-guided filter function, rjAndjrespectively the size of the filter window and the smoothing coefficient,the relationship with the initial filter parameters is as follows
j＝k·0/(j+1) (12)
rj＝r0/(j+1) (13)
Where k is a constant and is appropriately adjusted according to the filtering strength requirement.
Advantageous effects
1. The invention discloses a self-adaptive noise reduction method for a low-light-level video image based on gradient-guided filtering, which adopts gradient-guided filtering and iterative-guided filtering, so that the signal-to-noise ratio of the low-light-level video image subjected to noise reduction treatment is obviously improved, and the image quality is obviously improved.
2. The invention discloses a self-adaptive noise reduction method for low-light-level video images based on gradient-guided filtering, which is used for estimating the noise intensity of the low-light-level video images, so that the filtering parameters of the gradient-guided filtering and the iterative-guided filtering are self-adaptively set according to the noise intensity, further, the purpose of realizing rapid noise reduction almost without human intervention is realized.
3. The invention discloses a self-adaptive noise reduction method of a low-light-level video image based on gradient guided filtering, which is used for detecting a motion region of the low-light-level video image and further improving the precision of noise intensity estimation of the low-light-level video image, thereby better improving the filtering and noise reduction effects.
4. The adaptive noise reduction method for the low-light-level video image based on the gradient guide filtering has good practical application prospect in the field of low-light-level night vision imaging.
Drawings
FIG. 1 is a flow chart of a adaptive noise reduction method for low-light-level video images based on gradient-guided filtering according to the present invention;
FIG. 2 is a flow chart of a method of detecting a motion region in a low-light level video image according to the present invention;
FIG. 3 is a flow chart of a method of estimating noise intensity of low-light level video images according to the present invention;
fig. 4 is a flow chart of the method for adaptive gradient-guided filtering and iterative-guided filtering processing of low-light-level video images according to the invention.
Fig. 5 is a comparison graph of the adaptive noise reduction method of the embodiment of the present invention and the effect of the classical video image noise reduction method.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to specific embodiments and the accompanying drawings.
Example 1:
as shown in fig. 1, the adaptive noise reduction method for low-light-level video images based on gradient-guided filtering disclosed in this embodiment includes the following steps:
step 1, as shown in fig. 2, detecting a motion region in a low-light-level video image.
Step 1.1, respectively carrying out guide filtering on a current x-th frame low-light level video image g (x) and a previous frame low-light level video image g (x-1), taking an input image to be filtered as a guide image, and respectively obtaining low-frequency information g of the current frame and the previous frame of video image after filteringL(x) And gL(x-1). Wherein, the filter radius r is 6, and the filter smoothing coefficient λ is 0.4.
Step 1.2, obtaining low-frequency image information g of the current frameL(x) And the last frame of low-frequency image information gL(x-1) and taking the absolute value to obtain a low-frequency absolute difference image D according to the formula (1)L(x)。
Step 1.3, firstly, obtaining a low-frequency absolute difference image DL(x) Mean value of (2)LThen, the motion determination threshold value MGate is obtained according to equation (2), where the adjustment coefficient α is 4.
Step 1.4, low-frequency absolute difference image DL(x) Respectively, with the motion threshold value MGate, and obtains the motion region motionmap (x) of the current frame image according to equation (3).
And 2, estimating the noise intensity of the low-light-level video image as shown in FIG. 3.
Step 2.1, combining the motion area Motionmap (x) of the current x frame video image, and respectively obtaining the static areas g of the current x frame image and the previous frame image according to the formula (4)S(x) And gS(x-1)，Wherein the image matrix multiplication here is actually the respective multiplication of the values of the corresponding pixels.
Step 2.2, the current x-th frame of static area image gS(x) And the previous frame still region image gS(x-1) obtaining an absolute difference image D of the static area of the adjacent frame according to the formula (5)S(x)。
Step 2.3, obtaining the absolute difference image D of the static area of the adjacent frame according to the formula (6)S(x) Is taken as the noise intensity estimation value noissev (x) of the current xth frame video image.
And 3, as shown in fig. 4, performing adaptive gradient guided filtering and iterative guided filtering processing on the low-light-level video image so as to obtain a final filtered output image.
And 3.1, combining the motion region motionmap (x) of the current frame image, and obtaining an initial guide image I (x) by using weighted average of information of two adjacent frames according to the formula (7), wherein m is 0.5.
Step 3.2, performing adaptive gradient guiding filtering according to the formula (8) by combining the initial guiding image I (x), wherein a filtering parameter is set according to the noise intensity estimated value NoiseV (x) of the current frame video image, specifically as shown in the formula (9) and the formula (10), wherein k is taken00.8, for a common 8-bit image, Lrange 28-1-255, finding the initial filtering result image F0(x)。
Step 3.3, the initial filtering result image F0(x) Performing iterative guided filtering n times according to equation (11), where n is 2, and setting a filtering parameter according to a noise intensity estimation value noisev (x) of a current frame video image, as shown in equations (12) and (13), where k is 2, and a final nth filtering result is a final filtered output image Fn(x) As shown in fig. 5 (f).
As shown in fig. 5, a comparison example with the video image noise reduction method in the related art is performed under the same input conditions.
1. The image processing result of the classical median filtering noise reduction method is shown in fig. 5(c), in which the filtering window size is 3 × 3.
2. The image processing result of the classical bilateral filtering denoising method is shown in fig. 5(d), in which the size of the filtering window is 5 × 5, and the standard deviations of the definition domain and the value domain are 3 and 0.1, respectively.
3. The processing result of the non-local area averaging (NLMS) noise reduction method in document 1 is shown in fig. 5(e), where each parameter uses its recommended default value.
Document 1: buads A, Coll B, Morel J M.A view of Image DenoisingAlgorithms with a New One [ J ]. Sim Journal on Multiscale Modeling & Simulation,2010,4(2): 490-530).
The comparison of image effects in fig. 5 shows that the embodiment enables the signal-to-noise ratio of the low-light-level video image after noise reduction to be significantly improved, the image quality is the best, and the method has a good practical application prospect.
Example 2:
as shown in fig. 1, the adaptive noise reduction method for low-light-level video images based on gradient-guided filtering disclosed in this embodiment includes detecting a motion region of a low-light-level video image, estimating noise of the low-light-level video image, and performing adaptive gradient-guided filtering and iterative-guided filtering to obtain a final filtered output image. The signal-to-noise ratio of the low-light-level video image subjected to noise reduction processing is obviously improved, the image quality is obviously improved, meanwhile, human intervention is hardly needed, self-adaption noise reduction can be rapidly achieved, and the method has a good practical application prospect.
Fig. 2 shows a flow chart of a low-light-level video image motion region detection method according to the present embodiment. Firstly, respectively inputting current x-th frame and previous frame low-light level video image g (x) and g (x-1), respectively obtaining low-frequency information g of the two adjacent frames low-light level video image by guiding filteringL(x) And gL(x-1), wherein the input image to be filtered is taken as a guide image in guide filtering, the filtering radius r is 6, and the filtering smoothing coefficient λ is 0.4; then, the low frequency information g of two adjacent frames of imagesL(x) And gL(x-1) obtaining the Low frequency Absolute Difference image D according to equation (1)L(x) (ii) a Then, an absolute difference image D is obtainedL(x) Mean value of (2)LObtaining the motion judgment according to the formula (2)An off threshold value MGate, in which the adjustment factor α is 4, and finally a low-frequency absolute difference image DL(x) Each pixel of (2) is compared with a motion judgment threshold value MGate, and a specific comparison method is as shown in formula (3), and finally a motion region motionmap (x) of the current frame image is obtained.
Fig. 3 shows a flow chart of the low-light-level video image noise intensity estimation method according to the present embodiment. Firstly, respectively inputting the current x-th frame and the previous frame of low-light video images g (x) and g (x-1), combining the motion area Motionmap (x) of the current frame image, and respectively calculating the static areas g of the two adjacent frames of images according to the formula (4)S(x) And gS(x-1), wherein the image matrix multiplication here is actually the respective multiplication of the values of the corresponding pixels; then, the current x-th frame image still area gS(x) And the static area g of the previous frame imageS(x-1) obtaining a still region absolute difference image D according to equation (5)S(x) (ii) a Finally, the absolute difference image D of the static area is obtained according to the formula (6)S(x) Is taken as the noise intensity estimate value noissev (x) of the current xth frame low-light video image.
Fig. 4 shows a flowchart of a processing method of adaptive gradient guided filtering and iterative guided filtering according to the present embodiment. Firstly, inputting a current x-th frame and previous frame low-light video images g (x) and g (x-1), and simultaneously combining a current frame image motion region motionmap (x), calculating an initial guide image I (x) according to a formula (7), wherein m is 0.5; then, combining the initial guiding image i (x) to perform adaptive gradient guiding filtering according to the formula (8), wherein the filtering parameters are set according to the noise intensity estimation value noisev (x) of the current frame video image, as shown in the formulas (9) and (10), wherein k is taken00.8, for a common 8-bit image, Lrange 28-1-255, the filtering adaptability is realized in that the filtering parameter is automatically adjusted according to the noise intensity estimation, and the initial filtering result image F is obtained0(x) (ii) a Finally, the initial filtering result image F is processed0(x) Performing n iterative guided filtering according to equation (11), wherein the guided image and the input image of each filtering are the output images of the previous filtering, and n is generally 2, and the current frame video image is used as the basisSetting a filtering parameter by using the noise intensity estimated value noisevv (x), specifically as shown in formula (12) and formula (13), where k is 2, and the final nth filtering result is the final filtered output image Fn(x)。
The above detailed description is intended to illustrate the objects, aspects and advantages of the present invention, and it should be understood that the above detailed description is only exemplary of the present invention and is not intended to limit the scope of the present invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (4)
1. A self-adaptive noise reduction method for low-light-level video images based on gradient guided filtering is characterized by comprising the following steps: the method comprises the following steps:
step 1, detecting a motion area in a low-light-level video image;
step 1.1, respectively conducting guide filtering on a current frame low-light-level video image and a previous frame low-light-level video image to obtain low-frequency information of the video images;
step 1.2, taking an absolute value after the difference of the low-frequency information of two adjacent frames of video images to obtain a low-frequency absolute difference image;
step 1.3, firstly, obtaining the average value of the low-frequency absolute difference image, and then obtaining a motion judgment threshold value according to the average value;
step 1.4, comparing the pixels of the low-frequency absolute difference image with a motion judgment threshold respectively to obtain a motion area of the current frame image;
step 2, estimating the noise intensity of the low-light-level video image;
step 2.1, combining the motion area of the current frame image to respectively obtain the static areas of the current frame image and the previous frame image;
step 2.2, making a difference between the static area of the current frame image and the static area of the previous frame image and taking an absolute value to obtain an absolute difference image of the static areas of adjacent frames;
step 2.3, solving an effective average value of absolute difference images of static areas of adjacent frames as an estimated value of noise intensity;
step 3, performing adaptive gradient guided filtering and iterative guided filtering processing on the low-light-level video image so as to obtain a final filtered output image;
step 3.1, combining the motion area of the current frame image, and utilizing the weighted average of adjacent frames to obtain an initial guide image;
step 3.2, performing gradient guide filtering on the low-light-level video image according to the initial guide image, setting filtering parameters by combining the noise intensity estimation of the current frame video image, and obtaining an initial filtering result image;
and 3.3, performing iteration guide filtering on the initial filtering result image for multiple times, setting filtering parameters according to the noise intensity estimation value NoiseV (x) of the current frame video image, and obtaining a final filtered output image.
2. The adaptive noise reduction method for low-light-level video images based on gradient-guided filtering as claimed in claim 1, wherein:
the step 1.1 is realized by the following specific method: respectively conducting guide filtering on the current x-th frame low-light level video image g (x) and the previous frame low-light level video image g (x-1), taking an input image to be filtered as a guide image, and respectively obtaining low-frequency information g of the current frame and the previous frame video image after filteringL(x) And gL(x-1)；
The step 1.2 is realized by the following specific method: obtaining low-frequency information g of current frame video imageL(x) And low frequency information g of the previous frame video imageL(x-1) and taking the absolute value to obtain a low-frequency absolute difference image DL(x) I.e. by
DL(x)＝|gL(x)-gL(x-1)| (1)
The step 1.3 is realized by the following specific method: firstly, obtaining low-frequency absolute difference image DL(x) Mean value of (2)LThe threshold value MGate for judging the movement is obtained by the following formula
MGate＝α·mDL(2)
In the formula, alpha is a constant, and the value of alpha is properly adjusted according to different motion conditions according to experience;
the step 1.4 is realized by the following specific method: low frequency absolute difference image DL(x) Respectively judging each pixel value and the motion threshold value MGate to obtain a motion region Motionmap (x) of the current frame image, namely
3. The adaptive noise reduction method for low-light-level video images based on gradient-guided filtering as claimed in claim 1 or 2, wherein:
the step 2.1 is realized by the following specific method: combining motion areas Motionmap (x) of the current x frame video image to respectively obtain static areas g of the current x frame image and the previous frame imageS(x) And gS(x-1), i.e.
The step 2.2 is realized by the following specific method: the static area g of the current x frame imageS(x) And the static area g of the previous frame imageS(x-1) making difference and taking absolute value to obtain absolute difference image D of adjacent frame static areaS(x) I.e. by
DS(x)＝|gS(x)-gS(x-1)| (5)
The step 2.3 is realized by the following specific method: obtaining absolute difference image D of static area of adjacent frameS(x) As the noise intensity estimation value NoiseV (x) of the xth frame video image, i.e. the effective mean value of
Where sum () is the sum of all pixels in the image matrix, and M and N are the number of columns and rows, respectively, of the video image.
4. The adaptive noise reduction method for low-light-level video images based on gradient-guided filtering as claimed in claim 3, wherein:
the step 3.1 is realized by the following specific method: combining the motion region Motionmap (x) of the current frame image, and utilizing the weighted average of the information of two adjacent frames to obtain the initial guide image I (x), namely
I(x)＝MotionMap(x)·g(x)+(1-MotionMap(x))·(m·g(x)+(1-m)·g(x-1)) (7)
Wherein m is a constant and 0< m <1, and m is set as appropriate empirically from case to case;
the step 3.2 is realized by the following specific method: according to the initial guide image I (x), gradient guide filtering is firstly carried out, meanwhile, filtering parameters are set according to the noise intensity estimated value NoiseV (x) of the current frame video image, and an initial filtering result image F is obtained0(x)；
In which GDGF () is the image gradient guided filter function, r0Is the size of the initial filtering window and,0the initial filter smoothing coefficients, wherein the filter parameters are related to the noise intensity estimate NoiseV (x);
0＝k0·(NoiseV(x)/Lrange)2(9)
in the formula, k0The luminance is a constant and is properly adjusted according to the filtering strength requirement, and Lrange is the dynamic range of the input image luminance;
step 3.3 the method is implemented specifically as follows: for the initial filtering result image F0(x) Performing iterative guided filtering for n times, wherein a guide image and an input image of each filtering are output images of the previous filtering, simultaneously setting filtering parameters according to a noise intensity estimated value NoiseV (x) of a video image of the current frame, and the filtering result of the nth time is an output image F after final filteringn(x) For the jth filtering, the following expression is given
Where GF () is the image-guided filter function, rjAndjrespectively the size and the smooth coefficient of the filtering window;
j＝k·0/(j+1) (12)
rj＝r0/(j+1) (13)
where k is a constant and is appropriately adjusted according to the filtering strength requirement.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710932558.7A CN107610072B (en) | 2017-10-10 | 2017-10-10 | Adaptive noise reduction method for low-light-level video image based on gradient guided filtering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710932558.7A CN107610072B (en) | 2017-10-10 | 2017-10-10 | Adaptive noise reduction method for low-light-level video image based on gradient guided filtering |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107610072A CN107610072A (en) | 2018-01-19 |
| CN107610072B true CN107610072B (en) | 2020-08-11 |
Family
ID=61068101
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710932558.7A Active CN107610072B (en) | 2017-10-10 | 2017-10-10 | Adaptive noise reduction method for low-light-level video image based on gradient guided filtering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107610072B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109859126B (en) | 2019-01-17 | 2021-02-02 | 浙江大华技术股份有限公司 | Video noise reduction method and device, electronic equipment and storage medium |
| CN111402172B (en) * | 2020-03-24 | 2023-08-22 | 湖南国科微电子股份有限公司 | Image noise reduction method, system, equipment and computer readable storage medium |
| CN111652821B (en) * | 2020-06-05 | 2024-01-23 | 北京环境特性研究所 | Low-light video image noise reduction processing method, device and equipment based on gradient information |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103888638A (en) * | 2014-03-15 | 2014-06-25 | 浙江大学 | Time-space domain self-adaption denoising method based on guide filtering and non-local average filtering |
| CN104917933A (en) * | 2015-05-28 | 2015-09-16 | 南京理工大学 | Low-light video de-noising method based on phase consistency and scene motion estimation |
| CN105472204A (en) * | 2014-09-05 | 2016-04-06 | 南京理工大学 | Inter-frame noise reduction method based on motion detection |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8149336B2 (en) * | 2008-05-07 | 2012-04-03 | Honeywell International Inc. | Method for digital noise reduction in low light video |
-
2017
- 2017-10-10 CN CN201710932558.7A patent/CN107610072B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103888638A (en) * | 2014-03-15 | 2014-06-25 | 浙江大学 | Time-space domain self-adaption denoising method based on guide filtering and non-local average filtering |
| CN105472204A (en) * | 2014-09-05 | 2016-04-06 | 南京理工大学 | Inter-frame noise reduction method based on motion detection |
| CN104917933A (en) * | 2015-05-28 | 2015-09-16 | 南京理工大学 | Low-light video de-noising method based on phase consistency and scene motion estimation |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107610072A (en) | 2018-01-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2019091270A1 (en) | Image enhancement method and system | |
| 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 | |
| Kim et al. | A novel approach for denoising and enhancement of extremely low-light video | |
| CN107610072B (en) | Adaptive noise reduction method for low-light-level video image based on gradient guided filtering | |
| CN109658343B (en) | Underwater image enhancement method combining color conversion and adaptive exposure | |
| CN110675351B (en) | Marine image processing method based on global brightness adaptive equalization | |
| CN108133462B (en) | Single image restoration method based on gradient field region segmentation | |
| Liu et al. | Infrared image detail enhancement approach based on improved joint bilateral filter | |
| CN109934790A (en) | Infrared imaging system asymmetric correction method with adaptive threshold | |
| CN111861896A (en) | UUV-oriented underwater image color compensation and recovery method | |
| CN115578297A (en) | Generalized attenuation image enhancement method for self-adaptive color compensation and detail optimization | |
| Mu et al. | Low and non-uniform illumination color image enhancement using weighted guided image filtering | |
| CN109325918A (en) | Image processing method and device and computer storage medium | |
| CN109035228B (en) | X-ray image processing method of non-uniform-thickness component | |
| Zeng et al. | High dynamic range infrared image compression and denoising | |
| Lei et al. | Low-light image enhancement using the cell vibration model | |
| CN116029944B (en) | Self-adaptive contrast enhancement method and device for gray level image | |
| Wang et al. | Retinex algorithm on changing scales for haze removal with depth map | |
| Yoo et al. | Flicker removal for CMOS wide dynamic range imaging based on alternating current component analysis | |
| CN109859138B (en) | Infrared image enhancement method based on human visual characteristics | |
| CN110298796B (en) | Low-illumination image enhancement method based on improved Retinex and logarithmic image processing | |
| CN110992287B (en) | Method for clarifying non-uniform illumination video | |
| CN109636749B (en) | Image processing method | |
| CN113610734B (en) | Image edge processing method based on guided filtering and application |
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 |