CN104463813B - Infrared image noise reduction method based on noise recognition - Google Patents
Infrared image noise reduction method based on noise recognition Download PDFInfo
- Publication number
- CN104463813B CN104463813B CN201510004744.5A CN201510004744A CN104463813B CN 104463813 B CN104463813 B CN 104463813B CN 201510004744 A CN201510004744 A CN 201510004744A CN 104463813 B CN104463813 B CN 104463813B
- Authority
- CN
- China
- Prior art keywords
- noise
- membership
- pixel
- degree
- current pixel
- 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
Landscapes
- Image Processing (AREA)
- Facsimile Image Signal Circuits (AREA)
Abstract
The invention discloses an infrared image noise reduction method based on noise recognition. According to the method, the noise recognition basic thought is introduced, the degree of membership of current pixels based on the trimmed means and the degree of membership of the current pixels based on the gradient are calculated respectively, the degree of interference caused by noise on the current pixels is studied, whether the current pixels are noise pixels or not is judged by adopting united criteria, finally noise reduction is carried out according to judgment results, and infrared image noise reduction is achieved. The infrared image noise reduction method is small in calculation amount, easy to implement in real time and capable of protecting image borders and details more effectively compared with a traditional algorithm, considers image texture gradient information when noise reduction is carried out on noise points, and estimates original signals more accurately.
Description
Technical field
The invention belongs to infrared image processing technology field is and in particular to a kind of infrared image noise reduction based on Noise Identification
Method.
Background technology
Infrared imaging system strong antijamming capability, concealment is good, and air penetration capacity is strong, adapts to multiple special occasions,
In an increasingly wide range of applications in terms of scientific research, military affairs, medical science, industry, the many such as civilian.But due to Infrared Detectors life
The factor such as production. art, sensitivity and target and environmental radiation characteristic affects, and thermal-induced imagery compares visible images to comparison
Not high, present high background, low-contrast feature, noise is obvious, be unfavorable for that the later stage uses.Capture to make full use of
Information, suppresses noise, improves picture quality, be easy to higher level process it is necessary to carry out noise reduction process to infrared image.
Traditional image denoising method is broadly divided into 3 classes：Time domain noise reduction, spatial domain noise reduction, frequency domain noise reduction.Time domain noise reduction utilizes
In signal acquisition process, signal has stronger correlation, and noise has the characteristic of random distribution, the same pixel to interframe
Signal averagely to get the effect of noise reduction, but the scene in high-speed motion can cause image blurring and smear；Spatial domain drops
Make an uproar is to carry out noise reduction, typical method mean filter, medium filtering, dimension using the correlation that neighbor spatially has
Nanofiltration ripple etc., algorithm realizes simple, fast operation, shortcoming be can make while noise reduction image blurring, especially in object edge
At edge and details；Frequency domain noise reduction is to be converted image from spatial transform to frequency domain by image, filters representative with the method for filtering
The HFS of noise, but the noise close to some frequency contents and signal cannot remove, filtering threshold select bad to fall
Influential effect of making an uproar is very big.Additionally, also having some to combine the principle of above noise reduction, from many aspects, noise reduction is carried out to image, such as
Wavelet de-noising is just combined with the principle in spatial domain and frequency domain noise reduction, has the spy of good local character and multiscale analysis
Point, relatively effective signal of energy and noise separation are opened, but operand is big.
With the development of infrared imaging system, system imaging resolution ratio more and more higher, this allows for needing real-time processing
View data gets more and more.Due to limited system resources, some are computationally intensive, take storage resource more than noise reduction algorithm uncomfortable
With.In order to realize the real-time processing of infrared image, need to study a kind of amount of calculation little it is easy to the infrared image noise reduction of real-time implementation
Algorithm.
Content of the invention
In order to solve above-mentioned technical problem, the invention provides a kind of infrared image noise reduction side based on Noise Identification
Method.
The technical solution adopted in the present invention is：A kind of infrared image noise-reduction method based on Noise Identification, its feature exists
In comprising the following steps：
Step 1：Calculate the degree of membership based on trimmed mean for the current pixel, including noise degree of membership μ based on trimmed meann
(i, j) and signal degree of membership μ based on trimmed means(i, j), wherein i and j are current pixel place coordinate；
Step 2：Calculate the degree of membership based on gradient for the current pixel, including noise degree of membership S based on gradientn(i, j) with
And signal degree of membership S based on gradients(i,j)；
Step 3：According to step 1, step 2 calculated based on the degree of membership of trimmed mean with being subordinate to based on gradient
Degree, judges whether current pixel is noise pixel；
Step 4：If current pixel is noise pixel, noise reduction process is carried out to this pixel, after having processed, returns to step 1
Until having traveled through entire image；If current pixel is normal signal pixel, it is returned directly to step 1 until traveling through complete width figure
Picture.
Preferably, noise degree of membership μ based on trimmed mean described in step 1nThe computing formula of (i, j) is：
Wherein, f (i, j) is the grey scale pixel value of coordinate (i, j), and a, b are variable element, take empirical value, T according to experiment
(i, j) is the trimmed mean of 3 × 3 windows centered on current pixel, and computing formula is：
Wherein, Ai,jRepresent the set of all pixels gray scale in 3 × 3 windows centered on current pixel, pMax is set
Ai,jIn maximum gray scale, pMin be set Ai,jIn minimal gray；
Signal degree of membership μ based on trimmed mean described in step 1sThe computing formula of (i, j) is：
μs(i, j)=1- μn(i,j).
Preferably, noise degree of membership S based on gradient described in step 2nThe computing formula of (i, j) is：
Wherein, d represents direction, totally 8 directions, be respectively upper U, upper left LU, upper right RU, left L, right R, lower-left LD, lower D,
Bottom right RD, Fn d(i, j) is the noise degree of membership based on gradient on d direction for the pixel of coordinate (i, j), and computing formula is：
Wherein：
As d=U, LU, RU, L, R, LD, D, RD,Respectively equal to | f (i 1, j) f (i, j) |, | f (i 1, j 1) f
(i,j)|、|f(i–1,j+1)–f(i,j)|、|f(i,j–1)–f(i,j)|、|f(i,j+1)–f(i,j)|、|f(i+1,j–1)–f
(i,j)|、|f(i+1,j)–f(i,j)|、|f(i+1,j+1)–f(i,j)|；
As d=U, LU, RU, L, R, LD, D, RD,Respectively equal to | f (i 1, j 1) f (i, j 1) |, | f (i, j
1)–f(i+1,j)|、|f(i–1,j)–f(i,j–1)|、|f(i+1,j–1)–f(i+1,j)|、|f(i–1,j+1)–f(i–1,j)|、
|f(i+1,j)–f(i,j+1)|、|f(i+1,j+1)–f(i,j+1)|、|f(i,j+1)–f(i–1,j)|；
As d=U, LU, RU, L, R, LD, D, RD,Respectively equal to | f (i 1, j+1) f (i, j+1) |, | f (i 1, j
1)–f(i,j)|、|f(i,j+1)–f(i+1,j)|、|f(i–1,j–1)–f(i–1,j)|、|f(i+1,j+1)–f(i+1,j)|、|f
(i,j–1)–f(i–1,j)|、|f(i+1,j–1)–f(i,j–1)|、|f(i+1,j)–f(i,j–1)|；
Function β () is defined as follows：
Wherein, c, d are variable element, take empirical value according to experiment；
Signal degree of membership S based on gradient described in step 2sThe computing formula of (i, j) is：
Wherein, Fs d(i, j) is the signal degree of membership based on gradient on d direction for the current pixel, and computing formula is：
Preferably, judging whether current pixel is noise pixel described in step 3, its method is：Work as μn(i,j)·
Sn(i, j) is more than or equal to μs(i,j)·SsWhen (i, j), judge current pixel as noise pixel；Work as μn(i,j)·Sn(i, j) is less than
μs(i,j)·SsWhen (i, j), judge current pixel as normal signal pixel.
Preferably, noise reduction process is carried out to this pixel described in step 4, its concrete methods of realizing be U, LU, RU,
Order is found in this 8 directions of L, R, LD, D, RDMinimum direction, remembers that this direction is dmin, then current picture
The gray scale that plain gray scale equation below calculates is replaced：
Method introduces the basic thought of Noise Identification, calculate respectively current pixel based on trimmed mean and be based on
The degree of membership of gradient, investigates the degree that current pixel is subject to noise jamming, judges whether current pixel is noise using joint criterion
Pixel, carries out noise reduction finally according to judged result, realizes the noise reduction to infrared image.The present invention has advantages below：
1st, amount of calculation little it is easy to real-time implementation.Because algorithm only carries out statistical computation to single pixel and its 8 neighborhoods, calculate
Method complexity low it is not necessary to take a large amount of storage resources to be used for caching frequency domain data, therefore amount of calculation little it is easy to real-time implementation；
2nd, relatively conventional algorithm can more effectively protect image border and details.Due to primarily looking in noise reduction process
Current pixel is subject to the degree of noise jamming, judges whether current pixel is noise pixel using joint criterion, improves noise and sentence
Disconnected science and accuracy, it is to avoid image blurring to introduce after the unnecessary noise reduction of non-noise point；
3rd, have also contemplated that the texture gradient information of image during noise reduction is carried out to noise spot, more accurately to former
Signal is had to be estimated.
Brief description
Fig. 1：It is method of the present invention flow chart.
Fig. 2：It is the original infrared image of 512 × 640 resolution ratio of the embodiment of the present invention.
Fig. 3：It is that the artificial on the original infrared image of 512 × 640 resolution ratio of the embodiment of the present invention adds random noise
Infrared image afterwards.
Fig. 4：It is 3 × 3 windows and eight direction schematic diagrams of the embodiment of the present invention.
Fig. 5：It is the image after this method is processed of the embodiment of the present invention.
Fig. 6：It is the image after classical median filter method is processed of the embodiment of the present invention.
Specific embodiment
Understand for the ease of those of ordinary skill in the art and implement the present invention, below in conjunction with the accompanying drawings and embodiment is to this
Bright be described in further detail it will be appreciated that described herein enforcement example be merely to illustrate and explain the present invention, not
For limiting the present invention.
Ask for an interview Fig. 1, the present invention is mainly made up of 4 steps：Calculate the degree of membership based on trimmed mean for the current pixel, calculating
If the degree of membership based on gradient for the current pixel, judging according to calculated degree of membership whether current pixel is noise pixel
Judge that current pixel for noise pixel, then carries out noise reduction process to this pixel.
Ask for an interview Fig. 2 and Fig. 3, be the original infrared image of 512 × 640 resolution ratio of the embodiment of the present invention.And 512
Infrared image after artificial interpolation random noise on the original infrared image of × 640 resolution ratio.Below with this original infrared image
As a example, each step of the present invention is described in detail：
Step (1) begins stepping through entire image with 3 × 3 window.Calculate current window center pixel and be based on trimmed mean
Degree of membership, including noise degree of membership μ based on trimmed meann(i, j) and signal degree of membership μ based on trimmed means(i,
J), wherein i and j is current pixel place coordinate, as shown in Figure 4；
Noise degree of membership μ based on trimmed meann(i, j) adopts below equation to calculate：
Wherein, f (i, j) is the grey scale pixel value of coordinate (i, j), and a, b are variable element, take empirical value according to experiment, if
System noise levels are relatively low, and in order to retain details as far as possible, noise judges should be loose, so a, b can suitably take large values (greatly
In maximum gray scale 1%), a, b are to take empirical value 20,80 (maximum gray scale is 16384), T according to experiment respectively in this example
(i, j) is the trimmed mean of 3 × 3 windows centered on current pixel, and computing formula is as follows：
Wherein, Ai,jRepresent the set of all pixels gray scale in 3 × 3 windows centered on current pixel, pMax is set
Ai,jIn maximum gray scale, pMin be set Ai,jIn minimal gray；
Signal degree of membership μ based on trimmed means(i, j) adopts below equation to calculate：
μs(i, j)=1- μn(i,j)；
Step (2) calculates the degree of membership based on gradient for the current pixel, including noise degree of membership S based on gradientn(i, j) with
And signal degree of membership S based on gradients(i,j)；
Noise degree of membership S based on gradientn(i, j) adopts below equation to calculate：
Wherein, d represents direction, totally 8 directions, be respectively upper U, upper left LU, upper right RU, left L, right R, lower-left LD, lower D,
Bottom right RD, Fn d(i, j) is the noise degree of membership based on gradient on d direction for the current pixel, and computing formula is as follows：
Wherein, as d=U, LU, RU, L, R, LD, D, RD,Respectively equal to | f (i 1, j) f (i, j) |, | f (i 1,
j–1)–f(i,j)|、|f(i–1,j+1)–f(i,j)|、|f(i,j–1)–f(i,j)|、|f(i,j+1)–f(i,j)|、|f(i+1,
j–1)–f(i,j)|、|f(i+1,j)–f(i,j)|、|f(i+1,j+1)–f(i,j)|；As d=U, LU, RU, L, R, LD, D, RD
When,Respectively equal to | f (i 1, j 1) f (i, j 1) |, | f (i, j 1) f (i+1, j) |, | f (i 1, j) f (i, j 1) |, |
f(i+1,j–1)–f(i+1,j)|、|f(i–1,j+1)–f(i–1,j)|、|f(i+1,j)–f(i,j+1)|、|f(i+1,j+1)–f
(i,j+1)|、|f(i,j+1)–f(i–1,j)|；As d=U, LU, RU, L, R, LD, D, RD,Respectively equal to | f (i 1, j+
1)–f(i,j+1)|、|f(i–1,j–1)–f(i,j)|、|f(i,j+1)–f(i+1,j)|、|f(i–1,j–1)–f(i–1,j)|、|f
(i+1,j+1)–f(i+1,j)|、|f(i,j–1)–f(i–1,j)|、|f(i+1,j–1)–f(i,j–1)|、|f(i+1,j)–f(i,
J 1) |, function β () is defined as follows：
Wherein, c, d are variable element, take empirical value according to experiment, if system noise levels are relatively low, in order to protect as far as possible
Stay details, noise judges should be loose, so c, d can suitably take large values (more than the 0.5% of maximum gray scale), in this example c,
D takes empirical value 20,55 according to experiment respectively；
Signal degree of membership S based on gradients(i, j) adopts below equation to calculate：
Wherein, Fs d(i, j) is the signal degree of membership based on gradient on d direction for the current pixel, and computing formula is as follows：
According to step (1), the calculated degree of membership of step (2), step (3) judges whether current pixel is noise pixel；
According to step (1), calculated 4 degrees of membership of step (2), work as μn(i,j)·Sn(i, j) is more than or equal to μs(i,
j)·SsWhen (i, j), judge current pixel as noise pixel；Work as μn(i,j)·Sn(i, j) is less than μs(i,j)·SsWhen (i, j),
Judge current pixel as normal signal pixel；
Step (4), if step (3) judges current pixel for noise pixel, carries out noise reduction process to this pixel, processes
Step (1) is returned to until having traveled through entire image after complete；If step (3) judges current pixel for normal signal pixel, directly
Take back step (1) until having traveled through entire image.Wherein concrete noise-reduction method is as follows：
Order is found in this 8 directions of U, LU, RU, L, R, LD, D, RDMinimum direction, remembers this
Direction is dmin, then the gray scale replacement that current pixel gray scale equation below calculates：
Ask for an interview accompanying drawing 5, be the image processing through the present invention；Ask for an interview accompanying drawing 6, be to process through classical median filtering algorithm
Image afterwards.Contrast Fig. 2, Fig. 3 can see, Fig. 3 compares Fig. 2 and with the addition of a lot of noises, picture quality degradation；Comparison diagram
2nd, Fig. 5 and Fig. 6 can see, Fig. 5 has only remained small part noise, and image substantially returns to close with Fig. 2, and on Fig. 6 still
Can significantly see there are many noise residuals, image detail obscures, and it is good that noise reduction is not so good as Fig. 5.
It should be appreciated that the part that this specification does not elaborate belongs to prior art.
It should be appreciated that the above-mentioned description for preferred embodiment is more detailed, can not therefore be considered to this
The restriction of invention patent protection scope, those of ordinary skill in the art, under the enlightenment of the present invention, is weighing without departing from the present invention
Profit requires under protected ambit, can also make replacement or deform, each fall within protection scope of the present invention, this
Bright scope is claimed should be defined by claims.
Claims (3)
1. a kind of infrared image noise-reduction method based on Noise Identification is it is characterised in that comprise the following steps：
Step 1：Calculate the degree of membership based on trimmed mean for the current pixel, including noise degree of membership μ based on trimmed meann(i,j)
And signal degree of membership μ based on trimmed means(i, j), wherein i and j are current pixel place coordinate；
Wherein said noise degree of membership μ based on trimmed meannThe computing formula of (i, j) is：
Wherein, f (i, j) is the grey scale pixel value of coordinate (i, j), and a, b are variable element, take empirical value, T (i, j) according to experiment
It is the trimmed mean of 3 × 3 windows centered on current pixel, computing formula is：
Wherein, Ai,jRepresent the set of all pixels gray scale in 3 × 3 windows centered on current pixel, pMax is set Ai,j
In maximum gray scale, pMin be set Ai,jIn minimal gray；
Described signal degree of membership μ based on trimmed meansThe computing formula of (i, j) is：
μs(i, j)=1- μn(i,j)；
Step 2：Calculate the degree of membership based on gradient for the current pixel, including noise degree of membership S based on gradientn(i, j) and it is based on
Signal degree of membership S of gradients(i,j)；
Wherein said noise degree of membership S based on gradientnThe computing formula of (i, j) is：
Wherein, d represents direction, totally 8 directions, is upper U, upper left LU, upper right RU, left L, right R, lower-left LD, lower D, bottom right respectively
RD,For the noise degree of membership based on gradient on d direction for the pixel of coordinate (i, j), computing formula is：
Wherein：
As d=U, LU, RU, L, R, LD, D, RD,Respectively equal to | f (i 1, j) f (i, j) |, | f (i 1, j 1) f (i, j)
|、|f(i–1,j+1)–f(i,j)|、|f(i,j–1)–f(i,j)|、|f(i,j+1)–f(i,j)|、|f(i+1,j–1)–f(i,j)
|、|f(i+1,j)–f(i,j)|、|f(i+1,j+1)–f(i,j)|；
As d=U, LU, RU, L, R, LD, D, RD,Respectively equal to | f (i 1, j 1) f (i, j 1) |, | f (i, j 1) f (i+
1,j)|、|f(i–1,j)–f(i,j–1)|、|f(i+1,j–1)–f(i+1,j)|、|f(i–1,j+1)–f(i–1,j)|、|f(i+1,
j)–f(i,j+1)|、|f(i+1,j+1)–f(i,j+1)|、|f(i,j+1)–f(i–1,j)|；
As d=U, LU, RU, L, R, LD, D, RD,Respectively equal to | f (i 1, j+1) f (i, j+1) |, | f (i 1, j 1) f
(i,j)|、|f(i,j+1)–f(i+1,j)|、|f(i–1,j–1)–f(i–1,j)|、|f(i+1,j+1)–f(i+1,j)|、|f(i,
j–1)–f(i–1,j)|、|f(i+1,j–1)–f(i,j–1)|、|f(i+1,j)–f(i,j–1)|；
Function β () is defined as follows：
Wherein, c, d are variable element, take empirical value according to experiment；
Described signal degree of membership S based on gradientsThe computing formula of (i, j) is：
Wherein, Fs d(i, j) is the signal degree of membership based on gradient on d direction for the current pixel, and computing formula is：
Step 3：According to step 1, the calculated degree of membership based on trimmed mean of step 2 and the degree of membership based on gradient, sentence
Whether disconnected current pixel is noise pixel；
Step 4：If current pixel be noise pixel, noise reduction process is carried out to this pixel, return to after having processed step 1 until
Travel through entire image；If current pixel is normal signal pixel, it is returned directly to step 1 until having traveled through entire image.
2. the infrared image noise-reduction method based on Noise Identification according to claim 1 it is characterised in that：Institute in step 3
That states judges whether current pixel is noise pixel, and its method is：Work as μn(i,j)·Sn(i, j) is more than or equal to μs(i,j)·Ss
When (i, j), judge current pixel as noise pixel；Work as μn(i,j)·Sn(i, j) is less than μs(i,j)·SsWhen (i, j), judge to work as
Preceding pixel is normal signal pixel.
3. the infrared image noise-reduction method based on Noise Identification according to claim 1 it is characterised in that：Institute in step 4
That states carries out noise reduction process to this pixel, and its concrete methods of realizing is to find in this 8 directions of U, LU, RU, L, R, LD, D, RD
OrderMinimum direction, remembers that this direction is dmin, then the gray scale that current pixel gray scale is calculated with equation below
Replace：
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510004744.5A CN104463813B (en) | 2015-01-07 | 2015-01-07 | Infrared image noise reduction method based on noise recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510004744.5A CN104463813B (en) | 2015-01-07 | 2015-01-07 | Infrared image noise reduction method based on noise recognition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104463813A CN104463813A (en) | 2015-03-25 |
| CN104463813B true CN104463813B (en) | 2017-02-22 |
Family
ID=52909801
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510004744.5A Active CN104463813B (en) | 2015-01-07 | 2015-01-07 | Infrared image noise reduction method based on noise recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104463813B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102509265A (en) * | 2011-11-02 | 2012-06-20 | 天津理工大学 | Digital image denoising method based on gray value difference and local energy |
| CN103337053A (en) * | 2013-06-13 | 2013-10-02 | 华中科技大学 | Switching non-local total variation based filtering method for image polluted by salt and pepper noise |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100170010A1 (en) * | 2001-04-18 | 2010-07-01 | Alain Goossens | Use of genes encoding membrane transporter pumps to stimulate the production of secondary metabolites in biological cells |
-
2015
- 2015-01-07 CN CN201510004744.5A patent/CN104463813B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102509265A (en) * | 2011-11-02 | 2012-06-20 | 天津理工大学 | Digital image denoising method based on gray value difference and local energy |
| CN103337053A (en) * | 2013-06-13 | 2013-10-02 | 华中科技大学 | Switching non-local total variation based filtering method for image polluted by salt and pepper noise |
Non-Patent Citations (2)
| Title |
|---|
| 一种基于噪声检测的图像去噪算法;郭承湘 等;《计算机工程》;20121130;第38卷(第21期);第218-220页 * |
| 一种基于噪声点检测的自适应中值滤波方法;高克芳，郭建钢;《福建农林大学学报（自然科学版）》;20090531;第38卷(第3期);第333-335页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104463813A (en) | 2015-03-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Noise detection and image denoising based on fractional calculus | |
| Ma et al. | An effective fusion defogging approach for single sea fog image | |
| CN109377450B (en) | Edge protection denoising method | |
| CN110782477A (en) | Moving target rapid detection method based on sequence image and computer vision system | |
| CN103295225B (en) | Train bogie edge detection method under the conditions of low-light | |
| Wang et al. | Low-light image joint enhancement optimization algorithm based on frame accumulation and multi-scale Retinex | |
| CN107818547B (en) | A kind of minimizing technology towards the spiced salt and Gaussian mixed noise in twilight image sequence | |
| Zhang et al. | Decision-based non-local means filter for removing impulse noise from digital images | |
| Yan et al. | Method to Enhance Degraded Image in Dust Environment. | |
| CN111768350B (en) | Infrared image enhancement method and system | |
| CN110047055A (en) | A kind of enhancing of infrared image details and denoising method | |
| CN105427255A (en) | GRHP based unmanned plane infrared image detail enhancement method | |
| Yu et al. | Image and video dehazing using view-based cluster segmentation | |
| Lin et al. | Using a hybrid of fuzzy theory and neural network filter for single image dehazing | |
| Hsieh et al. | Fast and robust infrared image small target detection based on the convolution of layered gradient kernel | |
| Wu et al. | Research on rail surface defect detection method based on UAV images | |
| CN113793278A (en) | Improved remote sensing image denoising method with minimized weighted nuclear norm and selectively enhanced Laplace operator | |
| Pan et al. | Single-image dehazing via dark channel prior and adaptive threshold | |
| Kim et al. | Cross fusion-based low dynamic and saturated image enhancement for infrared search and tracking systems | |
| CN111652809A (en) | Infrared image noise suppression method for enhancing details | |
| CN111461999A (en) | SAR image speckle suppression method based on super-pixel similarity measurement | |
| Zhang et al. | Background suppression algorithm for infrared images based on Robinson guard filter | |
| CN104463813B (en) | Infrared image noise reduction method based on noise recognition | |
| Ni et al. | Online low-light sand-dust video enhancement using adaptive dynamic brightness correction and a rolling guidance filter | |
| Cheng et al. | Video dehazing for surveillance unmanned aerial vehicle |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |