CN107274365A - A kind of mine image intensification method based on unsharp masking and NSCT algorithms - Google Patents
A kind of mine image intensification method based on unsharp masking and NSCT algorithms Download PDFInfo
- Publication number
- CN107274365A CN107274365A CN201710449966.7A CN201710449966A CN107274365A CN 107274365 A CN107274365 A CN 107274365A CN 201710449966 A CN201710449966 A CN 201710449966A CN 107274365 A CN107274365 A CN 107274365A
- Authority
- CN
- China
- Prior art keywords
- image
- pixel
- value
- enhancing
- nsct
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 63
- 230000000873 masking effect Effects 0.000 title claims abstract description 14
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 13
- 230000002708 enhancing effect Effects 0.000 claims abstract description 29
- 239000003245 coal Substances 0.000 claims abstract description 24
- 238000006243 chemical reaction Methods 0.000 claims abstract description 8
- 238000000354 decomposition reaction Methods 0.000 claims description 5
- 239000003623 enhancer Substances 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 238000012937 correction Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000000205 computational method Methods 0.000 claims description 2
- 238000005070 sampling Methods 0.000 abstract description 4
- 230000000007 visual effect Effects 0.000 abstract description 4
- 238000003384 imaging method Methods 0.000 abstract description 2
- 238000004519 manufacturing process Methods 0.000 description 8
- 230000009466 transformation Effects 0.000 description 8
- 238000005286 illumination Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000004807 localization Effects 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 230000003321 amplification Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000011426 transformation method Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/73—Deblurring; Sharpening
- G06T5/75—Unsharp masking
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The present invention proposes a kind of mine image intensification method based on unsharp masking and NSCT algorithms, and this method is the image enchancing method that a kind of Unsharp Masking Method is combined with NSCT (the contourlet conversion of non-lower sampling), including：Image is divided into high, normal, basic three kinds of level of detail, replacement median filter process is weighted to low details, moderate enhancing is done to high details area, centering details area, which is done, largely to be strengthened；Algorithm is strengthened using the high frequency imaging based on NSCT to image again, high frequency coefficient is classified based on bayes threshold estimation methods, strong edge is determined, weak edge and noise are handled different coefficients respectively.This method avoid the image blurring same topic that denoising is brought, and human-eye visual characteristic is met to the enhancing of image, both underground coal mine low-light (level), the characteristics of image of low contrast had been improved, be not in overshoot again, it not only avoid the loss of image detail, enhancing effect preferably, and inhibits the enhancing of noise.
Description
Technical field
The present invention relates to field of image enhancement, and in particular to a kind of method for enhancing underground coal mine image.
Background technology
Coal is the most important energy of China, comes from its economic price and abundant reserves, in particular for generating electricity.
The energy of China 80% comes from fire coal.But the exploitation in colliery really has very big difficulty, reason mainly has：First, China
Natural calamity is serious；2nd, the technological process of production is complicated；3rd, production equipment and mode fall behind.First two reason is substantially can not
Change.The third reason can be by improving production and the difficulty of coal mining being reduced using sophisticated equipment.But it is due to
The small buesiness management technology shortcoming of the more especially many of enterprise of the coal production of China, the mode of production falls behind, so as to lead
The generation of the coal mining accident of many has been caused, useful monitoring information can not be provided afterwards preferably to implement rescue.Cause
This is necessary the video monitoring of underground coal mine, and this is the important leverage and emergency management and rescue necessary means of mine safety production, in spy
Under different subsurface environment, the even or even completely black environment of uneven illumination causes picture contrast small, image blurring unclear, Er Qie
Substantial amounts of noise is mixed into video image acquisition transmitting procedure, causes video image picture coarse, poor quality, video pictures matter
Amount directly affects the timely acquisition of mine disaster information, thus image enhaucament become it is particularly important.
Image enchancing method mainly includes spatial domain and the major class of transform domain two.Spatial-domain algorithm directly enters on the original image
Row computing.Conventional method has greyscale transformation method, histogram equalization method, based on Enhancement Method theoretical Retinex, gradient field
Image enchancing method, the image enchancing method based on wavelet transformation, the image enchancing method based on high-pass filter, unsharp cover
Mould image enchancing method.Above method is in the enhancing direction of image using quite varied.But because underground coal mine shooting environmental is disliked
Bad, the image of shooting has following characteristics：(1) dust concentration is big under mine, high humidity, and camera be difficult to it is automatic poly-
It is burnt.(2) middle illumination fluctuation is frequent under mine, and for example underground coal mine large scale equipment is a lot, and grid disturbance is big, causes illumination to fluctuate.
(3) because the image of collection is reflecting to form by light, if the uneven illumination being irradiated on scenery is even, it will be obtained on image
The stronger part of illumination is brighter, and the weaker part of illumination is than dark.Therefore, because the particular surroundings of underground coal mine, is commonly used
Image processing method be difficult to the requirement for meeting the authenticity of image, reliability, make information recognition occur because of difficulty, be unfavorable for mine
Under safe and stable production.In order to overcome problem above, people have induced one transform domain method, it is relatively more representational including
Fourier transformation, method of wavelet transformation etc..Fourier transformation is solved with the spectral characteristic of signal to be difficult to solve in many time domains
Certainly the problem of, but the conversion does not have the ability of Time-Frequency Localization, easily causes image detail information loss.And wavelet transformation has
There is Time-Frequency Localization characteristic while standby spectral characteristic, Fourier transformation presence is solved well only has frequency domain processing
Ability does not possess the single characteristic of temporal processing ability, and the gradient of image provides more direct than histogram, more spaces
Information.But wavelet transformation is more sensitive to a singularity, and limited to edge direction ability to express.It is many that Do et al. proposes one kind
Yardstick geometrical analysis instrument-contourlet is converted, and it is a kind of multiscale analysis method, can effectively portray high dimensional information
Feature, but be due to the presence of sampling operation, contourlet conversion lacks translation invariance, the meeting when carrying out image denoising enhancing
Produce Pseudo-Gibbs artifacts.
Therefore, because special image-context, the method for carrying out image using common image procossing mode is difficult to meet
The requirement of the authenticity, reliability of image, makes information recognition occur because of difficulty, is unfavorable for the safe and stable production under mine.
The content of the invention
It is an object of the invention to provide a kind of method that underground coal mine strengthens image, for solving existing coal mine figure below
In image intensifying method, unsharp masking technology exist and overshoot phenomenon very sensitive to noise the problem of, while making up high frequency
The problem of enhancing that part cannot get well.The image blurring same topic that denoising is brought is avoided, and human eye is met to the enhancing of image
Visual characteristic, had both improved underground coal mine low-light (level), the characteristics of image of low contrast, was not in overshoot again, it is to avoid image is thin
The loss of section, enhancing effect preferably, and inhibits the enhancing of noise.
To achieve the above object, the solution of the present invention is：A kind of Unsharp Masking Method and NSCT be (non-lower sampling
Contourlet is converted, Non-Subsampled Contourlet Transform) image enchancing method that is combined, step
It is as follows：
(1) level of detail to input picture judges, divides the image into high, medium and low three level of detail；
(2) image progress denoising of the median filter method to the low details area is substituted using based on weighting；
(3) increase using based on the unsharp mask method image different degrees of to high, medium and low three details areas progress
By force, i.e.,：Low details area, which is not done, strengthens or strengthens very little, and moderate enhancing, centering detail areas are done to high details area
Domain, which is done, largely to strengthen；
(4) image obtained to step (3) carries out NSCT conversion, and pixel is passed through to the high-frequency sub-band coefficient after decomposition
Average value and maximum are used enters row coefficient classification based on bayes threshold values (Bayes's threshold value), and high-frequency sub-band is divided into noise, by force
Edge and weak edge, are strengthened noise, strong edge and weak edge by correction function.
Further, it is to the method that image detail degree is judged in described step (1)：
The local variance v (i, j) of each pixel is calculated first, and two threshold values T1 and T2, and T1 ＜ T2, part side are set
Difference represents the level of detail of pixel；Then, basic, normal, high three detail areas are divided into according to v (i, j) big wisp image,
I.e.：If v (i, j) ＜ T1It is then low details area；If T1＜ v (i, j) ＜ T2It is then middle details area；If v (i, j) ＞ T2It is then
High details area.
Further, local variance v (i, j) computational methods of each pixel are：
Described local variance is defined as the variance of all pixels in a given window, i.e., one (2n+1) × (2n+
1) window, f (i, j) is the gray value of window center pixel, and the local variance of pixel (i, j) is：
Wherein, f (k, l) is the gray value at pixel (k, l) place,The local mean value of pixel (i, j) is represented, n is represented
Integer.
Further, the weighting replacement median filter method described in step (2) is：
(1) window size is set as (2n+1) × (2n+1), is slided along column direction to the right, when sliding into next pixel,
The row of the window left side one will be removed, and a row add new pixel value on the right of window, if the pixel value that the left side one is arranged is a1, a2,
a3.....a2n+1, it is b that the right one, which arranges the pixel value newly added,1, b2, b3.....b2n+1；
(2) if pixel value meets a1=b1, a2=b2, a3=b3.....a2n+1=b2n+1Relation, then in the window
It is worth for former intermediate value；Otherwise, the value newly added is substituted into non-equivalence, calculates the average and intermediate value of the window；
(3) intermediate value and average obtained to step (2) is weighted, wherein, the weight of intermediate value is 0.3, the weight of average
For 0.7, using the value after weighting as output valve, the value of window center is replaced by；
(4) sliding window, the denoising of whole image is completed according to step (1)-(3).
Further, the calculation formula of the unsharp masking algorithm described in step (3) is：Y (i, j)=x (i, j)+γ z
(i, j), wherein, x (i, j) is received image signal；Z (i, j) (only goes for the output of signal after denoising to low details area
Make an uproar), γ is a direct proportion factor, can control the intensity of image enhaucament, and y (i, j) is enhanced image.
Further, the method described in step (4) is：Calculate in same layer, different sub-band is in same position pixel
Average valueAnd the maximum Pmax of all pixels point, NSTC high frequency coefficients can be divided by choosing a suitable threshold value
Class：Wherein, nose represents noise, and ste represents strong edge, and wke represents weak side
Edge, TijCalculated using based on Bayes threshold estimations method, c is regulation parameter；
Described is based on Bayes threshold estimation methods：Wherein σ is the Noise Variance Estimation on first layer subband,
Formula can be usedRepresent.
Correction function is：
Wherein, t is the conversion coefficient of input original image, and f (t) is the enhancing at weak edge
Function, coefficient processing is carried out using improved Sigmoid functions (s sigmoid growth curves)：
Wherein, a takes the parameter between [0,1], and K is enhancer.
The beneficial effect that the present invention reaches：Because denoising can typically make image blur, the details of image is lost, and is made
Make an uproar and strengthen and be difficult to reach the effect relatively optimized therebetween, the present invention strengthens Wavelet Denoising Method and unsharp masking using a kind of
Image, is divided into basic, normal, high three regions by the new processing method that method is combined according to level of detail, only in the low thin of image
Save region (that is, flat site) and carry out denoising, because according to human-eye visual characteristic, human eye is made an uproar to image flat site
The noise of acoustic ratio detail section is more sensitive, and the subregion of feelings Condition hypographs as one is flat, the noise quilt of this sampled images
Relative " removal ", and details area, while introducing NSCT algorithms, is made up the above method and high frequency imaging is increased by intact reservation
It is strong not enough.This method avoids the image blurring same topic that denoising is brought, and meets human-eye visual characteristic to the enhancing of image, both changes
It has been apt to underground coal mine low-light (level), the characteristics of image of low contrast, has been not in overshoot again, it is to avoid the loss of image detail, enhancing
Effect preferably, and inhibits the enhancing of noise.
Brief description of the drawings
Fig. 1 is the enhancing underground coal mine image method flow diagram of the invention based on Unsharp Masking Method and NSCT.
Fig. 2 is that unsharp masking of the present invention strengthens the flow chart of image.
Embodiment
The present invention will be further described in detail below in conjunction with the accompanying drawings.
1st, image detail is judged
(1) the local variance v (i, j) of each pixel is calculated, i.e., one (2n+1) × (2n+1) windows, f (i, j) is window
The gray value of central pixel point, the local variance of pixel (i, j) is：
Wherein, f (k, l) is the gray value at pixel (k, l) place,The local mean value of pixel (i, j) is represented, n is represented
Integer.
The local mean value of pixelFor：
V (i, j) just represents the level of detail of pixel (i, j).
(2) two threshold values T1 and T2, and T1 ＜ T2 are set；
(3) basic, normal, high three detail areas are divided into according to v (i, j) big wisp image, i.e.,：If v (i, j) ＜ T1It is then
Low details area；If T1＜ v (i, j) ＜ T2It is then middle details area；If v (i, j) ＞ T2It is then high details area.
2nd, weighting replaces median filtering algorithm
Operation method is：
(1) window size is set as (2n+1) × (2n+1), is slided along column direction to the right, when sliding into next pixel,
The row of the window left side one will be removed, and a row add new pixel value on the right of window, if the pixel value that the left side one is arranged is a1, a2,
a3.....a2n+1, it is b that the right one, which arranges the pixel value newly added,1, b2, b3.....b2n+1；；
(2) if pixel value meets a1=b1, a2=b2, a3=b3.....a2n+1=b2n+1Relation, then in the window
It is worth for former intermediate value；Otherwise, the value newly added is substituted into non-equivalence, calculates the average and intermediate value of the window；
(3) intermediate value and average obtained to step (2) is weighted, wherein, the weight of intermediate value is 0.3, the weight of average
For 0.7, using the value after weighting as output valve, the value of window center is replaced by；
(4) sliding window, the denoising of whole image is completed according to step (1)-(3)
This method drastically increases arithmetic speed in terms of noise is handled, and computational complexity is reduced, to realtime graphic
Noise processed has bigger meaning.
3rd, Unsharp Masking Method
The calculation formula of unsharp masking algorithm is：Y (i, j)=x (i, j)+γ z (i, j)
Wherein, x (i, j) is received image signal；Z (i, j) (is only carried out for the output of signal after denoising to low details area
Denoising), γ is a direct proportion factor, can control the intensity of image enhaucament, and y (i, j) is enhanced image.
If input picture x (i, j) obtains image M (i, j) after being handled through median filtering algorithm.
Enhancer γ may be defined as the nonlinear function γ (i, j) of image detail degree, i.e.,：
In formula, γ 1, γ 2, γ 3 is the enhancer of the basic, normal, high details area of image respectively, and the γ of 0 ＜ γ, 1 γ 2
3 ＜ 1.
If input picture x (i, j) obtains image M (i, j) after being handled through median filtering algorithm.
Finally, the enhanced image of denoising is obtained, rewritable formula (1) is y (i, j)=M (i, j)+γ (i, j) z (i, j),
As shown in Figure 2.
4th, the HFS enhancing method based on NSCT
The image obtained to the above method carries out NSCT decomposition, to each high-frequency sub-band coefficient after decomposition, calculates same layer
It is interior, average value of the different sub-band in same position pixelAnd the maximum Pmax of all pixels point, choose one properly
Threshold value NSTC high frequency coefficients can be classified：
Wherein, nose represents noise, and ste represents strong edge, and wke represents weak edge, TijRepresent i-th layer, the son on j directions
Band threshold value.It is regulation parameter between one [1,5] that c, which is,.σ is the Noise Variance Estimation on first layer subband, can use formulaRepresent.
Wherein, median represents to take intermediate value, and x represents a high-pass filtering coefficient.X is expressed as input picture in NSCT domains
The coefficient of smallest dimension (decomposition first layer).
Wherein TijCalculating, estimated using sample, obtain a Bayes threshold estimation that can be adaptively adjusted with yardstick
Formula：
The target of enhancing underground coal mine image is the detailed information such as the weak edge of amplification, while suppress noise, therefore under acquisition is non-
Sampling contourlet transformation coefficient correction function be：
Wherein, t is the conversion coefficient of input original image, and f (t) is the enhancing function at weak edge, using improved
Sigmoid functions carry out coefficient processing：
Wherein, a takes the parameter between [0,1], and K is enhancer.
Claims (8)
1. a kind of mine image intensification method based on unsharp masking and NSCT algorithms, it is characterised in that described image increases
Strong method is the method being combined based on Unsharp Masking Method and NSCT, and step is as follows：
(1) level of detail to input picture judges, divides the image into high, medium and low three level of detail；
(2) image progress denoising of the median filter method to the low details area is substituted using based on weighting；
(3) different degrees of image enhaucament is carried out to high, medium and low three details areas using based on unsharp mask method,
I.e.：Low details area, which is not done, strengthens or strengthens very little, and moderate enhancing is done to high details area, and centering details area is done
Largely strengthen；
(4) image obtained to step (3) carries out NSCT conversion, to high-frequency sub-band coefficient being averaged by pixel after decomposition
Value and maximum are used enters row coefficient classification based on bayes threshold values, and high-frequency sub-band is divided into noise, strong edge and weak edge, led to
Crossing correction function strengthens noise, strong edge and weak edge.
2. enhancing underground coal mine image method according to claim 1, it is characterised in that in described step (1), to figure
As the method that level of detail is judged is：The local variance v (i, j) of each pixel is calculated first, and two threshold value T1 are set
And T2, and T1<T2, local variance is the level of detail for representing pixel；
Then, basic, normal, high three detail areas are divided into according to v (i, j) big wisp image, i.e.,：If v (i, j) ＜ T1Then to be low
Details area；If T1＜ v (i, j) ＜ T2It is then middle details area；If v (i, j) ＞ T2It is then high details area.
3. enhancing underground coal mine image method according to claim 2, it is characterised in that the part side of each pixel
Poor v (i, j) computational methods are：
Described local variance is defined as the variance of all pixels in a given window, i.e., one (2n+1) × (2n+1) windows
Mouthful, f (i, j) is the gray value of window center pixel, and the local variance of pixel (i, j) is：
Wherein, f (k, l) is the gray value at pixel (k, l) place,The local mean value of pixel (i, j) is represented, n represents whole
Number.
4. enhancing underground coal mine image method according to claim 1, it is characterised in that in described step (2), weighting
Substituting median filter method is：
(1) window size is set as (2n+1) × (2n+1), is slided along column direction to the right, when sliding into next pixel, window
The row of the left side one will be removed, and a row add new pixel value on the right of window, if the pixel value that the left side one is arranged is a1,a2,
a3.....a2n+1, it is b that the right one, which arranges the pixel value newly added,1,b2,b3.....b2n+1；
(2) if pixel value meets a1=b1,a2=b2,a3=b3.....a2n+1=b2n+1Relation, then the intermediate value of the window be
Former intermediate value；Otherwise, the value newly added is substituted into non-equivalence, calculates the average and intermediate value of the window；
(3) intermediate value and average obtained to step (2) is weighted, wherein, the weight of intermediate value is 0.3, and the weight of average is
0.7, using the value after weighting as output valve, it is replaced by the value of window center；
(4) sliding window, the denoising of whole image is completed according to step (1)-(3).
5. enhancing underground coal mine image method according to claim 1, it is characterised in that described step (3) it is anti-sharp
Change mask algorithm calculation formula be：Y (i, j)=x (i, j)+γ z (i, j), wherein, x (i, j) is received image signal；z(i,
J) it is the output of signal after denoising, wherein only carrying out denoising to low details area, γ is a direct proportion factor, can be with control figure
The intensity of image intensifying, y (i, j) is enhanced image.
6. enhancing underground coal mine image method according to claim 1, it is characterised in that the method for described step (4)
For：Calculate in same layer, average value of the different sub-band in same position pixelAnd the maximum P of all pixels pointmax,
NSTC high frequency coefficients can be classified by choosing a suitable threshold value：
Wherein, nose represents noise, and ste represents strong edge, and wke represents weak edge, TijUsing based on Bayes threshold estimation method meters
Calculate, c is regulation parameter.
7. enhancing underground coal mine image method according to claim 6, it is characterised in that described based on Bayes threshold values
The estimation technique is：Wherein σ is the Noise Variance Estimation on first layer subband, uses formulaRepresent, its
Middle x represents a high-pass filtering coefficient, is the coefficient that input picture smallest dimension in NSCT domains decomposes first layer.
8. enhancing underground coal mine image method according to claim 1, it is characterised in that in described step (4), amendment
Function is：
Wherein, t is the conversion coefficient of input original image, and f (t) is the enhancing function at weak edge, using improved Sigmoid letters
Number carries out coefficient processing, i.e.,：
Wherein, a takes the parameter between [0,1], and K is enhancer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710449966.7A CN107274365A (en) | 2017-06-15 | 2017-06-15 | A kind of mine image intensification method based on unsharp masking and NSCT algorithms |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710449966.7A CN107274365A (en) | 2017-06-15 | 2017-06-15 | A kind of mine image intensification method based on unsharp masking and NSCT algorithms |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107274365A true CN107274365A (en) | 2017-10-20 |
Family
ID=60067713
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710449966.7A Pending CN107274365A (en) | 2017-06-15 | 2017-06-15 | A kind of mine image intensification method based on unsharp masking and NSCT algorithms |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107274365A (en) |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107730473A (en) * | 2017-11-03 | 2018-02-23 | 中国矿业大学 | A kind of underground coal mine image processing method based on deep neural network |
| CN107846317A (en) * | 2017-11-10 | 2018-03-27 | 曹典 | Router Display panel style switching system |
| CN108961179A (en) * | 2018-06-19 | 2018-12-07 | 上海中和软件有限公司 | A kind of medical image after-treatment system and its application method |
| CN109325922A (en) * | 2018-09-12 | 2019-02-12 | 深圳开阳电子股份有限公司 | A kind of image self-adapting enhancement method, device and image processing equipment |
| CN109472755A (en) * | 2018-11-06 | 2019-03-15 | 武汉高德智感科技有限公司 | A kind of domain infrared image logarithm LOG Enhancement Method |
| CN109636736A (en) * | 2018-11-06 | 2019-04-16 | 中国航空工业集团公司洛阳电光设备研究所 | A kind of wide dynamic range infrared image contrast Enhancement Method |
| CN109787904A (en) * | 2017-11-10 | 2019-05-21 | 曹一典 | Router Display panel style switching method |
| CN109872289A (en) * | 2019-02-19 | 2019-06-11 | 重庆邮电大学 | Image enchancing method based on improved non-downsampling Contourlet conversion |
| CN109919865A (en) * | 2019-02-20 | 2019-06-21 | 中国科学院上海微***与信息技术研究所 | A kind of image multilayer detail enhancing method |
| CN110211058A (en) * | 2019-05-15 | 2019-09-06 | 南京极目大数据技术有限公司 | A kind of data enhancement methods of medical image |
| CN111145114A (en) * | 2019-12-19 | 2020-05-12 | 腾讯科技（深圳）有限公司 | Image enhancement method and device and computer readable storage medium |
| CN111738943A (en) * | 2020-06-12 | 2020-10-02 | 吉林大学 | Medical image enhancement method combining spatial domain and frequency domain |
| CN111986098A (en) * | 2020-05-14 | 2020-11-24 | 南京航空航天大学 | Passive terahertz image enhancement method containing fixed background |
| CN112200735A (en) * | 2020-09-18 | 2021-01-08 | 安徽理工大学 | Temperature identification method based on flame image and control method of low-concentration gas combustion system |
| CN112258430A (en) * | 2020-10-30 | 2021-01-22 | 长光卫星技术有限公司 | Universal correction method for remote sensing image radiation nonuniformity |
| CN113487517A (en) * | 2021-07-26 | 2021-10-08 | 中国人民解放军国防科技大学 | Unmanned aerial vehicle target detection method, device and equipment based on image enhancement |
| CN116681696A (en) * | 2023-07-27 | 2023-09-01 | 东莞雅达高精密塑胶模具有限公司 | Mold quality monitoring method for automatic production equipment |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103413279A (en) * | 2013-08-24 | 2013-11-27 | 西安电子科技大学 | SAR image denoising method based on AD-NSCT algorithm |
| CN103473748A (en) * | 2013-09-23 | 2013-12-25 | 中国矿业大学(北京) | Method for enhancing underground coal mine image |
-
2017
- 2017-06-15 CN CN201710449966.7A patent/CN107274365A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103413279A (en) * | 2013-08-24 | 2013-11-27 | 西安电子科技大学 | SAR image denoising method based on AD-NSCT algorithm |
| CN103473748A (en) * | 2013-09-23 | 2013-12-25 | 中国矿业大学(北京) | Method for enhancing underground coal mine image |
Non-Patent Citations (3)
| Title |
|---|
| 张林等: "基于非降采样Contourlet变换的非线性图像增强新算法", 电子与信息学报 * |
| 赵高长等: "改进的中值滤波算法在图像去噪中的应用", 应用光学 * |
| 阿依古力・吾布力;贾振红;覃锡忠;杨杰;NIKOLA KASABOV;: "基于剪切波变换的反锐化掩膜遥感图像增强", 计算机工程与设计 * |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107730473A (en) * | 2017-11-03 | 2018-02-23 | 中国矿业大学 | A kind of underground coal mine image processing method based on deep neural network |
| CN109787904A (en) * | 2017-11-10 | 2019-05-21 | 曹一典 | Router Display panel style switching method |
| CN107846317A (en) * | 2017-11-10 | 2018-03-27 | 曹典 | Router Display panel style switching system |
| CN108961179B (en) * | 2018-06-19 | 2022-10-18 | 上海中和软件有限公司 | Medical image post-processing system and using method thereof |
| CN108961179A (en) * | 2018-06-19 | 2018-12-07 | 上海中和软件有限公司 | A kind of medical image after-treatment system and its application method |
| CN109325922B (en) * | 2018-09-12 | 2022-03-11 | 深圳开阳电子股份有限公司 | Image self-adaptive enhancement method and device and image processing equipment |
| CN109325922A (en) * | 2018-09-12 | 2019-02-12 | 深圳开阳电子股份有限公司 | A kind of image self-adapting enhancement method, device and image processing equipment |
| CN109636736A (en) * | 2018-11-06 | 2019-04-16 | 中国航空工业集团公司洛阳电光设备研究所 | A kind of wide dynamic range infrared image contrast Enhancement Method |
| CN109472755A (en) * | 2018-11-06 | 2019-03-15 | 武汉高德智感科技有限公司 | A kind of domain infrared image logarithm LOG Enhancement Method |
| CN109636736B (en) * | 2018-11-06 | 2022-11-01 | 中国航空工业集团公司洛阳电光设备研究所 | Wide dynamic range infrared image contrast enhancement method |
| CN109872289A (en) * | 2019-02-19 | 2019-06-11 | 重庆邮电大学 | Image enchancing method based on improved non-downsampling Contourlet conversion |
| CN109919865A (en) * | 2019-02-20 | 2019-06-21 | 中国科学院上海微***与信息技术研究所 | A kind of image multilayer detail enhancing method |
| CN110211058A (en) * | 2019-05-15 | 2019-09-06 | 南京极目大数据技术有限公司 | A kind of data enhancement methods of medical image |
| CN111145114A (en) * | 2019-12-19 | 2020-05-12 | 腾讯科技（深圳）有限公司 | Image enhancement method and device and computer readable storage medium |
| CN111986098A (en) * | 2020-05-14 | 2020-11-24 | 南京航空航天大学 | Passive terahertz image enhancement method containing fixed background |
| CN111986098B (en) * | 2020-05-14 | 2024-04-30 | 南京航空航天大学 | Passive terahertz image enhancement method containing fixed background |
| CN111738943A (en) * | 2020-06-12 | 2020-10-02 | 吉林大学 | Medical image enhancement method combining spatial domain and frequency domain |
| CN111738943B (en) * | 2020-06-12 | 2023-12-05 | 吉林大学 | Medical image enhancement method combining spatial domain and frequency domain |
| CN112200735A (en) * | 2020-09-18 | 2021-01-08 | 安徽理工大学 | Temperature identification method based on flame image and control method of low-concentration gas combustion system |
| CN112258430A (en) * | 2020-10-30 | 2021-01-22 | 长光卫星技术有限公司 | Universal correction method for remote sensing image radiation nonuniformity |
| CN112258430B (en) * | 2020-10-30 | 2022-06-21 | 长光卫星技术股份有限公司 | Universal correction method for remote sensing image radiation nonuniformity |
| CN113487517A (en) * | 2021-07-26 | 2021-10-08 | 中国人民解放军国防科技大学 | Unmanned aerial vehicle target detection method, device and equipment based on image enhancement |
| CN116681696A (en) * | 2023-07-27 | 2023-09-01 | 东莞雅达高精密塑胶模具有限公司 | Mold quality monitoring method for automatic production equipment |
| CN116681696B (en) * | 2023-07-27 | 2023-10-20 | 东莞雅达高精密塑胶模具有限公司 | Mold quality monitoring method for automatic production equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107274365A (en) | A kind of mine image intensification method based on unsharp masking and NSCT algorithms | |
| Ma et al. | Contrast limited adaptive histogram equalization-based fusion in YIQ and HSI color spaces for underwater image enhancement | |
| Wang et al. | Dehazing for images with large sky region | |
| CN108564597B (en) | Video foreground object extraction method fusing Gaussian mixture model and H-S optical flow method | |
| CN110400267A (en) | A kind of preprocess method based on inspection image | |
| CN104574293A (en) | Multiscale Retinex image sharpening algorithm based on bounded operation | |
| CN104463804A (en) | Image enhancement method based on intuitional fuzzy set | |
| CN103295225B (en) | Train bogie edge detection method under the conditions of low-light | |
| Pan et al. | De-scattering and edge-enhancement algorithms for underwater image restoration | |
| CN103473748A (en) | Method for enhancing underground coal mine image | |
| CN101493939A (en) | Method for detecting cooked image based on small wave domain homomorphic filtering | |
| CN104657951A (en) | Multiplicative noise removal method for image | |
| CN106971166A (en) | The image pre-processing method and system of parking stall detection | |
| CN105427255A (en) | GRHP based unmanned plane infrared image detail enhancement method | |
| CN110298796A (en) | Based on the enhancement method of low-illumination image for improving Retinex and Logarithmic image processing | |
| Bansal et al. | A review of image restoration based image defogging algorithms | |
| CN117252773A (en) | Image enhancement method and system based on self-adaptive color correction and guided filtering | |
| CN115587945A (en) | High dynamic infrared image detail enhancement method, system and computer storage medium | |
| Fu et al. | An anisotropic Gaussian filtering model for image de-hazing | |
| Zhao et al. | A multi-scale U-shaped attention network-based GAN method for single image dehazing | |
| Wen et al. | A survey of image dehazing algorithm based on retinex theory | |
| Jia et al. | A wavelet-based approach to improve foggy image clarity | |
| CN116309233A (en) | Infrared and visible light image fusion method based on night vision enhancement | |
| Zhang | Image enhancement method based on deep learning | |
| Tang et al. | Sky-preserved image dehazing and enhancement for outdoor scenes |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20171020 |
|
| WD01 | Invention patent application deemed withdrawn after publication |