CN107871311A - A kind of image enhaucament and fusion method applied to cmos image sensor - Google Patents
A kind of image enhaucament and fusion method applied to cmos image sensor Download PDFInfo
- Publication number
- CN107871311A CN107871311A CN201711058317.0A CN201711058317A CN107871311A CN 107871311 A CN107871311 A CN 107871311A CN 201711058317 A CN201711058317 A CN 201711058317A CN 107871311 A CN107871311 A CN 107871311A
- Authority
- CN
- China
- Prior art keywords
- image
- pixel
- wavelet
- coefficient
- high frequency
- 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
- 238000007500 overflow downdraw method Methods 0.000 title claims abstract description 14
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 26
- 230000009467 reduction Effects 0.000 claims abstract description 15
- 230000009466 transformation Effects 0.000 claims abstract description 11
- 238000000034 method Methods 0.000 claims description 13
- 230000004927 fusion Effects 0.000 claims description 11
- 238000004458 analytical method Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 6
- 238000003709 image segmentation Methods 0.000 claims description 4
- 238000003672 processing method Methods 0.000 claims description 4
- 238000003384 imaging method Methods 0.000 claims description 3
- 230000015572 biosynthetic process Effects 0.000 claims 1
- 230000000149 penetrating effect Effects 0.000 claims 1
- 230000035699 permeability Effects 0.000 abstract description 4
- 230000004075 alteration Effects 0.000 abstract 1
- 230000000694 effects Effects 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000002708 enhancing effect Effects 0.000 description 3
- 238000011084 recovery Methods 0.000 description 3
- 230000011218 segmentation Effects 0.000 description 2
- 230000002123 temporal effect Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000000593 degrading effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
- 230000000007 visual effect Effects 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
- G06T5/40—Image enhancement or restoration using histogram techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20048—Transform domain processing
- G06T2207/20064—Wavelet transform [DWT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of image enhaucament and fusion method applied to cmos image sensor, specifically, for the view data of cmos image sensor output, dividing processing is carried out to image first, piece image is divided into two images, the image new to two width carries out picture quality enhancement processing again, modulated the contrast of image, then carry out image noise reduction processing, reduce the noise that prime image procossing is brought, to meet the needs of subsequent treatment is to picture quality, then wavelet decomposition is carried out respectively to two images, finally the image after small echo processing is merged and wavelet inverse transformation is handled, with the final output image after being reduced.A kind of image enhaucament and fusion method applied to cmos image sensor of the invention solve the image fault caused by noise or aberration problems, enhance image quality quality and image permeability, keeps image edge detailss, while realizes and realize on limited platform resource and handle in real time.
Description
Technical field
The present invention relates to digital image processing field, more particularly to a kind of image enhaucament applied to cmos image sensor
And fusion method.
Background technology
The optical signal of reality scene is converted into electric signal by cmos image sensor, during conversion and processing, no
Can avoid must be by noise pollution so that image quality decrease, can distort distortion when serious, influence visual effect.Therefore, in order to
Image fault is avoided, it is preferable to obtain restored image, it is necessary to noise reduction is carried out to image of being made an uproar and enhancing is handled.
At present the image procossing mode complicated algorithm in video source modeling field, be unfavorable for the system integration, take more hardware
Resource and image processing effect is bad.
The content of the invention
In view of this, the problem of present invention seek to address that in background technology, propose that one kind can apply to cmos image biography
The image enhaucament and fusion method of sensor.The present invention adopts the technical scheme that, for the image of input, is designed using pipelining-stage
Thought, split by image, image information reconstruct and fusion, realize enhancing, noise reduction and the reduction treatment of image, the technical office
Manage better performances and beneficial to realization.
In order to achieve the above object, the concrete technical scheme that the present invention takes is as follows:
A kind of image enhaucament and fusion method applied to cmos image sensor of the present invention, it is characterised in that the party
Method comprises the following steps:
1) for the view data of cmos image sensor output, dividing processing is carried out to image first, by piece image
It is divided into two images;
2) image new to two width carries out picture quality enhancement processing again, is modulated the contrast of image, improves image
Permeability;
3) image noise reduction processing is then carried out, reduces the noise that prime image procossing is brought;
4) and then to two images merge respectively, wavelet analysis and inversion process, with the output image after being restored.
Further, the image segmentation processing method in step 1 is as follows:The view data exported for imaging sensor,
Maximum gray scale numerical value is defined as max_pixel, and the scope of image pixel value falls between { 0, max_pixel };By pixel most
The half being worth greatly judged original image and split as cut off value, i.e., be divided into the pixel value of piece image 0,
Max_pixel/2 } and { max_pixel/2, max_pixel } two sections;By sentencing to the pixel value pixel of image
Disconnected, if pixel value pixel meets 0≤pixel≤max_pixel/2, all pixels in the range of this form image A,
If pixel value pixel meets max_pixel/2≤pixel≤max_pixel, all pixels in the range of this form figure
As B, image A and B are two images that original image is divided into, and array sizes are the half of original image.
Further, the image noise reduction processing step in step 3 is:A the small echo of the layer from l to N) is carried out to plus noise image
Decompose;B) each layer high frequency coefficient of the wavelet decomposition of artwork is modeled with normal distribution respectively;) pair C plus image of making an uproar high frequency system
Number histogram carries out Histogram Matching operation, makes its Histogram Matching in modeling;D) using the low frequency coefficient of n-th layer and from l
The high frequency coefficient progress wavelet reconstruction after Histogram Matching operation processing to n-th layer obtains restored image.
Further, it is divided into step 4 the step of image co-registration:
A) to the image M of two width difference noises1And M2Two layers of wavelet decomposition are carried out, if the coefficient of wavelet decomposition of two images
Respectively D1And D (P)2(P), wherein P=(i, j, k, l), (i, j) are the locus of decomposition coefficient, and k is Decomposition order, and l is
Subband frequency range, (l=1,2,3,4);
B) D is compared1And D (P)2(P), according to corresponding comparison rule, two wavelet coefficients are merged, drawn new small
Wave system number expression Dr (P), including two steps:The first step, by D1And D (P)2(P) high frequency coefficient that first layer decomposes is compared,
If two coefficients are identical, then retain in the wavelet coefficient Dr (P) after corresponding fusion, otherwise will be with Dr (P)
Part is set to 0 corresponding to the coefficient;Second step, compare D1And D (P)2(P) high frequency coefficient that the second layer decomposes, if two coefficients
The absolute value of difference is less than threshold value T ({ T=TH, TV, TD, wherein THFor horizontal direction high frequency coefficient threshold value, TVFor vertical direction high frequency
Coefficient threshold, TDFor diagonally opposed high frequency coefficient threshold value), then the wavelet coefficient after corresponding fusion takes being averaged for two coefficients
Value, otherwise equally sets to 0;
C) second layer low-frequency wavelet coefficients after closing take D1And D (P)2(P) low frequency coefficient that the second layer decomposes is averaged
It is low;
D) corresponding wavelet inverse transformation is carried out to the wavelet coefficient Dr (P) after fusion, you can the image that must be merged.
A kind of image enhaucament and fusion method applied to cmos image sensor of the present invention, possess following beneficial to effect
Fruit:
1) processing procedure of the algorithm includes image segmentation, picture quality enhancement, image noise reduction, small echo processing and image co-registration,
The advantages of algorithm is high treating effect, modularized design and easily transplanting.
2) algorithm is based on wavelet analysis, less beneficial to realizing and taking resource;Wavelet transformation is a kind of time scale point
Analysis and the new technology of multiresolution analysis, because its time window and frequency window are all variable, therefore all have in the domain of time-frequency two and characterize letter
The ability of number local feature.By space time and the partial transformation of frequency, information can be effectively extracted from signal, is utilized
The computings such as flexible and translation carry out multiple dimensioned refinement to function or signal and analyzed, and solve the indeterminable many of Fourier and ask
Topic.Wavelet transformation has relatively low frequency resolution and higher temporal resolution in low frequency part, HFS have compared with
Low temporal resolution and higher frequency resolution, it is well suited for detecting the transient state abnormal phenomena carried secretly in normal signal and shows
Its composition, so there is important application value in signal analysis, image procossing etc..
Image interfusion method based on the wavelet analysis effectively difference of same image must can degrade figure is carried out at fusion
Reason to realize the target of recovery, the advantages of this method be image restoration effect preferably, modularized design and easily transplanting.
3) dividing processing is carried out to image, the image new to two width carries out contrast picture quality enhancement processing, realized to image
Double-histogram modulation;Image noise reduction based on wavelet transformation is carried out to the image after segmentation, reduces the image quality of introducing
Noise, to meet the needs of subsequent treatment is to picture quality;
4) wavelet decomposition is carried out respectively to the image after segmentation, the image after finally handling small echo is merged and small echo
Inversion process, to realize the noise reduction of image and recovery.
Brief description of the drawings
The invention will be further described with reference to the accompanying drawings and examples.
Fig. 1 is the image enhaucament and blending algorithm flow chart of the present invention;
The image quality that Fig. 2 is the present invention strengthens flow chart;
Fig. 3 is the image wavelet decomposing schematic representation of the present invention;
Fig. 4 is the image co-registration schematic diagram of the present invention.
Embodiment
The specific implementation process of the present invention is described in detail below in conjunction with accompanying drawing, so that present disclosure is more clear
Chu is understandable.Certainly the invention is not limited in the specific embodiment, the general replacement known to those skilled in the art
Cover within the scope of the present invention.
A kind of image enhaucament and fusion method applied to cmos image sensor of the present invention, specifically, for
The view data of cmos image sensor output, carries out dividing processing to image first, piece image is divided into two images, then
The image new to two width carries out picture quality enhancement processing, is modulated the contrast of image, then carries out image noise reduction processing, drop
The noise that low prime image procossing is brought, to meet the needs of subsequent treatment is to picture quality, then two images are entered respectively
Row wavelet decomposition, finally the image after small echo processing is merged and wavelet inverse transformation is handled, with final after being reduced
Output image.It is described in detail below for processing method framework.
A kind of image enhaucament and fusion method applied to cmos image sensor of the present invention, this method include following step
Suddenly:
1) for the view data of cmos image sensor output, dividing processing is carried out to image first, by piece image
It is divided into two images;Image segmentation processing method is as follows:For the view data of imaging sensor output, maximum gray scale numerical value
Max_pixel is defined as, the scope of image pixel value falls between { 0, max_pixel };By the half of pixel maximum
As cut off value, original image is judged and split, i.e., by the pixel value of piece image be divided into { 0, max_pixel/2 } and
{ max_pixel/2, max_pixel } two sections;By judging the pixel value pixel of image, if pixel value pixel
Meet 0≤pixel≤max_pixel/2, then all pixels in the range of this form image A, if pixel value pixel meets
Max_pixel/2≤pixel≤max_pixel, then all pixels in the range of this form image B, image A and B are
Two images that original image is divided into, array sizes are the half of original image.
2) image new to two width carries out picture quality enhancement processing again, is modulated the contrast of image, improves image
Permeability;Picture quality enhancement uses improved histogram equalization method, i.e., the closeer part of intensity profile is carried out into section stretching,
Being distributed sparse part, then section can be compressed, so that picture contrast obtains larger enhancing on the whole.Specific handling process
For:As shown in Fig. 2 being first segmented each gray level of image, and count each section of pixel quantity;Then according to pel array
Size and gray level, calculate the width of unit brightness, then calculate each section of pixel quantity and the brightness width ratio of unit brightness
Example, and the accumulated value of this ratio is calculated, the tuned slope value K, i.e. brightness width ratio and block are then calculated according to total hop count
Several ratio;Finally grayscale pixel value F (x, y), F (x, y) meet formula in each section in statistics new histogram:F(x,y)
=(f (x, y)-f0)*K+g0(wherein f0For every section of initial value, K is brightness width ratio and the ratio of total hop count, g0It is bright for each section
Spend the accumulated value of width ratio), it is achieved thereby that section is stretched or compressed in brightness space distribution, complete brightness region
Between redistribution so that evenly, permeability is more preferable for image brightness distribution.
3) image noise reduction processing is then carried out, reduces the noise that prime image procossing is brought;Specifically, image noise reduction is handled
Step is:A the wavelet decomposition of the layer from l to N) is carried out to plus noise image;B) to each layer high frequency coefficient of the wavelet decomposition of artwork
Modeled respectively with normal distribution;C) pair high frequency coefficient histogram for adding image of making an uproar carries out Histogram Matching operation, makes its histogram
It is matched with modeling;D) the height after Histogram Matching operation processing using the low frequency coefficient of n-th layer and from l to n-th layer
Frequency coefficient carries out wavelet reconstruction and obtains restored image.Image noise reduction principle based on wavelet transformation is:Noise is not affected by a width
For the image of pollution, the distribution of its high frequency coefficient can be modeled with the less normal distribution of appropriate variance.When image is because of superposition
Noise and when degrading, its high frequency coefficient is distributed as the larger normal distribution of variance, thus pair plus figure of making an uproar wavelet decomposition high frequency coefficient
Histogram Matching operation is carried out, makes the high frequency coefficient histogram of its histogram approximation artwork, the purpose for restoring artwork can be reached.
Recovery effect can be better if more accurate to the modeling of original image high frequency coefficient, and the algorithm maintains figure as far as possible while denoising
As details.
4) and then to two images merge respectively, wavelet analysis and inversion process, with the output image after being restored;
Wavelet transformation is a kind of time scale analysis method of signal, signal decomposition into low frequency al and high frequency d1 two parts, is being decomposed
In, the information lost in low frequency al is captured by high frequency dl.Low frequency a2 and high frequency d2 are resolved into next layer of decomposition, and by al
Two parts, the information that loses is captured low frequency by high frequency d2 in the heart, and so on go down, deeper decomposition can be carried out.Such as
Image wavelet decomposing schematic representation shown in Fig. 3, L represent low frequency, H table high frequencies, and subscript 1,2 represents that one-level or two level are decomposed.
Image co-registration based on wavelet analysis can be handled to reach restored image the different figures of being made an uproar of same image
Purpose.According to wavelet decomposition theory, above-mentioned wavelet decomposition is passed through to two images, it can be deduced that a series of subband data,
Wherein except the data of low frequency sub-band are in addition to, the data of other subbands are distributed in null value or so, and wherein absolute value is larger
Coefficient correspond to gray scale mutation part, i.e., corresponding to the notable feature (such as edge, noise) in original image.Therefore to two
The coefficient of wavelet decomposition of width image is merged, then carries out inverse wavelet transform, you can the restored image after being merged, is treated
Journey is divided into as shown in figure 4, specifically, the step of image co-registration:
A) to the image M of two width difference noises1And M2Two layers of wavelet decomposition are carried out, if the coefficient of wavelet decomposition of two images
Respectively D1And D (P)2(P), wherein P=(i, j, k, l), (i, j) are the locus of decomposition coefficient, and k is Decomposition order, and l is
Subband frequency range, (l=1,2,3,4);
B) D is compared1And D (P)2(P), according to corresponding comparison rule, two wavelet coefficients are merged, drawn new small
Wave system number expression Dr (P), including two steps:The first step, by D1And D (P)2(P) high frequency coefficient that first layer decomposes is compared,
If two coefficients are identical, then retain in the wavelet coefficient Dr (P) after corresponding fusion, otherwise will be with Dr (P)
Part is set to 0 corresponding to the coefficient;Second step, compare D1And D (P)2(P) high frequency coefficient that the second layer decomposes, if two coefficients
The absolute value of difference is less than threshold value T ({ T=TH, TV, TD, wherein THFor horizontal direction high frequency coefficient threshold value, TVFor vertical direction high frequency
Coefficient threshold, TDFor diagonally opposed high frequency coefficient threshold value), then the wavelet coefficient after corresponding fusion takes being averaged for two coefficients
Value, otherwise equally sets to 0;
C) second layer low-frequency wavelet coefficients after closing take D1And D (P)2(P) low frequency coefficient that the second layer decomposes is averaged
It is low.
D) corresponding wavelet inverse transformation is carried out to the wavelet coefficient Dr (P) after fusion, you can the image that must be merged.
Finally illustrate, the above embodiments are merely illustrative of the technical solutions of the present invention and it is unrestricted, although with reference to compared with
The present invention is described in detail good embodiment, it will be understood by those within the art that, can be to the skill of the present invention
Art scheme is modified or equivalent substitution, and without departing from the objective and scope of technical solution of the present invention, it all should cover at this
Among the right of invention.
Claims (4)
1. a kind of image enhaucament and fusion method applied to cmos image sensor, it is characterised in that this method includes following
Step:
1) for the view data of cmos image sensor output, dividing processing is carried out to image first, piece image is divided into
Two images;
2) image new to two width carries out picture quality enhancement processing again, is modulated the contrast of image, improves the penetrating of image
Property;
3) image noise reduction processing is then carried out, reduces the noise that prime image procossing is brought;
4) and then to two images merge respectively, wavelet analysis and inversion process, with the output image after being restored.
2. it is applied to the image enhaucament and fusion method of cmos image sensor as claimed in claim 1, it is characterised in that step
Image segmentation processing method in rapid 1 is as follows:For the view data of imaging sensor output, maximum gray scale numerical value is defined as
Max_pixel, the scope of image pixel value fall between { 0, max_pixel };Using the half of pixel maximum as point
Dividing value, original image is judged and split, i.e., the pixel value of piece image is divided into { 0, max_pixel/2 } and { max_
Pixel/2, max_pixel } two sections;By judging the pixel value pixel of image, if pixel value pixel meets 0
≤ pixel≤max_pixel/2, then all pixels in the range of this form image A, if pixel value pixel meets max_
Pixel/2≤pixel≤max_pixel, then all pixels formation the image B, image A and B in the range of this are as original
Two images that image is divided into, array sizes are the half of original image.
3. it is applied to the image enhaucament and fusion method of cmos image sensor as claimed in claim 2, it is characterised in that step
Image noise reduction processing step in rapid 3 is:A the wavelet decomposition of the layer from l to N) is carried out to plus noise image;B) to the small echo of artwork
Each layer high frequency coefficient decomposed is modeled with normal distribution respectively;C) pair high frequency coefficient histogram for adding image of making an uproar enters column hisgram
With operation, make its Histogram Matching in modeling;D) histogram is passed through using the low frequency coefficient of n-th layer and from l to n-th layer
High frequency coefficient after being handled with operation carries out wavelet reconstruction and obtains restored image.
4. it is applied to the image enhaucament and fusion method of cmos image sensor as claimed in claim 3, it is characterised in that step
It is divided into rapid 4 the step of image co-registration:
A) to the image M of two width difference noises1And M2Two layers of wavelet decomposition are carried out, if the coefficient of wavelet decomposition difference of two images
For D1And D (P)2(P), wherein P=(i, j, k, l), (i, j) are the locus of decomposition coefficient, and k is Decomposition order, and l is subband
Frequency range, (l=1,2,3,4);
B) D is compared1And D (P)2(P), according to corresponding comparison rule, two wavelet coefficients are merged, draw new wavelet systems
Number expression Dr (P), including two steps:The first step, by D1And D (P)2(P) high frequency coefficient that first layer decomposes is compared, if
Two coefficients are identical, then retain in the wavelet coefficient Dr (P) after corresponding fusion, otherwise will be with this in Dr (P)
Part is set to 0 corresponding to number;Second step, compare D1And D (P)2(P) high frequency coefficient that the second layer decomposes, if two coefficient differences
Absolute value is less than threshold value T ({ T=TH, TV, TD, wherein THFor horizontal direction high frequency coefficient threshold value, TVFor vertical direction high frequency coefficient
Threshold value, TDFor diagonally opposed high frequency coefficient threshold value), then the wavelet coefficient after corresponding fusion takes the average value of two coefficients, no
Then equally set to 0;
C) second layer low-frequency wavelet coefficients after closing take D1And D (P)2(P) low frequency coefficient that the second layer decomposes it is average low.
D) corresponding wavelet inverse transformation is carried out to the wavelet coefficient Dr (P) after fusion, you can the image that must be merged.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711058317.0A CN107871311A (en) | 2017-11-01 | 2017-11-01 | A kind of image enhaucament and fusion method applied to cmos image sensor |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711058317.0A CN107871311A (en) | 2017-11-01 | 2017-11-01 | A kind of image enhaucament and fusion method applied to cmos image sensor |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107871311A true CN107871311A (en) | 2018-04-03 |
Family
ID=61753535
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711058317.0A Pending CN107871311A (en) | 2017-11-01 | 2017-11-01 | A kind of image enhaucament and fusion method applied to cmos image sensor |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107871311A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110674828A (en) * | 2018-07-03 | 2020-01-10 | 江威 | Method and device for normalizing fundus images |
CN111192212A (en) * | 2019-12-25 | 2020-05-22 | 北京东宇宏达科技有限公司 | Self-adaptive infrared digital image detail enhancement method |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104966272A (en) * | 2015-05-29 | 2015-10-07 | 中国农业大学 | Underwater sea cucumber image processing method and system |
CN106504281A (en) * | 2016-12-02 | 2017-03-15 | 中国电子科技集团公司第四十四研究所 | The image quality for being applied to cmos image sensor strengthens and filtering method |
-
2017
- 2017-11-01 CN CN201711058317.0A patent/CN107871311A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104966272A (en) * | 2015-05-29 | 2015-10-07 | 中国农业大学 | Underwater sea cucumber image processing method and system |
CN106504281A (en) * | 2016-12-02 | 2017-03-15 | 中国电子科技集团公司第四十四研究所 | The image quality for being applied to cmos image sensor strengthens and filtering method |
Non-Patent Citations (3)
Title |
---|
于萌: ""图像去雾方法的研究与设计"", 《中国优秀硕博士学位论文全文数据库(硕士)信息科技辑》 * |
孙水发 等 著: "《ImageJ图像处理与实践》", 31 December 2013, 国防工业出版社 * |
谢义方: ""数字图像复原算法研究"", 《中国优秀硕博士学位论文全文数据库信息科技辑》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110674828A (en) * | 2018-07-03 | 2020-01-10 | 江威 | Method and device for normalizing fundus images |
CN111192212A (en) * | 2019-12-25 | 2020-05-22 | 北京东宇宏达科技有限公司 | Self-adaptive infrared digital image detail enhancement method |
CN111192212B (en) * | 2019-12-25 | 2022-05-17 | 北京东宇宏达科技有限公司 | Self-adaptive infrared digital image detail enhancement method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN100550978C (en) | A kind of self-adapting method for filtering image that keeps the edge | |
CN111583123A (en) | Wavelet transform-based image enhancement algorithm for fusing high-frequency and low-frequency information | |
CN109685728B (en) | Digital image processing method based on local time-frequency domain transformation | |
CN103093433B (en) | Natural image denoising method based on regionalism and dictionary learning | |
Suganya et al. | Survey on image enhancement techniques | |
Liu et al. | A medical image enhancement method using adaptive thresholding in NSCT domain combined unsharp masking | |
CN108389163A (en) | A kind of self-adapting enhancement method based on half-light coloured image | |
CN104378636B (en) | A kind of video encoding method and device | |
CN104182939B (en) | Medical image detail enhancement method | |
Yazdi et al. | Impulsive noise suppression of images using adaptive median filter | |
CN103971340A (en) | High-bit-width digital image dynamic range compression and detail enhancement method | |
Chouhan et al. | Contrast enhancement of dark images using stochastic resonance in wavelet domain | |
CN112991199B (en) | Image high-low frequency decomposition noise removal method based on residual dense network | |
CN105427255A (en) | GRHP based unmanned plane infrared image detail enhancement method | |
CN104766289A (en) | Image denoising and compressing method | |
CN107871311A (en) | A kind of image enhaucament and fusion method applied to cmos image sensor | |
RU2448367C1 (en) | Method of increasing visual information content of digital greyscale images | |
CN104240208A (en) | Uncooled infrared focal plane detector image detail enhancement method | |
CN110175959B (en) | Typhoon cloud picture enhancement method | |
Umamaheswari et al. | An enhanced approach for medical brain image enhancement | |
Gopi et al. | Capsule endoscopic colour image denoising using complex wavelet transform | |
Panchaxri et al. | Image denoising using adaptive NL means filtering with method noise thresholding | |
CN115880173A (en) | Image processing method, image processing apparatus, and computer-readable storage medium | |
Singh et al. | Genetic algorithm based medical image denoising through sub band adaptive thresholding | |
CN106780398B (en) | A kind of image de-noising method based on noise prediction |
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 |