CN108876741B - Image enhancement method under complex illumination condition - Google Patents
Image enhancement method under complex illumination condition Download PDFInfo
- Publication number
- CN108876741B CN108876741B CN201810647661.1A CN201810647661A CN108876741B CN 108876741 B CN108876741 B CN 108876741B CN 201810647661 A CN201810647661 A CN 201810647661A CN 108876741 B CN108876741 B CN 108876741B
- Authority
- CN
- China
- Prior art keywords
- image
- fuzzy
- wavelet
- enhancement
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000005286 illumination Methods 0.000 title claims abstract description 19
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 24
- 230000000694 effects Effects 0.000 claims abstract description 9
- 230000005540 biological transmission Effects 0.000 claims abstract description 4
- 238000000605 extraction Methods 0.000 claims abstract description 4
- 239000011159 matrix material Substances 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 10
- 230000009466 transformation Effects 0.000 claims description 6
- 230000001131 transforming effect Effects 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 230000001276 controlling effect Effects 0.000 claims description 3
- 230000001105 regulatory effect Effects 0.000 claims description 3
- 239000003245 coal Substances 0.000 abstract description 6
- 238000001914 filtration Methods 0.000 abstract description 2
- 230000007423 decrease Effects 0.000 abstract 1
- 238000001514 detection method Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000002708 enhancing effect Effects 0.000 description 3
- 239000003623 enhancer Substances 0.000 description 2
- 235000009413 Ratibida columnifera Nutrition 0.000 description 1
- 241000510442 Ratibida peduncularis Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/10—Image enhancement or restoration using non-spatial domain filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- 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]
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an image enhancement method under a complex illumination condition, which mainly comprises an initialization stage, a fuzzy enhancement stage and an image reconstruction stage, wherein the image acquisition and transmission of an underground video image are completed, the video image is subjected to wavelet decomposition, a low-frequency coefficient histogram is equalized, and a high-frequency coefficient is subjected to coefficient extraction; designing a denoising model and a new PAL fuzzy enhancement algorithm to realize filtering processing and fuzzy enhancement of a high-frequency part and anti-fuzzy processing of a fuzzy enhancement characteristic image; and finally, performing wavelet reconstruction on the high-frequency coefficient subjected to the anti-fuzzy processing and the low-frequency coefficient subjected to histogram equalization, and obtaining an effect picture after image enhancement. The method can effectively overcome the problem of sharp decline of the image characteristic points caused by shadow, dark area, dark light and highlight of the image due to complex underground illumination conditions, and promotes further application of the image equipment in the underground coal mine.
Description
Technical Field
The invention belongs to the field of image recognition, relates to an image enhancement and image denoising technology, and particularly relates to an image preprocessing method applied to underground.
Background
The current detection of underground coal mine targets based on image enhancement has important significance for maintaining mine safety, but the current image enhancement technology is difficult to apply in actual complex environments. Due to the fact that the stealth capability of the underground target is strong, signals radiated by the target in the underground complex environment are weak and are easily interfered by the background, and therefore part of the target is mixed in clutter below the background limit. In engineering, the detection capability of a camera is improved mainly by means of increasing the gain of a photoelectric detector, increasing the photoelectric conversion efficiency of a device, increasing the focal length of an optical lens and the like, but targets in video images acquired under a complex underground environment are still weak, and the practical underground application is still difficult to meet.
The image enhancement under the current complex environment mostly adopts single spatial domain enhancement or frequency domain enhancement, and is mainly researched in the aspects of enhancing image contrast, highlighting image details, eliminating noise and the like. However, the underground illumination condition of the coal mine is complex, special conditions such as poor light, uneven illumination, much dust and the like exist, and image shadows, dark areas, dark light, high light and the like caused by the complex illumination condition cause that the method for directly processing the pixel gray value by adopting the spatial domain enhancement and the method for realizing the enhancement by adopting the frequency domain filter to process the frequency domain image cannot meet the requirements of eliminating noise and enhancing details of the underground video image. Therefore, the research on an image enhancement method suitable for complex illumination conditions under the coal mine is a problem that image equipment is urgently needed to be solved in the application of the coal mine.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide an underground image enhancement method, which solves the problem that the prior underground image identification technology cannot meet the problem of sharp reduction of image characteristic points caused by image shadow, bright and dark areas, dark light and high light due to complex underground illumination conditions.
The invention specifically adopts the following technical scheme to solve the technical problems:
1. the image enhancement method under the complex illumination condition is characterized by comprising an initialization stage, a fuzzy enhancement stage and an image reconstruction stage; in the initialization stage, the underground video image acquisition and transmission can be completed, the video image is subjected to wavelet decomposition, and the low-frequency coefficient histogram is equalized; extracting coefficients of a high-frequency coefficient obtained after wavelet decomposition in a fuzzy enhancement stage, and realizing filtering processing and fuzzy enhancement of a high-frequency part by introducing a denoising model of a fuzzy membership factor and a new PAL fuzzy enhancement algorithm so as to obtain fuzzy enhancement characteristic images in different scales and different directions and perform anti-fuzzy processing on the fuzzy enhancement characteristic images; and in the image reconstruction stage, performing wavelet reconstruction on the high-frequency coefficient subjected to anti-fuzzy processing and the low-frequency coefficient subjected to histogram equalization, and obtaining an image-enhanced effect image.
Further, the initialization phase comprises the steps of:
(1) a visible light camera or an infrared camera is used for acquiring video images of the underground complex environment;
(2) transmitting the video image acquired in the step 1 to an image processing unit through an RJ45 interface or a USB interface of a camera, and storing the video image;
(3) performing single-frame extraction on the video image transmitted to the image processing unit in the step 2, and performing gray processing on the obtained single-frame color image;
(4) performing multi-scale wavelet decomposition on the image subjected to the graying treatment in the step 3 to obtain a low-frequency coefficient under the nth scale and high-frequency coefficients in 3n directions;
(5) and (4) carrying out histogram equalization processing on the low-frequency coefficient under the nth scale obtained in the step (4).
Further, the blur enhancement stage comprises the steps of:
(1) firstly, extracting the 3-direction high-frequency coefficients decomposed when n is 1, then extracting the 3-direction high-frequency coefficients decomposed when n is 2, and then sequentially extracting the 3-direction high-frequency coefficients decomposed when n is 2;
(2) sequentially and respectively removing noise in the high-frequency coefficients of the image by the aid of a denoising model introducing a fuzzy membership factor s from the high-frequency coefficients in the 3 directions obtained in the step 1;
(3) sequentially transforming the high-frequency coefficients subjected to denoising treatment in the step 2 from a spatial domain to a fuzzy domain set by designing a fuzzy membership function;
(4) by designing a fuzzy enhancement operator, sequentially carrying out fuzzy set nonlinear transformation on the fuzzy set domain obtained in the step 3, namely carrying out fuzzy function calculation on pixel values of a high-frequency image in the fuzzy domain to obtain fuzzy enhancement characteristic images in different scales and different directions;
(5) and (4) according to the fuzzy membership function designed in the step (3), transforming the high-frequency coefficient processed in the step (4) from the fuzzy domain to the space domain again, namely performing anti-fuzzy processing on the fuzzy enhancement characteristic image.
Further, the reconstruction phase comprises the steps of:
(1) performing wavelet reconstruction on the nth scale high-frequency coefficient obtained after the anti-fuzzy processing and the nth scale lower histogram equalized low-frequency coefficient to reconstruct an image subjected to fuzzy enhancement on the nth scale;
(2) carrying out image matrix construction on the image pixel matrix after the blurring enhancement on the nth scale obtained in the step 1 and the high-frequency coefficient of the (n-1) th layer obtained after the anti-blurring processing, and reconstructing a blurring enhanced image on the (n-1) th scale;
(3) and (3) repeating the steps 1 and 2, and ending the repeated process when obtaining the image after the fuzzy enhancement processing with the same resolution as the original image.
Further, the design of the denoising model mainly comprises:
(1) in a wavelet threshold denoising model, a fuzzy membership factor s is introduced, so that a wavelet threshold can be adaptively adjusted according to the noise distribution condition in a high-frequency coefficient of an image after wavelet decomposition, and the wavelet threshold function structural expression in the wavelet denoising model is as follows:
in the formula, muTAs a function of wavelet threshold, ωijThe absolute value of a point (i, j) in a wavelet high-frequency coefficient is shown, sgn (·) is a sign function, s is a fuzzy membership factor, and T is a wavelet threshold;
(2) the fuzzy membership factor s in the model replaces a fixed threshold T in a wavelet soft threshold, the factor can be adaptively adjusted according to the distribution condition of image noise, the flexibility of the model is greatly improved, and the calculation formula of the factor is as follows:
wherein a is a regulating factor, and a belongs to (0, 1)],ωijIs the absolute value of the (i, j) point in the wavelet high-frequency coefficient, and T is the wavelet threshold;
(3) the coefficient of the noise information in the high-frequency coefficient of the image has a larger difference with the coefficient of the image information, the mean square error obtained by calculating the mean square error of the high-frequency part of the multi-scale wavelet decomposition is larger, then a certain coefficient value is determined, the mean square error of all the coefficients larger than the value is minimum, the effect is better, the coefficient value is the threshold value to be selected, and the calculation formula of the threshold value T is as follows:
T=2-n|ωij|σ2/σ'
where n is the highest wavelet decomposition level, σ is the image noise variance, and σ is mean (| ω ═ mijI)/0.6745, sigma' is the standard deviation of the wavelet decomposition coefficient of the image, andm is the maximum value of the row of the processed picture, and N is the maximum value of the column of the processed image.
Further, the PAL fuzzy enhancement algorithm includes:
(1) design membership function muij:
In the formula, ωijIs the absolute value of (i, j) point in wavelet high-frequency coefficient, omegamaxIs the absolute value of the maximum coefficient in the high frequency part;
ωminis the absolute value of the minimum coefficient of the high frequency part;
(2) designing a fuzzy enhancement operator:
in the formula, t is a parameter for controlling the convergence rate, and the larger t is, the faster the convergence rate is; variable parameter muc∈[0,1]。
The invention has the beneficial effects that:
according to the image enhancement method under the complex illumination condition, the multi-scale decomposition of the wavelet decomposition model is utilized, so that high-frequency information and low-frequency information are separated on each resolution after decomposition, and the problems of high-frequency noise of the image and the integral contrast of the image are respectively solved; after the low-frequency part is subjected to histogram equalization processing, the gray contrast of the image is improved, the overall brightness of the image is improved, the gray dynamic range is enlarged, and a better enhancement effect is achieved. Meanwhile, no noise signal is introduced into the low-frequency part of the picture processed by the algorithm, and the algorithm can effectively separate the noise signal from the overall outline information of the picture; the high-frequency noise under different scales can be effectively removed by designing a denoising mathematical model on each resolution of a high-frequency part, the detail information and the texture feature information of an image can be effectively reserved by adopting the nonlinear transformation processing of a fuzzy domain set in a fuzzy enhanced image, and the general appearance characteristic feature of the original image is reserved by a low-frequency coefficient.
Compared with the original image, the image enhanced by the algorithm has the advantages that the contrast of the image is obviously improved, the halation at the position with strong light and shade contrast can be well removed, the image enhancement effect is natural, the integrity of detail information in the underground image is well maintained by the fuzzy processing function, the key characteristic information of the image is enhanced, and the method can be applied to visual detection in the complex illumination environment of the underground coal mine.
Drawings
FIG. 1 is a flowchart of an initialization phase of image enhancement under complex illumination conditions according to the present invention;
FIG. 2 is a flow chart of a blur enhancement stage of image enhancement under complex lighting conditions according to the present invention;
fig. 3 is a flowchart of a reconstruction phase of image enhancement under a complex illumination condition according to the present invention.
FIG. 4 is a diagram illustrating the effect of image enhancement processing under complex illumination conditions according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions and specific implementation methods of the present invention are described in detail below with reference to the accompanying drawings.
A. A basic flow of an image enhancement method under a complex illumination condition is described with reference to fig. 1, which includes the following specific steps:
(1) a visible light camera or an infrared camera adopted by the image acquisition (101) is of an explosion-proof design, and the camera can acquire video images of a target object in a complex underground environment;
(2) transmitting the video image acquired in the step 1 to an image processing unit through an RJ45 interface or a USB interface of the camera by image transmission (102), wherein the image processing unit comprises a DSP image processing chip, an ISP image processing chip, a CPU or a GPU;
(3) the image storage (103) stores the video image transmitted by the acquisition equipment in the step (2), and the used image data storage can adopt an image acquisition card or directly upload the video image data to a cloud server;
(4) graying (104) to extract the single frame of the video image transmitted to the image processing unit in the step (2), store the image, and perform graying processing on the single frame color image obtained after storage;
(5) selecting a corresponding wavelet basis function from the grayed image obtained in the step 4 by using wavelet decomposition (105), designing the value of a scale n of the wavelet decomposition, wherein n is usually 1, 2, 3 or 4, and performing wavelet decomposition on the gray image with the wavelet basis function and the scale n determined to obtain a low-frequency coefficient under the nth scale and high-frequency coefficients in 3n directions; the wavelet basis functions mainly comprise a Haar wavelet basis Biorthogonal (biornr. Nd) wavelet system, a db series wavelet basis, a Coiflet (coifn) wavelet system, a Symlettsa (symn) wavelet system, a Molet (morl) wavelet, a Mexican Hat (mexh) wavelet and a Meyer wavelet;
(6) extracting low-frequency coefficients (106) and directly acquiring the low-frequency coefficients under the nth scale obtained in the step (5) through a wavelet low-frequency acquisition function;
(7) the low frequency processing (107) performs histogram equalization processing on the acquired low frequency coefficients of the image to uniformly distribute the pixels of the low frequency coefficient image in a certain image, thereby enhancing the contrast and brightness of the image.
B. Referring to fig. 2, a description is given to a blur enhancement stage of the image enhancement method under the complex illumination condition, which specifically includes the following steps:
(1) extracting high-frequency coefficients in 3n directions obtained by decomposing the multi-scale wavelet (201), wherein the high-frequency coefficients in 3 directions decomposed when n is 1 are firstly extracted, then the high-frequency coefficients in 3 directions decomposed when n is 2 are extracted, and then the high-frequency coefficients in 3 directions decomposed when n is the scale are sequentially extracted;
(2) and (2) designing a denoising model (202), sequentially carrying out denoising treatment on the high-frequency coefficients in 3 directions obtained in each scale in the step (1) by using a wavelet denoising model introducing a fuzzy membership factor s, wherein a wavelet threshold function in the wavelet denoising model is constructed by an expression:
in the formula, muTAs a function of wavelet threshold, ωijThe absolute value of a point (i, j) in a wavelet high-frequency coefficient is shown, sgn (·) is a sign function, s is a fuzzy membership factor, and T is a wavelet threshold; a fuzzy membership factor s in the model replaces a fixed threshold T in a wavelet soft threshold, the factor can be adaptively adjusted according to the distribution condition of image noise, the flexibility of the model is greatly improved, and the calculation formula of the factor is as follows:
wherein a is a regulating factor, and a belongs to (0, 1)],ωijIs the absolute value of the (i, j) point in the wavelet high-frequency coefficient, and T is the wavelet threshold; the coefficient of the noise information in the high-frequency coefficient of the image has a larger difference with the coefficient of the image information, the mean square error obtained by calculating the mean square error of the high-frequency part of the multi-scale wavelet decomposition is larger, then a certain coefficient value is determined, the mean square error of all the coefficients larger than the value is minimum, the effect is better, the coefficient value is the threshold value to be selected, and the calculation formula of the threshold value T is as follows:
T=2-n|ωij|σ2/σ'
in the formula, n is the highest wavelet decomposition layer number, and sigma is a graphLike variance of noise, and σ is mean (| ω)ijI)/0.6745, σ' is
The standard deviation of the wavelet decomposition coefficient of the image is expressed as:
in the formula, M is the maximum value of a row of a processed picture, N is the maximum value of a column of the processed picture, and the high-frequency coefficients under all scales are subjected to wavelet denoising according to the designed wavelet denoising function;
(3) designing a membership function (203) by designing a fuzzy membership function muijThe expression is as follows:
in the formula, ωijIs the absolute value of (i, j) point in wavelet high-frequency coefficient, omegamaxIs the absolute value of the maximum coefficient in the high frequency part; omegaminIs the absolute value of the minimum coefficient of the high frequency part; sequentially adopting fuzzy membership function mu for the high-frequency coefficient subjected to denoising treatment in the step 2ijFurther realizing the transformation of the high-frequency coefficient from the space domain to the fuzzy set domain;
(4) designing the blur enhancer (204) by designing the blur enhancer T (mu)ij) The expression is as follows:
in the formula, t is a parameter for controlling the convergence rate, and the larger t is, the faster the convergence rate is; variable parameter muc∈[0,1](ii) a Sequentially carrying out fuzzy set nonlinear transformation processing on the fuzzy set domain obtained in the step 3, namely carrying out fuzzy enhancement operator calculation on the pixel value of the high-frequency image in the fuzzy domain to further obtain fuzzy enhancement characteristic images in different scales and different directions;
(5) and (4) performing anti-fuzzy processing (205) according to the fuzzy membership function designed in the step (3), and transforming the high-frequency coefficient processed in the step (4) from the fuzzy set domain to the space domain again, namely performing anti-fuzzy processing on the fuzzy enhancement characteristic image.
C. Referring to fig. 3, an image reconstruction stage of the image enhancement method under the complex illumination condition according to the present invention is described, which includes the following steps:
(1) the nth layer wavelet coefficient construction (301) is used for constructing an image matrix of the low-frequency coefficient processed in the initialization stage and the nth layer high-frequency coefficient processed in the fuzzy enhancement stage;
(2) the nth layer of wavelet reconstruction (302) is used for reconstructing the wavelet reconstruction function input by the image matrix constructed in the step (1) and reconstructing an image after fuzzy enhancement on the nth scale, wherein the rows and columns of the matrix have the resolution of the reconstructed image;
(3) an m-th layer wavelet coefficient construction (303) is used for constructing an image matrix by using the pixel value corresponding to the image subjected to fuzzy enhancement on the nth scale obtained in the step (2) and the n-1-th layer high-frequency coefficient subjected to fuzzy enhancement stage processing, wherein the row and the column of the constructed image matrix are 2 times of the row and the column of the image matrix in the step (1);
(4) the m-th layer of wavelet reconstruction (304) carries out reconstruction on the wavelet reconstruction function input by the image matrix constructed in the step (3) and reconstructs the image after fuzzy enhancement on the (n-1) th scale, and the resolution of the image after the current reconstruction is 2 times of the resolution of the n-th layer of reconstruction ;
(5) and (6) judging whether m is 0(305), if m is not 0, returning to (303) and repeating the step (3) and the step (4), if m is 0, ending the repeating process, and outputting the enhanced effect image after the fuzzy enhancement processing with the same resolution as the original image.
Claims (5)
1. The image enhancement method under the complex illumination condition is characterized by comprising an initialization stage, a fuzzy enhancement stage and an image reconstruction stage; in the initialization stage, the underground video image acquisition and transmission are completed, the n-scale wavelet decomposition is carried out on the video image, and the low-frequency coefficient histogram is equalized; extracting coefficients of a high-frequency coefficient obtained after wavelet decomposition in a fuzzy enhancement stage, designing a wavelet denoising model introducing a fuzzy membership factor by introducing the fuzzy membership factor into the wavelet soft threshold denoising model, and removing noise in the high-frequency coefficient after coefficient extraction through the wavelet denoising model; transforming the high-frequency coefficient subjected to denoising processing from a space domain to a fuzzy set domain through a new PAL fuzzy enhancement algorithm, sequentially carrying out fuzzy set nonlinear transformation on the fuzzy set domain by adopting a fuzzy enhancement operator to obtain fuzzy enhancement characteristic images under different scales and different directions, and carrying out anti-fuzzy processing on the fuzzy enhancement characteristic images, wherein the new PAL fuzzy enhancement algorithm comprises the design of a new membership function and a fuzzy enhancement operator; performing wavelet reconstruction on the high-frequency coefficient subjected to anti-fuzzy processing and the low-frequency coefficient subjected to histogram equalization in an image reconstruction stage, and obtaining an image-enhanced effect graph;
the membership function muijComprises the following steps:ωmax-ωmin≤D≤2(ωmax-ωmin) (ii) a In the formula, ωijIs the absolute value of (i, j) point in wavelet high-frequency coefficient, omegamaxIs the absolute value of the maximum coefficient in the high frequency part; omegaminIs the absolute value of the minimum coefficient of the high frequency part;
2. The method of claim 1, wherein the initialization phase comprises the steps of:
1) a visible light camera or an infrared camera is used for acquiring video images of the underground complex environment;
2) transmitting the video image acquired in the step 1 to an image processing unit through an RJ45 interface or a USB interface of a camera, and storing the video image;
3) performing single-frame extraction on the video image transmitted to the image processing unit in the step 2, and performing gray processing on the obtained single-frame color image;
4) performing multi-scale wavelet decomposition on the image subjected to the graying treatment in the step 3 to obtain a low-frequency coefficient under the nth scale and high-frequency coefficients in 3n directions;
5) and (4) carrying out histogram equalization processing on the low-frequency coefficient under the nth scale obtained in the step (4).
3. The method of claim 1, wherein the blur enhancement stage comprises the steps of:
1) extracting high-frequency coefficients in 3n directions obtained after the n-scale wavelet decomposition, namely extracting the high-frequency coefficients in 3 directions decomposed when n is 1, then extracting the high-frequency coefficients in 3 directions decomposed when n is 2, and then sequentially extracting the high-frequency coefficients in 3 directions decomposed when n is the nth scale;
2) sequentially and respectively removing noise in the high-frequency coefficients in 3 directions obtained in the step 1 by adopting a wavelet denoising model introducing a fuzzy membership factor s;
3) sequentially transforming the high-frequency coefficients subjected to denoising treatment in the step 2 from a spatial domain to a fuzzy domain set by designing a fuzzy membership function;
4) by designing a fuzzy enhancement operator, sequentially carrying out fuzzy set nonlinear transformation on the fuzzy set domain obtained in the step 3, namely carrying out fuzzy function calculation on the pixel values of the de-noised high-frequency coefficient in the fuzzy domain to obtain fuzzy enhancement characteristic images in different scales and different directions;
5) and (4) according to the fuzzy membership function designed in the step (3), transforming the high-frequency coefficient processed in the step (4) from the fuzzy domain to the space domain again, namely performing anti-fuzzy processing on the fuzzy enhancement characteristic image.
4. The method of claim 1, wherein the image enhancement method under complex lighting conditions comprises: the image reconstruction stage comprises the following steps:
1) performing wavelet reconstruction on the nth scale high-frequency coefficient obtained after the anti-fuzzy processing and the nth scale lower histogram equalized low-frequency coefficient to reconstruct an image subjected to fuzzy enhancement on the nth scale;
2) carrying out image matrix construction on the image pixel matrix after the blurring enhancement on the nth scale obtained in the step 1 and the high-frequency coefficient of the (n-1) th layer obtained after the anti-blurring processing, and reconstructing a blurring enhanced image on the (n-1) th scale;
3) and (3) repeating the steps 1 and 2, and ending the repeated process when obtaining the image after the fuzzy enhancement processing with the same resolution as the original image.
5. The method of claim 1, wherein the image enhancement method under complex lighting conditions comprises: the design of the wavelet denoising model mainly comprises the following steps:
1) in a wavelet soft threshold denoising model, a fuzzy membership factor s is introduced to make wavelet threshold adaptively adjusted according to the noise distribution condition in the high-frequency coefficient of the wavelet decomposed image, and the wavelet threshold function in the wavelet denoising model is constructed by the following expression:
in the formula, muTAs a function of wavelet threshold, ωijThe absolute value of a point (i, j) in a wavelet high-frequency coefficient is shown, sgn (·) is a sign function, s is a fuzzy membership factor, and T is a wavelet threshold;
2) the fuzzy membership factor s in the wavelet denoising model replaces a fixed threshold T in a wavelet soft threshold, and is adaptively adjusted according to the distribution condition of image noise, and the calculation formula of the fuzzy membership factor is as follows:
wherein a is a regulating factor, and a belongs to (0, 1)],ωijIs the absolute value of the (i, j) point in the wavelet high-frequency coefficient, and T is the wavelet threshold;
3) the mean square error calculation is carried out on the high-frequency coefficient of the multi-scale wavelet decomposition, then a certain coefficient value is determined, the mean square error of all the coefficients which are larger than the certain coefficient value is enabled to be minimum, then the coefficient value is the threshold value to be selected, and the calculation formula of the threshold value T is as follows:
T=2-n|ωij|σ2/σ'
where n is the highest wavelet decomposition level, σ is the image noise variance, and σ is mean (| ω ═ mijI)/0.6745, sigma' is the standard deviation of the wavelet decomposition coefficient of the image, andm is the maximum value of the row of the processed picture, and N is the maximum value of the column of the processed image.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810647661.1A CN108876741B (en) | 2018-06-22 | 2018-06-22 | Image enhancement method under complex illumination condition |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810647661.1A CN108876741B (en) | 2018-06-22 | 2018-06-22 | Image enhancement method under complex illumination condition |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108876741A CN108876741A (en) | 2018-11-23 |
CN108876741B true CN108876741B (en) | 2021-08-24 |
Family
ID=64340849
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810647661.1A Active CN108876741B (en) | 2018-06-22 | 2018-06-22 | Image enhancement method under complex illumination condition |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108876741B (en) |
Families Citing this family (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP3785222B1 (en) | 2018-05-30 | 2024-04-17 | Shanghai United Imaging Healthcare Co., Ltd. | Systems and methods for image processing |
CN109658354B (en) * | 2018-12-20 | 2022-02-08 | 上海联影医疗科技股份有限公司 | Image enhancement method and system |
CN110187093A (en) * | 2019-06-26 | 2019-08-30 | 上海艾瑞德生物科技有限公司 | A kind of cross-platform detection device detecting fluorescence immunoassay and colloid gold reagent item |
CN112669222A (en) * | 2019-10-15 | 2021-04-16 | 国际关系学院 | Digital image defogging method based on wavelet domain histogram equalization |
CN111325680B (en) * | 2020-01-08 | 2023-04-25 | 深圳深知未来智能有限公司 | Image brightening method capable of inhibiting local overexposure |
CN113112435B (en) * | 2020-04-23 | 2023-09-22 | 江苏理工学院 | Variable contrast enhancement method and device for wavelet domain positive and negative image fusion |
CN112862716A (en) * | 2021-02-07 | 2021-05-28 | 江苏科技大学 | Sonar image target enhancement method based on dual-tree complex wavelet and fuzzy theory |
CN113128511B (en) * | 2021-03-31 | 2023-07-25 | 武汉钢铁有限公司 | Coke tissue identification method and device |
CN113246859B (en) * | 2021-05-28 | 2023-04-07 | 深圳天联星科技有限公司 | Electronic rearview mirror with driving auxiliary system warning function |
CN113212307B (en) * | 2021-05-28 | 2022-11-22 | 深圳天联星科技有限公司 | Electronic rearview mirror with 360-degree environment monitoring function |
CN113656462B (en) * | 2021-08-18 | 2022-05-17 | 北京奥康达体育产业股份有限公司 | Wisdom sports park data analysis system based on thing networking |
CN117422656B (en) * | 2023-12-14 | 2024-04-12 | 西南交通大学 | Low-illumination fuzzy traffic image enhancement method, device, equipment and medium |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5799112A (en) * | 1996-08-30 | 1998-08-25 | Xerox Corporation | Method and apparatus for wavelet-based universal halftone image unscreening |
CN101201937A (en) * | 2007-09-18 | 2008-06-18 | 上海医疗器械厂有限公司 | Digital image enhancement method and device based on wavelet restruction and decompose |
CN101430759A (en) * | 2008-12-04 | 2009-05-13 | 上海大学 | Optimized recognition pretreatment method for human face |
CN103310423A (en) * | 2013-07-08 | 2013-09-18 | 中国矿业大学(北京) | Mine image intensification method |
CN103886344A (en) * | 2014-04-14 | 2014-06-25 | 西安科技大学 | Image type fire flame identification method |
CN103927752A (en) * | 2014-04-21 | 2014-07-16 | 国家电网公司 | Method for monitoring operating state of electrical equipment |
CN105844601A (en) * | 2016-05-20 | 2016-08-10 | 中国矿业大学(北京) | Mine image enhancement method based on bilateral filtering and multi-scale Retinex algorithm |
-
2018
- 2018-06-22 CN CN201810647661.1A patent/CN108876741B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5799112A (en) * | 1996-08-30 | 1998-08-25 | Xerox Corporation | Method and apparatus for wavelet-based universal halftone image unscreening |
CN101201937A (en) * | 2007-09-18 | 2008-06-18 | 上海医疗器械厂有限公司 | Digital image enhancement method and device based on wavelet restruction and decompose |
CN101430759A (en) * | 2008-12-04 | 2009-05-13 | 上海大学 | Optimized recognition pretreatment method for human face |
CN103310423A (en) * | 2013-07-08 | 2013-09-18 | 中国矿业大学(北京) | Mine image intensification method |
CN103886344A (en) * | 2014-04-14 | 2014-06-25 | 西安科技大学 | Image type fire flame identification method |
CN103927752A (en) * | 2014-04-21 | 2014-07-16 | 国家电网公司 | Method for monitoring operating state of electrical equipment |
CN105844601A (en) * | 2016-05-20 | 2016-08-10 | 中国矿业大学(北京) | Mine image enhancement method based on bilateral filtering and multi-scale Retinex algorithm |
Also Published As
Publication number | Publication date |
---|---|
CN108876741A (en) | 2018-11-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108876741B (en) | Image enhancement method under complex illumination condition | |
Liu et al. | Efficient single image dehazing and denoising: An efficient multi-scale correlated wavelet approach | |
Gao et al. | Naturalness preserved nonuniform illumination estimation for image enhancement based on retinex | |
AU2020100462A4 (en) | Edge-preserving image super-resolution via low rank and total variation model | |
CN112767251B (en) | Image super-resolution method based on multi-scale detail feature fusion neural network | |
CN111583123A (en) | Wavelet transform-based image enhancement algorithm for fusing high-frequency and low-frequency information | |
CN108564597B (en) | Video foreground object extraction method fusing Gaussian mixture model and H-S optical flow method | |
EP3563339B1 (en) | Method and device for image processing | |
Khan et al. | Localization of radiance transformation for image dehazing in wavelet domain | |
Suganya et al. | Survey on image enhancement techniques | |
CN103295204B (en) | A kind of image self-adapting enhancement method based on non-down sampling contourlet transform | |
CN108932699B (en) | Three-dimensional matching harmonic filtering image denoising method based on transform domain | |
CN108875645B (en) | Face recognition method under complex illumination condition of underground coal mine | |
Janani et al. | Infrared image enhancement techniques—A review | |
CN114612359A (en) | Visible light and infrared image fusion method based on feature extraction | |
Yang et al. | Low-light image enhancement based on Retinex theory and dual-tree complex wavelet transform | |
CN114331937A (en) | Multi-source image fusion method based on feedback iterative adjustment under low illumination condition | |
CN113610730A (en) | Method and system for removing non-uniform thin cloud of satellite image | |
CN113592750A (en) | Infrared enhancement method based on gradient histogram | |
Yang et al. | A novel single image dehazing method | |
Wang et al. | Rapid nighttime haze removal with color-gray layer decomposition | |
Zohaib et al. | Image enhancement by using histogram equalization technique in Matlab | |
CN113066023B (en) | SAR image speckle removing method based on self-calibration convolutional neural network | |
Tsai et al. | Multiexposure image fusion using intensity enhancement and detail extraction | |
Tyagi et al. | Comparative study of image enhancement and analysis of thermal images using image processing and wavelet techniques |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |