CN115861117A - Mine infrared image enhancement method and device - Google Patents
Mine infrared image enhancement method and device Download PDFInfo
- Publication number
- CN115861117A CN115861117A CN202211633829.6A CN202211633829A CN115861117A CN 115861117 A CN115861117 A CN 115861117A CN 202211633829 A CN202211633829 A CN 202211633829A CN 115861117 A CN115861117 A CN 115861117A
- Authority
- CN
- China
- Prior art keywords
- image
- detail
- subgraph
- function
- mine
- 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
Images
Landscapes
- Image Processing (AREA)
Abstract
The invention discloses a mine infrared image enhancement method and a mine infrared image enhancement device, wherein the method comprises the following steps: decomposing the degraded image into a detail sub-image containing high-frequency information and a basic sub-image containing low-frequency information by adopting double-domain filtering; improving the brightness, the contrast and the definition of the basic subgraph by using a CLAHE algorithm so as to highlight the general appearance characteristics of the monitoring scene; carrying out noise suppression and edge sharpening on the detail subgraph by adopting the constructed ILoG operator and eliminating the gradient inversion phenomenon so as to improve the detail characteristics of the object in the monitored scene; adopting a self-adaptive reconstruction function to reconstruct the processed basic subgraph and the processed detail subgraph to obtain a reconstructed image with improved image quality; a Gamma correction function of gray level redistribution is designed, and the brightness of the reconstructed image is adjusted, so that the mine infrared enhanced image is obtained. The mine infrared image enhancement method and the mine infrared image enhancement device have the advantages of high instantaneity and robustness and the like, and can meet the actual requirements of intelligent mine construction.
Description
Technical Field
The invention relates to the technical field of image enhancement, in particular to a mine infrared image enhancement method and device.
Background
The infrared image can reflect the temperature field distribution characteristics of the surface of an object in the monitoring area, and has the advantages of strong anti-interference capability, sensitivity to the temperature of the object and the like. At present, infrared images are gradually used for mine monitoring video analysis and potential safety hazard monitoring, but due to the fact that temperature difference between a target and a background in an underground environment is small, transmission attenuation of infrared radiation is large, the acquired infrared images have the defects of being strong in spatial correlation, concentrated in gray distribution, poor in detail resolution and the like. In addition, due to random interference of a mine environment and self-disturbance of thermal imaging equipment, mine infrared images are full of noise, so that the signal to noise ratio of the infrared images is lower than that of visible light images. Therefore, enhancing mine infrared degraded images has become one of the important research directions in intelligent mine construction.
Aiming at the problem of low quality of infrared images in a coal mine monitoring video, the invention realizes a mine infrared image enhancement method and a mine infrared image enhancement device of ILoG operator with improved double-domain decomposition coupling and contrast limited adaptive histogram equalization CLAHE in combination with a special underground coal mine environment. Decomposing the mine infrared degradation image into a basic sub-image and a detail sub-image by using a dual-domain decomposition model; adjusting the brightness, contrast and definition of the basic subgraph by using a CLAHE algorithm; carrying out noise suppression and edge sharpening on the detail subgraph by adopting the constructed ILoG operator; and reconstructing the adjusted basic subgraph and the adjusted detailed subgraph, and adjusting the brightness of the reconstructed image by adopting a designed Gamma correction GCGR function with gray level redistribution to obtain the infrared enhanced image. Compared with the existing infrared image enhancement method, the infrared image enhancement method is faster, more accurate and more reliable, and is not only suitable for the complex environment with low illumination, high dust and water mist of a mine, but also suitable for infrared video monitoring of other underground spaces. In addition, in the mine emergency rescue process, the improved video images can also assist rescue workers to make a prediction on site disaster and make a corresponding emergency plan.
Disclosure of Invention
In order to solve the problems in the background art, the invention provides a mine infrared image enhancement method and a mine infrared image enhancement device, and the technical problems are specifically solved by adopting the following technical scheme:
in a first aspect, the invention provides a mine infrared image enhancement method, and the implementation process of the enhancement method comprises the following steps:
step 1: a 5 multiplied by 5 sliding window double-domain filtering model is selected to decompose the mine infrared image f into a basic sub-image f containing general appearance characteristics B And a detail sub-graph f containing detail features D ；
Step 2: the brightness, the contrast and the definition of the basic subgraph are improved by adopting a CLAHE algorithm;
and step 3: constructing an ILoG function, and performing noise suppression and edge sharpening on a detail sub-graph by adopting the ILoG function, and eliminating a gradient inversion phenomenon;
and 4, step 4: adopting a self-adaptive reconstruction function to carry out self-adaptive reconstruction on the basic subgraph and the detail subgraph processed in the step 3 and the step 4 so as to obtain a reconstructed image f with improved image quality R ；
The self-adaptive reconstruction function is f R (x,y)＝[1-α(x,y)]F B (x,y)+α(x,y)F D (x, y) wherein F B (x, y) is the basic subgraph F promoted in the step 2 B Pixel value at coordinate (x, y), F D (x, y) is a sharpened image F D A pixel value at coordinates (x, y), α (x, y) is an adaptive fusion coefficient, and
wherein
Respectively sharpened image F D A first-order difference operator in the row and column directions in the neighborhood with coordinates (x, y) as the center;
and 5: designing a Gamma correction function of gray redistribution, and adjusting the brightness of the reconstructed image to obtain the mine infrared enhanced image.
Further, the two-domain filtering process includes:
step 1: decomposing different image layers of the infrared image by adopting a bilateral filter to obtain a basic subgraph f representing the scene profile characteristics B (ii) a The bilateral filter is
Wherein f is B (k, l) is a base subgraph after bilateral filter decomposition, f (p, q) is a pixel value of the infrared image f at a coordinate (p, q) in a sliding window, and w (p, q, k, l) is a weight coefficient; (p, q) are coordinates of neighborhood pixel points; (k, l) is the coordinate of the central pixel point;
step 2: subtracting the base subgraph f from the infrared image f B I.e. f D (k,l)＝f(k,l)-ξf B (k, l) obtaining a detail sub-graph f representing the detail feature of the target D (ii) a Where ξ is an estimation coefficient ofAnd the loss of detail information of a small target in the infrared image is avoided, and xi is epsilon (0, 1).
Further, the process of improving the brightness, the contrast and the definition of the base sub-image by the CLAHE algorithm includes:
step 1: partitioning said base sub-graph by
Calculating the statistical probability of the ith level gray scale in each region; wherein p (i) is the statistical probability of the ith gray level, i is more than or equal to 0 and less than L, n i The total number of pixels of the ith gray level, n is the total number of pixels of each area, and L is the total number of gray levels in a single-channel image;
step 2: drawing a gray level histogram of a corresponding region according to the statistical probability p, and cutting a peak value of the gray level histogram to enable the amplitude value to be lower than the upper limit of a threshold value;
and step 3: and uniformly distributing the cut pixel points in the original gray level histogram, and when the cut parts exceed the cutting threshold value again after the pixel points are redistributed, repeatedly cutting each area by combining the corresponding gray level histogram distribution, and further performing repeated iteration processing until the exceeding parts are ignored.
Further, the specific process of performing noise suppression and edge sharpening on the detail subgraph by using the ILoG function and eliminating gradient inversion comprises the following steps:
step 1: using Gaussian filter to pair detail sub-graph f D Denoising;
step 2: substituting the Gaussian kernel function into a Laplacian operator to construct a LoG operator, wherein the LoG operator is
Wherein f is LoG Is a LoG kernel function; sigma is the standard deviation of the Gaussian kernel function, and the larger the sigma value is, the better the filtering effect is; />
For Laplacian operator, f G (x, y) is a Gaussian kernel function;
and 3, step 3: step 1 is performed by adopting LoG operatorNoisy detail image f p Carrying out edge sharpening to obtain a characteristic image f with sudden gray change in a detail subgraph C ；
And 4, step 4: constructing an ILoG function by adopting the LoG operator and the unit step function in the step 3, and eliminating gradient inversion in the characteristic image by adopting a detail sub-image denoising and sharpening strategy based on the ILoG function to obtain a sharpened image; the sharpened image F D By F D (x,y)＝f p (x,y)-usf(f C (x,y))f C (x, y) is obtained by calculation; where usf (-) is a unit step function; f D (x, y) is a sharpened image F D Pixel value at coordinate (x, y).
Further, the Gamma correction function of the gray scale redistribution for brightness adjustment of the reconstructed image is
Wherein F (x, y) is an infrared-enhanced image, F R (x, y) is a reconstructed image f R Pixel value at coordinates (x, y), f R,max For reconstructing an image f R Maximum pixel value of f R,min For reconstructing an image f R Is an adjustment factor, epsilon e (0, 1), for adjusting the brightness of the reconstructed image.
In a second aspect, the present invention provides a mine infrared image enhancement device, where the enhancement device obtains a mine infrared enhanced image by using a mine infrared image enhancement method, and the enhancement device includes:
the decomposition module is used for decomposing the mine infrared degradation image;
the lifting module is used for lifting the brightness, the contrast and the definition of the basic subgraph;
the sharpening module is used for sharpening the edge of the detail sub-image and eliminating the gradient inversion phenomenon;
the reconstruction module is used for reconstructing the basic subgraph and the detail subgraph;
and the correction module is used for correcting the brightness of the reconstructed image.
In the mine infrared image enhancement method and device provided by the embodiment of the invention, a dual-domain filtering is adopted to decompose a degraded image into a detail sub-image containing high-frequency information and a basic sub-image containing low-frequency information, the CLAHE algorithm is utilized to improve the brightness, contrast and definition of the basic sub-image, a constructed ILoG operator is adopted to carry out noise suppression and edge sharpening on the detail sub-image and eliminate the gradient inversion phenomenon, a self-adaptive reconstruction function is adopted to reconstruct the processed basic sub-image and detail sub-image, a reconstructed image with improved image quality is obtained, a Gamma correction function with distributed gray levels is designed, the brightness of the reconstructed image is adjusted, and the mine infrared enhanced image is further obtained. The comprehensive enhancement performance and robustness of the method are superior to those of the existing algorithm, and the method can be suitable for enhancement processing of the infrared image in the mine complex environment.
Drawings
FIG. 1 is a schematic flow diagram of a method for mine infrared image enhancement according to the present invention;
FIG. 2 is a block schematic diagram of the mine infrared image enhancement device of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings and specific implementation methods in the embodiments, and it is obvious that the described embodiments are a part of embodiments of the present invention, but not all embodiments. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.
The invention provides a mine infrared image enhancement method and device. Fig. 1 shows a schematic flow chart of a mine infrared image enhancement method according to an embodiment of the present invention. The mine infrared image enhancement method comprises the following steps:
and 110, decomposing the degraded image into a basic subgraph and a detail subgraph by adopting double-domain filtering.
In this embodiment, the core idea of the two-domain filtering is to define any pixel point and some neighboring pixel points as a linear relationship, and after performing local filtering processing in sequence, accumulate all local filtering results to derive a global filtering result.
And step 120, utilizing a CLAHE algorithm to improve the brightness, the contrast and the definition of the basic subgraph.
In this embodiment, the CLAHE algorithm is a variation of the AHE algorithm, and the CLAHE algorithm relates the improvement degree of the area contrast to the curve gradient of the gray histogram, and performs appropriate clipping on the sharp gray histogram before obtaining the cumulative distribution function of each area, and performs average distribution on the clipped pixels in the gray histogram, thereby overcoming the problem that the AHE algorithm is easy to cause the contrast overshoot adjustment of the uniform area.
And step 130, carrying out noise suppression and edge sharpening on the detail subgraph by adopting the constructed ILoG operator, and eliminating the gradient inversion phenomenon.
In this embodiment, a detail sub-image denoising and sharpening strategy based on ILoG transformation is adopted to realize enhancement of detail features and suppression of high-frequency noise. Firstly, denoising a detail subgraph by adopting a Gaussian filter; secondly, edge sharpening is carried out on the denoised image by adopting a Laplacian operator; finally, gradient inversion is eliminated by constructing a unit step function. The strategy can not only improve the signal-to-noise ratio of the detail subgraph, but also effectively improve the significance of detail features such as edges, textures and the like of the detail subgraph.
Step 140, the basic subgraph and the detail subgraph processed in the steps 120 and 130 are adaptively reconstructed by adopting an adaptive reconstruction function, and a reconstructed image f with improved image quality is obtained R ；
The self-adaptive reconstruction function is f R (x,y)＝[1-α(x,y)]F B (x,y)+α(x,y)F D (x, y) wherein F B (x, y) is the basic subgraph F promoted in the step 2 B Pixel value at coordinate (x, y), F D (x, y) is a sharpened image F D A pixel value at coordinates (x, y), α (x, y) is an adaptive fusion coefficient, and
wherein
Respectively sharpened image F D And (3) a first-order difference operator in the row and column directions in the neighborhood with coordinates (x, y) as the center.
And 150, designing a Gamma correction function of gray redistribution, and adjusting the brightness of the reconstructed image to obtain the mine infrared enhanced image.
In this embodiment, after the detail sub-image is subjected to noise reduction and sharpening by using ILoG operator, the obtained sharpened image is entirely dark, and the sharpened image F is directly subjected to noise reduction and sharpening D And contrast adjusted base sub-graph F B And if the reconstruction is carried out, the detail features of the dark areas in the obtained reconstructed image are not obvious. For this purpose, a Gamma corrected GCGR function of the gray redistribution is used to adjust the brightness of the reconstructed image and highlight the detail characteristics of the dark regions of the image.
In the mine infrared image enhancement method and device provided by the embodiment of the invention, a mine degraded image is decomposed into a basic sub-image containing a profile characteristic and a detail sub-image containing a detail characteristic by adopting double-domain filtering, the brightness, the contrast and the definition of the basic sub-image are improved by utilizing a CLAHE algorithm, the detail sub-image is subjected to noise suppression and edge sharpening by adopting a constructed ILoG operator, the gradient reversal phenomenon is eliminated, the processed basic sub-image and detail sub-image are reconstructed, a reconstructed image with improved image quality is obtained, a Gamma correction function with gray level redistribution is designed, the brightness of the reconstructed image is adjusted, and the mine infrared enhanced image is further obtained. The method has better comprehensive enhancement performance and robustness, and is suitable for enhancing the infrared image in the complex mine environment.
It should be noted that, for simplicity of description, the above-mentioned method embodiments are described as a series of acts or combination of acts, but those skilled in the art will recognize that the present invention is not limited by the order of acts, as some steps may occur in other orders or concurrently in accordance with the invention. Further, those skilled in the art should also appreciate that the embodiments described in the specification are exemplary embodiments and that the acts and modules referred to are not necessarily required to practice the invention.
The above is a description of method embodiments, and the embodiments of the present invention are further described below by way of apparatus embodiments.
Fig. 2 shows a schematic block diagram of a mine infrared image enhancement method according to an embodiment of the present invention. As shown in fig. 2, the mine infrared image enhancement device includes:
a decomposition module 210 for decomposing the mine infrared degraded image;
a lifting module 220, configured to lift brightness, contrast, and sharpness of the base sub-image;
a sharpening module 230, configured to sharpen an edge of the detail sub-image and eliminate a gradient inversion phenomenon;
a reconstruction module 240 for reconstructing the base subgraph and the detail subgraph;
a correcting module 250, configured to correct the brightness of the reconstructed image.
In some embodiments, the decomposition module 210 is specifically configured to:
separating different layers of the infrared image by adopting a double-domain decomposition model to obtain a basic subgraph representing the scene general appearance characteristics, and subtracting the basic subgraph from the original infrared image to obtain a detail subgraph representing the target detail characteristics; the method mainly comprises the following steps:
step 1: decomposing different image layers of the infrared image by adopting a bilateral filter to obtain a basic subgraph f representing the scene profile characteristics B (ii) a The bilateral filter is
Wherein f is B (k, l) is a base subgraph after bilateral filter decomposition, f (p, q) is a pixel value of the infrared image f at a coordinate (p, q) in a sliding window, and w (p, q, k, l) is a weight coefficient; (p, q) are coordinates of neighborhood pixel points; (k, l) is the coordinate of the central pixel point;
in some embodiments, the weighting factor w (p, q, k, l) is a product of a spatial kernel and a value-domain kernel, expressed as
In the formula>
Is a spatial domain variance, σ d R/30, R is the number of lines or columns of the original image; />
Is a value domain variance, σ r ＝(V max -V min )/50，V max Is the maximum gray level, V, of the original image min Is the minimum gray level of the original image.
And 2, step: subtracting the base subgraph f from the infrared image f B I.e. f D (k,l)＝f(k,l)-ξf B (k, l) obtaining a detail sub-graph f representing the detail feature of the target D (ii) a And xi is an estimation coefficient, and xi belongs to (0, 1) in order to avoid loss of detail information of small targets in the infrared image.
In some embodiments, the lifting module 220 is specifically configured to:
the Contrast Limited Adaptive Histogram (CLAHE) algorithm correlates the improvement degree of the regional contrast with the curve gradient of the gray histogram, properly cuts the sharp gray histogram before acquiring the cumulative distribution function of each region, and evenly distributes the cut pixels in the gray histogram, thereby overcoming the problem that the AHE algorithm is easy to cause the contrast over-adjustment of the uniform region. For this purpose, the CLAHE algorithm is used for the base subgraph f B Brightness and contrast adjustments are made. In some embodiments, a gray histogram of the corresponding region is plotted according to the statistical probability p, and the peak value of the gray histogram is clipped so that the amplitude value is lower than the upper threshold. The implementation process comprises the following steps:
step 1: partitioning said base sub-graph by
Calculating the statistical probability of the ith level of gray in each region; wherein p (i) is the statistical probability of the ith gray level, i is more than or equal to 0 and less than L, n i The total number of pixels of the ith gray level, n is the total number of pixels of each area, and L is the total number of gray levels in a single-channel image;
step 2: drawing a gray level histogram of a corresponding region according to the statistical probability p, and cutting a peak value of the gray level histogram to enable the amplitude value to be lower than the upper limit of a threshold value;
and step 3: and uniformly distributing the cut pixel points in the original gray level histogram, and when the cut parts exceed the cutting threshold value again after the pixel points are redistributed, repeatedly cutting each area by combining the corresponding gray level histogram distribution, and further performing repeated iteration processing until the exceeding parts are ignored.
In the implementation process, the cut pixels are uniformly distributed in the original gray level histogram to ensure that no pixel is lost. The iterative processing can reduce the image 'block' effect caused by distinct gray-scale intensity distribution property in non-adjacent areas to the maximum extent. In addition, the method can better reserve the fine scene profile characteristics and avoid being swallowed in the contrast enhancement process. Obtaining the cumulative distribution function of the basic subgraph by cutting the gray level histogram, and obtaining the basic subgraph F after the contrast adjustment by gray level remapping B 。
In some embodiments, the sharpening module 230 is specifically configured to:
denoising the detail subgraph by adopting a Gaussian filter; in some embodiments, said denoising said detail subgraph using a Gaussian filter comprises:
performing convolution operation on the 2D Gaussian filter and the detail subgraph to obtain a denoised smooth image f p . The Gaussian filter process is f P (x,y)＝f G (x,y)*f D (x, y); in the formula f p (x, y) is a smoothed image f P A pixel value at position (x, y); f. of D (x, y) is detail sub-diagram f D A pixel value at location (x, y); f. of G (x, y) is a Gaussian kernel function,
* Is the convolution operator.
Assuming that the 2D Gaussian kernel size is (2r + 1) × (2r + 1), the discretized Gaussian kernel function f G (x, y) is represented as:
wherein r is the radius of the Gaussian kernel and is used for controlling the size of the Gaussian kernel.
In some embodiments, the edge sharpening the denoised detail image by using the Laplacian operator includes:
the Laplacian transform of a 2D grayscale image is the isotropic second derivative, which is defined as
In the formula>
Is Laplacian operator; f. of L (x, y) is a Laplacian kernel function.
Substituting the Gaussian kernel function
And obtaining a LoG operator after derivation. The LoG operator is->
In the formula f LoG Is a LoG kernel function; sigma is the standard deviation of the Gaussian kernel function, and the larger the sigma value is, the better the filtering effect is.
Smoothing the image f p And carrying out convolution with the LoG operator to obtain a characteristic image of gray mutation in the detail subgraph. The convolution process is
In the formula f C And (x, y) is a characteristic image of the detail subgraph.
Gradient inversion is eliminated by constructing a unit step function. In some embodiments, said eliminating gradient inversion by constructing a unit step function comprises: constructing an ILoG function by adopting the LoG operator and the unit step function; eliminating gradient inversion in the characteristic image by adopting a detail sub-image denoising and sharpening strategy based on an ILoG function, and obtaining a sharpened image;
the sharpened image F D By F D (x,y)＝f p (x,y)-usf(f C (x,y))f C (x, y) is calculated; where usf (-) is the unit stepA function; f D (x, y) is a sharpened image F D Pixel value at coordinate (x, y).
In some embodiments, the reconstruction module 240 is specifically configured to: sampling self-adaptive reconstruction function pair to the basic subgraph F B And sharpening image F D Carrying out self-adaptive reconstruction to obtain a reconstructed image f R 。
In some embodiments, the correction module 250 is specifically configured to:
adjusting reconstructed image f with Gamma correction of gray scale redistribution (GCGR) function R And an infrared-enhanced image F is obtained.
In some embodiments, the GCGR function adjusts the reconstructed image f R The luminance of (b) includes: and adopting a GCGR function to adjust the brightness of the reconstructed image and highlight the detail characteristics of dark areas of the image. Having a GCGR function of
Wherein F (x, y) is the pixel value of the infrared enhanced image F at the coordinate (x, y), F R (x, y) is a reconstructed image f R Pixel value at coordinate (x, y), f R,max For reconstructing an image f R Maximum pixel value of f R,min For reconstructing an image f R Is an adjustment factor, epsilon e (0, 1), for adjusting the brightness of the reconstructed image.
It can be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working process of the described module may refer to the corresponding process in the foregoing method embodiment, and is not described herein again.
The above are all preferred embodiments of the present invention, and the protection scope of the present application is not limited thereby, so: all equivalent changes made according to the structure, shape and principle of the invention are covered by the protection scope of the invention.
Claims (6)
1. A mine infrared image enhancement method is characterized in that the implementation process of the enhancement method comprises the following steps:
step 1: decomposing the mine degraded image into a basic sub-image containing profile features and a detail sub-image containing detail features by adopting double-domain filtering;
step 2: the luminance, the contrast and the definition of the basic subgraph are improved by adopting a CLAHE algorithm;
and step 3: constructing an ILoG function, and performing noise suppression and edge sharpening on a detail sub-graph by adopting the ILoG function, and eliminating a gradient inversion phenomenon;
and 4, step 4: adopting a self-adaptive reconstruction function to carry out self-adaptive reconstruction on the basic subgraph and the detail subgraph processed in the step 3 and the step 4 so as to obtain a reconstructed image f with improved image quality R ；
The self-adaptive reconstruction function is f R (x,y)＝[1-α(x,y)]F B (x,y)+α(x,y)F D (x, y) wherein F B (x, y) is the basic subgraph F promoted in the step 2 B Pixel value at coordinate (x, y), F D (x, y) is a sharpened image F D A pixel value at coordinates (x, y), α (x, y) is an adaptive fusion coefficient, and
wherein ^ f x ，▽f y Respectively sharpened image F D A first-order difference operator in the row and column directions in the neighborhood with coordinates (x, y) as the center;
and 5: designing a Gamma correction function of gray redistribution, and adjusting the brightness of the reconstructed image to obtain the mine infrared enhanced image.
2. The mine infrared image enhancement method of claim 1, wherein the two-domain filtering process comprises:
step 1: decomposing different image layers of the infrared image by adopting a bilateral filter to obtain a basic subgraph f representing the scene profile characteristics B (ii) a The bilateral filter is
Wherein f is B (k, l) is a base subgraph after bilateral filter decomposition, and f (p, q) is the coordinate of the infrared image f in the sliding windowPixel value at (p, q), w (p, q, k, l) is a weight coefficient; (p, q) are coordinates of neighborhood pixel points; (k, l) is the coordinate of the central pixel point;
step 2: subtracting the base subgraph f from the infrared image f B I.e. f D (k,l)＝f(k,l)-ξf B (k, l) obtaining a detail sub-graph f representing the detail feature of the target D (ii) a And xi is an estimation coefficient, and xi is an element (0, 1) in order to avoid loss of detail information of small targets in the infrared image.
3. The mine infrared image enhancement method as claimed in claim 1, wherein the CLAHE algorithm for improving the brightness, contrast and sharpness of the base sub-image comprises:
step 1: partitioning said base sub-graph by
Calculating the statistical probability of the ith level of gray in each region; wherein p (i) is the statistical probability of the ith gray level, i is more than or equal to 0 and less than L, n i The total number of pixels of the ith gray level, n is the total number of pixels of each area, and L is the total number of gray levels in a single-channel image;
step 2: drawing a gray level histogram of a corresponding region according to the statistical probability p, and cutting a peak value of the gray level histogram to enable the amplitude value to be lower than the upper limit of a threshold value;
and step 3: and uniformly distributing the cut pixel points in the original gray level histogram, and when the cut parts exceed the cutting threshold value again after the pixel points are redistributed, repeatedly cutting each area by combining the corresponding gray level histogram distribution, and further performing repeated iteration processing until the exceeding parts are ignored.
4. The mine infrared image enhancement method of claim 1, wherein the specific process of performing noise suppression and edge sharpening on the detail subgraph by using the ILoG function and eliminating gradient inversion comprises:
step 1: using Gaussian filter to pair detail sub-graph f D Denoising;
step 2: substituting the Gaussian kernel function into a Laplacian operator to construct a LoG operator, wherein the LoG operator is
Wherein f is LoG Is a LoG kernel function; sigma is the standard deviation of Gaussian kernel function, and the larger the sigma value is, the better the filtering effect is; v 2 For Laplacian operator, f G (x, y) is a Gaussian kernel function;
and step 3: adopting LoG operator to denoise detail image f in step 1 p Carrying out edge sharpening to obtain a characteristic image f with sudden gray change in a detail sub-image C ；
And 4, step 4: constructing an ILoG function by adopting the LoG operator and the unit step function in the step 3, and eliminating gradient inversion in the characteristic image by adopting a detail sub-image denoising and sharpening strategy based on the ILoG function to obtain a sharpened image; the sharpened image F D By F D (x,y)＝f p (x,y)-usf(f C (x,y))f C (x, y) is calculated; where usf (-) is a unit step function; f D (x, y) is a sharpened image F D Pixel value at coordinate (x, y).
5. The mine infrared image enhancement method of claim 1, wherein the Gamma correction function for gray scale redistribution is
Wherein F (x, y), F R (x, y) are respectively the enhanced image F and the reconstructed image F R Pixel value at coordinate (x, y), f R,max For reconstructing an image f R Maximum pixel value of f R,min For reconstructing an image f R Is an adjustment factor, epsilon e (0, 1), for adjusting the brightness of the reconstructed image.
6. The mine infrared image enhancement device is characterized in that the enhancement device obtains a mine infrared enhanced image by adopting a mine infrared image enhancement method, and comprises the following steps:
the decomposition module is used for decomposing the mine infrared degradation image;
the lifting module is used for lifting the brightness, the contrast and the definition of the basic subgraph;
the sharpening module is used for sharpening the edge of the detail sub-image and eliminating the gradient inversion phenomenon;
the reconstruction module is used for reconstructing the basic subgraph and the detail subgraph;
and the correction module is used for correcting the brightness of the reconstructed image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211633829.6A CN115861117A (en) | 2022-12-19 | 2022-12-19 | Mine infrared image enhancement method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211633829.6A CN115861117A (en) | 2022-12-19 | 2022-12-19 | Mine infrared image enhancement method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115861117A true CN115861117A (en) | 2023-03-28 |
Family
ID=85674134
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211633829.6A Pending CN115861117A (en) | 2022-12-19 | 2022-12-19 | Mine infrared image enhancement method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115861117A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117437151A (en) * | 2023-12-21 | 2024-01-23 | 成都市晶林科技有限公司 | Pseudo-color mapping method for noise suppression |
-
2022
- 2022-12-19 CN CN202211633829.6A patent/CN115861117A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117437151A (en) * | 2023-12-21 | 2024-01-23 | 成都市晶林科技有限公司 | Pseudo-color mapping method for noise suppression |
| CN117437151B (en) * | 2023-12-21 | 2024-03-08 | 成都市晶林科技有限公司 | Pseudo-color mapping method for noise suppression |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105844601A (en) | Mine image enhancement method based on bilateral filtering and multi-scale Retinex algorithm | |
| CN111986120A (en) | Low-illumination image enhancement optimization method based on frame accumulation and multi-scale Retinex | |
| CN103077508A (en) | Transform domain non local and minimum mean square error-based SAR (Synthetic Aperture Radar) image denoising method | |
| CN110047055B (en) | Infrared image detail enhancement and denoising method | |
| CN107330870B (en) | A kind of thick fog minimizing technology accurately estimated based on scene light radiation | |
| CN115496695A (en) | High-dynamic infrared image self-adaptive enhancement and compression method | |
| CN104881847A (en) | Match video image enhancement method based on wavelet analysis and pseudo-color processing | |
| CN109949256B (en) | Astronomical image fusion method based on Fourier transform | |
| CN115861117A (en) | Mine infrared image enhancement method and device | |
| Weizheng et al. | Digital image denoising method based on mean filter | |
| Diwakar et al. | CT Image noise reduction based on adaptive wiener filtering with wavelet packet thresholding | |
| CN115393216A (en) | Image defogging method and device based on polarization characteristics and atmospheric transmission model | |
| CN113850744B (en) | Image enhancement method based on self-adaptive Retinex and wavelet fusion | |
| CN103679663A (en) | Image denoising method combining Tetrolet transform domain and PDE (Partial Differential Equation) and GCV (Generalized Cross Validation) theory | |
| CN106709876B (en) | Optical remote sensing image defogging method based on dark image element principle | |
| CN108629740B (en) | Image denoising processing method and device | |
| CN105096274B (en) | Infrared image noise-reduction method based on non-down sampling contourlet domain mixing statistical model | |
| CN113222833A (en) | Side-scan sonar image processing method and device | |
| CN108460736A (en) | A kind of low-light (level) power equipment image song wave zone Enhancement Method | |
| CN116579953A (en) | Self-supervision water surface image enhancement method and related equipment | |
| Gir et al. | Speckle reduction of synthetic aperture radar images using median filter and savitzky-golay filter | |
| CN111640160A (en) | CT image preprocessing method | |
| KR101583430B1 (en) | Method and apparatus for processing image | |
| Cheng et al. | Fast iterative adaptive reconstruction in low-dose CT imaging | |
| JP7427239B2 (en) | Signal processing device, signal processing method and program |
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 |