CN115293987A - Improved limited self-adaptive image equalization enhancement algorithm - Google Patents
Improved limited self-adaptive image equalization enhancement algorithm Download PDFInfo
- Publication number
- CN115293987A CN115293987A CN202210993348.XA CN202210993348A CN115293987A CN 115293987 A CN115293987 A CN 115293987A CN 202210993348 A CN202210993348 A CN 202210993348A CN 115293987 A CN115293987 A CN 115293987A
- Authority
- CN
- China
- Prior art keywords
- image
- gray level
- mapping
- enhancement
- adaptive
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/40—Scaling the whole image or part thereof
- G06T3/4007—Interpolation-based scaling, e.g. bilinear interpolation
-
- 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 by the use of more than one image, e.g. averaging, 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
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- 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
- G06T2207/20081—Training; Learning
-
- 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
Abstract
The invention belongs to the technical field of data processing, and provides an improved limited self-adaptive image equalization enhancement algorithm, which comprises the following steps: firstly, carrying out self-adaptive subimage decomposition on an image, decomposing an original image into a plurality of continuous non-overlapping subimages, and traversing each subimage; secondly, calculating a gray level mapping function corresponding to each pixel point in the image; step three, correcting the mapping gray level f; and step four, replacing the gray levels of all pixel points in the original image with the enhanced image. According to the invention, through a self-adaptive algorithm, the problems of too many or too few numbers and shapes of block images are solved, the defects of excessive local enhancement or insufficient local enhancement are reduced, the global information of the images is fully utilized, and the problems of gray scale aliasing, excessive enhancement, obvious block effect and the like of the enhanced images are prevented.
Description
Technical Field
The invention belongs to the technical field of image enhancement, and particularly relates to an improved limited self-adaptive image equalization enhancement algorithm.
Background
The image enhancement is an important preprocessing step in image processing, and can effectively improve the image quality and improve the subjective visual effect of the image. The method comprises the steps of firstly dividing an image into a plurality of continuous and non-overlapping image blocks, then respectively carrying out histogram correction and equalization to obtain corresponding gray mapping functions, and finally carrying out bilinear interpolation to relieve discontinuous effects among different image blocks. The local detail information of the image is effectively enhanced, and the subjective visual effect of the image is improved, but the method has the following problems:
1. the number and shape of the block images cannot be adapted. The number of block images has a large influence on the local enhancement effect: if the number of blocks is too large, local enhancement is excessive; if the number of the blocks is too small, the local enhancement is insufficient; the blocking image shape fixation will affect the final image enhancement effect.
2. The global information of the image is not fully utilized. The image is divided into a plurality of sub-images to be enhanced independently, so that the local information of the image can be effectively utilized, but the global information of the image is not fully utilized, which may cause the problems of gray aliasing, over-enhancement, obvious blocking effect and the like of the enhanced image.
Disclosure of Invention
The invention aims to overcome the defects in the prior art, and provides an improved limited self-adaptive image equalization enhancement algorithm to further improve the image enhancement effect.
In order to solve the above technical problem, the present invention provides an improved limited adaptive image equalization enhancement algorithm, including: firstly, carrying out self-adaptive subimage decomposition on an image, decomposing an original image into a plurality of continuous non-overlapping subimages, and traversing each subimage; secondly, calculating a gray level mapping function corresponding to each pixel point in the image; step three, correcting the mapping gray level f; and step four, replacing the gray levels of all pixel points in the original image with the enhanced image.
Further, mapping the image through a cumulative probability distribution function to obtain equalized gray levels.
Further, the pixel point gray value is reconstructed by using the equalized gray level.
Further, the gray scale mapping function is
Further, the mapping gray level corresponding to a certain pixel point in the image is a distance-based weighted value of the mapping gray level of all pixel points in the original image, which are the same as the gray level of the pixel points.
Has the beneficial effects that:
the invention solves the problems of too much or too little quantity and shape of the block images through the self-adaptive algorithm, reduces the defects of excessive local enhancement or insufficient local enhancement, fully utilizes the global information of the images and prevents the problems of gray level aliasing, excessive enhancement, obvious block effect and the like of the enhanced images.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:
FIG. 1 is a block diagram of the principle flow of the present algorithm;
FIG. 2 is an original image to be processed;
FIG. 3 is an image after processing by a conventional algorithm;
fig. 4 is an image after processing by the enhancement algorithm of the present invention.
Detailed Description
The preferred embodiments of the present invention will be described in conjunction with the accompanying drawings, and it will be understood that they are described herein for the purpose of illustration and explanation and not limitation.
As shown in fig. 1, the main steps of the improved enhancement algorithm proposed by this patent are as follows:
1. and performing adaptive sub-image decomposition on the image. Selecting an original image with size set to I M×N Decomposing into round (M/S) by round (N/S) consecutive non-overlapping sub-images of size S X S, traversing each sub-image if it and its right sideOr the lower neighboring subimage satisfies J (I) i ,I i_adjacent ) If T is greater than T, the sub-images are merged, wherein I i For the current sub-picture, I i_adjacent And J is a fusion discriminator based on deep learning, the input of the fusion discriminator is two sub-images, the output is a fusion probability, and T is a probability threshold value which is generally set to be 0.5. The network structure and the training method adopt the prior art, and training data comprises about 3000 groups of subimages needing to be fused by manual judgment. After fusion, we can self-adapt to get the best sub-image number and shape. When a background area (such as sky, ocean and the like) with a large number of close gray levels is enhanced, the traditional method decomposes the background area into a plurality of sub-images, only local information is considered, and the phenomenon of local over-enhancement is easy to occur. Through the fusion steps, the gray level near background regions can be fused into a sub-image in a self-adaptive mode to be processed, and the image enhancement effect is effectively improved.
2. And calculating a gray level mapping function corresponding to each pixel point in the image. The corrected histogram and the mapping function of the sub-image after the self-adaptive decomposition in the step one are firstly solved, then the mapping gray level f (i, j) corresponding to each pixel point is obtained based on a bilinear interpolation algorithm, (i =1,2, …, M, j =1,2, …, N), the size of M multiplied by N is the size of an input image, the image is mapped through an accumulation probability distribution function to obtain an equalized gray level, and the gray level of the pixel point is reconstructed by using the equalized gray level.
3. The mapping gray level f is corrected. Performing the following operation on the mapping gray level f obtained in the second step to obtain the final mapping gray level f out 。
for i=1:M
for j=1:N
end
end
And P { I (I, j) } is the positions of all pixel points with the gray values I (I, j) in the original image. σ is a variance, which controls the degree of influence of the distance on the weighting coefficients: the larger σ, the smaller the distance impact. It can be seen from the above formula that the mapping gray level corresponding to a certain pixel point in the image is a distance-based weighted value of the mapping gray level of all pixel points in the original image that have the same gray level, so that the closer the pixel points in the original image that have the same gray level are, the closer the mapped gray level is. Because the mapping gray level of each pixel in the traditional algorithm only considers the local information of the sub-block images nearby, the phenomena of gray aliasing (pixels with the same gray level in the image have a great gray level difference after enhancement) and over-enhancement easily occur. Through the weighting operation, the phenomena of gray level mixing and over-enhancement can be effectively relieved. In addition, if the difference of the mapping functions between the adjacent image blocks is large, the boundary blocking effect is still obvious after bilinear interpolation. By the weighting operation, the difference of the mapping function between adjacent blocks can be effectively relieved, and the boundary effect between adjacent blocks is avoided.
4. Replacing the gray levels of all pixel points (i, j) in the original image with f out (i, j) (i =1,2, …, M, j =1,2, …, N) results in an enhanced image.
In order to verify the effectiveness, reasonableness, feasibility and scientificity of the algorithm provided by the patent, the original image in the image data of the figure 2 is enhanced by adopting a traditional algorithm and an improved algorithm. FIG. 2 is an original image; FIG. 3 is an image after enhancement by a conventional algorithm; fig. 4 is an image after enhancement by applying the method of the present invention. σ =20,s =20. As can be seen from fig. 3, the gray levels of the large-area gray background areas in the original image are relatively close to each other, which results in a local over-enhancement phenomenon; in addition, since the difference between the sub-block histograms is significant, there is a small amount of discontinuity and blocking between sub-image boundaries. As can be seen from FIG. 4, the algorithm can effectively avoid the image over-enhancement and the boundary effect, and greatly improve the subjective visual effect of the image.
Finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that changes may be made in the embodiments and/or equivalents thereof without departing from the spirit and scope of the invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (5)
1. An improved constrained adaptive image equalization enhancement algorithm, comprising:
firstly, carrying out self-adaptive subimage decomposition on an image, decomposing an original image into a plurality of continuous non-overlapping subimages, and traversing each subimage;
secondly, calculating a gray level mapping function corresponding to each pixel point in the image;
step three, correcting the mapping gray level f;
and step four, replacing the gray levels of all pixel points in the original image with the enhanced image.
2. The improved limited adaptive image equalization enhancement algorithm of claim 1, wherein: and mapping the image through an accumulative probability distribution function to obtain the equalized gray level.
3. The improved limited adaptive image equalization enhancement algorithm of claim 2, wherein: and reconstructing the gray value of the pixel point by using the equalized gray level.
5. The improved limited adaptive image equalization enhancement algorithm of claim 4, wherein: the mapping gray level corresponding to a certain pixel point in the image is the weighted value of the mapping gray level of all the pixel points which are the same as the gray level of the pixel point in the original image based on the distance.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210993348.XA CN115293987A (en) | 2022-08-18 | 2022-08-18 | Improved limited self-adaptive image equalization enhancement algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210993348.XA CN115293987A (en) | 2022-08-18 | 2022-08-18 | Improved limited self-adaptive image equalization enhancement algorithm |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115293987A true CN115293987A (en) | 2022-11-04 |
Family
ID=83830356
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210993348.XA Pending CN115293987A (en) | 2022-08-18 | 2022-08-18 | Improved limited self-adaptive image equalization enhancement algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115293987A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117522870A (en) * | 2024-01-04 | 2024-02-06 | 陕西凯迈航空航天机电设备有限公司 | Intelligent defect detection method for aeroengine parts based on machine vision |
-
2022
- 2022-08-18 CN CN202210993348.XA patent/CN115293987A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117522870A (en) * | 2024-01-04 | 2024-02-06 | 陕西凯迈航空航天机电设备有限公司 | Intelligent defect detection method for aeroengine parts based on machine vision |
CN117522870B (en) * | 2024-01-04 | 2024-03-19 | 陕西凯迈航空航天机电设备有限公司 | Intelligent defect detection method for aeroengine parts based on machine vision |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110148095B (en) | Underwater image enhancement method and enhancement device | |
CN108876743B (en) | Image rapid defogging method, system, terminal and storage medium | |
CN105931201B (en) | A kind of image subjective vision effect reinforcing method based on wavelet transformation | |
CN114049283A (en) | Self-adaptive gray gradient histogram equalization remote sensing image enhancement method | |
CN114092361B (en) | Image local contrast enhancement method for optimizing CLAHE | |
EP2449524A1 (en) | Contrast enhancement | |
CN104253930A (en) | Real-time video defogging method | |
CN111612722B (en) | Low-illumination image processing method based on simplified Unet full-convolution neural network | |
CN111105376B (en) | Single-exposure high-dynamic-range image generation method based on double-branch neural network | |
CN102640485A (en) | Image processor, image processing method, and integrated circuit | |
CN115293987A (en) | Improved limited self-adaptive image equalization enhancement algorithm | |
CN111325685B (en) | Image enhancement algorithm based on multi-scale relative gradient histogram equalization | |
CN115456905A (en) | Single image defogging method based on bright and dark region segmentation | |
CN112488968B (en) | Image enhancement method for hierarchical histogram equalization fusion | |
CN114155173A (en) | Image defogging method and device and nonvolatile storage medium | |
CN110675345A (en) | Fuzzy completion processing method and device for remote sensing image to-be-repaired area | |
CN109767407A (en) | A kind of quadratic estimate method of atmospheric transmissivity image during defogging | |
CN109859138B (en) | Infrared image enhancement method based on human visual characteristics | |
CN109741276B (en) | Infrared image base layer processing method and system based on filtering layered framework | |
CN114331925A (en) | Multi-scale relative gradient histogram equalization method and device | |
CN114283076A (en) | Image processing method based on transmissivity and CLAHE algorithm | |
CN107945196A (en) | A kind of the image boundary detection method for correction image mosaic of taking photo by plane | |
CN115100083B (en) | Image brightness self-adaptive adjusting method for vehicle-mounted image | |
CN115660971B (en) | Method for realizing USM sharpening based on deep learning hardware accelerator | |
CN110717875A (en) | High-definition image processing method |
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 |