CN109919865B - Image multi-layer detail enhancement method - Google Patents
Image multi-layer detail enhancement method Download PDFInfo
- Publication number
- CN109919865B CN109919865B CN201910127062.1A CN201910127062A CN109919865B CN 109919865 B CN109919865 B CN 109919865B CN 201910127062 A CN201910127062 A CN 201910127062A CN 109919865 B CN109919865 B CN 109919865B
- Authority
- CN
- China
- Prior art keywords
- detail
- window
- windows
- auxiliary
- image
- 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
Landscapes
- Image Processing (AREA)
Abstract
The invention discloses a method for enhancing image multilayer details, which comprises the following steps: s1: selecting a window from an original image as a detail window; s2: selecting a plurality of auxiliary windows from detail windows, wherein the sizes of the auxiliary windows are smaller than those of the detail windows; s3: carrying out smoothing processing on the detail window to generate a smooth enhancement window; carrying out smoothing processing on the plurality of auxiliary windows to generate a plurality of smooth auxiliary windows; s4: respectively extracting the smooth enhancement window and the plurality of smooth auxiliary windows from the detail window to form a plurality of detail auxiliary windows; s5: and weighting and superposing the smooth enhanced window and the plurality of detail auxiliary windows to obtain a detail enhanced window. According to the image multilayer detail enhancement method, through the steps, not only the low-frequency image detail is considered, but also the high-frequency image detail is considered, and meanwhile, partial noise can be eliminated, so that the image detail enhancement result is more accurate.
Description
Technical Field
The invention relates to a computer image processing technology, in particular to an image multi-layer detail enhancement method.
Background
Image enhancement refers to a process of emphasizing certain information in an image as desired while attenuating or removing certain unwanted information. Image enhancement algorithms can be basically divided into two broad categories: spatial domain processing and transform domain processing. Most of traditional spatial domain algorithms such as histogram equalization, spatial filtering, image edge sharpening and the like enhance the contrast and improve the image noise. The wavelet enhancement method gives consideration to the characteristics of both the spatial domain and the frequency domain of an image signal, and the multi-directional characteristic and the multi-resolution characteristic of the wavelet enhancement method open a new way for solving the problem of spatial domain contradiction, but the visual nonlinear characteristic is not fully considered in the classical method, and the aim of image enhancement is to improve the visual effect of an image so as to improve the identification precision.
The existing image detail enhancement method generally cannot take the details of low-frequency and high-frequency images into consideration at the same time, so that the image detail enhancement result is not accurate enough and the noise is high.
Disclosure of Invention
The invention aims to solve the technical problems that the conventional image detail enhancement method generally cannot take into account the low-frequency and high-frequency image details at the same time, so that the image detail enhancement result is not accurate enough and the noise is high, and the invention aims to provide an image multilayer detail enhancement method to solve the problems.
The invention is realized by the following technical scheme:
an image multi-layer detail enhancement method comprises the following steps: s1: selecting a window from an original image as a detail window; s2: selecting a plurality of auxiliary windows from detail windows, wherein the sizes of the auxiliary windows are smaller than those of the detail windows; s3: smoothing the detail window to generate a smooth enhancement window; carrying out smoothing processing on the plurality of auxiliary windows to generate a plurality of smooth auxiliary windows; s4: respectively extracting the smooth enhancement window and the plurality of smooth auxiliary windows from the detail window to form a plurality of detail auxiliary windows; s5: and weighting and superposing the smooth enhanced window and the plurality of detail auxiliary windows to obtain a detail enhanced window.
The existing image detail enhancement method generally cannot take low-frequency and high-frequency image details into consideration at the same time, so that the image detail enhancement result is not accurate enough, and the noise is high. When the method is applied, the design core of the method is to represent the original image as a basic component and a detail component, enhance the detail component on the basis and obtain the enhanced image, so the key point of the design of the method is how to obtain the detail component.
The invention selects a window in an original image as a detail window, wherein the detail window is a window for detail enhancement, the detail window is spliced after detail enhancement to obtain a result after the detail enhancement of the original image, then a plurality of auxiliary windows are selected from the detail window, the sizes of the auxiliary windows are all smaller than the detail window, the plurality of auxiliary windows are used for extracting details, generally all the auxiliary windows are different in size, so that the auxiliary windows with various sizes can be covered, and the inventor finds that when the auxiliary window is smaller, the high-frequency fine image details, such as image texture details, can be well extracted, but the auxiliary window is not beneficial to extracting the low-frequency large image details, such as image edges; when the auxiliary window is large, large image details with low frequency can be well extracted, but the extraction of the image details with high frequency and small details is not facilitated; therefore, the adoption of the plurality of auxiliary windows can realize the extraction of the image characteristics of different frequencies in the image through different auxiliary windows, and the aim of multi-scale enhancement of the image is fulfilled.
In order to extract details of the detail window, smoothing the detail window to generate a smooth enhanced window, and smoothing the multiple auxiliary windows to generate multiple smooth auxiliary windows, wherein the smoothing is equivalent to filtering and erasing details in the detail window and the multiple auxiliary windows, so as to generate an image as a background image; in the process, some noise which is not obvious in the image can be kept in the background image after the filtering parameters are adjusted, so that the situation that the noise is enhanced does not occur in the subsequent operation.
In order to realize detail extraction, the method respectively extracts a smooth enhancement window and a plurality of smooth auxiliary windows from a detail window to form a plurality of detail auxiliary windows; the extraction here means that the two images are subtracted, and after the part of the detail window corresponding to the background of the auxiliary window is extracted, the remaining part is the detail part which can be enhanced. The number of detail auxiliary windows corresponds to the number of smooth auxiliary windows plus one.
And then weighting and superposing the smooth enhancement window and a plurality of detail auxiliary windows to be used as a detail enhanced window. The method is equivalent to the enhancement of the details on the smooth background based on the detail window, and the weight values of the auxiliary windows with different details can be changed according to the different requirements on the high-frequency details and the low-frequency details during the enhancement, so that the method has higher applicability. By setting the steps, the invention takes the details of the low-frequency image into account, takes the details of the high-frequency image into account, and can eliminate partial noise, so that the image detail enhancement result is more accurate.
Further, the center points of the plurality of auxiliary windows coincide with the center point of the detail window.
When the method is applied, in order to enable the detail enhancement to be more synchronous, the central points of the auxiliary windows and the central points of the detail windows are set to be coincident.
Further, the method also comprises the following steps: s6: sliding a detail window in the original image, and executing S1 until all points in the original image are used as the center point of the detail window; s7: and splicing the center points of all the detail-enhanced windows to form a detail-enhanced image.
When the method is applied, in order to realize the multi-layer enhancement of the details of the whole original image, the multi-layer detail enhancement of a detail window is executed firstly, then the detail window is slid by one pixel, the multi-layer detail enhancement is executed again, the process is circulated until the multi-layer detail enhancement of the whole original image is completed, and then the images are spliced to form the complete original image after the detail enhancement.
Further, the detail window is a square window of m × m, and m is an odd number; the auxiliary window is an n × n square window, and n is an odd number smaller than m.
When the method is applied, in order to better position the detail window and the auxiliary window, the detail window is set to be a square window of m multiplied by m, the auxiliary window is set to be a square window of n multiplied by n, and the center point of one square window is definitely determined because n and m are both odd numbers.
Further, the sizes of the plurality of auxiliary windows are sequentially increased.
When the method is applied, because the sizes of the plurality of auxiliary windows are sequentially increased, multi-level detail enhancement can be realized, and when the detail window is not large, even one auxiliary window can be arranged in each step to enhance the complete detail from low frequency to high frequency.
Further, in the step S3, the smoothing processing on the detail window adopts filtering on the gray value of the detail window; and the smoothing treatment on the plurality of auxiliary windows adopts the filtering of the gray values of the auxiliary windows.
When the method is applied, the background image can be well obtained by filtering the gray value of the detail window and filtering the gray value of the auxiliary window.
Further, the filtering adopts Gaussian filtering, bilateral filtering, guided filtering and/or low-pass filtering.
Further, step S5 includes the following substeps: the weight values of the detail auxiliary windows are the same, and the weight values of the detail auxiliary windows are larger than the weight value of the smooth enhancement window.
When the method is applied, in order to simplify the actual computation amount of multi-layer detail enhancement, the weight values of a plurality of detail auxiliary windows are set to be the same, so that the computation amount can be reduced by more than 40% on the basis of ensuring the multi-layer detail enhancement.
Further, the weight value of the smooth enhancement window is 1, and the weight values of the detail auxiliary windows are 2-5.
Compared with the prior art, the invention has the following advantages and beneficial effects:
according to the image multilayer detail enhancement method, through the steps, not only the low-frequency image detail is considered, but also the high-frequency image detail is considered, and meanwhile, partial noise can be eliminated, so that the image detail enhancement result is more accurate.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to examples, and the exemplary embodiments and descriptions thereof are only used for explaining the present invention and are not used as limitations of the present invention.
Example 1
The invention discloses an image multilayer detail enhancement method, which comprises the following steps: s1: selecting a window from an original image as a detail window; s2: selecting a plurality of auxiliary windows from detail windows, wherein the sizes of the auxiliary windows are smaller than those of the detail windows; s3: carrying out smoothing processing on the detail window to generate a smooth enhancement window; carrying out smoothing processing on the plurality of auxiliary windows to generate a plurality of smooth auxiliary windows; s4: respectively extracting the smooth enhancement window and the plurality of smooth auxiliary windows from the detail window to form a plurality of detail auxiliary windows; s5: and weighting and superposing the smooth enhanced window and the plurality of detail auxiliary windows to obtain a detail enhanced window.
In the implementation of this embodiment, the design core of the present invention is to represent the original image as the basic component and the detail component, and enhance the detail component on the basis of the basic component and the detail component to obtain the enhanced image. The invention selects a window in an original image as a detail window, wherein the detail window is a window for detail enhancement, the detail window is spliced after detail enhancement to obtain a result after the detail enhancement of the original image, then a plurality of auxiliary windows are selected from the detail window, the sizes of the auxiliary windows are all smaller than the detail window, the plurality of auxiliary windows are used for extracting details, generally all the auxiliary windows are different in size, so that the auxiliary windows with various sizes can be covered, and the inventor finds that when the auxiliary window is smaller, the high-frequency fine image details, such as image texture details, can be well extracted, but the auxiliary window is not beneficial to extracting the low-frequency large image details, such as image edges; when the auxiliary window is large, large image details with low frequency can be well extracted, but the extraction of the image details with high frequency and small details is not facilitated; therefore, the adoption of the plurality of auxiliary windows can realize the extraction of the image characteristics of different frequencies in the image through different auxiliary windows, and the aim of multi-scale enhancement of the image is fulfilled.
In order to extract details of the detail window, smoothing the detail window to generate a smooth enhanced window, and smoothing the multiple auxiliary windows to generate multiple smooth auxiliary windows, wherein the smoothing is equivalent to filtering and erasing the details in the detail window and the multiple auxiliary windows to generate an image as a background image; in the process, some noise which is not obvious in the image can be kept in the background image after the filtering parameters are adjusted, so that the situation that the noise is enhanced does not occur in the subsequent operation.
In order to realize detail extraction, the method respectively extracts a smooth enhancement window and a plurality of smooth auxiliary windows from a detail window to form a plurality of detail auxiliary windows; the extraction here means that after the two images are subtracted and the part of the detail window corresponding to the background of the auxiliary window is extracted, the remaining detail part is the detail part which can be enhanced. And then weighting and superposing the smooth enhancement window and a plurality of detail auxiliary windows to be used as a detail enhanced window. The method is equivalent to the enhancement of the details on the smooth background based on the detail window, and the weight values of the auxiliary windows with different details can be changed according to the different requirements on the high-frequency details and the low-frequency details during the enhancement, so that the method has higher applicability. By setting the steps, the invention considers the details of the low-frequency image and the high-frequency image, and can eliminate partial noise, so that the image detail enhancement result is more accurate.
Example 2
In this embodiment, based on embodiment 1, the center points of the plurality of auxiliary windows coincide with the center point of the detail window.
In this embodiment, in order to enhance the detail more synchronously, the center points of the plurality of auxiliary windows and the center point of the detail window are set to coincide with each other.
Example 3
The embodiment further includes the following steps based on the embodiment 2: s6: sliding a detail window in the original image, and executing S1 until all points in the original image are used as the center point of the detail window; s7: and splicing all the detail-enhanced windows to form a detail-enhanced image.
In this embodiment, in order to implement multi-layer enhancement on details of the entire original image, multi-layer detail enhancement of a detail window is performed first, then the detail window is slid by one pixel, and multi-layer detail enhancement is performed again, and this is repeated until the entire original image is subjected to multi-layer detail enhancement, and then the images are stitched to form the entire detail-enhanced original image.
Example 4
In this embodiment, on the basis of embodiment 2, the detail window is a square window of m × m, and m is an odd number; the auxiliary window is an n × n square window, and n is an odd number smaller than m.
In this embodiment, in order to better position the detail window and the auxiliary window, the detail window is set to be a square window of m × m, and the auxiliary window is set to be a square window of n × n.
Example 5
In this embodiment, the sizes of the plurality of auxiliary windows are sequentially increased based on embodiment 1.
When the embodiment is implemented, because the sizes of the plurality of auxiliary windows are sequentially increased, multi-level detail enhancement can be realized, and when the detail window is not large, even one auxiliary window can be arranged in each step to enhance the complete detail from low frequency to high frequency.
Example 6
In this embodiment, on the basis of embodiment 1, the step S3 of smoothing the detail window employs filtering the gray value of the detail window; and the smoothing treatment on the plurality of auxiliary windows adopts the filtering of the gray values of the auxiliary windows.
In the implementation of this embodiment, the background image can be well obtained by filtering the gray value of the detail window and filtering the gray value of the auxiliary window.
Example 7
In this embodiment, on the basis of embodiment 1, step S5 includes the following sub-steps: the weight values of the detail auxiliary windows are the same, and the weight values of the detail auxiliary windows are larger than the weight value of the smooth enhancement window.
In this embodiment, in order to simplify the actual computation amount of the multi-layer detail enhancement, the weight values of the multiple detail auxiliary windows are set to be the same, so that the computation amount can be reduced by more than 40% on the basis of ensuring the multi-layer detail enhancement.
Example 8
In this embodiment, on the basis of embodiments 1 to 7, the original image is an image of q × p, and the detail window image _ i is an 11 × 11 window, where S11 (x, y) is defined, and x =0,1,3,4.. Q; y =0,1,2,3.. P, x, y are integers, x denotes row coordinates, y denotes column coordinates, and S11 (x, y) denotes an 11 × 11 field centered around (x, y), so S11 (x, y) denotes a detail window.
In 11 × 11 detail windows, subdividing 3 × 3,5 × 5,7 × 7,9 × 9 four auxiliary windows, wherein the four auxiliary windows and the detail window are the same central point, and each window can represent details with different features in different frequency bands;
and then carrying out Gaussian filtering on the image gray levels of the four auxiliary windows and the detail window:
do _3 is a 3 x 3 auxiliary window filtered smoothed auxiliary window;
do _5 is the smoothed auxiliary window after 5 by 5 auxiliary window filtering;
do _7 is the smoothed auxiliary window after 7 x 7 auxiliary window filtering;
do _9 is a 9 x 9 auxiliary window filtered smoothed auxiliary window;
do _11 is the detail window filtered smooth enhancement window;
and then performing detail extraction:
Detail3＝image_i-Do_3；
Detail5＝image_i-Do_5；
Detail7＝image_i-Do_7；
Detail9＝image_i-Do_9；
Detail11＝image_i-Do_11；
detail auxiliary windows are Detail auxiliary windows of Detail3, detail5, detail7, detail9 and Detail 11;
then, detail enhancement is carried out:
Dout_enhance＝Do_11+gain*(Detail11+Detail9+Detail7+Detail5+Detail3)；
in the above formula, dout _ enhance is a window after detail enhancement, gain is a gain coefficient, and is also equivalent to a weight value of a detail auxiliary window, where the weight value of the smooth enhancement window is 1, and gain is 2.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (9)
1. An image multi-layer detail enhancement method is characterized by comprising the following steps:
s1: selecting a window from an original image as a detail window;
s2: selecting a plurality of auxiliary windows from detail windows, wherein the sizes of the auxiliary windows are smaller than those of the detail windows;
s3: carrying out smoothing processing on the detail window to generate a smooth enhancement window; carrying out smoothing processing on the plurality of auxiliary windows to generate a plurality of smooth auxiliary windows;
s4: respectively extracting the smooth enhancement window and the plurality of smooth auxiliary windows from the detail window to form a plurality of detail auxiliary windows;
s5: and weighting and superposing the smooth enhanced window and the plurality of detail auxiliary windows to obtain a detail enhanced window.
2. The method of claim 1, wherein the center points of the plurality of auxiliary windows coincide with the center point of the detail window.
3. The image multi-layer detail enhancement method according to claim 2, further comprising the steps of:
s6: sliding a detail window in the original image, and executing S1 until all points in the original image are used as the center point of the detail window;
s7: and splicing the central points of all the detail-enhanced windows to form a detail-enhanced image.
4. The image multi-layer detail enhancement method according to claim 2, wherein the detail window is a m x m square window, and m is an odd number; the auxiliary window is a square window of n × n, and n is an odd number smaller than m.
5. The method of claim 1, wherein the size of the plurality of auxiliary windows increases sequentially.
6. The method for enhancing image multilayer details according to claim 1, wherein the step S3 of smoothing the detail window employs filtering the gray value of the detail window; and the smoothing treatment on the plurality of auxiliary windows adopts the filtering of the gray values of the auxiliary windows.
7. The image multi-layer detail enhancement method according to claim 6, characterized in that said filtering employs Gaussian filtering, bilateral filtering, guided filtering and/or low-pass filtering.
8. The image multi-layer detail enhancement method according to claim 1, wherein the step S5 comprises the following sub-steps:
the weight values of the detail auxiliary windows are the same, and the weight values of the detail auxiliary windows are larger than the weight value of the smooth enhancement window.
9. The method as claimed in claim 8, wherein the weight value of the smooth enhancement window is 1, and the weight values of the detail auxiliary windows are 2-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910127062.1A CN109919865B (en) | 2019-02-20 | 2019-02-20 | Image multi-layer detail enhancement method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910127062.1A CN109919865B (en) | 2019-02-20 | 2019-02-20 | Image multi-layer detail enhancement method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109919865A CN109919865A (en) | 2019-06-21 |
| CN109919865B true CN109919865B (en) | 2022-10-11 |
Family
ID=66961838
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910127062.1A Active CN109919865B (en) | 2019-02-20 | 2019-02-20 | Image multi-layer detail enhancement method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109919865B (en) |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070223834A1 (en) * | 2006-03-23 | 2007-09-27 | Samsung Electronics Co., Ltd. | Method for small detail restoration in digital images |
| CN105957030B (en) * | 2016-04-26 | 2019-03-22 | 成都市晶林科技有限公司 | One kind being applied to the enhancing of thermal infrared imager image detail and noise suppressing method |
| CN107274365A (en) * | 2017-06-15 | 2017-10-20 | 中国矿业大学(北京) | A kind of mine image intensification method based on unsharp masking and NSCT algorithms |
-
2019
- 2019-02-20 CN CN201910127062.1A patent/CN109919865B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109919865A (en) | 2019-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Srinivas et al. | Low light image enhancement with adaptive sigmoid transfer function | |
| CN107358586A (en) | A kind of image enchancing method, device and equipment | |
| CN109493295B (en) | Non-local haar transform image denoising method | |
| CN109389560A (en) | A kind of adaptive weighted filter image denoising method, device and image processing equipment | |
| CN106920222A (en) | A kind of image smoothing method and device | |
| Kaur et al. | Performance evaluation of fuzzy and histogram based color image enhancement | |
| CN104616259B (en) | A kind of adaptive non-local mean image de-noising method of noise intensity | |
| Majeed et al. | Adaptive entropy index histogram equalization for poor contrast images | |
| Abdullah et al. | Deep CNN based skin lesion image denoising and segmentation using active contour method | |
| CN108492268A (en) | Enhancement algorithm for low-illumination image based on wavelet coefficient fusion | |
| Abdullah-Al-Wadud | A modified histogram equalization for contrast enhancement preserving the small parts in images | |
| CN106599891A (en) | Remote sensing image region-of-interest rapid extraction method based on scale phase spectrum saliency | |
| Srivastava et al. | Histogram equalization: A comparative analysis & a segmented approach to process digital images | |
| CN201726464U (en) | Novel video image sharpening processing device | |
| Das et al. | Histogram equalization techniques for contrast enhancement: a review | |
| CN109919865B (en) | Image multi-layer detail enhancement method | |
| CN110545414B (en) | Image sharpening method | |
| Cho et al. | Enhancement technique of image contrast using new histogram transformation | |
| CN111784610A (en) | Clustering-based side window filter optimization method | |
| CN108629740A (en) | A kind of processing method and processing device of image denoising | |
| Septiana et al. | X-ray image enhancement using a modified anisotropic diffusion | |
| CN116342504A (en) | Image processing method and device, electronic equipment and readable storage medium | |
| Chen et al. | An adaptive image dehazing algorithm based on dark channel prior | |
| CN113487505A (en) | Infrared image mixed noise reduction method based on noise identification | |
| Wang et al. | Adaptive switching anisotropic diffusion model for universal noise removal |
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 |