CN110211058A - A kind of data enhancement methods of medical image - Google Patents
A kind of data enhancement methods of medical image Download PDFInfo
- Publication number
- CN110211058A CN110211058A CN201910404304.7A CN201910404304A CN110211058A CN 110211058 A CN110211058 A CN 110211058A CN 201910404304 A CN201910404304 A CN 201910404304A CN 110211058 A CN110211058 A CN 110211058A
- Authority
- CN
- China
- Prior art keywords
- image
- block
- sub
- function
- pixel
- 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
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000005315 distribution function Methods 0.000 claims abstract description 19
- 230000000873 masking effect Effects 0.000 claims abstract description 11
- 238000001914 filtration Methods 0.000 claims abstract description 9
- 238000007781 pre-processing Methods 0.000 claims abstract description 6
- 230000009467 reduction Effects 0.000 claims abstract description 6
- 230000000052 comparative effect Effects 0.000 claims abstract description 4
- 238000002474 experimental method Methods 0.000 claims abstract description 4
- 230000008092 positive effect Effects 0.000 claims abstract description 4
- 230000006870 function Effects 0.000 claims description 19
- 230000002708 enhancing effect Effects 0.000 claims description 11
- 238000012886 linear function Methods 0.000 claims description 8
- 230000009466 transformation Effects 0.000 claims description 6
- 230000001131 transforming effect Effects 0.000 claims description 6
- 230000001186 cumulative effect Effects 0.000 claims description 4
- 238000009825 accumulation Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 claims description 3
- 238000003745 diagnosis Methods 0.000 description 5
- 238000010191 image analysis Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 239000004744 fabric Substances 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000003930 cognitive ability Effects 0.000 description 1
- 238000004195 computer-aided diagnosis Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
Classifications
-
- G06T5/70—
-
- G06T5/73—
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/40—ICT specially adapted for the handling or processing of medical images for processing medical images, e.g. editing
Abstract
The present invention discloses a kind of data enhancement methods of medical image, specifically includes the following steps: step 1: carrying out preprocessing noise reduction by guiding filtering algorithm to image；Step 2: piecemeal being carried out to image, the frequency distribution function of the sub-block is sought in each sub-block, frequency histogram is cut by the threshold value of predefined, carries out picture superposition；Step 3: by the unsharp masking algorithm with noise suppressed, image being divided into high, medium and low details area according to the number of detailed information to be sharpened processing to original image；Step 4: by comparative experiments, the parameter that image visually has positive effect being set, image clearly degree is effectively improved.Application image Enhancement Method of the present invention handles the digital medical images such as X-ray, preprocessing noise reduction is carried out by guiding filtering algorithm, further frequency histogram is cut using the threshold value of predefined, carries out picture superposition, recycles the unsharp masking algorithm with noise suppressed.
Description
Technical field
The invention belongs to field of medical image processing, the data enhancement methods of specifically a kind of medical image.
Background technique
In modern medical service diagnosis, computer-aided diagnosis is widely used, the medical image in computer medical diagnosis
It is capable of providing a large amount of medical informations, it has also become the important references of doctor.Medical image analysis refers at comprehensive utilization digitized video
The technology of the computerized informations subject such as reason, artificial intelligence analyzes medical image, assists diagnosis.In medical image analysis,
The unstable factors such as doctor's subjective experience and cognitive ability can be effectively reduced using computer recognizer, while can be added
Fast diagnosis efficiency.Develop and calculate based on the hardware bring such as GPU the promotion of power with the brings technology such as cloud computing platform,
Medical image analysis has stepped into big data era.Go out useful to know using technology minings such as deep learnings from medical image data
Know, has become the research hotspot of academia and industry.
In medical image, since X-ray projection imaging technique has fast, at low cost, high reliablity the good characteristics of speed,
It is in medical field using increasingly extensive.But since inside of human body tissue is intricate and X-ray has scattering properties etc.
The influence of factor, that often there is noise levels is high for collected X-ray digital medical image, contrast is lower, image more mould
The problem of paste, causes undesirable influence to medical image analysis and diagnosis so that medical image recognition rate is lower, so, have
Necessity targetedly handles the digital medical images such as X-ray.In current practical application, to X-ray digital medical figure
The method that the quality of picture is promoted is relatively simple, using some traditional image enchancing methods, the benefit of these methods
It is to realize simply, is readily integrated on Medical Devices, but actual effect is not ideal enough.
Summary of the invention
The purpose of this hair is pre-processed for medical image, and noise is removed, and passes through parameter adjustment and image procossing
Algorithm achievees the purpose that reduce picture noise level, promotes picture contrast, improving image definition, provides a kind of medicine shadow
The data enhancement methods of picture.
The present invention is achieved through the following technical solutions: a kind of data enhancement methods of medical image specifically include following
Step:
Step 1: preprocessing noise reduction is carried out by guiding filtering algorithm to image；
Step 2: piecemeal being carried out to image, the frequency distribution function of the sub-block is sought in each sub-block, by prior
The threshold value of definition cuts frequency histogram, carries out picture superposition；
Step 3: by the unsharp masking algorithm with noise suppressed, by image according to detailed information number be divided into it is high,
In, low details area processing is sharpened to original image；
Step 4: by comparative experiments, the parameter that image visually has positive effect being set, image clearly is effectively improved
Degree.
In a further embodiment, the step 1 specifically includes the following steps:
Step 1-1: being related to three images, input picture p, navigational figure I and output image q in guiding filtering,
Middle navigational figure has similar edge and detailed information with input picture；For piece image, it is considered as a two-dimentional letter
Number, it is assumed that in a two-dimentional window, outputting and inputting for the two-dimensional function is in a linear relationship:
Wherein, Ii is the value of navigational figure pixel, and qi is the value of output pixel, and k is the rectangular two dimension that a radius is r
Window, ak and bk are the coefficients of the linear function when the center of two-dimentional window is located at position K；
Step 1-2: gradient is sought on both sides, is obtained:
It can be seen that output image has similar gradient with navigational figure；
Step 1-3: in order to find out the two coefficients of ak and bk, the difference between output image q and input picture p is minimized
To solve:
Wherein, μkNavigational figure I is represented in two-dimentional window ωkIn mean value.pkRepresenting input images are in two-dimentional window ωk
In mean value.Navigational figure is represented in window ωkOn variance.
Step 1-4: for each pixel, if window ωkSize be AxB, then this pixel can be by AxB
Window is included, by partial model described above it is found that the value of each pixel is described jointly by multiple linear functions
, calculate the average value of all linear functions comprising the point:
In a further embodiment, the step 2 specifically includes the following steps:
Step 2-1: 16x16=256 8x8=64 block is divided the image into according to the size of input picture, in every height
The frequency distribution function of the sub-block is all sought in block；
Step 2-2: the slope by improving frequency distribution function, then contrast is just promoted: will be made when slope is excessively high
In contrast with degree of spending enhancing consequence, continue through the slope of limitation transforming function transformation function to limiting contrast；
Step 2-3: since transforming function transformation function is actually the accumulation of histogram frequency distribution diagram, so the threshold for passing through predefined
Value cuts frequency histogram；
Step 2-4: after carrying out threshold value cutting processing to frequency distribution function, accumulative point is asked to new frequency distribution function
Cloth function recycles cumulative distribution function to obtain the new straight of each sub-block in conjunction with traditional histogram equalization operation later
Fang Tu.
In a further embodiment, the step 3 specifically includes the following steps:
Step 3-1: the number of detailed information is measured by local variance in image region, it is assumed that each picture
Element is calculated using the variance in 5x5 neighborhood, then is had:
Wherein x (i, j) is the pixel in image, and x (m, n) is all pixels gray scale in pixel x (i, j) surrounding 5x5 neighborhood
The mean value of value.D (m, n) is variance, and which represent the numbers of detailed information at pixel (m, n).
Step 3-2: if setting two threshold value T1 and T2 (T1 < T2), so that it may to pass through T1 and T2 and D (m, n) for image
Be divided into high details area, medium detail region, low details area: if D (m, n) < T1, variance is small herein, relatively flat, details
Information is few；If T1 < D (m, n) < T2, variance is in median size herein, and details is relatively abundanter；If D (m, n) > T2, herein
Variance is big, and details is very rich；
Step 3-3: it for low details area, is directly denoised using mean filter；For high details area, in order to
Prevent overshoot phenomenon from leading to the artificial trace for occurring unnatural, operation algorithm carries out moderate enhancing；And for medium thin
Region is saved, then carries out enhancing by a relatively large margin.
In a further embodiment, the step 4 specifically includes the following steps:
Step 4-1: in traditional histogram equalization, the relationship of image and input picture is exported are as follows:
X ' (m, n)=x (m, n)+λ z (m, n)
And in the unsharp masking method with noise suppressed, definition:
Step 4-2: different values is had according to the abundant degree of pixel region detailed information:
Wherein, λL,λM,λHIt is successively the low details area of image, the enhancement factor in medium detail region and high details area.
In this way, can be kept away while becoming apparent from image by the unsharp masking algorithm with noise suppressed
Exempt from the excessive enhancing to noise, improves the quality of image to the full extent.
Beneficial effects of the present invention: the present invention proposes a kind of medical image preprocess method of multistage step by step, the party
Method cures the number such as X-ray by comparing result parameter of the different methods on digital medical image, application image Enhancement Method
It learns image to be handled, preprocessing noise reduction is carried out by guiding filtering algorithm, further using the threshold value of predefined come to frequency
Rate histogram is cut, and picture superposition is carried out, and recycles the unsharp masking algorithm with noise suppressed, image is pressed
Processing is sharpened to original image according to how much detailed information is divided into high, medium and low details area.
Detailed description of the invention
Fig. 1 is operational flowchart of the invention.
Fig. 2 is the effect contrast figure of image enhancement of the invention.
Specific embodiment
In the following description, a large amount of concrete details are given so as to provide a more thorough understanding of the present invention.So
And it is obvious to the skilled person that the present invention may not need one or more of these details and be able to
Implement.In other examples, in order to avoid confusion with the present invention, for some technical characteristics well known in the art not into
Row description.
As shown in Figure 1, the present invention provides a kind of data enhancement methods of medical image, comprising the following steps:
Step 1, preprocessing noise reduction is carried out by guiding filtering algorithm to image
Step 2, piecemeal is carried out to image, the frequency distribution function of the sub-block is sought in each sub-block, by prior
The threshold value of definition cuts frequency histogram, carries out picture superposition
Step 3, by the unsharp masking algorithm with noise suppressed, by image according to detailed information how much be divided into it is high, in,
Low details area to original image is sharpened processing.
Step 4, by comparative experiments, the parameter that image visually has positive effect is set, image clearly is effectively improved
Degree.
In order to be better understood by this method, in which:
The step 1 the following steps are included:
Step 1-1, is related to three images in guiding filtering, input picture (image to be filtered) p, navigational figure I and
Export image q.Wherein navigational figure should have similar edge and detailed information with input picture.It, can for piece image
To regard a two-dimensional function as, it is assumed that in a two-dimentional window, outputting and inputting for the two-dimensional function is in a linear relationship,
Wherein, Ii is the value of navigational figure pixel, and qi is the value of output pixel, and k is the rectangular two dimension that a radius is r
Window, ak and bk are the coefficients of the linear function when the center of two-dimentional window is located at position K.
Step 1-2: gradient is sought on both sides, is obtained:
It can be seen that output image has similar gradient with navigational figure.
Step 1-3: it in order to find out the two coefficients of ak and bk, can be minimized between output image q and input picture p
Difference solves.
Wherein, μkNavigational figure I is represented in two-dimentional window ωkIn mean value.pkRepresenting input images are in two-dimentional window ωk
In mean value.Navigational figure is represented in window ωkOn variance.
Step 1-4: for each pixel, if window ωkSize be AxB, then this pixel can be by AxB
Window included.By partial model described above it is found that the value of each pixel is described jointly by multiple linear functions
, it is only necessary to calculate the average value of all linear functions comprising the point:
The step 2 the following steps are included:
Step 2-1: 16x16=256 8x8=64 block is divided the image into according to the size of input picture, in every height
The frequency distribution function of the sub-block is all sought in block.
Step 2-2: the slope by improving frequency distribution function, then contrast is just promoted.It will be made when slope is excessively high
In contrast with degree of spending enhancing consequence, continue through the slope of limitation transforming function transformation function to limiting contrast.
Step 2-3: since transforming function transformation function is actually the accumulation of histogram frequency distribution diagram, so the threshold for passing through predefined
Value cuts frequency histogram.
Step 2-4: after carrying out threshold value cutting processing to frequency distribution function, accumulative point is asked to new frequency distribution function
Cloth function recycles cumulative distribution function to obtain the new straight of each sub-block in conjunction with traditional histogram equalization operation later
Fang Tu.
Step 3 specifically includes the following steps:
Step 3-1: the number of detailed information is measured by local variance in image region, it is assumed that each picture
Element is calculated using the variance in 5x5 neighborhood, then is had:
Wherein x (I, j) is the pixel in image, and x (m, n) is all pixels gray scale in pixel x (I, j) surrounding 5x5 neighborhood
The mean value of value.D (m, n) is variance, and which represent the numbers of detailed information at pixel (m, n).
Step 3-2: if setting two threshold value T1 and T2 (T1 < T2), so that it may to pass through T1 and T2 and D (m, n) for image
It is divided into high details area, medium detail region, low details area.If D (m, n) < T1, variance is small herein, relatively flat, details
Information is few；If T1 < D (m, n) < T2, variance is in median size herein, and details is relatively abundanter；If D (m, n) > T2, herein
Variance is big, and details is very rich.
Step 3-3: it for low details area, is directly denoised using mean filter.For high details area, in order to
Prevent overshoot phenomenon from leading to the artificial trace for occurring unnatural, operation algorithm carries out moderate enhancing.And for medium thin
Region is saved, then carries out enhancing by a relatively large margin.
Step 4 the following steps are included:
Step 4-1: in traditional histogram equalization, the relationship of image and input picture is exported are as follows:
X ' (m, n)=x (m, n)+λ z (m, n)
And in the unsharp masking method with noise suppressed, definition:
Step 4-2: different values is had according to the abundant degree of pixel region detailed information:
Wherein, λL,λM,λHIt is successively the low details area of image, the enhancement factor in medium detail region and high details area.
In this way, can be kept away while becoming apparent from image by the unsharp masking algorithm with noise suppressed
Exempt from the excessive enhancing to noise, improves the quality of image to the full extent.This patent is directed to how to improve X-ray digital medical figure
Image quality amount is studied, and so that each algorithm is intercoupled by parameter regulation, forms organic digital medical image processing
Process.By this patent method treated medical image in terms of all have greatly improved.
As shown in Fig. 2, the figure in (a) (b) (c) (d) (e) in Fig. 2 is respectively original image, DHE, GHE, ESIHE and we
Method treated figure, it is evident that this method processing rear figure be high-visible, clarity and texture to its collected information
There are apparent promotion, significant effect.
It is further to note that specific technical features described in the above specific embodiments, in not lance
In the case where shield, it can be combined in any appropriate way.In order to avoid unnecessary repetition, the present invention to it is various can
No further explanation will be given for the combination of energy.
Claims (5)
1. a kind of data enhancement methods of medical image, which is characterized in that specifically includes the following steps:
Step 1: preprocessing noise reduction is carried out by guiding filtering algorithm to image；
Step 2: piecemeal being carried out to image, the frequency distribution function of the sub-block is sought in each sub-block, passes through predefined
Threshold value frequency histogram is cut, carry out picture superposition；
Step 3: by the unsharp masking algorithm with noise suppressed, image being divided into according to the number of detailed information high, medium and low
Details area to original image is sharpened processing；
Step 4: by comparative experiments, the parameter that image visually has positive effect being set, image clearly degree is effectively improved.
2. a kind of data enhancement methods of medical image according to claim 1, which is characterized in that the step 1 is specific
The following steps are included:
Step 1-1: it is related to three images, input picture p, navigational figure I and output image q in guiding filtering, wherein drawing
Leading image has similar edge and detailed information with input picture；For piece image, it is considered as a two-dimensional function, it is false
It is located in a two-dimentional window, outputting and inputting for the two-dimensional function is in a linear relationship:
Step 1-2: gradient is sought on both sides, is obtained:
▽ q=a ▽ I；
Step 1-3: in order to find out the two coefficients of ak and bk, the difference between output image q and input picture p is minimized to ask
Solution:
Step 1-4: for each pixel, if window ωkSize be AxB, then this pixel can be by AxB window
Included, by partial model described above it is found that the value of each pixel is described jointly by multiple linear functions, meter
Calculate the average value of all linear functions comprising the point:
3. a kind of data enhancement methods of medical image according to claim 1, which is characterized in that the step 2 is specific
The following steps are included:
Step 2-1: 16x16=256 8x8=64 block is divided the image into according to the size of input picture, in each sub-block
All seek the frequency distribution function of the sub-block；
Step 2-2: the slope by improving frequency distribution function, then contrast is just promoted: will result in when slope is excessively high pair
Than the consequence of degree of spending enhancing, the slope of limitation transforming function transformation function is continued through to limit contrast；
Step 2-3: since transforming function transformation function is actually the accumulation of histogram frequency distribution diagram, so by the threshold value of predefined come
Frequency histogram is cut；
Step 2-4: after carrying out threshold value cutting processing to frequency distribution function, cumulative distribution letter is asked to new frequency distribution function
Number recycles cumulative distribution function to obtain the new histogram of each sub-block in conjunction with traditional histogram equalization operation later
Figure.
4. a kind of data enhancement methods of medical image according to claim 1, which is characterized in that the step 3 is specific
The following steps are included:
Step 3-1: the number of detailed information is measured by local variance in image region, it is assumed that is adopted to each pixel
It is calculated, is then had with the variance in 5x5 neighborhood:
Step 3-2: if setting two threshold value T1 and T2 (T1 < T2), so that it may to be divided the image by T1 and T2 and D (m, n)
High details area, medium detail region, low details area: if D (m, n) < T1, variance is small herein, relatively flat, detailed information
It is few；If T1 < D (m, n) < T2, variance is in median size herein, and details is relatively abundanter；If D (m, n) > T2, variance herein
Greatly, details is very rich；
Step 3-3: it for low details area, is directly denoised using mean filter；For high details area, in order to prevent
Overshoot phenomenon leads to the artificial trace for occurring unnatural, and operation algorithm carries out moderate enhancing；And for medium detail area
Domain then carries out enhancing by a relatively large margin.
5. a kind of data enhancement methods of medical image according to claim 1, which is characterized in that the step 4 is specific
The following steps are included:
Step 4-1: in traditional histogram equalization, the relationship of image and input picture is exported are as follows:
X ' (m, n)=x (m, n)+λ z (m, n)
And in the unsharp masking method with noise suppressed, definition:
Step 4-2: different values is had according to the abundant degree of pixel region detailed information:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910404304.7A CN110211058A (en) | 2019-05-15 | 2019-05-15 | A kind of data enhancement methods of medical image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910404304.7A CN110211058A (en) | 2019-05-15 | 2019-05-15 | A kind of data enhancement methods of medical image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110211058A true CN110211058A (en) | 2019-09-06 |
Family
ID=67787352
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910404304.7A Pending CN110211058A (en) | 2019-05-15 | 2019-05-15 | A kind of data enhancement methods of medical image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110211058A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110766639A (en) * | 2019-10-30 | 2020-02-07 | 北京迈格威科技有限公司 | Image enhancement method and device, mobile equipment and computer readable storage medium |
| CN111407430A (en) * | 2020-03-27 | 2020-07-14 | 中国人民解放军东部战区总医院 | Anti-infection dilator |
| CN111738943A (en) * | 2020-06-12 | 2020-10-02 | 吉林大学 | Medical image enhancement method combining spatial domain and frequency domain |
| CN112150437A (en) * | 2020-09-23 | 2020-12-29 | 南昌航空大学 | DR (digital radiography) detection image processing method for crack defects of diffuser in laser additive manufacturing |
| CN112801167A (en) * | 2021-01-25 | 2021-05-14 | 河北北方学院 | Analysis method of medical image big data |
| CN113112438A (en) * | 2021-04-29 | 2021-07-13 | 中国船舶重工集团公司第七二三研究所 | Image enhancement method based on cutting histogram |
| CN113764073A (en) * | 2021-09-02 | 2021-12-07 | 宁波权智科技有限公司 | Medical image analysis method and device |
| CN114241022A (en) * | 2022-02-28 | 2022-03-25 | 北京艾尔思时代科技有限公司 | Unmanned aerial vehicle image automatic registration method and system |
| CN116309187A (en) * | 2023-05-15 | 2023-06-23 | 济南科汛智能科技有限公司 | Intelligent enhancement method for medical image of children pneumonia |
| CN113724145B (en) * | 2021-07-13 | 2024-04-26 | 西姆高新技术（江苏）有限公司 | X-ray image processing method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103034979A (en) * | 2012-11-30 | 2013-04-10 | 声泰特（成都）科技有限公司 | Ultrasonic image definition improving method |
| CN106897972A (en) * | 2016-12-28 | 2017-06-27 | 南京第五十五所技术开发有限公司 | A kind of self-adapting histogram underwater picture Enhancement Method of white balance and dark primary |
| 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 |
| CN108062746A (en) * | 2016-11-09 | 2018-05-22 | 深圳市优朋普乐传媒发展有限公司 | A kind of method of video image processing and device, video coding system |
-
2019
- 2019-05-15 CN CN201910404304.7A patent/CN110211058A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103034979A (en) * | 2012-11-30 | 2013-04-10 | 声泰特（成都）科技有限公司 | Ultrasonic image definition improving method |
| CN108062746A (en) * | 2016-11-09 | 2018-05-22 | 深圳市优朋普乐传媒发展有限公司 | A kind of method of video image processing and device, video coding system |
| CN106897972A (en) * | 2016-12-28 | 2017-06-27 | 南京第五十五所技术开发有限公司 | A kind of self-adapting histogram underwater picture Enhancement Method of white balance and dark primary |
| 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 |
Non-Patent Citations (3)
| Title |
|---|
| AIPIANO: "《双边滤波与引导滤波》", 7 March 2014 * |
| LUO1873626242: "《对比度受限的自适应直方图均衡化(CLAHE)》", 2 November 2015 * |
| 谢明果等: "带噪声抑制的反锐化掩模图像增强算法", 《现代电子技术》 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110766639B (en) * | 2019-10-30 | 2022-09-27 | 北京迈格威科技有限公司 | Image enhancement method and device, mobile equipment and computer readable storage medium |
| CN110766639A (en) * | 2019-10-30 | 2020-02-07 | 北京迈格威科技有限公司 | Image enhancement method and device, mobile equipment and computer readable storage medium |
| CN111407430A (en) * | 2020-03-27 | 2020-07-14 | 中国人民解放军东部战区总医院 | Anti-infection dilator |
| CN111738943A (en) * | 2020-06-12 | 2020-10-02 | 吉林大学 | Medical image enhancement method combining spatial domain and frequency domain |
| CN111738943B (en) * | 2020-06-12 | 2023-12-05 | 吉林大学 | Medical image enhancement method combining spatial domain and frequency domain |
| CN112150437A (en) * | 2020-09-23 | 2020-12-29 | 南昌航空大学 | DR (digital radiography) detection image processing method for crack defects of diffuser in laser additive manufacturing |
| CN112150437B (en) * | 2020-09-23 | 2023-05-09 | 南昌航空大学 | Laser additive manufacturing diffuser crack defect DR detection image processing method |
| CN112801167A (en) * | 2021-01-25 | 2021-05-14 | 河北北方学院 | Analysis method of medical image big data |
| CN113112438A (en) * | 2021-04-29 | 2021-07-13 | 中国船舶重工集团公司第七二三研究所 | Image enhancement method based on cutting histogram |
| CN113724145B (en) * | 2021-07-13 | 2024-04-26 | 西姆高新技术（江苏）有限公司 | X-ray image processing method and device |
| CN113764073A (en) * | 2021-09-02 | 2021-12-07 | 宁波权智科技有限公司 | Medical image analysis method and device |
| CN114241022A (en) * | 2022-02-28 | 2022-03-25 | 北京艾尔思时代科技有限公司 | Unmanned aerial vehicle image automatic registration method and system |
| CN114241022B (en) * | 2022-02-28 | 2022-06-03 | 北京艾尔思时代科技有限公司 | Unmanned aerial vehicle image automatic registration method and system |
| CN116309187A (en) * | 2023-05-15 | 2023-06-23 | 济南科汛智能科技有限公司 | Intelligent enhancement method for medical image of children pneumonia |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110211058A (en) | A kind of data enhancement methods of medical image | |
| Zhou et al. | Retinex-based laplacian pyramid method for image defogging | |
| CN102800063B (en) | Image enhancement and abstraction method based on anisotropic filtering | |
| CN104463804B (en) | Image enhancement method based on intuitional fuzzy set | |
| CN106530237A (en) | Image enhancement method | |
| CN108921800A (en) | Non-local mean denoising method based on form adaptive search window | |
| US20200364500A1 (en) | Training image enhancement model and enhancing image | |
| Yan et al. | Method to Enhance Degraded Image in Dust Environment. | |
| Mahmoud et al. | Segmentation of skin cancer images based on gradient vector flow (GVF) snake | |
| Yang | Recursive approximation of the bilateral filter | |
| Xu et al. | FPGA-based low-visibility enhancement accelerator for video sequence by adaptive histogram equalization with dynamic clip-threshold | |
| Chen et al. | A new watershed algorithm for cellular image segmentation based on mathematical morphology | |
| CN108305268B (en) | Image segmentation method and device | |
| CN103593829A (en) | Hand vein pattern developing method based on mathematical morphology operation | |
| Zhou et al. | Underwater image enhancement method based on color correction and three-interval histogram stretching | |
| Vyavahare et al. | Segmentation using region growing algorithm based on CLAHE for medical images | |
| CN103226808B (en) | A kind of image enchancing method and system | |
| Ma et al. | Noise-against skeleton extraction framework and application on hand gesture recognition | |
| Maurya et al. | An efficient method for brain image preprocessing with anisotropic diffusion filter & tumor segmentation | |
| Wang et al. | Multifeature contrast enhancement algorithm for digital media images based on the diffusion equation | |
| Mohanapriya et al. | Comparative study of different enhancement techniques for medical images | |
| Jin et al. | Review of applications of partial differential equations for image enhancement | |
| Ariateja et al. | A review of contrast enhancement techniques in digital image processing | |
| Xia et al. | X-ray image enhancement base on the improved adaptive low-pass filtering | |
| He et al. | W-LDMM: A Wasserstein driven low-dimensional manifold model for noisy image restoration |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190906 |