CN113706405B - Image striping method combining feature extraction and nonlinear fitting - Google Patents
Image striping method combining feature extraction and nonlinear fitting Download PDFInfo
- Publication number
- CN113706405B CN113706405B CN202110913535.8A CN202110913535A CN113706405B CN 113706405 B CN113706405 B CN 113706405B CN 202110913535 A CN202110913535 A CN 202110913535A CN 113706405 B CN113706405 B CN 113706405B
- Authority
- CN
- China
- Prior art keywords
- image
- processed
- column
- vector
- nonlinear fitting
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000000605 extraction Methods 0.000 title claims abstract description 19
- 238000012545 processing Methods 0.000 claims abstract description 19
- 230000009466 transformation Effects 0.000 claims abstract description 7
- 238000001914 filtration Methods 0.000 claims description 25
- 238000012935 Averaging Methods 0.000 claims description 12
- 238000009499 grossing Methods 0.000 claims description 11
- 230000002146 bilateral effect Effects 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 4
- 230000035772 mutation Effects 0.000 claims description 4
- 238000012544 monitoring process Methods 0.000 abstract description 4
- 230000000694 effects Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000005587 bubbling Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000007429 general method Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
-
- 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
- G06T2207/20024—Filtering details
- G06T2207/20028—Bilateral filtering
-
- 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
- G06T2207/20024—Filtering details
- G06T2207/20032—Median filtering
-
- 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
- G06T2207/20224—Image subtraction
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
Aiming at the problems that the current stripe removing method has high utilization rate of hardware resources and high time cost and is difficult to realize real-time application on a platform with limited resources such as video monitoring and the like, the invention discloses an image stripe removing method combining feature extraction and nonlinear fitting, which comprises the following steps: (1) Acquiring an image to be processed, and obtaining a one-dimensional vector set through column average transformation; (2) Solving the smooth characteristic of the vector to obtain a nonlinear fitting processing result; (3) Performing row-wise subtraction operation on the image to be processed by utilizing the feature vector to obtain a preliminary reconstructed image; (4) Carrying out fine processing on the preliminary reconstructed image, and eliminating noise omitted in the step 3; (5) And performing row-wise subtraction operation on the preliminary reconstructed image by using the new feature vector to obtain a final reconstructed image.
Description
Technical Field
The invention belongs to the field of image denoising, and relates to the problems of feature extraction and matrix recovery based on an image convolution filtering theory and statistical filtering. And solving a nonlinear model through extracting and counting noise characteristics so as to achieve the purpose of final denoising.
Background
Streak noise is one of the most dominant noises affecting image quality such as visible light monitoring and infrared. Such noise is generally caused by imperfections in the readout circuitry process, and over time and environmental changes, such as changes in the ambient temperature of the infrared sensor, changes in the response of the image sensor at different locations, etc., streak noise may gradually change, continuously affecting image quality, and subsequent detection and recognition algorithms. Although some hardware pre-processing of analog electrical signals is performed prior to digital-to-analog signal conversion, such pre-processing is not always effective due to calibration and lack of parameter accuracy. Therefore, image-based streak noise removal is becoming increasingly important, particularly under the premise of rapid development in the current fields of video monitoring, intelligent recognition and the like.
The algorithms studied at present are mainly divided into three types, the first type is a method based on scene statistics, and the principle of the method is that firstly, a black body with a constant temperature field is shot as a reference to correct an image, so that the output of each sensor is kept consistent. And updating correction information in real time through stripe estimation in the real scene. Typical algorithms include the minimum mean square error finite impulse response filtering algorithm proposed by Hayat et al, and the method has simple operation and certain effect. But is difficult to adjust adaptively when encountering abrupt changes in the fringes in the image, while the scene estimation algorithm fails in the presence of dead pixels. The second type is a filtering-based method, which generally adopts modes such as Fourier transformation and wavelet transformation, and utilizes spatial and frequency domain features to extract stripes and remove the stripes, so the method is also called a feature-based method, and is mainly typically a wavelet analysis-based and self-adaptive Fourier zero frequency amplitude normalization algorithm proposed by Roshan et al, the method is not influenced by stripe mutation, the instantaneity is good, but part of edge details can be removed by mistake because the algorithm belongs to global processing. The final classification is based on an optimized variation method, and the main idea is to construct an energy functional related to noise and constraint conditions, convert the functional into Euler-Lagrange partial differential equations according to the constraint conditions, and finally calculate extremum under the non-constraint conditions, namely denoising. Typical algorithms have the anisotropic total variation method proposed by Bouali et al, and the algorithm has good effect of removing stripes, but has difficult influence on the non-uniformity of gain, and has the condition of uncertain convergence speed.
Disclosure of Invention
The invention solves the technical problems that: the image striping method combining feature extraction and nonlinear fitting is provided, the defects of the prior art are overcome, the problem that the existing striping method is high in utilization rate of hardware resources, high in time cost and difficult to realize real-time application on a platform with limited resources such as video monitoring is solved.
The technical scheme of the invention is as follows: an image striping method combining feature extraction and nonlinear fitting comprises the following steps:
step 1, acquiring an image to be processed, and obtaining a one-dimensional vector set through column average transformation;
step 2, solving the smooth characteristic of the vector to obtain a nonlinear fitting processing result;
Step 3, performing row-wise subtraction operation on the image to be processed by using the feature vector C S (1, N) to obtain a preliminary reconstructed image I rou (M, N);
Step 4, carrying out fine processing on the preliminary reconstructed image I rou (M, N) to eliminate the noise omitted in the step 3;
and 5, performing row-wise subtraction operation on the I rou (M, N) by using the feature vector to obtain a final reconstructed image.
The step 1 specifically includes: acquiring an image I ref (M, N) to be processed by a camera, wherein (M, N) represents the height and width of pixels of the image to be processed; then, the images are averaged by columns to obtain a column average value, and a one-dimensional vector set C N (1, N) with a length of N is obtained.
The formula for obtaining the column average value by carrying out column averaging on the image is as follows:
C(1,t)＝mean(Iref(M,t))，t＝1,…,N；
wherein mean represents the number of columns of the current image, t represents the number of columns of the current image, the value ranges from 1 to the image width N, and the deformation I ref (M, t) of the image I ref (M, N) to be processed is represented as a vector formed by all pixels in a certain column of the image, so as to obtain a column average result C (1, t).
The step 2 specifically includes: the initial one-dimensional vector set is processed by a polynomial smoothing algorithm to obtain a new smoothed vector set; and subtracting the processed vector set from the original vector set to obtain a finally output nonlinear stripe characteristic vector C S (1, N).
The specific process of the step2 is as follows:
step 2.1, performing smoothing operation on partial data mutation existing in C N (1, N) by using polynomial smoothing filtering with window function length of k=11, and correcting dislocation of data trend to obtain a nonlinear fitting vector C SG (1, N);
Step 2.2, C N (1, N) is subtracted from C SG (1, N) to obtain the final output nonlinear stripe feature vector C S (1, N).
The step3 specifically includes:
By using the characteristic vector C S (1, N), the line-by-line subtraction operation is carried out on the image to be processed
Irou(m,N)＝Iref(m,N)-CS(1,N)，m＝1,…,M；
Wherein M represents the number of lines of the current processed image, the numerical value is from 1 to the image height M, the formula also comprises deformation I ref (M, N) of the image I ref (M, N) to be processed, the deformation I ref (M, N) is represented as a vector formed by all pixels in a certain line of the image, and a preliminary reconstructed image I rou (M, N) is obtained.
The step 4 specifically includes:
Step 4.1, extracting noise characteristics through bilateral filtering with a kernel function size of 11×11 to obtain a spatial characteristic image I bif (M, N) of noise and part of edges;
Step 4.2, performing the bilateral filtering processing on the I rou (M, N) to obtain a processed image I bi (M, N);
Step 4.3, subtracting I rou (M, N) from I bi (M, N) to obtain a noise and partial edge space feature image I bif (M, N);
and 4.4, respectively carrying out column-wise averaging, column-wise median-calculating and column-wise kernel-calculating on the space feature images to obtain a median filter with the kernel function of 5*1, and then carrying out column-wise averaging on the filtering result to obtain three one-dimensional vector sets with the width of N.
The specific process of the step 4.1 is as follows: setting the size of the bilateral filtering kernel function to 11 x 11, and setting the spatial domain variance and the pixel range domain variance to 5 and 7 respectively; meanwhile, the non-central pixel point in the stripe direction is weighted, namely multiplied by 0.8, so that further extraction of the features is realized.
The specific process of the step 4.4 is as follows:
Step 4.4.1, performing column-wise averaging on the acquired stripe noise characteristic images I bif (M, N) to obtain a one-dimensional vector set C1;
Step 4.4.2, obtaining a one-dimensional vector set C2 by solving a median value according to columns, wherein the specific flow comprises the following steps: firstly, sorting the values of each column according to the size, wherein the sorting result is as follows:
Ibif(1,t)＜Ibif(2,t)＜…＜Ibif(M,t)
C2 is obtainable from the formula:
Median filtering result I mid (M, N) with kernel size 5*1:
Where i and j represent the length M and width N of the image, w k represents: when the current value is a median value, the weight value is 1, otherwise, the weight value is 0;
Step 4.4.3, processing the I mid (M, N) by using a C2 computing method to obtain C3, namely: firstly, sorting the values of each column according to the size, wherein the sorting result is as follows:
Imid(1,t)＜Imid(2,t)＜…＜Imid(M,t)
Next, C3 can be obtained by:
Step 4.4.4, reorder C1, C2, C3 to obtain a new vector set C4:
The formula for solving the one-dimensional vector set C1 is as follows:
Compared with the prior art, the invention has the advantages that: the image de-striping method combining feature extraction and nonlinear fitting provided by the invention is characterized in that firstly, aiming at the characteristic of stripe having directivity, the weight of a filtering kernel function is modified so as to increase the recognition capability of stripe noise on the premise of not influencing edge features; and secondly, carrying out median processing on the characteristic image, and eliminating mismatching under the condition of mixing edges and stripes. In particular, the present invention relates to an improved method having the following advantages:
(1) Targeted stripe removal. Under the condition that the edge characteristics are overlapped with the stripes, the scheme corrects the value of the overlapped area to obtain more accurate values.
(2) Facing more types of streak noise. Aiming at the characteristics of different stripe noises, the processes from rough denoising to fine denoising are adopted for processing respectively. The recognition and removal capability for more types of streak noise is obtained.
(3) The processing speed is increased. Because the original image is directly processed, time-consuming algorithms such as transformation and iteration of a calculation domain are not involved, and the processing speed is higher than that of a general method.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
fig. 2 is a flow chart of feature vector estimation according to the present invention.
Detailed Description
The present invention will be described in detail with reference to the accompanying drawings. As shown in fig. 1 and 2, the image striping algorithm combining feature extraction and nonlinear fitting provided by the present invention includes the following steps:
step 1, acquiring an image to be processed, and obtaining a one-dimensional vector set through column average transformation;
An image I ref (M, N) to be processed is acquired by the camera, where (M, N) represents the height and width of the pixels of the image to be processed. Then, the images are subjected to column-wise averaging operation to obtain a one-dimensional vector set C N (1, N) with a length of N, and the column average value is obtained as follows:
C(1,t)＝mean(Iref(M,t)) (t＝1,…,N)
And step 2, solving the smooth characteristic of the vector to obtain a nonlinear fitting processing result.
Polynomial smoothing algorithms are proposed by Savizkg and Golag. This method was originally proposed because the chemical spectrum was sampled with an entire row of dead spots at random locations. The algorithm is a weighted fit over a window using a window function, and is therefore also known as convolutional smoothing. The method uses the following formula to obtain a fit of the smoothing vector:
Wherein Y represents an input vector, denoted C N (1, N) in the present invention, Representing the output of the smoothed vector, i.e. the final output of the nonlinear fitting process,A split formula representing a least squares solution is shown as follows:
wherein G is in matrix form, as follows:
Where k=2g+1 represents the length of the window function.
Step 2.1, performing smoothing operation on partial data mutation existing in C N (1, N) by using polynomial smoothing filtering with window function length of k=11, and correcting dislocation of data trend to obtain a nonlinear fitting vector C SG (1, N);
Step 2.2, C N (1, N) is subtracted from C SG (1, N) to obtain the final output nonlinear stripe feature vector C S (1, N).
And 3, performing row-wise subtraction operation on the image to be processed by using the feature vector C S (1, N).
Irou(m,)＝Iref(m,)-CS(1,N) (m＝1,…,M)
A preliminary reconstructed image I rou (M, N) is obtained.
And 4, performing fine processing on the preliminary reconstructed image I rou (M, N) to eliminate the noise omitted in the step 3.
The bilateral filtering is different from the traditional filter, is a nonlinear weighting filter, and can achieve the effect of maintaining edges while reducing noise. The intensity values of surrounding pixels are sampled and averaged using a kernel function of fixed size to represent the intensity value of the center pixel, wherein the weights are based on a gaussian distribution. Meanwhile, in the pixel range domain, the bilateral filtering also considers the difference between the gray level of the pixel in the convolution kernel and the gray level of the central pixel, namely, the comprehensive result of the spatial domain and the pixel range domain is formed. Bilateral filtering is as follows:
Where p=1, …, m×n, I q represents the central pixel point of the current kernel function, I and j refer to the coordinates of the central pixel point, S represents the kernel function size, 11×11 in the present invention, q represents all pixels except the central pixel point, k and l refer to the coordinate positions of these pixel points, ω (I, j, k, l) is also included in the formula, AndThe normalization function, the spatial domain weighting function and the pixel range domain weighting function are respectively represented, and the formula is as follows:
Wherein σ s、σr and σ d represent spatial domain variance and pixel domain variance, respectively.
In the region where the image is flat, the change of the pixel value is small, the weight of the pixel range domain is close to 1, and the weight of the space domain plays a main role at the moment and is similar to Gaussian filtering; at the edge of the image, the pixel value changes greatly, and the weight of the pixel range domain is increased, so that the maintenance of the edge information is guided.
Step 4.1, setting the bilateral filtering kernel function to 11 x 11, and setting the spatial domain variance and the pixel range domain variance to 5 and 7 respectively. Meanwhile, the non-central pixel point in the stripe direction is weighted, namely multiplied by 0.8, so that further extraction of the features is realized.
And 4.2, performing bilateral filtering processing on the I rou (M, N) to obtain a processed image I bi (M, N).
Step 4.3, subtracting I rou (M, N) from I bi (M, N) to obtain a noise and partial edge spatial feature image I bif (M, N).
And 4.4, respectively carrying out column-wise averaging, column-wise median-calculating and column-wise kernel-calculating on the space feature images to obtain a median filter with the kernel function of 5*1, and then carrying out column-wise averaging on the filtering result to obtain three one-dimensional vector sets with the width of N. The specific implementation method is as follows:
step 4.4.1, performing column-wise averaging on the acquired stripe noise characteristic image I bif (M, N) to obtain a one-dimensional vector set C1, wherein the formula is as follows:
Step 4.4.2, obtaining a one-dimensional vector set C2 by solving a median value according to columns, wherein the specific flow comprises the following steps: firstly, sorting the values of each column according to the size, wherein the sorting result is as follows:
Ibif(1,t)＜Ibif(2,t)＜…＜Ibif(M,t)
next, C2 can be obtained by:
Median filtering result I mid (M, N) with kernel size 5*1:
Where i and j represent the length M and width N of the image, w k represents: when the current value is the median value, the weight value is 1, otherwise, the weight value is 0.
Step 4.4.3, processing the I mid (M, N) by using a C2 computing method to obtain C3, namely: firstly, sorting the values of each column according to the size, wherein the sorting result is as follows:
Imid(1,t)＜Imid(2,t)＜…＜Imid(M,t)
Next, C3 can be obtained by:
Step 4.4.4, reorder C1, C2, C3 to obtain a new vector set C4:
And 4.5, performing bubbling sequencing on the vector sets according to columns, and selecting the maximum value vector set as a new noise characteristic vector.
And 5, performing row-wise traversal subtraction operation on the I rou (M, N) by using the feature vector to obtain a final reconstructed image.
Although the present invention has been described with respect to the preferred embodiments, it is not intended to be limited thereto, and any person skilled in the art can make any possible variations and modifications to the technical solution of the present invention by using the technical matters disclosed above without departing from the spirit and scope of the present invention, so any simple modifications, equivalent variations and modifications to the above embodiments according to the technical matters of the present invention fall within the scope of the technical solution of the present invention.
Claims (8)
1. An image striping method combining feature extraction and nonlinear fitting is characterized by comprising the following steps:
step 1, acquiring an image to be processed, and obtaining a one-dimensional vector set through column average transformation;
step 2, solving the smooth characteristic of the vector to obtain a nonlinear fitting processing result;
Step 3, performing row-wise subtraction operation on the image to be processed by using the feature vector C S (1, N) to obtain a preliminary reconstructed image I rou (M, N);
Step 4, carrying out fine processing on the preliminary reconstructed image I rou (M, N) to eliminate the noise omitted in the step 3; the method specifically comprises the following steps:
step 4.1, setting the bilateral filtering kernel function to be 11 x 11 in size, and setting the spatial domain variance and the pixel range domain variance to be 5 and 7 respectively; meanwhile, weighting the non-central pixel point in the stripe direction, namely multiplying by 0.8, so as to realize further extraction of the features;
Step 4.2, performing the bilateral filtering processing on the I rou (M, N) to obtain a processed image I bi (M, N);
Step 4.3, subtracting I rou (M, N) from I bi (M, N) to obtain a noise and partial edge space feature image I bif (M, N);
Step 4.4, respectively carrying out column-wise averaging, column-wise median-calculating and column-wise kernel-calculating on the space feature images to obtain a median filter with a kernel function of 5*1, and then carrying out column-wise averaging on the filtering result to obtain three one-dimensional vector sets with a width of N;
and 5, performing row-wise subtraction operation on the I rou (M, N) by using the feature vector to obtain a final reconstructed image.
2. The method for image striping in combination with feature extraction and nonlinear fitting of claim 1, wherein: the step 1 specifically includes: acquiring an image I ref (M, N) to be processed by a camera, wherein (M, N) represents the height and width of pixels of the image to be processed; then, the images are averaged by columns to obtain a column average value, and a one-dimensional vector set C N (1, N) with a length of N is obtained.
3. The method for image striping in combination with feature extraction and nonlinear fitting of claim 2, wherein: the formula for obtaining the column average value by carrying out column averaging on the image is as follows:
C(1,t)＝mean(Iref(M,t))，t＝1,…,N；
wherein mean represents the number of columns of the current image, t represents the number of columns of the current image, the value ranges from 1 to the image width N, and the deformation I ref (M, t) of the image I ref (M, N) to be processed is represented as a vector formed by all pixels in a certain column of the image, so as to obtain a column average result C (1, t).
4. A method of image striping in combination with feature extraction and nonlinear fitting as claimed in claim 3, wherein: the step 2 specifically includes: the initial one-dimensional vector set is processed by a polynomial smoothing algorithm to obtain a new smoothed vector set; and subtracting the processed vector set from the original vector set to obtain a finally output nonlinear stripe characteristic vector C S (1, N).
5. The method for image striping in combination with feature extraction and nonlinear fitting of claim 4, wherein: the specific process of the step 2 is as follows:
step 2.1, performing smoothing operation on partial data mutation existing in C N (1, N) by using polynomial smoothing filtering with window function length of k=11, and correcting dislocation of data trend to obtain a nonlinear fitting vector C SG (1, N);
Step 2.2, C N (1, N) is subtracted from C SG (1, N) to obtain the final output nonlinear stripe feature vector C S (1, N).
6. A method of image striping in combination with feature extraction and nonlinear fitting as claimed in claim 3, wherein: the step 3 specifically includes:
By using the characteristic vector C S (1, N), the line-by-line subtraction operation is carried out on the image to be processed
Irou(m,N)＝Iref(m,N)-CS(1,N)，m＝1,…,M；
Wherein M represents the number of lines of the current processed image, the numerical value is from 1 to the image height M, the formula also comprises deformation I ref (M, N) of the image I ref (M, N) to be processed, the deformation I ref (M, N) is represented as a vector formed by all pixels in a certain line of the image, and a preliminary reconstructed image I rou (M, N) is obtained.
7. The method for image striping in combination with feature extraction and nonlinear fitting of claim 6, wherein: the specific process of the step 4.4 is as follows:
Step 4.4.1, performing column-wise averaging on the acquired stripe noise characteristic images I bif (M, N) to obtain a one-dimensional vector set C1;
Step 4.4.2, obtaining a one-dimensional vector set C2 by solving a median value according to columns, wherein the specific flow comprises the following steps: firstly, sorting the values of each column according to the size, wherein the sorting result is as follows:
Ibif(1,t)<Ibif(2,t)<…<Ibif(M,t)
C2 is obtainable from the formula:
Median filtering result I mid (M, N) with kernel size 5*1:
Where i and j represent the length M and width N of the image, w k represents: when the current value is a median value, the weight value is 1, otherwise, the weight value is 0;
Step 4.4.3, processing the I mid (M, N) by using a C2 computing method to obtain C3, namely: firstly, sorting the values of each column according to the size, wherein the sorting result is as follows:
Imid(1,t)<Imid(2,t)<…<Imid(M,t)
Next, C3 can be obtained by:
Step 4.4.4, reorder C1, C2, C3 to obtain a new vector set C4:
8. The method for image striping in combination with feature extraction and nonlinear fitting of claim 7, wherein: the formula for solving the one-dimensional vector set C1 is as follows:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110913535.8A CN113706405B (en) | 2021-08-10 | 2021-08-10 | Image striping method combining feature extraction and nonlinear fitting |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110913535.8A CN113706405B (en) | 2021-08-10 | 2021-08-10 | Image striping method combining feature extraction and nonlinear fitting |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113706405A CN113706405A (en) | 2021-11-26 |
| CN113706405B true CN113706405B (en) | 2024-08-09 |
Family
ID=78652100
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110913535.8A Active CN113706405B (en) | 2021-08-10 | 2021-08-10 | Image striping method combining feature extraction and nonlinear fitting |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113706405B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114187198B (en) * | 2021-11-30 | 2025-06-27 | 北京时代民芯科技有限公司 | A method for image stripe removal based on fuzzy matrix |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102779279A (en) * | 2012-06-14 | 2012-11-14 | 南京航空航天大学 | Image identification method of stirring friction welding arc stripe interval |
| CN111161172A (en) * | 2019-12-18 | 2020-05-15 | 北京波谱华光科技有限公司 | Infrared image column direction stripe eliminating method, system and computer storage medium |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| SE533650C2 (en) * | 2009-04-22 | 2010-11-16 | Flir Systems Ab | Imaging method for suppressing column or row noise in an IR-detected image |
| JP2019016865A (en) * | 2017-07-04 | 2019-01-31 | キヤノン株式会社 | Image processing device and image processing method |
-
2021
- 2021-08-10 CN CN202110913535.8A patent/CN113706405B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102779279A (en) * | 2012-06-14 | 2012-11-14 | 南京航空航天大学 | Image identification method of stirring friction welding arc stripe interval |
| CN111161172A (en) * | 2019-12-18 | 2020-05-15 | 北京波谱华光科技有限公司 | Infrared image column direction stripe eliminating method, system and computer storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113706405A (en) | 2021-11-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110211056B (en) | Adaptive infrared image stripping algorithm based on local median histogram | |
| CN114049283B (en) | An adaptive grayscale gradient histogram equalization remote sensing image enhancement method | |
| CN108230264B (en) | A single image dehazing method based on ResNet neural network | |
| CN101620727B (en) | Infrared Image Weighted Histogram Adaptive Enhancement Algorithm | |
| CN111861896A (en) | A UUV-Oriented Color Compensation and Restoration Method for Underwater Images | |
| CN105844601A (en) | Mine image enhancement method based on bilateral filtering and multi-scale Retinex algorithm | |
| Nezhadarya et al. | A new scheme for robust gradient vector estimation in color images | |
| CN109903254B (en) | An Improved Bilateral Filtering Method Based on Poisson Kernel | |
| CN111325688B (en) | Unmanned aerial vehicle image defogging method for optimizing atmosphere light by fusion morphology clustering | |
| CN102289819A (en) | Method for detecting infrared motion target in real time for background adaptive estimation | |
| CN113592750B (en) | Infrared enhancement method based on gradient histogram | |
| CN103489168A (en) | Enhancing method and system for infrared image being converted to pseudo color image in self-adaptive mode | |
| CN113487509B (en) | A remote sensing image fog removal method based on pixel clustering and transmittance fusion | |
| CN113706405B (en) | Image striping method combining feature extraction and nonlinear fitting | |
| CN114627011A (en) | Infrared sequence image noise reduction method based on improved bilateral filtering and multi-frame averaging | |
| CN114519676B (en) | Device and method for denoising raw images based on Bayer format | |
| CN113269687A (en) | Non-local average infrared image denoising method based on parameter optimization | |
| CN109685729A (en) | A kind of Poisson denoising method of Pixel-level threshold adaptive | |
| CN111598886A (en) | Pixel-level transmittance estimation method based on single image | |
| CN104766287A (en) | Blurred image blind restoration method based on significance detection | |
| CN106506999B (en) | Moment matching-based TDI-CMOS (time delay integration-complementary metal oxide semiconductor) image sensor FPN (field-programmable gate array) correction method | |
| CN117911419A (en) | Method and device for detecting steel rotation angle enhancement of medium plate, medium and equipment | |
| CN114187198B (en) | A method for image stripe removal based on fuzzy matrix | |
| CN113822823B (en) | Method and system for point-nearest neighbor restoration of spatially variable blur kernel of aero-optic effect images | |
| CN116862809A (en) | An image enhancement method under low exposure conditions |
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 |