CN112446838B - Image noise detection method and device based on local statistical information - Google Patents
Image noise detection method and device based on local statistical information Download PDFInfo
- Publication number
- CN112446838B CN112446838B CN202011338644.3A CN202011338644A CN112446838B CN 112446838 B CN112446838 B CN 112446838B CN 202011338644 A CN202011338644 A CN 202011338644A CN 112446838 B CN112446838 B CN 112446838B
- Authority
- CN
- China
- Prior art keywords
- pixel
- noise
- image
- value
- calculating
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 66
- 230000002159 abnormal effect Effects 0.000 claims description 14
- 238000000034 method Methods 0.000 claims description 14
- 238000001914 filtration Methods 0.000 claims description 6
- 230000000694 effects Effects 0.000 claims description 5
- 230000005484 gravity Effects 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 3
- 230000035945 sensitivity Effects 0.000 description 3
- 238000013527 convolutional neural network Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000000265 homogenisation Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 230000011218 segmentation Effects 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
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
- Testing, Inspecting, Measuring Of Stereoscopic Televisions And Televisions (AREA)
Abstract
The invention provides an image noise detection method and device based on local statistical information, and the technical scheme comprises the following steps: s1, calculating a local statistical information value of each pixel in an image to be detected; s2, judging whether each pixel in the image to be detected is in a flat area or a complex area; s3, calculating a first noise detection threshold of the flat area, and calculating a second noise detection threshold of the complex area; s4, under the condition that a certain pixel is in a flat area and the local statistical information value of the pixel is smaller than a first noise detection threshold value, judging the pixel as a noise pixel, otherwise, judging the pixel as a clean pixel; and under the condition that a certain pixel is in a complex area and the local statistical information value of the pixel is smaller than a second noise detection threshold value, judging the pixel as a noise pixel, and otherwise, judging the pixel as a clean pixel.
Description
Technical Field
The present invention relates to the field of image noise detection technologies, and in particular, to an image noise detection method and apparatus based on local statistical information.
Background
During the acquisition and transmission of images, digital images are often corrupted by impulse noise due to the sensor equipment. Random Value Impulse Noise (RVIN) is one of impulse noise, and noise pixel values thereof are randomly located between 0 and 255, so that it is difficult to handle. In order to perform operations such as contour extraction, region segmentation, and object recognition on the image later, it is necessary to restore the noise image.
The current mainstream denoising algorithm mainly can be divided into a block matching-based method, a convolutional neural network-based method and a fuzzy rule-based method, and in the recent popular denoising algorithm, the fuzzy rule and the convolutional neural network are introduced, so that although a good filtering effect is achieved, the algorithm complexity is increased, the running time is prolonged, and the equipment cost is high.
Disclosure of Invention
The invention aims to provide an image noise detection method and device based on local statistical information, which are simple to realize and have higher detection accuracy and sensitivity compared with the prior art
The invention is realized by the following technical scheme: the first aspect of the present invention provides an image noise detection method based on local statistical information, comprising the steps of:
s1, calculating a local statistical information value of each pixel in an image to be detected;
s2, judging whether each pixel in the image to be detected is in a flat area or a complex area;
s3, calculating a first noise detection threshold of the flat area, and calculating a second noise detection threshold of the complex area;
S4, under the condition that a certain pixel is in a flat area and the local statistical information value of the pixel is smaller than a first noise detection threshold value, judging the pixel as a noise pixel, otherwise, judging the pixel as a clean pixel;
and under the condition that a certain pixel is in a complex area and the local statistical information value of the pixel is smaller than a second noise detection threshold value, judging the pixel as a noise pixel, and otherwise, judging the pixel as a clean pixel.
Preferably, in step S1, a local statistical information value of each pixel in the image to be measured is calculated, including
Constructing a neighborhood by taking any given pixel x in an image to be detected as a centerCalculating pixel x and neighborhoodEuclidean distance and gray level difference for any pixel y:
based on Euclidean distance and gray level difference, calculating pixel x and neighborhood Similarity of any other pixel y: s (x, y) =d (x, y) ×i (x, y)
Calculating pixel x and neighborhoodThe sum of the similarity of all pixels:
ζx is normalized to be constrained in the [0.1] interval:
Will be Normalized to [0.1] interval:
in the formula, D (x, y) is the Euclidean distance between the pixel x and the pixel y, I (x, y) is the gray level difference between the pixel x and the pixel y, and (s, t) represents that the pixel x is in the neighborhood (M, n) represents the position of pixel y in the neighborhoodIn (2), σ D is an adjustment parameter of euclidean distance, σ I is an adjustment parameter of gray level difference, ζx is a sum of similarity, and LS X is a local statistical information value of pixel x.
Preferably, determining whether each pixel in the image to be measured is in a flat area or a complex area includes:
Computing the neighborhood Estimated mean μ x of intensities of all pixels within:
Calculating a neighborhood based on the estimated mean value Standard deviation of intensities of all pixels in a display
According to the standard deviation, whether the given pixel x is in a flat area or a complex area is judged:
Wherein W1 and W2 are LS y weights used to adjust the specific gravity of clean and noise pixels to calculate the local variance effect, a, b are normalized parameters, T σ is a threshold to distinguish whether a pixel is in a complex region or a flat region, LS y is a neighborhood The maximum value of the local statistics values for all pixels within, u y, is the gray value of the pixel y having the maximum value of the local statistics values.
Preferably, calculating a first noise detection threshold for a flat region, and calculating a second noise detection threshold for a complex region, includes:
Selecting a plurality of flat areas with the size of M from the image to be detected, and judging abnormal pixels and non-abnormal pixels in the flat areas:
estimating the noise level of each region:
the overall noise level of the image is obtained by performing a weighted average operation on the noise level of each region:
Calculating a first noise detection threshold for the flat region:
θf＝-0.12σ3+0.07σ2+0.75σ+0.19
calculating a second noise detection threshold for the complex region:
θc＝0.31σ3+0.63σ2+0.52σ+0.03
Where Q n is the number of outlier pixels, Q c is the number of non-outlier pixels, d is the number of flat regions, I x is the intensity of pixel x, I y is the intensity of pixel y, and θ is the empirical threshold.
Preferably, in step S4, when the pixel x is in the flat area, the LS x value of the pixel x is compared with the first noise detection threshold value:
When LS x≤θf, pixel x is a noise pixel, and when LS x>θf, pixel x is a clean pixel.
Preferably, when the pixel x is in the complex region, the LS x value of the pixel x is compared with the size of the second noise detection threshold:
When LS x≤θc, pixel x is a noise pixel, and when LS x>θc, pixel x is a clean pixel.
Preferably, the step S4 further includes, when the pixel x is determined as a noise pixel, performing filtering preprocessing on the image to be detected to obtain a filtered image of the image to be detected, and comparing the pixels x located at the same coordinates of the two images:
When |i x-Ix'|>TP, pixel x is a clean pixel, and when |i x-Ix'|≤TP, pixel x is a noise pixel, where I x' is the intensity value of the corresponding point of pixel x in the filtered image, and T P is the decision threshold.
Preferably, the θ is in the range of [5,8].
Preferably, the value of T P is 15.
The second aspect of the present invention provides an image noise detection apparatus, including an acquisition module, further including:
the calculating module is used for calculating the local statistical information value of each pixel in the image to be measured;
and the first judging module is used for judging whether each pixel in the image to be detected is in a flat area or a complex area.
The second judging module judges the pixel as a noise pixel under the condition that a certain pixel is in a flat area and the local statistical information value of the pixel is smaller than a first noise detection threshold value, and otherwise, the pixel is a clean pixel;
and under the condition that a certain pixel is in a complex area and the local statistical information value of the pixel is smaller than a second noise detection threshold value, judging the pixel as a noise pixel, and otherwise, judging the pixel as a clean pixel.
Compared with the prior art, the invention has the following beneficial effects:
The image noise detection method and the device based on the local statistical information provided by the invention are characterized in that whether the pixel is noise or not is represented by the local statistical information value of each pixel in the image to be detected, and noise pixels and clean pixels can be screened out by solving the local statistical information value of each pixel point in the image and setting a proper threshold value, so that the accuracy and the sensitivity of the image noise detection method are higher, and the problem of lower accuracy and sensitivity of the impulse noise detection method in the prior art is solved; the invention does not involve complex multiplication operation in the implementation process, so that the implementation method is simple, and the problem of complex detection method caused by adopting the multiplication operation in the prior art is solved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only preferred embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of an image noise detection method based on local statistics according to embodiment 1 of the present invention;
fig. 2 is a flowchart of an image noise detection method based on local statistics according to embodiment 2 of the present invention;
Fig. 3 is a schematic structural diagram of an image noise detecting device according to embodiment 3 of the present invention.
Detailed Description
For a better understanding of the technical content of the present invention, specific examples are provided below and the present invention is further described with reference to the accompanying drawings.
In the description of embodiments of the present application, the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more such feature. In the description of the present embodiment, unless otherwise specified, the meaning of "plurality" is two or more.
Example 1
Referring to fig. 1, as a first embodiment of the present application, the present application provides an image noise detection method based on local statistical information, comprising the steps of:
S1, calculating a local statistical information value of each pixel in an image to be detected, wherein the method comprises the following steps:
constructing a neighborhood by taking any given pixel x in an image to be detected as a center Calculating pixel x and neighborhoodEuclidean distance and gray level difference for any pixel y:
based on Euclidean distance and gray level difference, calculating pixel x and neighborhood Similarity of any other pixel y: s (x, y) =d (x, y) ×i (x, y)
Calculating pixel x and neighborhoodThe sum of the similarity of all pixels:
ζx is normalized to be constrained in the [0.1] interval:
Here, the Representation pairIs performed by the homogenization process.
From observation, it can be found that ζx of each pixel in the noise image is substantially dispersed in [0,2.5 ]. For more convenience, the data can be processed more quickly, the following formula can be used to apply any pixel to be processedNormalized to [0.1] interval:
In the formula, D (x, y) is the euclidean distance between pixel x and pixel y, I (x, y) is the gray level difference between pixel x and pixel y, and when the distance between two pixels and the gray level difference become large, they both decrease, which also means that if the gray level difference between two pixels is large or the distance is far, their similarity is small, and even the euclidean distance can be omitted;
(s, t) represents that pixel x is in the neighborhood (M, n) represents the position of pixel y in the neighborhoodIn (2), σ D is an adjustment parameter of euclidean distance, σ I is an adjustment parameter of gray level difference, the influence of the two parameters on D (x, y) and I (x, y) can be changed by adjusting the values of the two parameters, ζx is the sum of similarity, LS X is a local statistical information value of the pixel x, and the probability of whether the pixel is noise can be represented. If the LS X value is smaller, this means that the pixel x has a smaller similarity with the pixels in its neighborhood, which means that the pixel x has a greater probability of being noise.
In a preferred implementation of the present embodiment, the neighborhood constructedIs a 5 x 5 neighborhood.
S2, judging whether each pixel in the image to be detected is in a flat area or a complex area, wherein the method comprises the following steps:
Computing the neighborhood Estimated mean μ x of intensities of all pixels within:
Calculating a neighborhood based on the estimated mean value Standard deviation of intensities of all pixels in a display
According to the standard deviation, whether the given pixel x is in a flat area or a complex area is judged:
Wherein W1 and W2 are LS y weights used to adjust the specific gravity of clean and noise pixels to calculate the local variance effect, a, b are normalized parameters, T σ is a threshold to distinguish whether a pixel is in a complex region or a flat region, LS y is a neighborhood The maximum value of the local statistics values for all pixels within, u y, is the gray value of the pixel y having the maximum value of the local statistics values.
In a preferred implementation of this example, the T σ is in the range of [0.3,8]
S3, calculating a first noise detection threshold of the flat area, and calculating a second noise detection threshold of the complex area, wherein the method comprises the following steps:
Selecting a plurality of flat areas with the size of M from the image to be detected, and judging abnormal pixels and non-abnormal pixels in the flat areas:
When I x-Iy > theta, the pixel x is an abnormal pixel, when I x-Iy is less than or equal to theta, the pixel x is a non-abnormal pixel, I x is the intensity of the pixel x, I y is the intensity of the pixel y, the intensity is the gray value of the pixel, and the gray value of the pixel of the image to be detected can be obtained by leading the image to be detected into a corresponding Matlab program;
in a preferred implementation of this example, the θ is in the range of [5,8].
In yet another preferred implementation of this embodiment, the θ is in the range of [0,20].
Estimating the noise level of each region:
the overall noise level of the image is obtained by performing a weighted average operation on the noise level of each region:
Calculating a first noise detection threshold for the flat region:
θf＝-0.12σ3+0.07σ2+0.75σ+0.19
calculating a second noise detection threshold for the complex region:
θc＝0.31σ3+0.63σ2+0.52σ+0.03
where Q n is the number of outlier pixels, Q c is the number of non-outlier pixels, d is the number of flat regions, and θ is the empirical threshold.
S4, under the condition that a certain pixel is in a flat area and the local statistical information value of the pixel is smaller than a first noise detection threshold value, judging the pixel as a noise pixel, otherwise, judging the pixel as a clean pixel;
under the condition that a certain pixel is in a complex area and the local statistical information value of the pixel is smaller than a second noise detection threshold value, the pixel is judged to be a noise pixel, otherwise, the pixel is a clean pixel, and the specific method is as follows:
when the pixel x is in the flat region, comparing the LS x value of the pixel x with the magnitude of the first noise detection threshold:
When LS x≤θc, pixel x is a noise pixel, and when LS x>θc, pixel x is a clean pixel.
When the pixel x is in the complex region, comparing the LS x value of the pixel x with the size of the second noise detection threshold:
When LS x≤θf, pixel x is a noise pixel, and when LS x>θf, pixel x is a clean pixel.
Example 2
Referring to fig. 2, as a second embodiment of the present invention, when a clean pixel is on an edge or contour of an image, an intensity difference between the clean pixel and a pixel in the vicinity thereof is more remarkable, which easily results in that the pixel on the edge and the contour is regarded as a noise pixel in a noise detection process, and in order to further improve the accuracy of a detection result, the present invention adds a limitation condition on the basis of step S4 to avoid erroneously detecting the edge pixel as a noise pixel:
when the pixel x is judged to be a noise pixel, carrying out median filtering and Gaussian filtering pretreatment on the image to be detected to obtain a filtered image of the image to be detected, and comparing the pixels x positioned at the same coordinates of the two images:
When |i x-Ix'|>TP, pixel x is a clean pixel, and when |i x-Ix'|≤TP, pixel x is a noise pixel, where I x' is the intensity value of the corresponding point of pixel x in the filtered image, and T P is the decision threshold.
In a preferred implementation manner of this embodiment, the value of T P is 15.
Example 3
Referring to fig. 3, as a third embodiment of the present invention, the present invention provides an image noise detection apparatus, including an acquisition module, the acquisition apparatus being configured to acquire a noise image to be detected, further including:
the calculating module is used for calculating the local statistical information value of each pixel in the image to be measured;
and the first judging module is used for judging whether each pixel in the image to be detected is in a flat area or a complex area.
The second judging module judges the pixel as a noise pixel under the condition that a certain pixel is in a flat area and the local statistical information value of the pixel is smaller than a first noise detection threshold value, and otherwise, the pixel is a clean pixel;
and under the condition that a certain pixel is in a complex area and the local statistical information value of the pixel is smaller than a second noise detection threshold value, judging the pixel as a noise pixel, and otherwise, judging the pixel as a clean pixel.
Further, the present embodiment provides an image noise detection apparatus, which implements the methods described in embodiment 1 and embodiment 2 when executed.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather to enable any modification, equivalent replacement, improvement or the like to be made within the spirit and principles of the invention.
Claims (4)
1. An image noise detection method based on local statistical information is characterized by comprising the following steps:
s1, calculating a local statistical information value of each pixel in an image to be detected;
s2, judging whether each pixel in the image to be detected is in a flat area or a complex area;
s3, calculating a first noise detection threshold of the flat area, and calculating a second noise detection threshold of the complex area;
S4, under the condition that a certain pixel is in a flat area and the local statistical information value of the pixel is smaller than a first noise detection threshold value, judging the pixel as a noise pixel, otherwise, judging the pixel as a clean pixel;
Under the condition that a certain pixel is in a complex area and the local statistical information value of the pixel is smaller than a second noise detection threshold value, judging the pixel as a noise pixel, otherwise, judging the pixel as a clean pixel;
in step S1, a local statistical information value of each pixel in the image to be measured is calculated, including
Constructing a neighborhood by taking any given pixel x in an image to be detected as a centerCalculating pixel x and neighborhoodEuclidean distance and gray level difference for any pixel y:
based on Euclidean distance and gray level difference, calculating pixel x and neighborhood Similarity of any other pixel y: s (x, y) =d (x, y) ×i (x, y);
Calculating pixel x and neighborhood The sum of the similarity of all pixels:
ζx is normalized to be constrained in the [0.1] interval:
Will be Normalized to [0.1] interval:
in the formula, D (x, y) is the Euclidean distance between the pixel x and the pixel y, I (x, y) is the gray level difference between the pixel x and the pixel y, and (s, t) represents that the pixel x is in the neighborhood (M, n) represents the position of pixel y in the neighborhoodIn (2), sigma D is an adjusting parameter of Euclidean distance, sigma I is an adjusting parameter of gray level difference, ζx is a sum of similarity, LS X is a local statistical information value of pixel x;
The judging of whether each pixel in the image to be detected is in a flat area or a complex area comprises the following steps:
Computing the neighborhood Estimated mean μ x of intensities of all pixels within:
Calculating a neighborhood based on the estimated mean value Standard deviation of intensities of all pixels in a display
According to the standard deviation, whether the given pixel x is in a flat area or a complex area is judged:
Wherein W1 and W2 are LS y weights used to adjust the specific gravity of clean and noise pixels to calculate the local variance effect, a, b are normalized parameters, T σ is a threshold to distinguish whether a pixel is in a complex region or a flat region, LS y is a neighborhood The maximum value of the local statistics values for all pixels within, u y is the gray value of pixel y having the maximum value of the local statistics values;
calculating a first noise detection threshold for a flat region, calculating a second noise detection threshold for a complex region, comprising:
Selecting a plurality of flat areas with the size of M from the image to be detected, and judging abnormal pixels and non-abnormal pixels in the flat areas:
estimating the noise level of each region:
the overall noise level of the image is obtained by performing a weighted average operation on the noise level of each region:
Calculating a first noise detection threshold for the flat region:
θf＝-0.12σ3+0.07σ2+0.75σ+0.19
calculating a second noise detection threshold for the complex region:
θc＝0.31σ3+0.63σ2+0.52σ+0.03
Wherein, Q n is the number of abnormal pixels, Q c is the number of non-abnormal pixels, d is the number of flat areas, I x is the intensity of pixel x, I y is the intensity of pixel y, θ is the empirical threshold;
In step S4, when the pixel x is in the flat region, the LS x value of the pixel x is compared with the first noise detection threshold value:
When LS x≤θf, pixel x is a noise pixel, and when LS x>θf, pixel x is a clean pixel;
When the pixel x is in the complex region, comparing the LS x value of the pixel x with the size of the second noise detection threshold:
When LS x≤θc, pixel x is a noise pixel, and when LS x>θc, pixel x is a clean pixel;
The step S4 further includes, when the pixel x is determined as a noise pixel, performing filtering preprocessing on the image to be detected to obtain a filtered image of the image to be detected, and comparing the pixels x located at the same coordinates of the two images:
When |i x-Ix'|>TP, pixel x is a clean pixel, and when |i x-Ix'|≤TP, pixel x is a noise pixel, where I x' is the intensity value of the corresponding point of pixel x in the filtered image, and T P is the decision threshold.
2. The method for detecting image noise based on local statistical information according to claim 1, wherein the range of θ is [5,8].
3. The method for detecting image noise based on local statistical information according to claim 1, wherein the value of T P is 15.
4. An image noise detection apparatus, comprising an acquisition module, characterized by further comprising:
A calculation module for calculating local statistical information value of each pixel in the image to be measured, and constructing a neighborhood by taking any given pixel x in the image to be measured as the center Calculating pixel x and neighborhoodEuclidean distance and gray level difference for any pixel y:
based on Euclidean distance and gray level difference, calculating pixel x and neighborhood Similarity of any other pixel y: s (x, y) =d (x, y) ×i (x, y);
Calculating pixel x and neighborhood The sum of the similarity of all pixels:
ζx is normalized to be constrained in the [0.1] interval:
Will be Normalized to [0.1] interval:
in the formula, D (x, y) is the Euclidean distance between the pixel x and the pixel y, I (x, y) is the gray level difference between the pixel x and the pixel y, and (s, t) represents that the pixel x is in the neighborhood (M, n) represents the position of pixel y in the neighborhoodIn (2), sigma D is an adjusting parameter of Euclidean distance, sigma I is an adjusting parameter of gray level difference, ζx is a sum of similarity, LS X is a local statistical information value of pixel x;
The first judging module is used for judging whether each pixel in the image to be detected is in a flat area or a complex area;
Computing the neighborhood Estimated mean μ x of intensities of all pixels within:
Calculating a neighborhood based on the estimated mean value Standard deviation of intensities of all pixels in a display
According to the standard deviation, whether the given pixel x is in a flat area or a complex area is judged:
Wherein W1 and W2 are LS y weights used to adjust the specific gravity of clean and noise pixels to calculate the local variance effect, a, b are normalized parameters, T σ is a threshold to distinguish whether a pixel is in a complex region or a flat region, LS y is a neighborhood The maximum value of the local statistics values for all pixels within, u y is the gray value of pixel y having the maximum value of the local statistics values;
The second judging module judges the pixel as a noise pixel under the condition that a certain pixel is in a flat area and the local statistical information value of the pixel is smaller than a first noise detection threshold value, and otherwise, the pixel is a clean pixel;
Under the condition that a certain pixel is in a complex area and the local statistical information value of the pixel is smaller than a second noise detection threshold value, judging the pixel as a noise pixel, otherwise, judging the pixel as a clean pixel;
calculating a first noise detection threshold for a flat region, calculating a second noise detection threshold for a complex region, comprising:
Selecting a plurality of flat areas with the size of M from the image to be detected, and judging abnormal pixels and non-abnormal pixels in the flat areas:
estimating the noise level of each region:
the overall noise level of the image is obtained by performing a weighted average operation on the noise level of each region:
Calculating a first noise detection threshold for the flat region:
θf＝-0.12σ3+0.07σ2+0.75σ+0.19
calculating a second noise detection threshold for the complex region:
θc＝0.31σ3+0.63σ2+0.52σ+0.03
Wherein, Q n is the number of abnormal pixels, Q c is the number of non-abnormal pixels, d is the number of flat areas, I x is the intensity of pixel x, I y is the intensity of pixel y, θ is the empirical threshold;
In step S4, when the pixel x is in the flat region, the LS x value of the pixel x is compared with the first noise detection threshold value:
When LS x≤θf, pixel x is a noise pixel, and when LS x>θf, pixel x is a clean pixel;
When the pixel x is in the complex region, comparing the LS x value of the pixel x with the size of the second noise detection threshold:
When LS x≤θc, pixel x is a noise pixel, and when LS x>θc, pixel x is a clean pixel;
The step S4 further includes, when the pixel x is determined as a noise pixel, performing filtering preprocessing on the image to be detected to obtain a filtered image of the image to be detected, and comparing the pixels x located at the same coordinates of the two images:
When |i x-Ix'|>TP, pixel x is a clean pixel, and when |i x-Ix'|≤TP, pixel x is a noise pixel, where I x' is the intensity value of the corresponding point of pixel x in the filtered image, and T P is the decision threshold.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011338644.3A CN112446838B (en) | 2020-11-24 | 2020-11-24 | Image noise detection method and device based on local statistical information |
| PCT/CN2021/103352 WO2022110804A1 (en) | 2020-11-24 | 2021-06-30 | Image noise measurement method and device based on local statistical information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011338644.3A CN112446838B (en) | 2020-11-24 | 2020-11-24 | Image noise detection method and device based on local statistical information |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112446838A CN112446838A (en) | 2021-03-05 |
| CN112446838B true CN112446838B (en) | 2024-07-12 |
Family
ID=74737927
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011338644.3A Active CN112446838B (en) | 2020-11-24 | 2020-11-24 | Image noise detection method and device based on local statistical information |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN112446838B (en) |
| WO (1) | WO2022110804A1 (en) |
Families Citing this family (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112446838B (en) * | 2020-11-24 | 2024-07-12 | 海南大学 | Image noise detection method and device based on local statistical information |
| CN114764803B (en) * | 2022-06-16 | 2022-09-20 | 深圳深知未来智能有限公司 | Noise evaluation method and device based on real noise scene and storage medium |
| CN115272684B (en) * | 2022-09-29 | 2022-12-27 | 山东圣点世纪科技有限公司 | Method for processing pseudo noise in vein image enhancement process |
| CN115526890B (en) * | 2022-11-25 | 2023-03-24 | 深圳市腾泰博科技有限公司 | Method for identifying fault factors of record player head |
| CN115981589B (en) * | 2023-03-22 | 2023-06-02 | 东莞锐视光电科技有限公司 | Software system for generating stripe light |
| CN116342638B (en) * | 2023-03-31 | 2023-10-31 | 西南大学 | An image element extraction method |
| CN116168026B (en) * | 2023-04-24 | 2023-06-27 | 山东拜尔检测股份有限公司 | Water quality detection method and system based on computer vision |
| CN116188462B (en) * | 2023-04-24 | 2023-08-11 | 深圳市翠绿贵金属材料科技有限公司 | A quality detection method and system for precious metals based on visual identification |
| CN116342610B (en) * | 2023-05-31 | 2023-08-15 | 山东恒海钢结构有限公司 | Steel structure assembly type building welding abnormality detection method |
| CN116883370B (en) * | 2023-07-18 | 2024-02-20 | 西藏净微检测技术有限公司 | Agricultural product appearance quality detecting system |
| CN117437600B (en) * | 2023-12-20 | 2024-03-26 | 山东海纳智能装备科技股份有限公司 | Coal flow monitoring system based on image recognition technology |
| CN117689917B (en) * | 2024-02-02 | 2024-04-30 | 广州中海电信有限公司 | Cabin safety state monitoring method based on thermal imaging and laser detection technology |
| CN117952856B (en) * | 2024-03-15 | 2024-06-07 | 深圳市星能计算机有限公司 | Method for monitoring ventilation and heat dissipation state of server based on infrared thermal imaging |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103971343A (en) * | 2014-05-21 | 2014-08-06 | 浙江宇视科技有限公司 | Image denoising method based on similar pixel detection |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2002102086A2 (en) * | 2001-06-12 | 2002-12-19 | Miranda Technologies Inc. | Apparatus and method for adaptive spatial segmentation-based noise reducing for encoded image signal |
| CN105809630B (en) * | 2014-12-30 | 2019-03-12 | 展讯通信(天津)有限公司 | A kind of picture noise filter method and system |
| CN106373098B (en) * | 2016-08-30 | 2019-04-23 | 天津大学 | Random impulse noise removal method based on dissimilar pixel statistics |
| CN111784605B (en) * | 2020-06-30 | 2024-01-26 | 珠海全志科技股份有限公司 | Image noise reduction method based on region guidance, computer device and computer readable storage medium |
| CN112446838B (en) * | 2020-11-24 | 2024-07-12 | 海南大学 | Image noise detection method and device based on local statistical information |
-
2020
- 2020-11-24 CN CN202011338644.3A patent/CN112446838B/en active Active
-
2021
- 2021-06-30 WO PCT/CN2021/103352 patent/WO2022110804A1/en active Application Filing
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103971343A (en) * | 2014-05-21 | 2014-08-06 | 浙江宇视科技有限公司 | Image denoising method based on similar pixel detection |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2022110804A9 (en) | 2023-09-21 |
| CN112446838A (en) | 2021-03-05 |
| WO2022110804A1 (en) | 2022-06-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112446838B (en) | Image noise detection method and device based on local statistical information | |
| CN109870461B (en) | Electronic components quality detection system | |
| US9361672B2 (en) | Image blur detection | |
| CN117541588B (en) | Printing defect detection method for paper product | |
| CN116777941B (en) | Profile contour detection method and system based on machine vision | |
| CN115294137B (en) | Cloth surface color bleeding defect detection method | |
| CN116823824B (en) | Underground belt conveyor dust fall detecting system based on machine vision | |
| CN107220962B (en) | An image detection method and device for tunnel cracks | |
| CN105550694B (en) | Method for measuring fuzzy degree of face image | |
| CN110866503A (en) | Abnormality detection method and system for finger vein equipment | |
| CN118570072B (en) | A burr detection method and system during titanium metal processing | |
| CN112417955A (en) | Patrol video stream processing method and device | |
| CN115601368A (en) | Method for detecting defects of sheet metal parts of building material equipment | |
| WO2024016632A1 (en) | Bright spot location method, bright spot location apparatus, electronic device and storage medium | |
| CN117788470B (en) | Tire defect detection method based on artificial intelligence | |
| CN116883412B (en) | Graphene far infrared electric heating equipment fault detection method | |
| CN111340041A (en) | License plate recognition method and device based on deep learning | |
| CN116129195A (en) | Image quality evaluation device, image quality evaluation method, electronic device, and storage medium | |
| CN116152261B (en) | Visual inspection system for quality of printed product | |
| CN117745552A (en) | Self-adaptive image enhancement method and device and electronic equipment | |
| CN113269005B (en) | Safety belt detection method and device and electronic equipment | |
| CN106254723A (en) | A kind of method of real-time monitoring video noise interference | |
| CN113221691A (en) | Fingerprint segmentation method and device, readable storage medium and terminal | |
| CN112613456A (en) | Small target detection method based on multi-frame differential image accumulation | |
| CN116883401B (en) | Industrial product production quality detection system |
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 |