CN114926360A - Image noise reduction processing working method based on noise estimation - Google Patents

Image noise reduction processing working method based on noise estimation Download PDF

Info

Publication number
CN114926360A
CN114926360A CN202210578415.1A CN202210578415A CN114926360A CN 114926360 A CN114926360 A CN 114926360A CN 202210578415 A CN202210578415 A CN 202210578415A CN 114926360 A CN114926360 A CN 114926360A
Authority
CN
China
Prior art keywords
image block
image
pixel
difference
value
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
Application number
CN202210578415.1A
Other languages
Chinese (zh)
Inventor
张文理
周鹏
陈柯帆
秦玉鑫
李玲玲
赵青
王毅
孙要伟
梁坤
陈宇
刘兆瑜
王恒
赵建海
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhengzhou University of Aeronautics
Original Assignee
Zhengzhou University of Aeronautics
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Zhengzhou University of Aeronautics filed Critical Zhengzhou University of Aeronautics
Priority to CN202210578415.1A priority Critical patent/CN114926360A/en
Publication of CN114926360A publication Critical patent/CN114926360A/en
Pending legal-status Critical Current

Links

Images

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
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
    • 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
    • G06T2207/20021—Dividing image into blocks, subimages or windows

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Image Processing (AREA)

Abstract

The invention provides an image noise reduction processing working method based on noise estimation, which comprises the following steps: s1, acquiring a 3D scene image, dividing the image to be denoised to form an image block set, and carrying out noise estimation on each image block; s2, performing image block information indexing on the estimated image block, and performing Gaussian filtering according to the image pixel composition in the image block; and S3, traversing the image block region according to the set gray level association after filtering, and carrying out denoising by replacing the difference pixel points according to the approximate estimated value of the image block, thereby realizing the denoising processing of the 3D image.

Description

Image noise reduction processing working method based on noise estimation
Technical Field
The invention relates to the field of image analysis, in particular to an image denoising processing working method based on noise estimation.
Background
No matter a video image or a photo image, information in the image can meet the requirements of a user only by performing optimization processing, so that the image needs to be subjected to noise reduction processing, and in order to reduce system overhead, there are many technical means for completing noise reduction, such as weighted mean filtering, wiener filtering, wavelet noise reduction, bilateral filtering, non-local mean filtering (NLM), block matching three-dimensional collaborative filtering (BM3D), and the like. The above method can basically implement noise reduction processing, but has the problem of large calculation amount, and the processing model of the image is too complex, and the captured image features are not comprehensive, so that a person skilled in the art needs to solve the corresponding technical problem.
Disclosure of Invention
The invention aims to at least solve the technical problems in the prior art, and particularly creatively provides an image denoising processing working method based on noise estimation.
In order to achieve the above object, the present invention provides an image denoising processing method based on noise estimation, comprising the following steps:
s1, acquiring a 3D scene image, dividing the image to be denoised to form an image block set, and carrying out noise estimation on each image block;
s2, indexing image block information of the estimated image block, and performing Gaussian filtering according to the image pixel composition in the image block;
and S3, traversing the image block region according to the set gray level association after filtering, and denoising by replacing the difference pixel points according to the approximate estimated value of the image block, thereby realizing the denoising processing of the 3D image.
The S1 includes:
s1-1, calibrating the pixel difference of an image in a 3D scene image, carrying out image segmentation, and collecting the pixel position of the calibration difference in the formed image block;
s1-2, forming an image block set M for collecting difference pixels, and sequentially comparing the difference pixels C of each image block p containing the difference pixels p Sequentially searching reference pixels and searching range S of candidate pixel points p ＝C′ p ∩M(Q p ,q 0 ,M p ,M all )，C′ p As candidate pixel point, Q p For a set of alternative pixels, q 0 Number of alternative difference pixels, M p Is an arbitrary pixel of an image block, M all For all the pixels of the image block,
when the content of the candidate pixel point search in the image block is too huge, the candidate pixel point needs to be converged and a reference threshold value, | q 0 -q|≤M mid Where q is the number of difference pixels of the current image block, M mid Candidate pixel intermediate values of the image block;
s1-3, performing noise estimation on each image block to form a noise repair sample area, extracting the boundary of the image block noise area of the corresponding pixel point according to the specific position of the pixel point (x, y) in the image block,
the boundary extraction estimation rule is as follows:
Figure BDA0003661331080000021
beta is the pixel change degree adjustment value, sigma x,y c (x, y) is a set of pixel points in the image block, N valid Effective pixel points in the image block;
carrying out correlation noise estimation on the image blocks, extracting texture attribute characteristics,
Figure BDA0003661331080000022
for the
Figure BDA0003661331080000023
Representing pixels R on all x-axes x X-axis accorded gray scale quantization value P x The product of (a) is subjected to deviation value calculation by the gray level adjustment value eta,
Figure BDA0003661331080000024
representing pixels R in all y-axes y Y-axis accorded gray scale quantization value P y The product of (a) is subjected to deviation value calculation by the gray scale adjustment value eta,
the S2 includes:
s2-1, performing information indexing on the image block subjected to noise estimation to form specific weights of noise pixel points of the image block and a candidate pixel point set, and recording pixel gray level deviation values of an x axis and a y axis respectively;
s2-2, marking the pixels with overlarge pixel gray value deviation value one by scanning each pixel point of the image block, and accurately positioning the position of the pixel deviation value;
the scanning of the pixel points is performed by gaussian filtering,
Figure BDA0003661331080000031
sigma is the standard deviation of Gaussian distribution, and the image block is scanned line by line;
after the noise of the image block is estimated, acquiring a primary pixel deviation value, and obtaining an accurate pixel deviation value after Gaussian filtering, thereby preparing for noise reduction of the image block;
the S3 includes:
s3-1, performing gray level correlation analysis on the filtered image block according to the shading and shading of the pixels of the image block and the gradient direction, wherein the x axis of the image block comprises a difference pixel i and the y axis of the image block comprises the minimum value of the image pixel of the characteristic factor extracted by the difference pixel j
Figure BDA0003661331080000032
The classification map is used as an influence factor of the classification map of the region where the candidate pixel point is located and the region where the difference pixel point is located in the image block; the classification map can partition corresponding noise regions in the image block,
the classification chart E of the gray level correlation analysis is
Figure BDA0003661331080000033
Wherein mu is a difference pixel gradient adjustment coefficient, lambda is a difference pixel characteristic adjustment coefficient, and F (i, j) is a difference pixel characteristic function;
s3-2, traversing the image block through the classification map, searching the difference pixel points needing to be repaired, selecting the replaced pixels from the candidate pixel points according to the characteristic factors to carry out the process of denoising and replacing the difference pixel points,
obtaining approximate estimation values of candidate pixel points through approximate pixel calculation function calculation
Figure BDA0003661331080000041
During similarity measurement, locating a difference pixel binarization gradient value Z (i, j) in an image block in which an x axis contains a difference pixel i and a y axis contains a difference pixel j, and replacing a difference pixel point in the image block; u is a binarization set representing all difference pixels in the image block; selecting a candidate pixel binarization gradient value w (k, l) in an x-axis containing candidate pixels k and a y-axis containing candidate pixels l in an image block, wherein V is a binarization set representing all candidate pixel points in the image block;
Figure BDA0003661331080000042
the denoising estimated value of the difference pixel i for carrying out binarization replacement denoising on the image block is obtained;
Figure BDA0003661331080000043
the method comprises the steps of obtaining a denoising estimation value of a difference pixel j for binarization replacement denoising of an image block;
Figure BDA0003661331080000044
selecting an estimated value of a candidate pixel k for binarization replacement denoising of an image block;
Figure BDA0003661331080000045
selecting an estimated value of a candidate pixel l for binarization replacement denoising of an image block; a. the find An improved non-local mean calculation is performed for the sum of the weight reconstruction using the binarization of the difference pixels and the weight reconstruction of the candidate pixels,
s3-3, calculating the signal to noise ratio of the difference pixel points, and if the candidate pixel points used in the image blocks exceed the signal to noise ratio of the difference pixel points of the image blocks, calculating the signal to noise ratio of the repaired and denoised image blocks, wherein the larger the value is, the better the denoising effect is.
In summary, due to the adoption of the technical scheme, the invention has the beneficial effects that:
according to the invention, after the image denoising operation is carried out, a clearer texture image can be extracted from a natural image, and the robustness is very strong.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
The above and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
FIG. 1 is a schematic diagram of a pixel point of image block difference according to the present invention;
FIG. 2 is a schematic diagram illustrating image block classification diagram extraction according to the present invention;
FIG. 3 is an effect diagram of the present invention;
fig. 4 is an overall flow chart of the present invention.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention, and are not to be construed as limiting the present invention.
As shown in FIG. 4, the invention especially proposes a working method of image denoising processing based on noise estimation, which comprises the following steps:
s1, acquiring a 3D scene image, dividing the image to be denoised to form an image block set, and carrying out noise estimation on each image block;
s2, performing image block information indexing on the estimated image block, and performing Gaussian filtering according to the image pixel composition in the image block;
and S3, traversing the image block region according to the set gray level association after filtering, and denoising by replacing the difference pixel points according to the approximate estimated value of the image block, thereby realizing the denoising processing of the 3D image.
The S1 includes:
as shown in fig. 1, S1-1, calibrating pixel differences of an image in a 3D scene image, performing image segmentation, and collecting pixel positions of the calibrated differences in the formed image blocks;
s1-2, forming an image block set M collecting difference pixels, sequentially for the difference pixels C of each image block p containing difference pixels p Sequentially searching reference pixels and searching range S of candidate pixel points p ＝C′ p ∩M(Q p ,q 0 ,M p ,M all )，C′ p As candidate pixel point, Q p For a set of alternative pixels, q 0 Number of alternative difference pixels, M p For any pixel of the image block, M all For all the pixels of the image block,
when the content of the candidate pixel point search in the image block is too huge, the candidate pixel point needs to be converged and a reference threshold value, | q 0 -q|≤M mid Where q is the number of difference pixels of the current image block, M mid Selecting the intermediate value of the alternative pixels of the image block;
the purpose of setting reference pixel searching is to obtain pixels with closer colors in the later period in the image denoising process, so that similar pixel points are grabbed, and preparation is made for filling difference pixel points.
S1-3, carrying out noise estimation on each image block so as to form a noise repair sample area, extracting the boundary of the noise area of the image block of the corresponding pixel point according to the specific position of the pixel point (x, y) in the image block,
the boundary extraction estimation rule is as follows:
Figure BDA0003661331080000061
beta is the pixel change degree adjustment value, sigma x,y c (x, y) is a set of pixel points in the image block, N valid Effective pixel points in the image block;
carrying out correlation noise estimation on the image blocks, extracting texture attribute characteristics,
Figure BDA0003661331080000062
wherein the pixel gray level deviation values for the x-axis and the y-axis in the image block are wx and wy respectively,
for the
Figure BDA0003661331080000063
Representing pixels R at all x-axes x X-axis accorded gray scale quantization value P x The product of (a) is subjected to deviation value calculation by the gray level adjustment value eta,
Figure BDA0003661331080000064
representing pixels R on all y-axes y Y-axis accorded gray scale quantization value P y The product of (a) is subjected to deviation value calculation by the gray level adjustment value eta,
the S2 includes:
s2-1, performing information indexing on the image block subjected to noise estimation to form specific weights of noise pixel points of the image block and a candidate pixel point set, and recording pixel gray level deviation values of an x axis and a y axis respectively;
s2-2, marking the pixels with overlarge pixel gray value deviation value one by scanning each pixel point of the image block, and accurately positioning the position of the pixel deviation value;
the scanning of the pixel points is performed by gaussian filtering,
Figure BDA0003661331080000071
sigma is the standard deviation of Gaussian distribution, and the image block is scanned line by line;
after the noise of the image block is estimated, acquiring a preliminary pixel deviation value, and obtaining an accurate pixel deviation value after Gaussian filtering so as to prepare for noise reduction of the image block;
the S3 includes:
s3-1, performing gray level correlation analysis on the filtered image block according to the shading and shading of the pixels of the image block and the gradient direction, wherein the x axis of the image block comprises a difference pixel i and the y axis of the image block comprises the minimum value of the image pixel of the characteristic factor extracted by the difference pixel j
Figure BDA0003661331080000072
The classification map is used as an influence factor of the classification map of the region where the candidate pixel points are located and the region where the difference pixel points are located in the image block; the classification map can partition corresponding noise regions in the image block,
as shown in FIGS. 2 and 3, classification chart E of the gray level correlation analysis is
Figure BDA0003661331080000073
Wherein mu is a difference pixel gradient adjustment coefficient, lambda is a difference pixel characteristic adjustment coefficient, and F (i, j) is a difference pixel characteristic function;
s3-2, traversing the image blocks through the classification map, searching the difference pixel points needing to be repaired, selecting the replaced pixels from the candidate pixel points according to the characteristic factors to carry out the process of denoising and replacing the difference pixel points,
obtaining approximate estimation values of candidate pixel points through approximate pixel calculation function calculation
Figure BDA0003661331080000081
During similarity measurement, locating a difference pixel binarization gradient value Z (i, j) in an image block in which an x axis contains a difference pixel i and a y axis contains a difference pixel j, and replacing a difference pixel point in the image block; u is a binarization set representing all difference pixels in the image block; selecting a candidate pixel binarization gradient value w (k, l) in an x-axis containing candidate pixels k and a y-axis containing candidate pixels l in an image block, wherein V is a binarization set representing all candidate pixel points in the image block;
Figure BDA0003661331080000082
the denoising estimated value of the difference pixel i for carrying out binarization replacement denoising on the image block is obtained;
Figure BDA0003661331080000083
the image block is subjected to binarization, replacing and denoising to obtain a denoising estimated value of a difference pixel j;
Figure BDA0003661331080000084
is to binarize the image blockReplacing the selection estimation value of the candidate pixel k for denoising;
Figure BDA0003661331080000085
selecting an estimated value of a candidate pixel l for binarization replacement denoising of an image block; a. the find Performing improved non-local mean calculation for the sum of weight reconstruction of binarization of the difference pixel points and weight reconstruction of the candidate pixel points, thereby estimating an approximate estimation value of denoising, selecting and replacing the difference pixel points through the improved candidate pixel points, knowing the optimal candidate pixel points, improving the image quality of the image block, and achieving the purpose of improving the denoising quality;
s3-3, calculating the signal to noise ratio of the difference pixel points, and if the candidate pixel points used in the image blocks exceed the signal to noise ratio of the difference pixel points of the image blocks, calculating the signal to noise ratio of the repaired and denoised image blocks, wherein the larger the value is, the better the denoising effect is.
When the image is subjected to blocking processing, the efficiency and the accuracy of image processing are improved, a noise estimation function is set to realize the detail capture of the image, a noise area of the image is extracted, a color difference part is scanned after Gaussian filtering is carried out on the noise area, accurate positioning is carried out, image denoising is carried out in a weight reconstruction mode, and the denoising effect is evaluated through a peak signal-to-noise ratio.
While embodiments of the invention have been shown and described, it will be understood by those of ordinary skill in the art that: various changes, modifications, substitutions and alterations can be made to the embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims (7)

1. An image noise reduction processing working method based on noise estimation is characterized by comprising the following steps:
s1, acquiring a 3D scene image, dividing the image to be denoised to form an image block set, and carrying out noise estimation on each image block;
s2, performing image block information indexing on the estimated image block, and performing Gaussian filtering according to the image pixel composition in the image block;
and S3, traversing the image block region according to the set gray level association after filtering, and denoising by replacing the difference pixel points according to the approximate estimated value of the image block, thereby realizing the denoising processing of the 3D image.
2. The method for image denoising processing based on noise estimation according to claim 1, wherein the S1 includes:
s1-1, calibrating the pixel difference of an image in a 3D scene image, carrying out image segmentation, and collecting the pixel position of the calibration difference in the formed image block;
s1-2, forming an image block set M for collecting difference pixels, and sequentially comparing the difference pixels C of each image block p containing the difference pixels p Sequentially searching reference pixels and searching range S of candidate pixel points p ＝C′ p ∩M(Q p ,q 0 ,M p ,M all )，C′ p As candidate pixel point, Q p For a set of alternative pixels, q 0 Number of alternative difference pixels, M p For any pixel of the image block, M all For all the pixels of the image block,
when the search content of the candidate pixel points in the image block is too huge, the candidate pixel points need to be converged and a reference threshold value, | q 0 -q|≤M mid Where q is the number of difference pixels of the current image block, M mid And candidate pixel intermediate values of the image block.
3. The method for noise estimation based image noise reduction processing according to claim 2, wherein the S1 further includes:
s1-3, carrying out noise estimation on each image block so as to form a noise repair sample area, extracting the boundary of the noise area of the image block of the corresponding pixel point according to the specific position of the pixel point (x, y) in the image block,
the boundary extraction estimation rule is as follows:
Figure FDA0003661331070000021
beta is the pixel change degree adjustment value, sigma x,y c (x, y) is a set of pixel points in the image block, N valid Effective pixel points in the image block;
carrying out correlation noise estimation on the image blocks, extracting texture attribute characteristics,
Figure FDA0003661331070000022
for
Figure FDA0003661331070000023
Representing pixels R at all x-axes x X-axis accorded gray scale quantization value P x The product of (a) is subjected to deviation value calculation by the gray scale adjustment value eta,
Figure FDA0003661331070000024
representing pixels R in all y-axes y Y-axis accorded gray scale quantization value P y The product of (c) is calculated by the gray scale adjustment value η as a deviation value.
4. The method for image denoising processing based on noise estimation according to claim 1, wherein the S2 includes:
s2-1, performing information indexing on the image block subjected to noise estimation to form specific weights of noise pixel points of the image block and a candidate pixel point set, and recording pixel gray level deviation values of an x axis and a y axis respectively;
s2-2, marking the pixels with overlarge pixel gray value deviation value one by scanning each pixel point of the image block, and accurately positioning the position of the pixel deviation value;
the scanning of the pixel points is performed by gaussian filtering,
Figure FDA0003661331070000025
sigma is the standard deviation of Gaussian distribution, and the image block is scanned line by line;
after the noise of the image block is estimated, a preliminary pixel deviation value is obtained, and an accurate pixel deviation value is obtained after Gaussian filtering, so that preparation is made for noise reduction of the image block.
5. The method for image denoising processing based on noise estimation according to claim 1, wherein the S3 includes:
s3-1, performing gray level correlation analysis on the filtered image block according to the shading and shading of the pixels of the image block and the gradient direction, wherein the x axis of the image block comprises a difference pixel i and the y axis of the image block comprises the minimum value of the image pixel of the characteristic factor extracted by the difference pixel j
Figure FDA0003661331070000031
The classification map is used as an influence factor of the classification map of the region where the candidate pixel points are located and the region where the difference pixel points are located in the image block; the classification map can divide corresponding noise regions in the image blocks.
6. The image denoising processing working method based on noise estimation according to claim 5, wherein the classification map E of the gray correlation analysis is
Figure FDA0003661331070000032
Where μ is the difference pixel gradient adjustment coefficient, λ is the difference pixel characteristic adjustment coefficient, and F (i, j) is the difference pixel characteristic function.
7. The method for noise estimation based image noise reduction processing according to claim 5, wherein said S3 further includes:
s3-2, traversing the image block through the classification map, searching the difference pixel points needing to be repaired, selecting the replaced pixels from the candidate pixel points according to the characteristic factors to carry out the process of denoising and replacing the difference pixel points,
obtaining an approximate estimation value of a candidate pixel point through approximate pixel calculation function calculation
Figure FDA0003661331070000041
During similarity measurement, locating a difference pixel binarization gradient value Z (i, j) in an image block in which an x axis contains a difference pixel i and a y axis contains a difference pixel j, and replacing a difference pixel point in the image block; u is a binarization set representing all difference pixels in the image block; selecting a candidate pixel binarization gradient value w (k, l) in an x-axis containing candidate pixels k and a y-axis containing candidate pixels l in an image block, wherein V is a binarization set representing all candidate pixel points in the image block;
Figure FDA0003661331070000042
the method comprises the steps of obtaining a denoising estimation value of a difference pixel i for binarization replacement denoising of an image block;
Figure FDA0003661331070000043
the method comprises the steps of obtaining a denoising estimation value of a difference pixel j for binarization replacement denoising of an image block;
Figure FDA0003661331070000044
the candidate pixel k is a selected estimation value of the candidate pixel k for carrying out binarization replacement denoising on the image block;
Figure FDA0003661331070000045
the candidate pixel l is selected to be an estimated value of the candidate pixel l for carrying out binarization replacement denoising on the image block; a. the find Performing improved non-local mean calculation for the sum of weight reconstruction by adopting difference pixel point binarization and weight reconstruction of candidate pixel points;
s3-3, calculating the signal to noise ratio of the difference pixel points, and if the candidate pixel points used in the image blocks exceed the signal to noise ratio of the difference pixel points of the image blocks, calculating the signal to noise ratio of the repaired and denoised image blocks, wherein the larger the value is, the better the denoising effect is.
CN202210578415.1A 2022-05-25 2022-05-25 Image noise reduction processing working method based on noise estimation Pending CN114926360A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210578415.1A CN114926360A (en) 2022-05-25 2022-05-25 Image noise reduction processing working method based on noise estimation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210578415.1A CN114926360A (en) 2022-05-25 2022-05-25 Image noise reduction processing working method based on noise estimation

Publications (1)

Publication Number Publication Date
CN114926360A true CN114926360A (en) 2022-08-19

Family

ID=82809740

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210578415.1A Pending CN114926360A (en) 2022-05-25 2022-05-25 Image noise reduction processing working method based on noise estimation

Country Status (1)

Country Link
CN (1) CN114926360A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116630447A (en) * 2023-07-24 2023-08-22 成都海风锐智科技有限责任公司 Weather prediction method based on image processing

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116630447A (en) * 2023-07-24 2023-08-22 成都海风锐智科技有限责任公司 Weather prediction method based on image processing
CN116630447B (en) * 2023-07-24 2023-10-20 成都海风锐智科技有限责任公司 Weather prediction method based on image processing

Similar Documents

Publication Publication Date Title
CN110490914B (en) Image fusion method based on brightness self-adaption and significance detection
CN111383209B (en) Unsupervised flaw detection method based on full convolution self-encoder network
CN109580630B (en) Visual inspection method for defects of mechanical parts
CN109191459B (en) Automatic identification and rating method for continuous casting billet macrostructure center segregation defect
CN107507208B (en) Image feature point extraction method based on curvature estimation on contour
CN109658381B (en) Method for detecting copper surface defects of flexible IC packaging substrate based on super-pixels
CN109064418B (en) Non-local mean value-based non-uniform noise image denoising method
CN107301661A (en) High-resolution remote sensing image method for registering based on edge point feature
CN114782432B (en) Edge detection method of improved canny operator based on textural features
WO2021109697A1 (en) Character segmentation method and apparatus, and computer-readable storage medium
CN109636784A (en) Saliency object detection method based on maximum neighborhood and super-pixel segmentation
CN116758059A (en) Visual nondestructive testing method for roadbed and pavement
CN110598613B (en) Expressway agglomerate fog monitoring method
CN114118144A (en) Anti-interference accurate aerial remote sensing image shadow detection method
CN116703251B (en) Rubber ring production quality detection method based on artificial intelligence
CN111738931B (en) Shadow removal algorithm for aerial image of photovoltaic array unmanned aerial vehicle
CN109211918B (en) Fabric bow weft detection method based on weft trend
CN115100206B (en) Printing defect identification method for textile with periodic pattern
CN111462027B (en) Multi-focus image fusion method based on multi-scale gradient and matting
CN109359604B (en) Method for identifying instrument under shadow interference facing inspection robot
CN104200434B (en) Non-local mean image denoising method based on noise variance estimation
CN111724376B (en) Paper disease detection method based on texture feature analysis
CN111753749A (en) Lane line detection method based on feature matching
CN108399614B (en) Fabric defect detection method based on non-sampling wavelet and Gumbel distribution
CN117094914A (en) Smart city road monitoring system based on computer vision

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