CN116385307A - Picture information filtering effect identification system - Google Patents
Picture information filtering effect identification system Download PDFInfo
- Publication number
- CN116385307A CN116385307A CN202310385998.0A CN202310385998A CN116385307A CN 116385307 A CN116385307 A CN 116385307A CN 202310385998 A CN202310385998 A CN 202310385998A CN 116385307 A CN116385307 A CN 116385307A
- Authority
- CN
- China
- Prior art keywords
- picture
- filtering
- denoised
- value
- current
- 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.)
- Granted
Links
- 238000001914 filtration Methods 0.000 title claims abstract description 163
- 230000000694 effects Effects 0.000 title claims abstract description 44
- 238000013528 artificial neural network Methods 0.000 claims abstract description 36
- 230000007246 mechanism Effects 0.000 claims abstract description 31
- 238000007405 data analysis Methods 0.000 claims abstract description 11
- 230000000875 corresponding effect Effects 0.000 claims description 40
- 238000000034 method Methods 0.000 claims description 27
- 238000013527 convolutional neural network Methods 0.000 claims description 8
- 238000004458 analytical method Methods 0.000 claims description 6
- 238000012163 sequencing technique Methods 0.000 claims description 5
- 238000011156 evaluation Methods 0.000 claims description 4
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 230000002596 correlated effect Effects 0.000 claims description 2
- 238000012545 processing Methods 0.000 abstract description 9
- 238000010586 diagram Methods 0.000 description 6
- 230000010287 polarization Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 239000013598 vector Substances 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
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
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- 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
-
- 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
- G06T2207/20084—Artificial neural networks [ANN]
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention relates to a picture information filtering effect identification system, which comprises: a mode selection mechanism for selecting a filtering mode from various filtering modes for a current picture to be denoised as a current filtering mode; the network building device builds a corresponding convolution neural network for executing denoising data analysis for the current filtering mode to obtain a filtering effect of a picture after filtering processing is executed by adopting the current filtering mode; and the filtering identification device uses the network for the current picture to be denoised so as to obtain a filtered effect corresponding to the current picture to be denoised. The picture information filtering effect identification system has compact logic and intelligent operation. Because various corresponding filtering effect prediction models can be established for various filtering modes, when a new frame of picture arrives, various filtering effect prediction models are adopted for the frame of picture to analyze the filtering effect respectively, and therefore, a filtering scheme with optimized quality can be selected for each frame of picture without actual filtering processing.
Description
Technical Field
The invention relates to the field of image processing, in particular to a picture information filtering effect identification system.
Background
The image filtering process is a branch of the image enhancement process. The aim of image enhancement is to improve the quality of the picture, such as increasing contrast, removing blur and noise, correcting geometric distortion, etc.; image restoration is a technique that attempts to estimate the original image assuming a model of known blur or noise.
Image enhancement can be classified into a frequency domain method and a spatial domain method according to the method used. The former regards the image as a two-dimensional signal, which is subjected to a two-dimensional fourier transform based signal enhancement. The noise in the graph can be removed by adopting a low-pass filtering (namely only passing low-frequency signals) method; by adopting the high-pass filtering method, high-frequency signals such as edges and the like can be enhanced, so that a blurred picture becomes clear. Representative spatial domain algorithms are local averaging and median filtering (taking intermediate pixel values in local neighborhoods) and the like, which can be used to remove or attenuate noise.
However, in practical use, the image filtering modes are various, including but not limited to median filtering modes (for example, the invention of application publication No. CN115205156a discloses a "distortion-free median filtering boundary filling method and apparatus, electronic device, storage medium", including acquiring a preset dimension image of a preset pixel point; selecting a median filter window with a specified size, sliding the median filter window through a preset dimensional image, wherein the number of pixel points in the median filter window is odd, selecting the pixel points which are overlapped in the median filter window and the preset dimensional image as effective pixel points when the filter window slides to the boundary of the preset dimensional image, wherein the non-overlapped pixel points are used as ineffective pixel points, marking the number of the ineffective pixel points, filling the preset pixel points into an effective pixel point sequence, replacing the ineffective pixel points to obtain a corresponding target filter window, wherein the number of the preset pixel points is the number of the ineffective pixel points, performing median filter processing on the target filter window after boundary filling), adaptively recursion filter mode, statistical ordering filter mode and high-pass filter mode (for example, the invention of application publication number CN112326033A discloses a method for demodulating polarized image high-frequency information by using high-pass filter, and the method comprises the steps of A1, performing space modulation polarization imaging on incident light to obtain an interference image containing polarization information, A2, performing transformation on the image to obtain polarization information represented by Stokes vectors in a frequency domain, performing median filter processing on the target filter window after boundary filling, performing subsequent filter processing on the rest position after Stokes vector processing on the rest position after Stokes is required to obtain high-frequency vectors A low-pass filtering mode (for example, the invention with application publication number CN112949669A discloses a method for estimating Gaussian low-pass filtering parameters in a digital image, which comprises the following steps of 1) carrying out gray level conversion on image information; 2) Performing Gaussian low-pass filtering treatment to obtain a training set; 3) Constructing a convolutional neural network; 4) Optimizing network advanced parameters; 5) Training a convolutional neural network; 6) Classification based on softmax; 7) The gaussian low pass filter parameters are estimated. ) A band pass filtering mode and a gradient sharpening filtering mode. Therefore, when a frame of picture arrives, which filtering mode is selected for the frame of picture can improve the signal to noise ratio of the frame of picture to the greatest extent, and obtain the best quality of picture, which is one of the key problems of people. If all the filtering algorithms try one by one on the frame picture, and quality comparison is performed on each obtained filtering picture, a great deal of time cost and operation cost are required for the task.
Disclosure of Invention
In order to overcome the technical problems in the prior art, the invention provides a picture information filtering effect identification system, which can establish various corresponding filtering effect prediction models for various filtering modes so as to respectively analyze the filtering effect of a new picture frame by adopting various filtering effect prediction models when the new picture frame arrives, thereby selecting a filtering scheme with optimized quality for each picture frame without actual filtering processing and avoiding the complicated and loaded image signal operation and comparison process.
According to an aspect of the present invention, there is provided a picture information filtering effect authentication system, the system including:
the picture analysis mechanism is used for obtaining various picture information of a current picture to be denoised, wherein the various picture information of the current picture to be denoised comprises picture resolution, signal-to-noise ratio before denoising, main noise quantity and picture contrast of the current picture to be denoised;
the content capturing mechanism is connected with the picture analyzing mechanism and is used for acquiring the integral gradient value of the current picture to be denoised;
a mode selection mechanism for selecting a filtering mode from various pre-stored filtering modes for the current picture to be denoised as a current filtering mode;
the network establishing device is connected with the mode selecting mechanism and is used for establishing a corresponding convolution neural network for executing denoising data analysis for the current filtering mode and outputting the convolution neural network as a target neural network, wherein the target neural network takes picture resolution, signal-to-noise ratio before denoising, main noise quantity, picture contrast and integral gradient value of a picture as various input data, and takes the signal-to-noise ratio after denoising after filtering the picture by adopting the current filtering mode as single output data;
the filtering identification device is connected with the network establishing device and is used for acquiring a signal-to-noise ratio after denoising, which is performed by adopting a current filtering mode on a current picture to be denoised, by adopting the target neural network;
wherein selecting a filter mode for the current picture to be denoised from among various filter modes stored in advance as a current filter mode includes: the pre-stored various filtering modes comprise a median filtering mode, an adaptive recursive filtering mode, a statistical ordering filtering mode, a band-pass filtering mode and a gradient sharpening filtering mode;
the method for obtaining the image information of the current image to be denoised comprises the steps of: the main noise quantity is the total number of noise types of which the corresponding noise amplitude in the current picture to be denoised is greater than or equal to a set amplitude threshold value;
the step of obtaining the overall gradient value of the current picture to be denoised comprises the following steps: and acquiring the pixel value gradient of each pixel point in the current picture to be denoised, and taking the intermediate value of each pixel value gradient corresponding to each pixel point in the current picture to be denoised as the integral gradient value of the current picture to be denoised.
The picture information filtering effect identification system has compact logic and intelligent operation. Because various corresponding filtering effect prediction models can be established for various filtering modes, when a new frame of picture arrives, various filtering effect prediction models are adopted for the frame of picture to analyze the filtering effect respectively, and therefore, a filtering scheme with optimized quality can be selected for each frame of picture without actual filtering processing.
Drawings
Embodiments of the present invention will be described below with reference to the accompanying drawings, in which:
fig. 1 is a block diagram showing the structure of a screen information filtering effect authentication system according to an embodiment of the present invention.
Fig. 2 is a block diagram showing the structure of a screen information filtering effect authentication system according to the B embodiment of the present invention.
Fig. 3 is a block diagram showing the structure of a screen information filtering effect authentication system according to the C embodiment of the present invention.
Detailed Description
An embodiment of the screen information filtering effect authentication system of the present invention will be described in detail below with reference to the accompanying drawings.
Embodiment A
Fig. 1 is a block diagram showing a structure of a screen information filtering effect authentication system according to an embodiment of the present invention, the system including:
the picture analysis mechanism is used for obtaining various picture information of a current picture to be denoised, wherein the various picture information of the current picture to be denoised comprises picture resolution, signal-to-noise ratio before denoising, main noise quantity and picture contrast of the current picture to be denoised;
the content capturing mechanism is connected with the picture analyzing mechanism and is used for acquiring the integral gradient value of the current picture to be denoised;
a mode selection mechanism for selecting a filtering mode from various pre-stored filtering modes for the current picture to be denoised as a current filtering mode;
the network establishing device is connected with the mode selecting mechanism and is used for establishing a corresponding convolution neural network for executing denoising data analysis for the current filtering mode and outputting the convolution neural network as a target neural network, wherein the target neural network takes picture resolution, signal-to-noise ratio before denoising, main noise quantity, picture contrast and integral gradient value of a picture as various input data, and takes the signal-to-noise ratio after denoising after filtering the picture by adopting the current filtering mode as single output data;
the filtering identification device is connected with the network establishing device and is used for acquiring a signal-to-noise ratio after denoising, which is performed by adopting a current filtering mode on a current picture to be denoised, by adopting the target neural network;
wherein selecting a filter mode for the current picture to be denoised from among various filter modes stored in advance as a current filter mode includes: the pre-stored various filtering modes comprise a median filtering mode, an adaptive recursive filtering mode, a statistical ordering filtering mode, a high-pass filtering mode, a low-pass filtering mode, a band-pass filtering mode and a gradient sharpening filtering mode;
the method for obtaining the image information of the current image to be denoised comprises the steps of: the main noise quantity is the total number of noise types of which the corresponding noise amplitude in the current picture to be denoised is greater than or equal to a set amplitude threshold value;
the step of obtaining the overall gradient value of the current picture to be denoised comprises the following steps: and acquiring the pixel value gradient of each pixel point in the current picture to be denoised, and taking the intermediate value of each pixel value gradient corresponding to each pixel point in the current picture to be denoised as the integral gradient value of the current picture to be denoised.
B embodiment
Fig. 2 is a block diagram showing a structure of a screen information filtering effect authentication system according to an embodiment of the present invention, including:
the picture analysis mechanism is used for obtaining various picture information of a current picture to be denoised, wherein the various picture information of the current picture to be denoised comprises picture resolution, signal-to-noise ratio before denoising, main noise quantity and picture contrast of the current picture to be denoised;
the content capturing mechanism is connected with the picture analyzing mechanism and is used for acquiring the integral gradient value of the current picture to be denoised;
a mode selection mechanism for selecting a filtering mode from various pre-stored filtering modes for the current picture to be denoised as a current filtering mode;
the network establishing device is connected with the mode selecting mechanism and is used for establishing a corresponding convolution neural network for executing denoising data analysis for the current filtering mode and outputting the convolution neural network as a target neural network, wherein the target neural network takes picture resolution, signal-to-noise ratio before denoising, main noise quantity, picture contrast and integral gradient value of a picture as various input data, and takes the signal-to-noise ratio after denoising after filtering the picture by adopting the current filtering mode as single output data;
the filtering identification device is connected with the network establishing device and is used for acquiring a signal-to-noise ratio after denoising, which is performed by adopting a current filtering mode on a current picture to be denoised, by adopting the target neural network;
and the mode storage device is connected with the mode selection mechanism and is used for storing various filtering modes in advance for selection by the mode selection mechanism.
C embodiment
Fig. 3 is a block diagram showing a structure of a screen information filtering effect authentication system according to an embodiment of the present invention, including:
the picture analysis mechanism is used for obtaining various picture information of a current picture to be denoised, wherein the various picture information of the current picture to be denoised comprises picture resolution, signal-to-noise ratio before denoising, main noise quantity and picture contrast of the current picture to be denoised;
the content capturing mechanism is connected with the picture analyzing mechanism and is used for acquiring the integral gradient value of the current picture to be denoised;
a mode selection mechanism for selecting a filtering mode from various pre-stored filtering modes for the current picture to be denoised as a current filtering mode;
the network establishing device is connected with the mode selecting mechanism and is used for establishing a corresponding convolution neural network for executing denoising data analysis for the current filtering mode and outputting the convolution neural network as a target neural network, wherein the target neural network takes picture resolution, signal-to-noise ratio before denoising, main noise quantity, picture contrast and integral gradient value of a picture as various input data, and takes the signal-to-noise ratio after denoising after filtering the picture by adopting the current filtering mode as single output data;
the filtering identification device is connected with the network establishing device and is used for acquiring a signal-to-noise ratio after denoising, which is performed by adopting a current filtering mode on a current picture to be denoised, by adopting the target neural network;
and the network learning device is connected with the filtering identification device and is used for carrying out repeated learning on the convolutional neural network which is established for the current filtering mode and is used for carrying out denoising data analysis, and then the convolutional neural network after the repeated learning is sent to the filtering identification device for use.
Next, a specific configuration of the screen information filtering effect identification system of the present invention will be further described.
In the picture information filtering effect authentication system according to various embodiments of the present invention:
after multiple learning, the convolutional neural network for performing denoising data analysis corresponding to the current filtering mode comprises: the number of times of learning is monotonically and positively correlated with the complexity of the filtering algorithm corresponding to the current filtering mode.
In the picture information filtering effect authentication system according to various embodiments of the present invention:
the step of obtaining the signal-to-noise ratio after denoising after filtering the current picture to be denoised by adopting the current filtering mode by using the target neural network comprises the following steps: and taking the picture resolution, the signal-to-noise ratio before denoising, the main noise quantity, the picture contrast and the overall gradient value of the current picture to be denoised as various input data of the target neural network, and operating the target neural network to obtain the signal-to-noise ratio after denoising of the current picture to be denoised after filtering is performed by adopting the current filtering mode.
In the picture information filtering effect authentication system according to various embodiments of the present invention:
the method for obtaining the pixel value gradient of each pixel point in the current picture to be denoised comprises the following steps of: sequencing the pixel value gradients corresponding to the pixel points in the current picture to be denoised in sequence from small to large in value, and taking the pixel value gradient corresponding to the central sequence number as the whole gradient value of the current picture to be denoised;
the method for sorting the pixel value gradients corresponding to the pixel points in the current picture to be denoised in order from small to large in value, taking the pixel value gradient corresponding to the central sequence number as the whole gradient value of the current picture to be denoised comprises the following steps: when the sequence obtained by sequencing is an even sequence, taking the average value of the gradients of the two pixel values corresponding to the two serial numbers at the central position as the integral gradient value of the current picture to be denoised.
And in a picture information filtering effect authentication system according to various embodiments of the present invention:
the step of obtaining the overall gradient value of the current picture to be denoised comprises the following steps: the method for obtaining the pixel value gradient of each pixel point in the current picture to be denoised comprises the following steps of: determining the pixel value gradient of each pixel point in the current picture to be denoised based on the difference value between the gray value of the pixel point and the representative value of each gray value corresponding to each pixel point in the neighborhood of the pixel point;
the step of obtaining the overall gradient value of the current picture to be denoised comprises the following steps: the method for obtaining the pixel value gradient of each pixel point in the current picture to be denoised comprises the following steps of: determining a pixel value gradient of each pixel point in the current picture to be denoised based on the difference value of the arithmetic average value of the gray value of each pixel point corresponding to each pixel point in the neighborhood of the pixel point;
alternatively, obtaining the overall gradient value of the current picture to be denoised includes: the method for obtaining the pixel value gradient of each pixel point in the current picture to be denoised comprises the following steps of: and determining the pixel value gradient of each pixel point in the current picture to be denoised based on the difference value of the gray value of each pixel point and the intermediate value of each gray value corresponding to each pixel point in the neighborhood of the pixel point.
In addition, in the screen information filtering effect identification system, for each pixel point in the current to-be-denoised screen, determining the pixel value gradient based on the difference value between the gray value and the representative value of each gray value corresponding to each pixel point in the neighborhood of the pixel point comprises: for each pixel point in a current picture to be denoised, each pixel point in the neighborhood of the pixel point refers to a plurality of pixel points covered by a square pixel point window taking the pixel point window as a center in the current picture to be denoised.
The technical points and the technical advantages of the invention are as follows:
1: the method comprises the steps of obtaining various relevant information of a current picture to be denoised by adopting a customized analysis mechanism, wherein the various relevant information comprises main noise quantity and overall gradient values, the main noise quantity is the total number of noise types, the corresponding noise amplitude of which is greater than or equal to a set amplitude threshold value, in the current picture to be denoised, and obtaining the overall gradient values of the current picture to be denoised comprises the following steps: acquiring a pixel value gradient of each pixel point in the current picture to be denoised, and taking a middle value of each pixel value gradient corresponding to each pixel point in the current picture to be denoised as an integral gradient value of the current picture to be denoised, so as to provide key data for prediction of each filtering effect corresponding to the current picture to be denoised in various filtering modes;
2: the method comprises the steps of establishing a corresponding convolution neural network for executing denoising data analysis for each filtering mode and outputting the convolution neural network as a target neural network, wherein the target neural network takes picture resolution, signal-to-noise ratio before denoising, main noise quantity, picture contrast and overall gradient value of a current picture to be denoised as various input data, and takes the signal-to-noise ratio after denoising after filtering the current picture to be denoised by adopting the current filtering mode as single output data, so that reliable evaluation of the filtering effect corresponding to each filtering mode of the current picture to be denoised is realized.
It is evident that various modifications and improvements may be made to the invention by those skilled in the art. Accordingly, it is intended that the present invention encompass all modifications and improvements as fall within the scope of the appended claims and their equivalents.
Claims (10)
1. A picture information filtering effect authentication system, the system comprising:
the picture analysis mechanism is used for obtaining various picture information of a current picture to be denoised, wherein the various picture information of the current picture to be denoised comprises picture resolution, signal-to-noise ratio before denoising, main noise quantity and picture contrast of the current picture to be denoised;
the content capturing mechanism is connected with the picture analyzing mechanism and is used for acquiring the integral gradient value of the current picture to be denoised;
a mode selection mechanism for selecting a filtering mode from various pre-stored filtering modes for the current picture to be denoised as a current filtering mode;
the network establishing device is connected with the mode selecting mechanism and is used for establishing a corresponding convolution neural network for executing denoising data analysis for the current filtering mode and outputting the convolution neural network as a target neural network, wherein the target neural network takes picture resolution, signal-to-noise ratio before denoising, main noise quantity, picture contrast and integral gradient value of a picture as various input data, and takes the signal-to-noise ratio after denoising after filtering the picture by adopting the current filtering mode as single output data;
the filtering identification device is connected with the network establishing device and is used for acquiring a signal-to-noise ratio after denoising, which is performed by adopting a current filtering mode on a current picture to be denoised, by adopting the target neural network;
wherein selecting a filter mode for the current picture to be denoised from among various filter modes stored in advance as a current filter mode includes: the pre-stored various filtering modes comprise a median filtering mode, an adaptive recursive filtering mode, a statistical ordering filtering mode, a high-pass filtering mode, a low-pass filtering mode, a band-pass filtering mode and a gradient sharpening filtering mode;
the method for obtaining the image information of the current image to be denoised comprises the steps of: the main noise quantity is the total number of noise types of which the corresponding noise amplitude in the current picture to be denoised is greater than or equal to a set amplitude threshold value;
the step of obtaining the overall gradient value of the current picture to be denoised comprises the following steps: and acquiring the pixel value gradient of each pixel point in the current picture to be denoised, and taking the intermediate value of each pixel value gradient corresponding to each pixel point in the current picture to be denoised as the integral gradient value of the current picture to be denoised.
2. The picture information filtering effect evaluation system of claim 1, wherein the system further comprises:
and the mode storage device is connected with the mode selection mechanism and is used for storing various filtering modes in advance for selection by the mode selection mechanism.
3. The picture information filtering effect evaluation system of claim 1, wherein the system further comprises:
and the network learning device is connected with the filtering identification device and is used for carrying out repeated learning on the convolutional neural network which is established for the current filtering mode and is used for carrying out denoising data analysis, and then the convolutional neural network after the repeated learning is sent to the filtering identification device for use.
4. A picture information filtering effect evaluation system as claimed in claim 3, characterized in that:
after multiple learning, the convolutional neural network for performing denoising data analysis corresponding to the current filtering mode comprises: the number of times of learning is monotonically and positively correlated with the complexity of the filtering algorithm corresponding to the current filtering mode.
5. The picture information filtering effect identification system of any one of claims 1-4, wherein:
the step of obtaining the signal-to-noise ratio after denoising after filtering the current picture to be denoised by adopting the current filtering mode by using the target neural network comprises the following steps: and taking the picture resolution, the signal-to-noise ratio before denoising, the main noise quantity, the picture contrast and the overall gradient value of the current picture to be denoised as various input data of the target neural network, and operating the target neural network to obtain the signal-to-noise ratio after denoising of the current picture to be denoised after filtering is performed by adopting the current filtering mode.
6. The picture information filtering effect identification system of any one of claims 1-4, wherein:
the method for obtaining the pixel value gradient of each pixel point in the current picture to be denoised comprises the following steps of: and sequencing the pixel value gradients corresponding to the pixel points in the current picture to be denoised in sequence from small to large in value, and taking the pixel value gradient corresponding to the central sequence number as the whole gradient value of the current picture to be denoised.
7. The picture information filtering effect identification system of claim 6, wherein:
sequencing the pixel value gradients corresponding to the pixel points in the current picture to be denoised in sequence from small to large in value, and taking the pixel value gradient corresponding to the central sequence number as the whole gradient value of the current picture to be denoised comprises the following steps: when the sequence obtained by sequencing is an even sequence, taking the average value of the gradients of the two pixel values corresponding to the two serial numbers at the central position as the integral gradient value of the current picture to be denoised.
8. The picture information filtering effect identification system of any one of claims 1-4, wherein:
the step of obtaining the overall gradient value of the current picture to be denoised comprises the following steps: the method for obtaining the pixel value gradient of each pixel point in the current picture to be denoised comprises the following steps of: and determining the pixel value gradient of each pixel point in the current picture to be denoised based on the difference value between the gray value of the pixel point and the representative value of each gray value corresponding to each pixel point in the neighborhood of the pixel point.
9. The picture information filtering effect identification system of claim 8, wherein:
the step of obtaining the overall gradient value of the current picture to be denoised comprises the following steps: the method for obtaining the pixel value gradient of each pixel point in the current picture to be denoised comprises the following steps of: and determining the pixel value gradient of each pixel point in the current picture to be denoised based on the difference value of the arithmetic average value of the gray value of each pixel point corresponding to each pixel point in the neighborhood of the pixel point.
10. The picture information filtering effect identification system of claim 8, wherein:
the step of obtaining the overall gradient value of the current picture to be denoised comprises the following steps: the method for obtaining the pixel value gradient of each pixel point in the current picture to be denoised comprises the following steps of: and determining the pixel value gradient of each pixel point in the current picture to be denoised based on the difference value of the gray value of each pixel point and the intermediate value of each gray value corresponding to each pixel point in the neighborhood of the pixel point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310385998.0A CN116385307B (en) | 2023-04-11 | 2023-04-11 | Picture information filtering effect identification system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310385998.0A CN116385307B (en) | 2023-04-11 | 2023-04-11 | Picture information filtering effect identification system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116385307A true CN116385307A (en) | 2023-07-04 |
| CN116385307B CN116385307B (en) | 2024-05-03 |
Family
ID=86976651
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310385998.0A Active CN116385307B (en) | 2023-04-11 | 2023-04-11 | Picture information filtering effect identification system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116385307B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180240219A1 (en) * | 2017-02-22 | 2018-08-23 | Siemens Healthcare Gmbh | Denoising medical images by learning sparse image representations with a deep unfolding approach |
| EP3576049A2 (en) * | 2018-05-31 | 2019-12-04 | Canon Medical Systems Corporation | An apparatus for denoising an image, a method for denosing an image and a computer readable storage medium |
| WO2019228456A1 (en) * | 2018-05-31 | 2019-12-05 | 杭州海康威视数字技术股份有限公司 | Image processing method, apparatus and device, and machine-readable storage medium |
| CN112652000A (en) * | 2020-12-30 | 2021-04-13 | 南京航空航天大学 | Method for judging small-scale motion direction of image |
| CN113313641A (en) * | 2021-04-28 | 2021-08-27 | 北京理工大学 | CT image denoising method with self-adaptive median filtering |
| US20210295474A1 (en) * | 2020-03-23 | 2021-09-23 | GE Precision Healthcare LLC | Methods and system for selective removal of streak artifacts and noise from images using deep neural networks |
| CN115471709A (en) * | 2022-09-28 | 2022-12-13 | 刘鹏 | Directional signal intelligent analysis platform |
-
2023
- 2023-04-11 CN CN202310385998.0A patent/CN116385307B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180240219A1 (en) * | 2017-02-22 | 2018-08-23 | Siemens Healthcare Gmbh | Denoising medical images by learning sparse image representations with a deep unfolding approach |
| EP3576049A2 (en) * | 2018-05-31 | 2019-12-04 | Canon Medical Systems Corporation | An apparatus for denoising an image, a method for denosing an image and a computer readable storage medium |
| WO2019228456A1 (en) * | 2018-05-31 | 2019-12-05 | 杭州海康威视数字技术股份有限公司 | Image processing method, apparatus and device, and machine-readable storage medium |
| US20210295474A1 (en) * | 2020-03-23 | 2021-09-23 | GE Precision Healthcare LLC | Methods and system for selective removal of streak artifacts and noise from images using deep neural networks |
| CN112652000A (en) * | 2020-12-30 | 2021-04-13 | 南京航空航天大学 | Method for judging small-scale motion direction of image |
| CN113313641A (en) * | 2021-04-28 | 2021-08-27 | 北京理工大学 | CT image denoising method with self-adaptive median filtering |
| CN115471709A (en) * | 2022-09-28 | 2022-12-13 | 刘鹏 | Directional signal intelligent analysis platform |
Non-Patent Citations (2)
| Title |
|---|
| SI CHO, ET AL: "Gradient prior-aided CNN denoiser with separable convolution-based optimization of feature dimension", 《IEEE TRANSCATIONS ON MULTIMEDIA》, 25 June 2018 (2018-06-25), pages 484 - 493 * |
| 左峥嵘: "基于深度卷积神经网络的红外小目标检测", 《红外与毫米波学报》, 15 June 2019 (2019-06-15), pages 371 - 380 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116385307B (en) | 2024-05-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Svoboda et al. | Compression artifacts removal using convolutional neural networks | |
| George et al. | A survey on various median filtering techniques for removal of impulse noise from digital image | |
| CN108805840B (en) | Image denoising method, device, terminal and computer readable storage medium | |
| Nimisha et al. | Blur-invariant deep learning for blind-deblurring | |
| Maggioni et al. | Video denoising, deblocking, and enhancement through separable 4-D nonlocal spatiotemporal transforms | |
| WO2014070273A1 (en) | Recursive conditional means image denoising | |
| CN112862753B (en) | Noise intensity estimation method and device and electronic equipment | |
| Mosleh et al. | Explicit ringing removal in image deblurring | |
| CN104899842B (en) | The adaptive extreme value median filter method of sequence for remote line-structured light image | |
| Patil et al. | Bilateral filter for image denoising | |
| CN110111261B (en) | Adaptive balance processing method for image, electronic device and computer readable storage medium | |
| CN112233037B (en) | Image enhancement system and method based on image segmentation | |
| CN116385307B (en) | Picture information filtering effect identification system | |
| CN117911267B (en) | Back-scattered X-ray image noise reduction method based on wavelet transformation | |
| Palacios-Enriquez et al. | Sparse technique for images corrupted by mixed Gaussian-impulsive noise | |
| CN116109878B (en) | Image reproduction identification method, system, device and storage medium | |
| CN100367770C (en) | Method for removing isolated noise point in video | |
| Kerouh et al. | Wavelet-based blind blur reduction | |
| Juneja et al. | Design and development of an improved adaptive median filtering method for impulse noise detection | |
| RU2364937C1 (en) | Method and device of noise filtering in video signals | |
| RU2405200C2 (en) | Method and device for fast noise filtration in digital images | |
| CN111311610A (en) | Image segmentation method and terminal equipment | |
| Mittal et al. | Enhancement and removal of noise from medical images by wavelet transform method | |
| Sari et al. | Development of denoising method for digital image in low-light condition | |
| Pathak et al. | Comparative analysis of image denoising techniques |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20231204 Address after: Room 508, Building 21, Zhenjiang Science and Technology Industrial Park, No. 669 Jinrun Avenue, Dantu District, Zhenjiang City, Jiangsu Province, 212000 Applicant after: Zhenjiang Ronggao Imaging Technology Co.,Ltd. Address before: Baiyun District of Guangzhou City, Guangdong province 510450 Pearl River east of the village Xian Science Park No. 1 Applicant before: Ren Chengfu |
|
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20231214 Address after: Room 1017, 1st Floor, No. 17, No. 9 Liyuan Road, Huangcun, Tianhe District, Guangzhou City, Guangdong Province, 510000 Applicant after: Guangzhou Haoyi E-commerce Co.,Ltd. Address before: Room 508, Building 21, Zhenjiang Science and Technology Industrial Park, No. 669 Jinrun Avenue, Dantu District, Zhenjiang City, Jiangsu Province, 212000 Applicant before: Zhenjiang Ronggao Imaging Technology Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20240410 Address after: No. 0441, 4th Floor, Building A, Innovation Center, High tech Zone, Hengyang City, Hunan Province, 421000 Applicant after: Hengyang Xinjia Media Co.,Ltd. Country or region after: China Address before: Room 1017, 1st Floor, No. 17, No. 9 Liyuan Road, Huangcun, Tianhe District, Guangzhou City, Guangdong Province, 510000 Applicant before: Guangzhou Haoyi E-commerce Co.,Ltd. Country or region before: China |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |