CN115034972B - Image denoising method, device and equipment - Google Patents
Image denoising method, device and equipment Download PDFInfo
- Publication number
- CN115034972B CN115034972B CN202111602102.7A CN202111602102A CN115034972B CN 115034972 B CN115034972 B CN 115034972B CN 202111602102 A CN202111602102 A CN 202111602102A CN 115034972 B CN115034972 B CN 115034972B
- Authority
- CN
- China
- Prior art keywords
- image
- layer
- denoising
- processing module
- original
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000012545 processing Methods 0.000 claims abstract description 69
- 230000004927 fusion Effects 0.000 claims abstract description 16
- 230000008569 process Effects 0.000 claims abstract description 6
- 230000006870 function Effects 0.000 claims description 46
- 230000004913 activation Effects 0.000 claims description 28
- 238000012549 training Methods 0.000 claims description 23
- 238000006243 chemical reaction Methods 0.000 claims description 21
- 238000010606 normalization Methods 0.000 claims description 11
- 238000010276 construction Methods 0.000 claims description 6
- 238000010586 diagram Methods 0.000 description 7
- 238000004590 computer program Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012795 verification Methods 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G06T2207/20081—Training; Learning
-
- 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
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an image denoising method, device and equipment, belonging to the technical field of image denoising, and being capable of removing image noise points in an original image, thereby indirectly improving the accuracy of image recognition and avoiding the occurrence of image recognition failure caused by the image noise points. According to the invention, the original image is denoised by constructing the image denoising model, the complexity is low, the execution efficiency is high, and the image denoising can be rapidly carried out. The image denoising model adopts the first image processing module and the second image processing module to process an original image, then carries out noise characteristic point fusion on the image output by the first image processing module and the image output by the second image processing module to obtain a noise characteristic image, and finally removes corresponding noise points in the original image according to the noise characteristic points in the noise characteristic image, thereby obtaining a noise-free image.
Description
Technical Field
The invention belongs to the technical field of image denoising, and particularly relates to an image denoising method, device and equipment.
Background
With the development of the information age, image recognition is becoming more and more common in work as well as life. Image recognition, which refers to a technique for processing, analyzing and understanding images by a computer to recognize various different patterns of objects and objects, is a practical application of applying a deep learning algorithm. Image recognition technology at present is generally divided into face recognition and commodity recognition, and the face recognition is mainly applied to security inspection, identity verification and mobile payment; the commodity identification is mainly applied to the commodity circulation process, in particular to the field of unmanned retail such as unmanned shelves and intelligent retail cabinets.
In the prior art, when image recognition is performed, an original image is usually directly recognized, and the original image may have image noise, which causes an image recognition error, thereby reducing the accuracy of the image recognition.
Disclosure of Invention
Aiming at the defects in the prior art, the image denoising method, the image denoising device and the image denoising equipment provided by the invention solve the problems in the prior art.
In order to achieve the purpose of the invention, the invention adopts the technical scheme that: an image denoising method, comprising:
acquiring a sample data set, wherein the sample data set comprises a plurality of original images and noiseless images corresponding to the original images;
constructing an image denoising model, wherein the image denoising model is used for removing image noise of an original image;
training the image denoising model according to the sample data set to obtain the trained image denoising model;
and acquiring an original image to be denoised, and denoising the original image to be denoised by adopting the trained image denoising model to obtain an image denoising result.
Further, the acquiring the sample data set includes:
acquiring a plurality of sample images, and taking the sample images as original images, wherein the sample images contain image noise;
denoising the sample image by adopting an image denoising algorithm to obtain a noiseless image corresponding to the original image;
and constructing a sample data set by using the original image and the noiseless image corresponding to the original image.
Further, the image denoising model comprises an input layer, a first image processing module, a second image processing module, a noiseless image acquisition module and an output layer;
the input layer is respectively connected with the first image processing module, the second image processing module and the noiseless image acquisition module, the first image processing module and the second image processing module are both connected with the noiseless image acquisition module, and the noiseless image acquisition module is connected with the output layer.
Further, the first image processing module comprises a first convolution layer, a first activation function layer, N first feature conversion units and a second convolution layer which are connected in sequence;
the N first feature conversion units have the same structure and respectively comprise a third convolution layer, a first batch of normalization layers and a second activation function which are sequentially connected.
Further, the second image processing module comprises a first flipping layer, a fourth convolution layer, a third activation function layer, N second feature conversion units, a fifth convolution layer, a second flipping layer and a sixth convolution layer which are connected in sequence, the feature conversion unit of the second image processing module has the same structure as the feature conversion unit of the first image processing module, and the first flipping layer and the second flipping layer are both used for mirror flipping of an image;
the second feature conversion unit has the same structure as the first feature conversion unit.
Further, the noiseless image acquisition module comprises a fusion layer, a second batch normalization layer, a deformation convolution layer, a normalization layer, a fourth activation function layer, a seventh convolution layer and a noise removal layer which are sequentially connected, wherein the fusion layer is respectively connected with the first image processing module and the second image processing module, the noise removal layer is also respectively connected with the input layer and the output layer, and the fusion layer is used for performing noise feature point fusion on an image output by the first image processing module and an image output by the second image processing module to acquire a noise feature image; and the noise removing layer is used for removing corresponding noise points in the original image according to the noise characteristic points in the noise characteristic image to obtain a noise-free image.
Further, the activation functions adopted by the first activation function layer, the second activation function layer, the third activation function layer and the fourth activation function layer are all ReLU functions.
Further, the training the image denoising model according to the sample data set includes:
the constructed loss function L is:
where i =1,2, \ 8230;, M, M represents the total number of original images in the sample data set, y i Representing the ith original image, x i Representing a noiseless image, R (y), corresponding to the ith original image i (ii) a Theta) represents that the input original image y is in the condition that the parameter of the image denoising model is theta i Obtaining a de-noised image;
and training the image denoising model by adopting the sample data set with the minimum loss function as a target until the loss function is smaller than a set threshold value to obtain the trained image denoising model.
In a second aspect, the present application provides an image denoising apparatus, including: the device comprises an acquisition module, a construction module, a training module and a denoising module;
the acquisition module is used for acquiring a sample data set, wherein the sample data set comprises a plurality of original images and noiseless images corresponding to the original images;
the construction module is used for constructing an image denoising model, and the image denoising model is used for removing image noise of an original image;
the training module is used for training the image denoising model according to the sample data set to obtain the trained image denoising model;
the denoising module is used for acquiring an original image to be denoised, and denoising the original image to be denoised by adopting the trained image denoising model to obtain an image denoising result.
In a third aspect, the present application provides an image denoising device, including a memory and a processor, where the memory and the processor are connected to each other through a bus;
the memory stores computer-executable instructions;
the processor executes computer-executable instructions stored by the memory to cause the processor to perform the image denoising method according to the first aspect.
The invention has the beneficial effects that:
(1) The invention provides an image denoising method, device and equipment, which can remove image noise points in an original image, thereby indirectly improving the accuracy of image recognition and avoiding the occurrence of image recognition failure caused by the image noise points.
(2) According to the invention, the original image is denoised by constructing the image denoising model, the complexity is low, the execution efficiency is high, and the image denoising can be rapidly carried out.
(3) According to the method, the loss function is constructed, and the loss function is adopted to train the image denoising model, so that the image denoising model can accurately denoise the image.
(4) The image denoising model adopts the first image processing module and the second image processing module to process an original image, then carries out noise characteristic point fusion on the image output by the first image processing module and the image output by the second image processing module to obtain a noise characteristic image, and finally removes corresponding noise points in the original image according to the noise characteristic points in the noise characteristic image, thereby obtaining a noise-free image.
Drawings
Fig. 1 is a flowchart of an image denoising method according to an embodiment of the present disclosure.
Fig. 2 is a schematic structural diagram of an image denoising model provided in the embodiment of the present application.
Fig. 3 is a schematic structural diagram of a first image processing module according to an embodiment of the present disclosure.
Fig. 4 is a schematic structural diagram of a feature conversion unit according to an embodiment of the present application.
Fig. 5 is a schematic structural diagram of a second image processing module according to an embodiment of the present disclosure.
Fig. 6 is a schematic structural diagram of a noise-free image acquisition module according to an embodiment of the present application.
Fig. 7 is a schematic structural diagram of an image denoising device according to an embodiment of the present application.
Fig. 8 is a schematic structural diagram of an image denoising device according to an embodiment of the present application.
The system comprises a 21-acquisition module, a 22-construction module, a 23-training module, a 24-denoising module, an 80-image denoising device, an 81-memory, an 82-processor and an 83-bus.
With the above figures, there are shown specific embodiments of the present application, which will be described in more detail below. These drawings and written description are not intended to limit the scope of the inventive concepts in any manner, but rather to illustrate the inventive concepts to those skilled in the art by reference to specific embodiments.
Detailed Description
The following description of the embodiments of the present invention is provided to facilitate the understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and it will be apparent to those skilled in the art that various changes may be made without departing from the spirit and scope of the invention as defined and defined in the appended claims, and all matters produced by the invention using the inventive concept are protected.
Embodiments of the present invention are described in detail below with reference to the accompanying drawings.
As shown in fig. 1, the present embodiment provides an image denoising method, including:
s11, obtaining a sample data set, wherein the sample data set comprises a plurality of original images and noiseless images corresponding to the original images.
S12, constructing an image denoising model, wherein the image denoising model is used for removing image noise of the original image.
And S13, training the image denoising model according to the sample data set to obtain the trained image denoising model.
S14, obtaining an original image to be denoised, and denoising the original image to be denoised by adopting the trained image denoising model to obtain an image denoising result.
In one possible implementation, the obtaining the sample data set includes:
acquiring a plurality of sample images, and taking the sample images as original images, wherein the sample images contain image noise.
And denoising the sample image by adopting an image denoising algorithm to obtain a noiseless image corresponding to the original image.
And constructing a sample data set by using the original image and the noiseless image corresponding to the original image.
In the present embodiment, the image denoising algorithm may include a gaussian filter algorithm or a median filter algorithm.
Optionally, a noiseless image may be downloaded on the internet as a noiseless image, noise may be added to the noiseless image, a noise image corresponding to the noiseless image is obtained, and the noise image is used as an original image, so that a sample data set may be obtained according to the original image and the corresponding noiseless image.
As shown in fig. 2, the image denoising model includes an input layer, a first image processing module, a second image processing module, a noise-free image obtaining module, and an output layer; the input layer is respectively connected with the first image processing module, the second image processing module and the noiseless image acquisition module, the first image processing module and the second image processing module are both connected with the noiseless image acquisition module, and the noiseless image acquisition module is connected with the output layer.
As shown in fig. 3, the first image processing module includes a first convolution layer, a first activation function layer, N first feature conversion units, and a second convolution layer, which are sequentially connected.
As shown in fig. 4, the N first feature transformation units have the same structure, and each of the N first feature transformation units includes a third convolution layer (convolution layers), a first Normalization layer (BN), and a second activation function (next activation function) connected in sequence.
As shown in fig. 5, the second image processing module includes a first flipping layer, a fourth convolution layer, a third activation function layer, N second feature conversion units, a fifth convolution layer, a second flipping layer, and a sixth convolution layer, which are connected in sequence, the feature conversion unit of the second image processing module has the same structure as the feature conversion unit of the first image processing module, and the first flipping layer and the second flipping layer are both used for mirror flipping of an image. The second feature conversion unit has the same structure as the first feature conversion unit.
Optionally, the value range of N is set to be a positive integer greater than or equal to 2.
As shown in fig. 6, the noiseless image acquiring module includes a fusion layer, a second batch of Normalization layer, a deformation convolution layer (Deformable Normalization layer), a Normalization layer (Switchable Normalization, SN), a fourth activation function layer, a seventh convolution layer, and a noise removing layer, which are connected in sequence, where the fusion layer is connected to the first image processing module and the second image processing module, the noise removing layer is further connected to the input layer and the output layer, and the fusion layer is used to perform noise feature point fusion on an image output by the first image processing module and an image output by the second image processing module to acquire a noise feature image; the noise removing layer is used for removing corresponding noise points in the original image according to the noise characteristic points in the noise characteristic image to obtain a noise-free image.
In this embodiment, an image output by the first image processing module is a first residual image, an image output by the second image processing module is also a second residual image, the residual images output by the first image processing module and the second image processing module are fused (i.e., noise feature points are added), a noise feature map is obtained, the noise feature map at this time is a fused residual image, and residual characteristics of all noise feature points in the fused residual image are respectively superimposed on corresponding feature points in the original image, so that noise in the original image is removed, and a noise-free image is obtained.
In a possible embodiment, the activation functions used by the first activation function layer, the second activation function layer, the third activation function layer, and the fourth activation function layer are all ReLU (Rectified linear unit) functions.
In a possible implementation, the training an image denoising model according to a sample data set includes:
the constructed loss function L is:
where i =1,2, \ 8230;, M, M represents the total number of original images in the sample data set, y i Representing the ith original image, x i Representing a noiseless image, R (y), corresponding to the ith original image i (ii) a Theta) represents that the input original image y is in the condition that the parameter of the image denoising model is theta i Obtaining a de-noised image;
and training the image denoising model by adopting the sample data set with the minimum loss function as a target until the loss function is smaller than a set threshold value to obtain the trained image denoising model.
Optionally, the parameter θ of the image denoising model may include a weight and a bias of each convolution layer, and in the training process, the weight and the bias of each convolution layer are updated with the minimum loss function as a target, so as to train the image denoising model.
Updating the parameter theta of the image denoising model as follows:
wherein, theta k-1 Representing the value of the parameter theta at the time of the (k-1) th training,
denotes a differential term, α denotes a learning rate, and θ k Representing the value of the parameter theta at the kth training, i.e. theta k-1 The value obtained after updating.
In this embodiment, obtaining an original image to be denoised, and denoising the original image to be denoised by using a trained image denoising model to obtain an image denoising result, including:
acquiring an original image to be denoised, and respectively inputting the original image to be denoised into a first image processing module and a second image processing module;
processing an original image to be denoised by a first image processing module to obtain a first residual image; processing the original image to be denoised by a second image processing module to obtain a second residual image; and after the denoised original image enters the second image processing module, turning over the original image, processing the original image to obtain a residual image, and turning over the residual image again to obtain a second residual image.
And fusing the first residual image and the second residual image through a noise-free image acquisition module, performing noise characteristic point fusion to acquire a noise characteristic image, and removing corresponding noise points from the original image according to the noise characteristic points in the noise characteristic image so as to acquire the noise-free image.
The method can remove the image noise in the original image, thereby indirectly improving the accuracy of image identification and avoiding the occurrence of image identification failure caused by the image noise.
According to the invention, the original image is denoised by constructing the image denoising model, the complexity is low, the execution efficiency is high, and the image denoising can be rapidly carried out. According to the method, the loss function is constructed, and the loss function is adopted to train the image denoising model, so that the image denoising model can accurately denoise the image.
The image denoising model adopts the first image processing module and the second image processing module to process an original image, then carries out noise characteristic point fusion on the image output by the first image processing module and the image output by the second image processing module to obtain a noise characteristic image, and finally removes corresponding noise points in the original image according to the noise characteristic points in the noise characteristic image, thereby obtaining a noise-free image.
Example 2
As shown in fig. 7, the present application provides an image denoising apparatus, including: the device comprises an acquisition module 21, a construction module 22, a training module 23 and a denoising module 24;
the obtaining module 21 is configured to obtain a sample data set, where the sample data set includes a plurality of original images and a noiseless image corresponding to the original images;
the building module 22 is configured to build an image denoising model, where the image denoising model is used to remove image noise of an original image;
the training module 23 is configured to train the image denoising model according to the sample data set, and obtain the trained image denoising model;
the denoising module 24 is configured to obtain an original image to be denoised, and denoise the original image to be denoised by using the trained image denoising model to obtain an image denoising result.
The image denoising device in the embodiment of fig. 7 can implement the technical solution in the embodiment 1, and the implementation principle and the beneficial effect are similar, which are not described herein again.
Example 3
As shown in fig. 8, the present application provides an image denoising apparatus, where the image denoising apparatus 80 includes a memory 81 and a processor 82, where the memory 81 and the processor 82 are connected to each other through a bus 83;
the memory 81 stores computer-executable instructions;
the processor 82 executes computer-executable instructions stored in the memory to cause the processor to perform the image denoising method as described in embodiment 1.
Example 4
The present embodiment provides a computer-readable storage medium, in which computer-executable instructions are stored, and when the computer-executable instructions are executed by a processor, the computer-readable storage medium is used for implementing the image denoising method according to embodiment 1.
Example 5
The embodiment of the present application may further provide a computer program product, which includes a computer program, and when the computer program is executed by a processor, the computer program implements the image denoising method described in embodiment 1.
Other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the application and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains.
It will be understood that the present application is not limited to the precise arrangements that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the application is limited only by the appended claims.
Claims (4)
1. An image denoising method, comprising:
acquiring a sample data set, wherein the sample data set comprises a plurality of original images and noiseless images corresponding to the original images;
constructing an image denoising model, wherein the image denoising model is used for removing image noise of an original image;
training the image denoising model according to the sample data set to obtain the trained image denoising model;
acquiring an original image to be denoised, and denoising the original image to be denoised by adopting a trained image denoising model to obtain an image denoising result;
the image denoising model comprises an input layer, a first image processing module, a second image processing module, a noiseless image acquisition module and an output layer;
the input layer is respectively connected with a first image processing module, a second image processing module and a noiseless image acquisition module, the first image processing module and the second image processing module are both connected with the noiseless image acquisition module, and the noiseless image acquisition module is connected with the output layer;
the first image processing module comprises a first convolution layer, a first activation function layer, N first feature conversion units and a second convolution layer which are sequentially connected;
the N first feature conversion units have the same structure and respectively comprise a third convolution layer, a first batch of normalization layers and a second activation function which are sequentially connected;
the second image processing module comprises a first overturning layer, a fourth convolution layer, a third activation function layer, N second feature conversion units, a fifth convolution layer, a second overturning layer and a sixth convolution layer which are sequentially connected, the feature conversion unit of the second image processing module has the same structure as the feature conversion unit of the first image processing module, and the first overturning layer and the second overturning layer are both used for mirroring and overturning images;
the second feature conversion unit has the same structure as the first feature conversion unit;
the noise-free image acquisition module comprises a fusion layer, a second batch normalization layer, a deformation convolution layer, a normalization layer, a fourth activation function layer, a seventh convolution layer and a noise removal layer which are sequentially connected, the fusion layer is respectively connected with the first image processing module and the second image processing module, the noise removal layer is also respectively connected with the input layer and the output layer, and the fusion layer is used for carrying out noise characteristic point fusion on an image output by the first image processing module and an image output by the second image processing module to acquire a noise characteristic image; the noise removal layer is used for removing corresponding noise points in the original image according to the noise characteristic points in the noise characteristic image to obtain a noise-free image;
the activation functions adopted by the first activation function layer, the second activation function layer, the third activation function layer and the fourth activation function layer are ReLU functions;
the training of the image denoising model according to the sample data set comprises the following steps:
the loss function L is constructed as:
where i =1,2, \ 8230;, M, M represents the total number of original images in the sample data set, y i Representing the ith original image, x i Representing a noiseless image, R (y), corresponding to the ith original image i (ii) a Theta) represents that the input original image y is in the condition that the parameter of the image denoising model is theta i Obtaining a de-noised image;
training the image denoising model by adopting the sample data set with the minimum loss function as a target until the loss function is smaller than a set threshold value to obtain a trained image denoising model;
the parameters theta of the image denoising model comprise the weight and the bias of each convolution layer, and in the training process, the parameters theta of the image denoising model are updated by taking the minimum loss function as a target, so that the image denoising model is trained; when the image denoising model is trained, updating a parameter theta of the image denoising model into:
wherein, theta k-1 Representing the value of the parameter theta at the time of the (k-1) th training,
denotes a differential term, α denotes a learning rate, and θ k Denotes theta k-1 The value obtained after updating.
2. The image denoising method of claim 1, wherein the acquiring a sample data set comprises:
acquiring a plurality of sample images, and taking the sample images as original images, wherein the sample images contain image noise;
denoising the sample image by adopting an image denoising algorithm to obtain a noiseless image corresponding to the original image;
and constructing a sample data set by using the original image and the noiseless image corresponding to the original image.
3. An image denoising apparatus for performing the image denoising method of claim 1, the apparatus comprising: the device comprises an acquisition module, a construction module, a training module and a denoising module;
the acquisition module is used for acquiring a sample data set, wherein the sample data set comprises a plurality of original images and noiseless images corresponding to the original images;
the construction module is used for constructing an image denoising model, and the image denoising model is used for removing image noise of an original image;
the training module is used for training the image denoising model according to the sample data set to obtain the trained image denoising model;
the denoising module is used for acquiring an original image to be denoised, and denoising the original image to be denoised by adopting the trained image denoising model to obtain an image denoising result.
4. The image denoising device is characterized by comprising a memory and a processor, wherein the memory and the processor are connected with each other through a bus;
the memory stores computer-executable instructions;
the processor executes computer-executable instructions stored in the memory to cause the processor to perform the image denoising method of any one of claims 1 through 2.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111602102.7A CN115034972B (en) | 2021-12-24 | 2021-12-24 | Image denoising method, device and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111602102.7A CN115034972B (en) | 2021-12-24 | 2021-12-24 | Image denoising method, device and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115034972A CN115034972A (en) | 2022-09-09 |
| CN115034972B true CN115034972B (en) | 2023-04-07 |
Family
ID=83118320
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111602102.7A Active CN115034972B (en) | 2021-12-24 | 2021-12-24 | Image denoising method, device and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115034972B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110930332A (en) * | 2019-11-22 | 2020-03-27 | 河北工程大学 | Artificial intelligence-based digital holographic image denoising method |
| CN113191983A (en) * | 2021-05-18 | 2021-07-30 | 陕西师范大学 | Image denoising method and device based on deep learning attention mechanism |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110211069B (en) * | 2019-06-03 | 2021-09-03 | 广东工业大学 | Image denoising model training method, system, equipment and computer medium |
| CN111861930A (en) * | 2020-07-27 | 2020-10-30 | 京东方科技集团股份有限公司 | Image denoising method and device, electronic equipment and image hyper-resolution denoising method |
| CN112801889A (en) * | 2021-01-06 | 2021-05-14 | 携程旅游网络技术（上海）有限公司 | Image denoising method, system, device and storage medium |
| CN112801909B (en) * | 2021-02-05 | 2022-06-14 | 福州大学 | Image fusion denoising method and system based on U-Net and pyramid module |
| CN113538281B (en) * | 2021-07-21 | 2023-07-11 | 深圳大学 | Image denoising method, image denoising device, computer equipment and storage medium |
-
2021
- 2021-12-24 CN CN202111602102.7A patent/CN115034972B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110930332A (en) * | 2019-11-22 | 2020-03-27 | 河北工程大学 | Artificial intelligence-based digital holographic image denoising method |
| CN113191983A (en) * | 2021-05-18 | 2021-07-30 | 陕西师范大学 | Image denoising method and device based on deep learning attention mechanism |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115034972A (en) | 2022-09-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108345827B (en) | Method, system and neural network for identifying document direction | |
| CN109492674B (en) | Generation method and device of SSD (solid State disk) framework for target detection | |
| CN110046622B (en) | Targeted attack sample generation method, device, equipment and storage medium | |
| CN110782406B (en) | Image denoising method and device based on information distillation network | |
| CN112860675B (en) | Big data processing method under online cloud service environment and cloud computing server | |
| CN110189260A (en) | A kind of image denoising method based on multiple dimensioned parallel gate neural network | |
| CN111915486A (en) | Confrontation sample defense method based on image super-resolution reconstruction | |
| CN109726195A (en) | A kind of data enhancement methods and device | |
| CN114862861B (en) | Lung lobe segmentation method and device based on few-sample learning | |
| CN114266894A (en) | Image segmentation method and device, electronic equipment and storage medium | |
| CN111275051A (en) | Character recognition method, character recognition device, computer equipment and computer-readable storage medium | |
| CN111738080A (en) | Face detection and alignment method and device | |
| CN115034972B (en) | Image denoising method, device and equipment | |
| CN112200789B (en) | Image recognition method and device, electronic equipment and storage medium | |
| CN116778164A (en) | Semantic segmentation method for improving deep V < 3+ > network based on multi-scale structure | |
| CN113723431B (en) | Image recognition method, apparatus and computer readable storage medium | |
| CN112990225B (en) | Image target identification method and device in complex environment | |
| CN114022475A (en) | Image anomaly detection and anomaly positioning method and system based on self-supervision mask | |
| CN112733670A (en) | Fingerprint feature extraction method and device, electronic equipment and storage medium | |
| CN114693987A (en) | Model generation method, model generation device, storage medium, face recognition method and face recognition device | |
| CN116188918B (en) | Image denoising method, training method of network model, device, medium and equipment | |
| CN116416212B (en) | Training method of road surface damage detection neural network and road surface damage detection neural network | |
| CN116645727B (en) | Behavior capturing and identifying method based on Openphase model algorithm | |
| CN113887300A (en) | Method and device for detecting target, human face and human face key point and storage medium | |
| CN113642470A (en) | Workpiece model identification method, device, equipment, storage medium and computer program |
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 |