CN112241940B - Fusion method and device for multiple multi-focus images - Google Patents
Fusion method and device for multiple multi-focus images Download PDFInfo
- Publication number
- CN112241940B CN112241940B CN202011036730.9A CN202011036730A CN112241940B CN 112241940 B CN112241940 B CN 112241940B CN 202011036730 A CN202011036730 A CN 202011036730A CN 112241940 B CN112241940 B CN 112241940B
- Authority
- CN
- China
- Prior art keywords
- image
- images
- fused
- fusion
- focus
- 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
- 238000007500 overflow downdraw method Methods 0.000 title claims abstract description 16
- 230000004927 fusion Effects 0.000 claims abstract description 75
- 238000000034 method Methods 0.000 claims abstract description 28
- 238000000605 extraction Methods 0.000 claims abstract description 18
- 238000012937 correction Methods 0.000 claims abstract description 11
- 238000010586 diagram Methods 0.000 claims description 30
- 238000013527 convolutional neural network Methods 0.000 claims description 7
- 238000012706 support-vector machine Methods 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 3
- 239000000654 additive Substances 0.000 claims description 2
- 230000000996 additive effect Effects 0.000 claims description 2
- 238000012545 processing Methods 0.000 abstract description 6
- 238000013473 artificial intelligence Methods 0.000 abstract description 3
- 230000015654 memory Effects 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000007792 addition Methods 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a fusion method of a plurality of multi-focus images, and belongs to the technical field of image processing and artificial intelligence. Comprising the following steps: extracting the characteristics of all images in the image set to be fused by using an image characteristic extraction algorithm, and selecting the characteristics of any two images as a first-stage baseline characteristic and a second-stage baseline characteristic; respectively carrying out feature fusion on the first-stage baseline features and the features of the rest images in the image set to be fused by adopting an image feature fusion algorithm, and forming a plurality of focusing level images; correcting the rest focus level graphs based on the focus level graphs formed by the first-stage baseline characteristic and the second-stage baseline characteristic by adopting a correction algorithm, and splicing the corrected focus level graphs into a focus level set; converting the focusing level set into a decision graph by adopting a decision algorithm; and fusing all the images in the image set to be fused into a final single Zhang Rongge result based on the decision graph by adopting an image pixel fusion algorithm. By adopting the method and the device, the fusion efficiency of a plurality of multi-focus images can be improved.
Description
Technical Field
The invention relates to the technical field of image processing and artificial intelligence, in particular to a method and a device for fusing multiple multi-focus images.
Background
The multi-focus image fusion is an important research branch in the fields of image analysis and fusion, and plays an important role in the fields of scientific research, military, medical treatment, digital camera shooting, microstructure analysis and the like. Due to the inherent characteristics of the optical sensor, a single shooting can only ensure that a target in a certain range in front of and behind a focusing area presents a clear image, and a non-focusing area presents a blurred image, so that objects with large depth distances can not be fully focused in one lens through physical operation. Therefore, a multi-focus image fusion method based on image processing is often adopted to finally obtain a fully focused image by fusing clear areas in each of a plurality of images.
With the breakthrough progress of artificial intelligence theory and computer vision technology in the field of image processing, deep learning gradually becomes a mainstream method in the field of multi-focus image fusion, most methods adopt feature extraction branches to extract high-dimensional features of images respectively, and then a feature fusion module is used for fusing the high-dimensional features of each image and outputting a final fusion result.
However, most of the current multi-focus image processing methods generally only focus on the fusion scene of two images, and use an intuitive two-by-two fusion strategy to fuse multiple images, i.e. fuse the 1 st image with the 2 nd image to obtain a fusion result, then fuse the fusion result with the 3 rd image, and so on. In practical applications, it is usually required to face a fusion scene of tens of images, and the image processing efficiency is seriously reduced by using the above-mentioned image processing method.
Therefore, a method for effectively improving the efficiency of multi-image fusion is urgently needed in the field of multi-focus image fusion.
Disclosure of Invention
The invention provides a method and a device for fusing a plurality of multi-focus images, which can improve the efficiency of fusing the plurality of multi-focus images. The technical scheme is as follows:
in one aspect, a method for fusing multiple multi-focus images is provided, where the method is applied to an electronic device, and includes:
extracting the characteristics of all images in an image set to be fused by adopting an image characteristic extraction algorithm, and selecting the characteristics of any two images in the image set to be fused as a first-stage baseline characteristic and a second-stage baseline characteristic;
respectively carrying out feature fusion on the first-stage baseline features and the features of the rest images in the image set to be fused by adopting an image feature fusion algorithm, and forming a plurality of focusing level diagrams;
correcting the rest focus level graphs by adopting a correction algorithm based on the focus level graphs formed by the first-stage baseline characteristic and the second-stage baseline characteristic, and splicing the corrected focus level graphs into a focus level set;
converting the focusing level set into a decision graph by adopting a decision algorithm;
and fusing all images in the image set to be fused into a final single Zhang Rongge result based on the decision graph by adopting an image pixel fusion algorithm.
Optionally, the image feature extraction algorithm includes: one or more of spatial frequency operators, gradient operators, convolutional neural networks, support vector machines.
Optionally, the image set to be fused is at least two registered images with different focusing areas shot for the same scene, and all images in the image set to be fused are the same in h×w, where H is the number of column pixels and W is the number of row pixels.
Optionally, the image feature fusion algorithm includes: one or more of spatial frequency operators, gradient operators, convolutional neural networks, support vector machines, additive fusion, maximum value fusion, and channel dimension stitching; the focus level diagram is the same as H×W of the image to be fused, where H is the number of column pixels and W is the number of row pixels.
Optionally, the correction algorithm has the following expression:
wherein p is i j Represents the probability that the 1 st image is clearer than the j-th image at the position of element i, and p i j ∈[0,1]；Is the probability that the 1 st image is clearer at the corrected element i relative to the j-th image.
Optionally, the stitching the corrected plurality of focus level maps into a focus level set includes: and splicing the corrected plurality of focusing level graphs into a focusing level set according to the channel direction, wherein the focusing level set refers to a three-dimensional matrix with the size of H multiplied by W multiplied by N, H is the number of column pixels, W is the number of row pixels, N is the number of images, and the value of N is larger than 1.
Optionally, the decision algorithm includes:
index Img for maximum value in channel direction of each element i in the focus level set i k Wherein the index Img i k Recording the clearest image sequence number k at the element i;
the index of each pixel is assembled into a decision graph.
Optionally, the image pixel fusion algorithm includes: and (5) selecting one or more of an index value algorithm, a weighted fusion algorithm and a guided filtering algorithm.
Optionally, the indexing algorithm refers to indexing the value Img according to the decision chart i j The pixels of the image j in the element i are assigned to the pixels of the final individual fusion result position i.
In one aspect, there is provided a multi-sheet multi-focus image fusion apparatus, the apparatus being applied to an electronic device, the apparatus comprising:
the feature extraction unit is used for extracting features of all images in the image set to be fused, and selecting features of any two images in the image set to be fused as a first-stage baseline feature and a second-stage baseline feature;
the feature fusion unit is used for respectively carrying out feature fusion on the first-stage baseline features and the features of the rest images in the image set to be fused, and forming a plurality of focusing level diagrams;
a correction unit for correcting the rest focus level graphs based on the focus level graphs formed by the first-stage baseline characteristic and the second-stage baseline characteristic, and splicing the corrected focus level graphs into a focus level set;
a decision unit for converting the focus level set into a decision graph;
and the pixel fusion unit fuses all the images in the image set to be fused into a final single Zhang Rongge result based on the decision graph.
The technical scheme provided by the invention has the beneficial effects that at least:
according to the invention, a few image baseline characteristics are extracted to be fused into a focusing level diagram, then the focusing level diagram is used for correcting other focusing level diagrams, the corrected focusing level diagram is formed into a focusing level set, then the focusing level set is converted into a decision diagram, and the fusion of a plurality of multi-focusing images is carried out according to the method of fusing all images, so that redundant characteristic extraction operation of the existing method (the workload of the characteristic extraction operation in the existing method is reduced by 50%) is avoided, the problem that the fusion efficiency of the multi-image multi-focusing images is lower under the current pairwise fusion strategy is solved, and the advantages are obvious under the application scene that the fusion of a large number of images and the requirement on the fusion speed of the images are higher.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a diagram of an implementation environment provided by an embodiment of the present invention;
fig. 2 is a flow chart of a method for fusing multiple multi-focus images according to an embodiment of the present invention;
FIG. 3 is a schematic view of a plurality of multi-focus images provided by an embodiment of the present invention;
FIG. 4 is a flow chart of a method for fusing multiple multi-focus images according to an embodiment of the present invention compared with a conventional two-by-two fusion method;
FIG. 5 is a schematic diagram of a multi-sheet multi-focus image fusion apparatus according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the embodiments of the present invention will be described in further detail with reference to the accompanying drawings.
First embodiment
A first embodiment of the present invention provides a method for fusing multiple multi-focus images, and fig. 1 is a schematic diagram of an implementation environment provided by the first embodiment of the present invention. The implementation environment may include at least one terminal 101, and a server 102 for providing services to the plurality of terminals 101. At least one terminal 101 is connected to the server 102 through a wireless or wired network, and the plurality of terminals 101 may be computer devices or intelligent terminals or the like capable of accessing the server 102. The terminal 101 may be provided with an image feature value extraction program, an image feature value fusion program, a correction program, a decision program, an image pixel fusion program, and other computing-related application programs. When a user wants to generate a fusion image for a certain image set to be fused, the image set to be fused can be selected in a local storage area of the terminal, the image set to be fused can be acquired in real time through the terminal, and the image set to be fused sent by the server can be received through the terminal. Server 102 may also provide the set of images to be fused for the application described above.
In addition, the terminal 101 may also serve as a demander to send the image set to be fused to the server 102, and request the server 102 to generate a fused image for the image set to be fused. In this case, the server 102 may further have at least one database for storing an image feature value extraction program, an image feature value fusion program, a correction program, a decision program, an image pixel fusion program, and the like. The server 102 may be a single terminal or a group of terminals, and when the server 102 is a group of terminals, each terminal may share and each provide a portion of the image set to be fused, etc.
The first embodiment of the present invention provides a method for fusing multiple multi-focus images, which may be implemented by an electronic device, which may be a terminal or a server. As shown in the flowchart of the multi-sheet multi-focus image fusion method in fig. 2, the process flow of the method may include the following steps:
in step 201, an image feature extraction algorithm is adopted, including one or more of algorithms such as a spatial frequency operator, a gradient operator, a convolutional neural network, a support vector machine and the like, to extract features of all images in an image set to be fused, and features of any two images in the image set to be fused are selected as a first-stage baseline feature and a second-stage baseline feature. Preferably, a convolutional neural network algorithm is selected to extract high-dimensional features in the image.
As shown in fig. 3, in step 201, the set of images to be fused mainly refers to at least two registered images with different focusing areas, which are shot for the same scene, where H is the number of column pixels and W is the number of row pixels, and all images in the set of images to be fused are the same.
In step 202, an image feature fusion algorithm is adopted, including one or more of spatial frequency operators, gradient operators, convolutional neural networks, support vector machines, addition fusion, maximum value fusion and channel dimension stitching, and feature fusion is performed on the first-stage baseline features and features of other images in the image set to be fused, so as to form a plurality of focus level graphs. The focusing level diagram is the same as H×W of the image to be fused, wherein H is the number of column pixels, W is the number of row pixels, each element in the focusing level diagram corresponds to each pixel in the image to be fused, and p is i j Representing the probability that the first image is sharper at the position of element i relative to the j-th image, and p i j ∈[0,1]。
In the present embodiment, since the focus level map of the 1 st image and the j-th image are fused reflects the sharpness information of the 1 st image compared with the j-th image, if the 1 st image is sharp compared with the j-th image in the vicinity of the pixel i, p i j The closer to 1, otherwise p i j The closer to 0. Further, p i j The larger the value of (c) is, the greater the sharpness of the 1 st image compared to the j-th image in the vicinity of the pixel i is. Therefore, the focusing level diagrams of the 1 st image and the 2 nd image can be used as the reference to correct the focusing level diagrams of the different images to be fused and the 1 st image according to the mode, and the focusing level diagram with unified standard is obtained for the subsequent clear region judgment of the different images.
In step 203, a correction algorithm is used to correct the remaining focus level maps based on the focus level maps formed by the first level baseline feature and the second level baseline feature, and the corrected focus level maps are combined into a focus level set.
Preferably, the expression of the correction algorithm is as follows:
wherein p is i j Represents the probability that the 1 st image is clearer than the j-th image at the position of element i, and p i j ∈[0,1]；Is the probability that the 1 st image is clearer at the corrected element i relative to the j-th image.
In step 203, the stitching the corrected multiple focus level maps into a focus level set may be stitching the corrected multiple focus level maps into a focus level set according to a channel direction, where the focus level set refers to a three-dimensional matrix with a size of h×w×n, where H is the number of column pixels, W is the number of row pixels, N is the number of images, and the value of N is greater than 1.
In step 204, a decision algorithm is used to convert the focus level set into a decision graph.
The decision algorithm in step 204 may preferably include the steps of:
s2041, index Img for maximum value of each element i in the focusing level set in the channel direction i k Wherein the index Img i k Recording the clearest image sequence number k at the element i;
s2042, the index of each pixel is assembled into a decision graph.
In step 205, an image pixel fusion algorithm is employed, comprising: and fusing all images in the image set to be fused into a final single Zhang Rongge result based on a decision diagram according to one or more of an index value algorithm, a weighted fusion algorithm, a guided filtering algorithm and the like.
Preferably, the indexing algorithm in step 205 refers to indexing the value Img in the decision graph i j The pixels of the image j in the element i are assigned to the pixels of the final individual fusion result position i.
Fig. 4 is a flow chart comparing a multi-sheet multi-focus image fusion method according to an embodiment of the present invention with a conventional two-by-two fusion method.
The traditional two-by-two fusion strategy fuses a plurality of images to be fused one by one in series, namely, firstly fuses the 1 st image and the 2 nd image, and uses an image feature fusion algorithm to obtain a fusion result. And further fusing the fusion result with the 3 rd image, and so on. The use of the pairwise fusion method requires the use of 2 (N-1) times the image feature extraction algorithm. The method for fusing the multiple multi-focus images provided by the embodiment of the invention has the advantages that the characteristic of the 1 st image is pre-stored as the baseline characteristic, so that only N times of image characteristic extraction algorithms are needed. Therefore, compared with the traditional pairwise fusion method, the method can reduce the operation times of feature extraction by 50%, thereby avoiding redundant feature extraction operation and improving the operation efficiency of fusion of a plurality of multi-focus images.
Second embodiment
As shown in fig. 5, a second embodiment of the present invention provides a multi-sheet multi-focus image fusion apparatus, including:
the feature extraction unit 501 is configured to extract features of all images in the image set to be fused, and select features of any two images in the image set to be fused as a first-stage baseline feature and a second-stage baseline feature;
the feature fusion unit 502 is configured to perform feature fusion on the first-level baseline feature and features of the rest of the images in the image set to be fused, and form a plurality of focus level graphs;
a correction unit 503 that corrects the remaining focus level charts based on the focus level charts formed by the first-stage baseline feature and the second-stage baseline feature, and that merges the corrected plurality of focus level charts into a focus level set;
a decision unit 504 for converting the focus level set into a decision graph;
the pixel fusion unit 505 fuses all the images in the image set to be fused into a final single Zhang Rongge result based on the decision graph.
The multi-sheet multi-focus image fusion apparatus of the present embodiment corresponds to the multi-sheet multi-focus image fusion method of the first embodiment; the functions realized by the units in the device in this embodiment are in one-to-one correspondence with the flow steps in the multi-sheet multi-focus image fusion method in the first embodiment; therefore, the description is omitted here.
Third embodiment
As shown in fig. 6, a schematic structural diagram of an electronic device 600 according to a third embodiment of the present invention, where the electronic device 600 may have relatively large differences according to different configurations or performances, may include one or more processors (central processing units, CPU) 601 and one or more memories 602, where at least one instruction is stored in the memories 602, and the at least one instruction is loaded and executed by the processors 601 to implement the multi-focus image fusion method according to the first embodiment of the present invention.
Fourth embodiment
In a fourth embodiment of the present invention, there is also provided a computer-readable storage medium, such as a memory including instructions executable by a processor in a terminal to perform the multi-sheet multi-focus image fusion method according to the first embodiment of the present invention. For example, the computer readable storage medium may be ROM, random Access Memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.
The technical scheme provided by the embodiment of the invention has the beneficial effects that at least the technical scheme comprises the following steps:
according to the invention, a few image baseline characteristics are extracted to be fused into a focusing level diagram, then the focusing level diagram is used for correcting other focusing level diagrams, the corrected focusing level diagram is formed into a focusing level set, then the focusing level set is converted into a decision diagram, and the fusion of a plurality of multi-focusing images is carried out according to the method of fusing all images, so that redundant characteristic extraction operation of the existing method (the workload of the characteristic extraction operation in the existing method is reduced by 50%) is avoided, the problem that the fusion efficiency of the multi-image multi-focusing images is lower under the current pairwise fusion strategy is solved, and the advantages are obvious under the application scene that the fusion of a large number of images and the requirement on the fusion speed of the images are higher.
It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program for instructing relevant hardware, where the program may be stored in a computer readable storage medium, and the storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.
The foregoing description of the preferred embodiments of the invention is not intended to limit the invention to the precise form disclosed, and any such modifications, equivalents, and alternatives falling within the spirit and scope of the invention are intended to be included within the scope of the invention.
Claims (6)
1. A method of fusing a plurality of multi-focus images, the method comprising:
extracting the characteristics of all images in an image set to be fused by adopting an image characteristic extraction algorithm, and selecting the characteristics of any two images in the image set to be fused as a first-stage baseline characteristic and a second-stage baseline characteristic;
respectively carrying out feature fusion on the first-stage baseline features and the features of the rest images in the image set to be fused by adopting an image feature fusion algorithm, and forming a plurality of focusing level diagrams;
correcting the rest focus level graphs by adopting a correction algorithm based on the focus level graphs formed by the first-stage baseline characteristic and the second-stage baseline characteristic, and splicing the corrected focus level graphs into a focus level set;
converting the focusing level set into a decision graph by adopting a decision algorithm;
fusing all images in the image set to be fused into a final single Zhang Rongge result based on the decision graph by adopting an image pixel fusion algorithm;
the method comprises the steps that an image set to be fused is at least two registered images with different focusing areas, wherein the images are shot for the same scene, and H multiplied by W of all the images in the image set to be fused are the same, wherein H is the number of column pixels, and W is the number of row pixels;
wherein the focusing level diagram is the same as H×W of the image to be fused, H is the number of column pixels, W is the number of row pixels, each element in the focusing level diagram corresponds to each pixel in the image to be fused, wherein p i j Representing the probability that the first image is sharper at the position of element i relative to the j-th image, and p i j ∈[0,1]；
Wherein, the correction algorithm has the expression as follows:
wherein p is i j Represents the probability that the 1 st image is clearer than the j-th image at the position of element i, and p i j ∈[0,1]；The probability that the 1 st image is clearer relative to the j th image at the corrected element i is given;
wherein the stitching the corrected plurality of focus level maps into a focus level set includes: splicing the corrected multiple focusing level graphs into a focusing level set according to the channel direction, wherein the focusing level set refers to a three-dimensional matrix with the size of H multiplied by W multiplied by N, H is the number of column pixels, W is the number of row pixels, N is the number of images, and the value of N is larger than 1;
wherein the decision algorithm comprises:
index Img for maximum value in channel direction of each element i in the focus level set i k Wherein the index Img i k Recording the clearest image sequence number k at the element i;
the index of each pixel is assembled into a decision graph.
2. The multi-sheet multi-focus image fusion method of claim 1, wherein the image feature extraction algorithm comprises: one or more of spatial frequency operators, gradient operators, convolutional neural networks, support vector machines.
3. The method of claim 1, wherein the image feature fusion algorithm comprises: spatial frequency operators, gradient operators, convolutional neural networks, support vector machines, additive fusion, maximum fusion, and channel dimension concatenation.
4. The multi-sheet multi-focus image fusion method of claim 1, wherein the image pixel fusion algorithm comprises: and (5) selecting one or more of an index value algorithm, a weighted fusion algorithm and a guided filtering algorithm.
5. The method of claim 4, wherein the indexing algorithm is based on an index value Img in the decision map i j The pixels of the image j in the element i are assigned to the pixels of the final individual fusion result position i.
6. A multi-sheet multi-focus image fusion apparatus, the apparatus comprising:
the feature extraction unit is used for extracting features of all images in the image set to be fused, and selecting features of any two images in the image set to be fused as a first-stage baseline feature and a second-stage baseline feature;
the feature fusion unit is used for respectively carrying out feature fusion on the first-stage baseline features and the features of the rest images in the image set to be fused, and forming a plurality of focusing level diagrams;
a correction unit for correcting the rest focus level graphs based on the focus level graphs formed by the first-stage baseline characteristic and the second-stage baseline characteristic, and splicing the corrected focus level graphs into a focus level set;
a decision unit for converting the focus level set into a decision graph;
the pixel fusion unit fuses all the images in the image set to be fused into a final single Zhang Rongge result based on the decision graph;
the method comprises the steps that an image set to be fused is at least two registered images with different focusing areas, wherein the images are shot for the same scene, and H multiplied by W of all the images in the image set to be fused are the same, wherein H is the number of column pixels, and W is the number of row pixels;
wherein the focusing level diagram is the same as H×W of the image to be fused, H is the number of column pixels, W is the number of row pixels, each element in the focusing level diagram corresponds to each pixel in the image to be fused, wherein p i j Representing the probability that the first image is sharper at the position of element i relative to the j-th image, and p i j ∈[0,1]；
Wherein, the correction algorithm has the following expression:
wherein p is i j Represents the probability that the 1 st image is clearer than the j-th image at the position of element i, and p i j ∈[0,1]；The probability that the 1 st image is clearer relative to the j th image at the corrected element i is given;
wherein the stitching the corrected plurality of focus level maps into a focus level set includes: splicing the corrected multiple focusing level graphs into a focusing level set according to the channel direction, wherein the focusing level set refers to a three-dimensional matrix with the size of H multiplied by W multiplied by N, H is the number of column pixels, W is the number of row pixels, N is the number of images, and the value of N is larger than 1;
wherein the decision algorithm comprises:
index Img for maximum value in channel direction of each element i in the focus level set i k Wherein the index Img i k Recording the clearest image sequence number k at the element i;
the index of each pixel is assembled into a decision graph.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011036730.9A CN112241940B (en) | 2020-09-28 | 2020-09-28 | Fusion method and device for multiple multi-focus images |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011036730.9A CN112241940B (en) | 2020-09-28 | 2020-09-28 | Fusion method and device for multiple multi-focus images |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112241940A CN112241940A (en) | 2021-01-19 |
| CN112241940B true CN112241940B (en) | 2023-12-19 |
Family
ID=74171774
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011036730.9A Active CN112241940B (en) | 2020-09-28 | 2020-09-28 | Fusion method and device for multiple multi-focus images |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112241940B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116848547A (en) | 2021-12-07 | 2023-10-03 | 宁德时代新能源科技股份有限公司 | Image processing method and system |
| CN116883461B (en) * | 2023-05-18 | 2024-03-01 | 珠海移科智能科技有限公司 | Method for acquiring clear document image and terminal device thereof |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006017233A1 (en) * | 2004-07-12 | 2006-02-16 | Lehigh University | Image fusion methods and apparatus |
| CN104700383A (en) * | 2012-12-16 | 2015-06-10 | 吴凡 | Multi-focus image generating device and multi-focus image file handling method |
| CN105894483A (en) * | 2016-03-30 | 2016-08-24 | 昆明理工大学 | Multi-focusing image fusion method based on multi-dimensional image analysis and block consistency verification |
| CN108230282A (en) * | 2017-11-24 | 2018-06-29 | 洛阳师范学院 | A kind of multi-focus image fusing method and system based on AGF |
| CN109300096A (en) * | 2018-08-07 | 2019-02-01 | 北京智脉识别科技有限公司 | A kind of multi-focus image fusing method and device |
| CN109389576A (en) * | 2018-10-10 | 2019-02-26 | 青岛大学 | Pulse-couple image interfusion method based on multichannel mechanism |
| CN110717879A (en) * | 2019-10-16 | 2020-01-21 | 北京京东尚科信息技术有限公司 | Multi-focus image processing method and device, storage medium and electronic equipment |
| CN111105346A (en) * | 2019-11-08 | 2020-05-05 | 同济大学 | Full-scanning microscopic image splicing method based on peak value search and gray template registration |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105100579B (en) * | 2014-05-09 | 2018-12-07 | 华为技术有限公司 | A kind of acquiring and processing method and relevant apparatus of image data |
| CN105335950B (en) * | 2014-05-28 | 2019-02-12 | 华为技术有限公司 | Image processing method and image processing apparatus |
| US10186023B2 (en) * | 2016-01-25 | 2019-01-22 | Qualcomm Incorporated | Unified multi-image fusion approach |
| EP3490241B1 (en) * | 2016-07-22 | 2021-07-07 | Sony Corporation | Image processing device and image processing method |
-
2020
- 2020-09-28 CN CN202011036730.9A patent/CN112241940B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006017233A1 (en) * | 2004-07-12 | 2006-02-16 | Lehigh University | Image fusion methods and apparatus |
| CN104700383A (en) * | 2012-12-16 | 2015-06-10 | 吴凡 | Multi-focus image generating device and multi-focus image file handling method |
| CN105894483A (en) * | 2016-03-30 | 2016-08-24 | 昆明理工大学 | Multi-focusing image fusion method based on multi-dimensional image analysis and block consistency verification |
| CN108230282A (en) * | 2017-11-24 | 2018-06-29 | 洛阳师范学院 | A kind of multi-focus image fusing method and system based on AGF |
| CN109300096A (en) * | 2018-08-07 | 2019-02-01 | 北京智脉识别科技有限公司 | A kind of multi-focus image fusing method and device |
| CN109389576A (en) * | 2018-10-10 | 2019-02-26 | 青岛大学 | Pulse-couple image interfusion method based on multichannel mechanism |
| CN110717879A (en) * | 2019-10-16 | 2020-01-21 | 北京京东尚科信息技术有限公司 | Multi-focus image processing method and device, storage medium and electronic equipment |
| CN111105346A (en) * | 2019-11-08 | 2020-05-05 | 同济大学 | Full-scanning microscopic image splicing method based on peak value search and gray template registration |
Non-Patent Citations (1)
| Title |
|---|
| 一种新的结合SVM和FNN的多聚焦图像融合算法;徐海;安吉尧;;计算技术与自动化(第01期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112241940A (en) | 2021-01-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7542740B2 (en) | Image line of sight correction method, device, electronic device, and computer program | |
| CN112241940B (en) | Fusion method and device for multiple multi-focus images | |
| CN113284073B (en) | Image restoration method, device and storage medium | |
| CN109859314A (en) | Three-dimensional rebuilding method, device, electronic equipment and storage medium | |
| CN113239875B (en) | Method, system and device for acquiring face characteristics and computer readable storage medium | |
| CN114694185B (en) | Cross-modal target re-identification method, device, equipment and medium | |
| CN113344794B (en) | Image processing method and device, computer equipment and storage medium | |
| CN112215788A (en) | Multi-focus image fusion algorithm based on improved generation countermeasure network | |
| CN115115552B (en) | Image correction model training method, image correction device and computer equipment | |
| CN115171199B (en) | Image processing method, image processing device, computer equipment and storage medium | |
| CN107463932A (en) | A kind of method that picture feature is extracted using binary system bottleneck neutral net | |
| US20190066264A1 (en) | Device and method for generating panorama image | |
| CN113658091A (en) | Image evaluation method, storage medium and terminal equipment | |
| WO2023246392A1 (en) | Image acquisition method, apparatus and device, and non-transient computer storage medium | |
| WO2023149135A1 (en) | Image processing device, image processing method, and program | |
| CN110399881A (en) | A kind of quality enhancement method and device based on binocular stereo image end to end | |
| CN112203023B (en) | Billion pixel video generation method and device, equipment and medium | |
| CN113343931A (en) | Training method for generating countermeasure network, image sight correction method and device | |
| KR102630844B1 (en) | Artificial intelligence camera system, method of transforming image therein, and computer-readable medium | |
| CN114118203A (en) | Image feature extraction and matching method and device and electronic equipment | |
| CN110795972A (en) | Pedestrian identity recognition method, device, equipment and storage medium | |
| CN115589456A (en) | Shooting method and device | |
| CN113313646B (en) | Image processing method and device, electronic equipment and computer readable storage medium | |
| CN113344997B (en) | Method and system for rapidly acquiring high-definition foreground image only containing target object | |
| CN111985510B (en) | Generative model training method, image generation device, medium, and terminal |
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 |