CN117689760A - OCT axial super-resolution method and system based on histogram information network - Google Patents
OCT axial super-resolution method and system based on histogram information network Download PDFInfo
- Publication number
- CN117689760A CN117689760A CN202410145241.9A CN202410145241A CN117689760A CN 117689760 A CN117689760 A CN 117689760A CN 202410145241 A CN202410145241 A CN 202410145241A CN 117689760 A CN117689760 A CN 117689760A
- Authority
- CN
- China
- Prior art keywords
- resolution
- axial
- image
- histogram information
- super
- 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
- 238000000034 method Methods 0.000 title claims abstract description 24
- 238000012545 processing Methods 0.000 claims abstract description 42
- 230000004927 fusion Effects 0.000 claims description 18
- 238000010586 diagram Methods 0.000 claims description 15
- 230000008878 coupling Effects 0.000 claims description 13
- 238000010168 coupling process Methods 0.000 claims description 13
- 238000005859 coupling reaction Methods 0.000 claims description 13
- 238000001228 spectrum Methods 0.000 claims description 12
- 238000005728 strengthening Methods 0.000 claims description 11
- 230000006870 function Effects 0.000 claims description 10
- 238000012549 training Methods 0.000 claims description 10
- 230000001965 increasing effect Effects 0.000 claims description 8
- 239000000284 extract Substances 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 6
- 230000003595 spectral effect Effects 0.000 claims description 5
- 239000011159 matrix material Substances 0.000 claims description 4
- 230000004913 activation Effects 0.000 claims description 3
- 230000002776 aggregation Effects 0.000 claims description 3
- 238000004220 aggregation Methods 0.000 claims description 3
- 230000004931 aggregating effect Effects 0.000 claims description 2
- 230000001131 transforming effect Effects 0.000 claims description 2
- 230000001953 sensory effect Effects 0.000 abstract description 9
- 238000009826 distribution Methods 0.000 abstract description 8
- 238000011158 quantitative evaluation Methods 0.000 abstract description 6
- 230000000694 effects Effects 0.000 abstract description 3
- 238000012014 optical coherence tomography Methods 0.000 description 42
- 230000008569 process Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000007634 remodeling Methods 0.000 description 1
- 238000003325 tomography Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 239000012780 transparent material Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/003—Reconstruction from projections, e.g. tomography
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4053—Scaling of whole images or parts thereof, e.g. expanding or contracting based on super-resolution, i.e. the output image resolution being higher than the sensor resolution
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration using histogram techniques
-
- 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
- G06T2207/10072—Tomographic images
- G06T2207/10101—Optical tomography; Optical coherence tomography [OCT]
-
- 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
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention provides an OCT axial super-resolution method and system based on a histogram information network, relates to the field of image processing, and aims at the problem that the super-resolution effect is affected by the amplified learning error when histogram information is directly used, and a selective feature kernel is designed to strengthen axial features so as to adapt to the introduction of the histogram information. The histogram information is introduced to carry out statistical distribution of pixels, so that a high quantitative evaluation index is obtained, and meanwhile, the sensory difference problem between the result and the true value is considered, so that the result is more similar to the sensory effect of high axial resolution.
Description
Technical Field
The invention relates to the field of image processing, in particular to an OCT axial super-resolution method and system based on a histogram information network.
Background
The optical coherence tomography is a novel imaging mode (Optical coherence tomography, OCT), and the advantages of high axial resolution and three-dimensional imaging in the depth direction are utilized to make breakthrough progress in the industrial field, such as three-dimensional tomography of memory cards, scratch defect detection of transparent materials and the like. High axial resolution is achieved by both increasing the spectral width of the light source and reducing the center wavelength, so there are researchers that achieve high axial resolution by improvements in hardware, such as the use of supercontinuum light sources, which significantly increases the complexity of the device design and the cost of the product.
In the prior art, deep learning is mostly used for directly carrying out axial super-resolution on OCT images, so that an image effect close to that acquired by high-performance OCT equipment is obtained. Most of the current methods for improving the image in the spatial domain or the frequency domain only consider the structure of the neural network, but often neglect the statistical information of the pixel distribution of the image, and the histogram can reflect various information of OCT gray images through the pixel type distribution, so that the histogram information is introduced to conduct statistical guidance of the pixel distribution on different areas of the image, not only can the quantitative evaluation index of the OCT image be improved, but also the light and shadow structures such as the contrast of a specific area can be adjusted, the problem that the super-resolution and the high axial resolution of the existing OCT image have sensory difference is solved, and the qualitative visual effect approaching to the image with the high axial resolution is obtained. Besides introducing statistical information of a histogram, the roll-off of the OCT equipment can cause the characteristic that the light intensity is continuously reduced in the axial direction when the equipment collects images, so that the fluctuation of pixel distribution variation of high and low axial resolutions along the A line direction (axial direction) is large and the similarity is reduced, and the direct introduction of the histogram information guides the pixel distribution of the OCT with low axial resolution, so that learning errors can be amplified, learning efficiency is reduced, and quantitative evaluation indexes and qualitative visual effects are influenced.
Disclosure of Invention
Aiming at the defects existing in the prior art, the invention provides an OCT axial super-resolution method and system based on a histogram information network, designs a selective feature kernel to strengthen axial features, further adapts to the introduction of histogram information, introduces the statistical distribution of pixels by the histogram information, thereby obtaining a high quantitative evaluation index, and simultaneously gives consideration to the problem of sensory difference between a result and a true value, so that the result is more similar to the sensory effect of high axial resolution.
The first object of the present invention is to provide an OCT axial super-resolution method based on a histogram information network, which adopts the following scheme:
comprising the following steps:
acquiring OCT image data to be subjected to axial super-resolution, and inputting an axial super-resolution processing model;
outputting an OCT image with improved axial resolution by the axial super-resolution processing model;
wherein, establishing the axial super-resolution processing model comprises:
acquiring original image data, and respectively carrying out image reconstruction and image reconstruction with increased spectrum clipping to obtain a high axial resolution image and a low axial resolution image;
the low axial resolution image input is based on a histogram information network, selective feature strengthening and histogram information coupling are carried out, a final feature map is obtained after multiple times of fusion, the final feature map is reconstructed, and a final reconstructed image is output;
and using the final reconstructed image and the high axial resolution image for training an axial super resolution processing model to obtain the trained axial super resolution processing model.
Further, the spectral clipping is an image reconstruction that performs incremental spectral clipping on each a-line of the original image data through a gaussian window to obtain a low axial resolution image.
Further, the selective feature enhancement includes:
extracting the characteristics of the input low axial resolution images by three groups of convolutions in parallel, and executing channel fusion on the obtained three groups of characteristic images;
and aggregating the feature map information in different directions, and outputting a feature map F1 after processing.
Further, when channel fusion is performed on the three sets of feature images, parameters are set for the number of the feature images in each set to obtain feature images with different numbers of weights, and the feature images in the axial direction are strengthened.
Further, after aggregation, vector Z is obtained through convolution and activation functions; and combining the vector Z and the feature map to obtain a feature map F1.
Further, the histogram information coupling includes:
taking the characteristic diagram F1 of the selective characteristic strengthening output as an input image, and transforming to obtain a frequency spectrum diagram of a frequency domain;
the spectrogram is processed in parallel by four paths;
extracting histogram information from the spectrogram through a first path and a second path, performing information coupling, multiplying the histogram information with the spectrogram subjected to the medium-size reshaping in a third path, performing re-size reshaping, performing residual combination on the re-size reshaped result and the initial spectrogram from a fourth path, and then converting to output a feature map;
and fusing the output characteristic diagram with the characteristic diagram F1 to obtain a characteristic diagram F2.
Further, the input spectrogram extracts histogram information in the first path, and uses linear mapping to expand and transpose the size of the histogram information; the input spectrogram extracts the histogram information in the second path, expands the size of the histogram information by using linear mapping but not transposes the histogram information, and then multiplies the histogram information by a matrix and activates the histogram information by a function to realize information coupling.
Further, the feature map F2 is used as an input image, the processing flow of the feature map F1 is repeated, and the feature map F3 is obtained by fusing the feature map F2 with the feature map F2 which is input after the feature map F2 is processed.
Further, the low axial resolution image, the feature map F1, the feature map F2 and the feature map F3 are combined in a channel mode to obtain a feature map F4 as a final feature map, the final feature map is reconstructed through convolution, and a final reconstructed image is output.
A second object of the present invention is to provide an OCT axial super-resolution system based on a histogram information network, comprising:
a data acquisition module configured to: acquiring OCT image data to be subjected to axial super-resolution, and inputting an axial super-resolution processing model;
an axial super-resolution processing module configured to: outputting an OCT image with improved axial resolution by the axial super-resolution processing model;
wherein, establishing the axial super-resolution processing model comprises:
acquiring original image data, and respectively carrying out image reconstruction and image reconstruction spectrum clipping with increased spectrum clipping to obtain a high axial resolution image and a low axial resolution image;
the low axial resolution image input is based on a histogram information network, selective feature strengthening and histogram information coupling are carried out, a final feature map is obtained after multiple times of fusion, the final feature map is reconstructed, and a final reconstructed image is output;
and using the final reconstructed image and the high axial resolution image for training an axial super resolution processing model to obtain the trained axial super resolution processing model.
Compared with the prior art, the invention has the advantages and positive effects that:
(1) Aiming at the problem that the super-resolution effect is affected by the amplified learning error when the histogram information is directly used, a selective feature kernel is designed to strengthen the axial features so as to adapt to the introduction of the histogram information. The histogram information is introduced to carry out statistical distribution of pixels, so that a high quantitative evaluation index is obtained, and meanwhile, the sensory difference problem between the result and the true value is considered, so that the result is more similar to the sensory effect of high axial resolution.
(2) The trained model can be applied to a low axial resolution OCT system to improve the axial resolution of the OCT image it outputs.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the invention.
Fig. 1 is a flow chart of the OCT axial super-resolution method based on the histogram information network in embodiments 1 and 2 of the present invention.
Fig. 2 is a flow chart of the low axial resolution image input histogram information based network in embodiments 1 and 2 of the present invention.
Fig. 3 is a flow chart of the axial features of the enhanced OCT images of embodiments 1 and 2 of the present invention.
Fig. 4 is a schematic flow chart of the fusion of histogram information in embodiments 1 and 2 of the present invention.
Detailed Description
Example 1
In an exemplary embodiment of the present invention, as shown in fig. 1-4, an OCT axial super-resolution method based on a histogram information network is presented.
In this embodiment, aiming at the problem of sensory difference existing in the super-resolution of the existing OCT image, the histogram information is introduced and the real physical factors are considered to obtain a high quantitative evaluation index, so that the super-resolved image is more similar to the sensory effect of high axial resolution.
The following describes the OCT axial super-resolution method based on the histogram information network in detail with reference to the accompanying drawings.
Referring to fig. 1, the OCT axial super-resolution method based on the histogram information network includes:
acquiring OCT image data to be subjected to axial super-resolution, and inputting an axial super-resolution processing model;
outputting an OCT image with improved axial resolution by the axial super-resolution processing model;
wherein, establishing the axial super-resolution processing model comprises:
acquiring original image data, and respectively carrying out image reconstruction and image reconstruction with increased spectrum clipping to obtain a high axial resolution image and a low axial resolution image;
the low axial resolution image input is based on a histogram information network, selective feature strengthening and histogram information coupling are carried out, a final feature map is obtained after multiple times of fusion, the final feature map is reconstructed, and a final reconstructed image is output;
and using the final reconstructed image and the high axial resolution image for training an axial super resolution processing model to obtain the trained axial super resolution processing model.
In the present embodiment, as shown in fig. 1, first, raw image data is acquired by the OCT apparatus. An image reconstruction algorithm is directly performed on the raw data to obtain a high axial resolution image. And performing spectrum clipping on each A line of the original data through a Gaussian window to obtain a low axial resolution image, and manufacturing image data pairs of high and low axial resolution images as a data set. Then inputting the low axial resolution image into a histogram information network, as shown in fig. 2, inputting the image into a selective characteristic strengthening module, obtaining a characteristic diagram F1, inputting the characteristic diagram F1 into the histogram information module, carrying out channel fusion on the output result of the characteristic diagram F1 and F1 to obtain a characteristic diagram F2, inputting the F2 into the histogram information module, carrying out channel fusion on the output result of the characteristic diagram F2 and F2, and finally carrying out channel fusion on the input image, the output result of the selective characteristic strengthening module and the result obtained by the two fusion, and obtaining an output result after 3X 3 convolution reconstruction.
The axial super-resolution processing model is trained through the process, then the training result is applied to the low axial resolution OCT system, OCT image data to be subjected to axial super-resolution is obtained and is input into the low axial resolution OCT system loaded with the trained axial super-resolution processing model, and the axial resolution of an OCT image output by the low axial resolution OCT system is improved.
The foregoing will be described in detail with reference to the accompanying drawings.
Step 1, data set preparation
Raw image data is acquired by the OCT apparatus. An image reconstruction algorithm is directly performed on the raw data to obtain a high axial resolution image. And performing spectrum clipping on each A line of the original data through a Gaussian window, and then reconstructing an image to obtain a low axial resolution image so as to prepare a data pair consisting of the high axial resolution image and the low axial resolution image.
Step 2, enhancing axial features of OCT images
Step 2.1
As shown in fig. 2, the low axial resolution image is input, features of the input image are extracted through three groups of convolutions of 3×1, 1×3 and 3×3 in parallel, channel fusion is performed on the three groups of feature images through Concat operation, parameters are set for the number of each group of feature images during fusion to obtain feature images with different weights so as to strengthen the feature images in the axial direction, the weights of the three groups of convolutions are set as K1, K2 and K3 respectively, the weights are set as K1> K2 and K1> K3 corresponding to the requirement of strengthening the axial direction, and the three groups of convolutions are continuously learned and adjusted to an optimal state during network training.
Step 2.2
And reducing the size of the feature map to 1 multiplied by 1 through the maximum global pooling size, so as to realize aggregation of the feature map information in different directions.
Step 2.3
The vector Z is then obtained by a 1 x 1 convolution and softmax activation function.
Step 2.4
Then, the element-wise multiplication of Z with the feature map subjected to the 3×3 convolution kernel is performed, and then an output image, i.e., F1 in fig. 2, is obtained.
Step 3, introducing histogram information
Step 3.1
The specific process is as shown in fig. 3, the output image of the last step is taken as the input image of the step, the input image is subjected to fast fourier transformation to obtain a spectrogram of a frequency domain, and the spectrogram is processed in parallel through four paths.
Step 3.2
According to the histogram feature information extracted as shown in formula 1, the spectrogram f (x, y) is input, the number of pixel values i in f (x, y) is calculated, the result is H (i), and the histogram vector v is composed of 256H (i).
Equation 1:
where f (x, y) is the input spectrogram, x, y is the pixel coordinates,is a dirac function, i is the pixel value size, and H (i) is the number of pixel values i.
Where H (i) is the number of pixel values i and v denotes a total of 256 data constituting a histogram vector.
The above histogram information extraction process is collectively called histogram vector extraction F hist A function.
First, from top to bottom, the first path and the second path are used for extracting histogram information and performing information coupling. The first path is to pass the input spectrogram through F hist The function extracts the histogram information and then expands and transposes its size using a linear mapping, resulting in (1, H)W), where H is the image height and W is the image width, is the same as described below. The second path passes the input spectrum image through F hist The function extracts the histogram information and then expands its size using a linear mapping but not transposes, resulting in (h×w, 1). The two were then multiplied by a matrix and activated by a softmax function, yielding the result (h×w ).
Step 3.3
The third path reshapes the spectrogram 1 size to (1, H W), where H is the image height and W is the image width. Matrix multiplication and size remodeling are carried out on the result of the step 3.2, and a result with the image size of (H, W) is obtained.
Step 3.4
And then carrying out residual combination on the third path of result and the initial spectrogram 1 to increase image information, and finally obtaining a result through inverse Fourier transform as output.
Step 4, fusion of histogram information and feature map
And F1 and the result of the step 3 are fused through a channel to obtain an input image F2 of the next step.
Step 5
F2 as the input image repeats the same operations of step 3 and step 4, resulting in an output image F3.
Step 6
The input image, F1, F2 and F3 are channel-combined to obtain a feature map F4, which is then reconstructed by a 3X 3 convolution to finally output an image.
Step 7, network training
The network training process of the axial super-resolution processing model is consistent with that of a conventional deep learning neural network.
Example 2
In another exemplary embodiment of the present invention, as shown in fig. 1-4, an OCT axial super-resolution system based on a histogram information network is presented.
The OCT axial super-resolution system based on the histogram information network comprises:
a data acquisition module configured to: acquiring OCT image data to be subjected to axial super-resolution, and inputting an axial super-resolution processing model;
an axial super-resolution processing module configured to: outputting an OCT image with improved axial resolution by the axial super-resolution processing model;
wherein, establishing the axial super-resolution processing model comprises:
acquiring original image data, and respectively carrying out image reconstruction and image reconstruction with increased spectrum clipping to obtain a high axial resolution image and a low axial resolution image;
the low axial resolution image input is based on a histogram information network, selective feature strengthening and histogram information coupling are carried out, a final feature map is obtained after multiple times of fusion, the final feature map is reconstructed, and a final reconstructed image is output;
and using the final reconstructed image and the high axial resolution image for training an axial super resolution processing model to obtain the trained axial super resolution processing model.
The operation of the OCT axial super-resolution system based on the histogram information network is described in embodiment 1, and will not be described herein.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. The OCT axial super-resolution method based on the histogram information network is characterized by comprising the following steps of:
acquiring OCT image data to be subjected to axial super-resolution, and inputting an axial super-resolution processing model;
outputting an OCT image with improved axial resolution by the axial super-resolution processing model;
wherein, establishing the axial super-resolution processing model comprises:
acquiring original image data, and respectively carrying out image reconstruction and image reconstruction with increased spectrum clipping to obtain a high axial resolution image and a low axial resolution image;
the low axial resolution image input is based on a histogram information network, selective feature strengthening and histogram information coupling are carried out, a final feature map is obtained after multiple times of fusion, the final feature map is reconstructed, and a final reconstructed image is output;
and using the final reconstructed image and the high axial resolution image for training an axial super resolution processing model to obtain the trained axial super resolution processing model.
2. The OCT axial super-resolution method based on a histogram information network of claim 1, wherein the spectral clipping is an image reconstruction by increasing spectral clipping of each a-line of the original image data through a gaussian window to obtain a low axial resolution image.
3. The histogram information network based OCT axial super-resolution method of claim 1, wherein the selective feature enhancement comprises:
extracting the characteristics of the input low axial resolution images by three groups of convolutions in parallel, and executing channel fusion on the obtained three groups of characteristic images;
and aggregating the feature map information in different directions, and outputting a feature map F1 after processing.
4. A histogram information network based OCT axial super-resolution method according to claim 3, wherein when channel fusion is performed on three sets of feature maps, parameters are set for the number of feature maps in each set to obtain feature maps with different number weights, and feature maps in the axial direction are enhanced.
5. A histogram information network based OCT axial super resolution method according to claim 3, characterized in that after aggregation, vector Z is obtained by convolution and activation functions; and combining the vector Z and the feature map to obtain a feature map F1.
6. The histogram information network based OCT axial super resolution method of claim 1, wherein the histogram information coupling comprises:
taking the characteristic diagram F1 of the selective characteristic strengthening output as an input image, and transforming to obtain a frequency spectrum diagram of a frequency domain;
the spectrogram is processed in parallel by four paths;
extracting histogram information from the spectrogram through a first path and a second path, performing information coupling, multiplying the histogram information with the spectrogram subjected to the medium-size reshaping in a third path, performing re-size reshaping, performing residual combination on the re-size reshaped result and the initial spectrogram from a fourth path, and then converting to output a feature map;
and fusing the output characteristic diagram with the characteristic diagram F1 to obtain a characteristic diagram F2.
7. The OCT axial super-resolution method based on a histogram information network of claim 6, wherein the input spectrogram extracts the histogram information in the first path, expands and transposes its size using linear mapping; the input spectrogram extracts the histogram information in the second path, expands the size of the histogram information by using linear mapping but not transposes the histogram information, and then multiplies the histogram information by a matrix and activates the histogram information by a function to realize information coupling.
8. The method for axially super-resolution OCT based on a histogram information network according to claim 6, wherein the feature map F2 is used as an input image, the processing procedure of the feature map F1 is repeated, and the feature map F3 is obtained by fusing the feature map F2 with the feature map F2 which is input after the feature map F2 is processed.
9. The method for axially super-resolution OCT based on a histogram information network according to claim 8, wherein the low-axial-resolution image, the feature map F1, the feature map F2, and the feature map F3 are channel-combined to obtain a feature map F4 as a final feature map, and the final feature map is reconstructed by convolution to output a final reconstructed image.
10. OCT axial super-resolution system based on histogram information network, characterized by comprising:
a data acquisition module configured to: acquiring OCT image data to be subjected to axial super-resolution, and inputting an axial super-resolution processing model;
an axial super-resolution processing module configured to: outputting an OCT image with improved axial resolution by the axial super-resolution processing model;
wherein, establishing the axial super-resolution processing model comprises:
acquiring original image data, and respectively carrying out image reconstruction and image reconstruction with increased spectrum clipping to obtain a high axial resolution image and a low axial resolution image;
the low axial resolution image input is based on a histogram information network, selective feature strengthening and histogram information coupling are carried out, a final feature map is obtained after multiple times of fusion, the final feature map is reconstructed, and a final reconstructed image is output;
and using the final reconstructed image and the high axial resolution image for training an axial super resolution processing model to obtain the trained axial super resolution processing model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410145241.9A CN117689760B (en) | 2024-02-02 | 2024-02-02 | OCT axial super-resolution method and system based on histogram information network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410145241.9A CN117689760B (en) | 2024-02-02 | 2024-02-02 | OCT axial super-resolution method and system based on histogram information network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117689760A true CN117689760A (en) | 2024-03-12 |
| CN117689760B CN117689760B (en) | 2024-05-03 |
Family
ID=90126881
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410145241.9A Active CN117689760B (en) | 2024-02-02 | 2024-02-02 | OCT axial super-resolution method and system based on histogram information network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117689760B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2012150730A (en) * | 2011-01-20 | 2012-08-09 | Panasonic Corp | Feature extraction device, feature extraction method, feature extraction program and image processing device |
| CN113362223A (en) * | 2021-05-25 | 2021-09-07 | 重庆邮电大学 | Image super-resolution reconstruction method based on attention mechanism and two-channel network |
| CN113396440A (en) * | 2019-02-14 | 2021-09-14 | 卡尔蔡司医疗技术公司 | System for OCT image conversion and ophthalmic image denoising and neural network thereof |
| CN115272072A (en) * | 2022-07-06 | 2022-11-01 | 大连海事大学 | Underwater image super-resolution method based on multi-feature image fusion |
| CN115619737A (en) * | 2022-10-18 | 2023-01-17 | 深圳华焊生物科技有限公司 | Biological tissue welding effect detection device |
| CN116051382A (en) * | 2023-03-02 | 2023-05-02 | 浙江工业大学 | Data enhancement method based on deep reinforcement learning generation type antagonistic neural network and super-resolution reconstruction |
-
2024
- 2024-02-02 CN CN202410145241.9A patent/CN117689760B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2012150730A (en) * | 2011-01-20 | 2012-08-09 | Panasonic Corp | Feature extraction device, feature extraction method, feature extraction program and image processing device |
| CN113396440A (en) * | 2019-02-14 | 2021-09-14 | 卡尔蔡司医疗技术公司 | System for OCT image conversion and ophthalmic image denoising and neural network thereof |
| CN113362223A (en) * | 2021-05-25 | 2021-09-07 | 重庆邮电大学 | Image super-resolution reconstruction method based on attention mechanism and two-channel network |
| CN115272072A (en) * | 2022-07-06 | 2022-11-01 | 大连海事大学 | Underwater image super-resolution method based on multi-feature image fusion |
| CN115619737A (en) * | 2022-10-18 | 2023-01-17 | 深圳华焊生物科技有限公司 | Biological tissue welding effect detection device |
| CN116051382A (en) * | 2023-03-02 | 2023-05-02 | 浙江工业大学 | Data enhancement method based on deep reinforcement learning generation type antagonistic neural network and super-resolution reconstruction |
Non-Patent Citations (4)
| Title |
|---|
| DAS, V ET AL.: "Diagnostic Information based Super-Resolution of Retinal Optical Coherence Tomography Images", IEEE, 1 January 2018 (2018-01-01) * |
| 周雨薇: "基于多分辨率分析和变分的遥感图像增强与融合算法研究", 中国优秀硕士学位论文集工程科技Ⅱ辑, 15 July 2014 (2014-07-15) * |
| 杨思晨等: "深度学习机制与小波融合的超分辨率重建算法", 北京航空航天大学学报, no. 01, 12 August 2019 (2019-08-12) * |
| 杨飘: "基于区域分析的图像超分辨与降噪技术研究", 中国优秀硕士学位论文集信息科技辑, 15 July 2020 (2020-07-15) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117689760B (en) | 2024-05-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111709902B (en) | Infrared and visible light image fusion method based on self-attention mechanism | |
| CN110660038B (en) | Multispectral image and full-color image fusion method based on generation countermeasure network | |
| CN109766858A (en) | Three-dimensional convolution neural network hyperspectral image classification method combined with bilateral filtering | |
| CN113139904B (en) | Image blind super-resolution method and system | |
| CN114862731B (en) | Multi-hyperspectral image fusion method guided by low-rank priori and spatial spectrum information | |
| Li et al. | A novel approach for multi-focus image fusion based on SF-PAPCNN and ISML in NSST domain | |
| CN111583285A (en) | Liver image semantic segmentation method based on edge attention strategy | |
| CN113066025B (en) | Image defogging method based on incremental learning and feature and attention transfer | |
| CN115147315B (en) | Neural network fluorescence microscopic image denoising method based on transformer module | |
| CN112070727A (en) | Metal surface defect detection method based on machine learning | |
| Sanghvi et al. | Photon limited non-blind deblurring using algorithm unrolling | |
| CN116402691B (en) | Image super-resolution method and system based on rapid image feature stitching | |
| CN116977651B (en) | Image denoising method based on double-branch and multi-scale feature extraction | |
| CN111275732B (en) | Foreground object image segmentation method based on depth convolution neural network | |
| CN103854265A (en) | Novel multi-focus image fusion technology | |
| You et al. | A simple and effective multi-focus image fusion method based on local standard deviations enhanced by the guided filter | |
| CN115205308A (en) | Fundus image blood vessel segmentation method based on linear filtering and deep learning | |
| CN103413351B (en) | Three-dimensional face fast reconstructing method based on compressive sensing theory | |
| CN117474765B (en) | DEM super-resolution reconstruction system based on reference image texture transfer | |
| CN117690161B (en) | Pedestrian detection method, device and medium based on image fusion | |
| CN117058507B (en) | Fourier convolution-based visible light and infrared image multi-scale feature fusion method | |
| CN109558880A (en) | A kind of whole profile testing method with Local Feature Fusion of view-based access control model | |
| CN117689760B (en) | OCT axial super-resolution method and system based on histogram information network | |
| CN116468083A (en) | Transformer-based network generation countermeasure method | |
| Park et al. | Image super-resolution using dilated window transformer |
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 |