Adaptive motion artifact detection method and three-dimensional reconstruction method using the same

Through the adaptive motion artifact detection method, the multi-view angle rearrangement image and energy difference map are used to automatically detect and remove motion artifacts in the scanning light field system, solving the problems of detection difficulties and inefficiency in the prior art, and realizing the Efficient and accurate three-dimensional reconstruction process.

CN118570398BActive Publication Date: 2024-11-01ZHEJIANG HEHU TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411036847.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-31
Publication Date
2024-11-01
Estimated Expiration
2044-07-31

AI Technical Summary

Technical Problem

When shooting dynamic objects, the existing scanning light field system reduces the imaging quality due to difficulty in detecting motion artifacts, and requires human intervention and is inefficient, making it impossible to achieve end-to-end unified processing.

Method used

Adaptive motion artifact detection method is adopted, by obtaining multi-view rearrangement images, decoupling and chunking, determining the reference scanning sub-graph, performing motion artifact detection, using energy difference maps for artifact detection, generating three-dimensional matrix output detection results , as a pre-module for three-dimensional reconstruction, automatically removes motion artifacts.

Benefits of technology

It realizes fast and accurate automatic motion detection, improves the accuracy and efficiency of motion artifact detection, enhances the compactness and integrity of scanning light field image processing, and realizes an end-to-end processing process.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118570398B_ABST
    Figure CN118570398B_ABST
Patent Text Reader

Abstract

The present invention relates to an adaptive motion artifact detection method and a three-dimensional reconstruction method using the same. The adaptive motion artifact detection method comprises: obtaining a multi-view rearranged image, decoupling the multi-view rearranged image to obtain a scanning sub-image sequence; dividing each scanning sub-image in the scanning sub-image sequence into blocks; and determining a reference scanning sub-image, and sequentially performing motion artifact detection on each sub-block in the reference scanning sub-image and the sub-blocks in the remaining scanning sub-images. The method of the present invention can replace human judgment, reduce labor costs, and improve detection efficiency; and realize an end-to-end scanning light field image processing process.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of image output technology, and more particularly to an adaptive motion artifact detection method and a three-dimensional reconstruction method using the same. Background Art

[0002] In the field of optical imaging, two-dimensional imaging technology has always played a vital role. However, due to the limitations of camera pixel resolution and optical system aberrations, traditional two-dimensional imaging methods have difficulty achieving higher resolutions. In recent years, meta-imaging technology has attracted much attention in the field of optical research due to its unique multi-angle image acquisition method. Unlike traditional two-dimensional imaging, meta-imaging technology can not only obtain spatial information, but also capture angle information, so it actually contains four-dimensional information.

[0003] As a branch of meta-imaging technology, light field imaging technology can record the four-dimensional light field information of the system and reconstruct the three-dimensional shape of the photographed target based on this information. Based on light field imaging, researchers further proposed a scanning light field system. This system introduces high-frequency predetermined trajectory scanning at the sub-pixel level, and generates an ultra-high-resolution light field image by synthesizing multiple scanned images, thereby significantly improving the spatial resolution of light field imaging.

[0004] However, the introduction of the scanning mode also brings a problem: when shooting dynamic objects, the movement of the object may cause the relative displacement of the images captured by each scanning point to exceed the scanning sub-pixel interval, thereby generating motion artifacts in the high-resolution scanning light field image, which in turn affects the subsequent 3D reconstruction process. Therefore, the detection of motion artifacts is crucial for the subsequent removal of motion artifacts and 3D reconstruction.

[0005] However, when actually shooting moving samples, factors such as fluorescence quenching, system noise, and background contamination will be coupled with various unpredictable motion patterns, causing some movements in the sample to be difficult to detect. The ignored sample spatiotemporal motion artifacts will be highlighted in the subsequent three-dimensional reconstruction process, affecting the final imaging quality.

[0006] At the same time, the existing scanning light field system requires human intervention to start the motion artifact removal method for samples with different motion modes, which not only interrupts the integrity of the processing flow and makes it impossible to uniformly process samples of each mode end to end, but also makes human judgment difficult and inefficient due to the above-mentioned coupling problem, and the detection stability is poor. Summary of the invention

[0007] In order to solve these problems, the present invention provides an adaptive motion artifact detection method based on multi-view scanning light field images. This method can replace human judgment, reduce labor costs, and improve detection efficiency. In addition, this method can be used as a pre-detection module for various scanning light field system motion artifact removal methods, improve the compactness of the scanning light field system, and realize an end-to-end scanning light field image processing process without human intervention.

[0008] In order to achieve the above object, the present invention adopts the following technical solution:

[0009] An adaptive motion artifact detection method, comprising:

[0010] S1, obtaining a multi-view rearranged image, and decoupling the multi-view rearranged image to obtain a scanned sub-image sequence;

[0011] S2, dividing each scanning sub-image in the scanning sub-image sequence into blocks;

[0012] S3. Determine a reference scanning sub-image, and perform motion artifact detection on each sub-block in the reference scanning sub-image and the sub-blocks in the remaining scanning sub-images in turn.

[0013] Preferably, before decoupling, denoising is performed on the multi-view rearranged image, including:

[0014] Gaussian noise reduction via multi-view averaging, and

[0015] Reduce background noise by subtracting a fixed value from the background noise intensity.

[0016] Preferably, in step S1, the scan sub-image sequence is s*h*w, where s is the total number of scan sub-images, h and w are the pixel height and width of each scan sub-image, and the values ​​are , .

[0017] Preferably, in step S1, motion compensation is performed after obtaining the scan sub-image, including determining the coordinate offset of each scan point in the scan sub-image relative to the central scan point, and performing affine transformation according to the coordinate offset; the expression is:

[0018]

[0019] In the formula, represents the scanned sub-image sequence after motion compensation, , are the offsets of the scanning points in the column and row directions relative to the central scanning point, is the affine transformation operation, is the scan sub-image sequence.

[0020] Preferably, in S3, motion artifact detection is performed on each sub-block in the reference scan sub-image and the sub-blocks in the remaining scan sub-images in sequence; including:

[0021] S31, according to the jth sub-block in the reference scanning sub-graph, traverse the sub-blocks with the same sequence number j in the remaining s-1 scanning sub-graphs, where s is the total number of scanning sub-graphs;

[0022] S32, respectively obtaining an energy difference map between the j-th sub-block in the reference scanning sub-image and the j-th sub-block in each scanning sub-image;

[0023] S33, obtaining a motion artifact detection result of the j-th sub-block in each scanning sub-image relative to the j-th sub-block in the reference scanning sub-image based on the energy difference map;

[0024] S34, set j=j+1, and repeat steps S31-S33 until all sub-blocks in the scanned sub-image are traversed.

[0025] Preferably, in S33, obtaining the motion artifact detection result of the j-th sub-block in each scanning sub-image relative to the j-th sub-block in the reference scanning sub-image based on the energy difference map includes:

[0026] Determine the maximum value of the energy difference map, determine the first pixel value of the maximum value of the energy difference map in the j-th sub-block of the reference scan sub-image and the second pixel value of the energy difference map in the j-th sub-block of the corresponding scan sub-image according to the spatial coordinates of the maximum value; perform motion artifact detection based on the maximum value of the energy difference map, the first pixel value and the second pixel value in the following manner;

[0027]

[0028] in, Indicates the test results. represents the maximum value of the energy difference map, represents the first pixel value, represents the second pixel value, represents the OR logical operation, and σ represents the preset threshold.

[0029] Preferably, in S33, obtaining the motion artifact detection result of the j-th sub-block in each scanning sub-image relative to the j-th sub-block in the reference scanning sub-image based on the energy difference map includes:

[0030] Determine the maximum value of the energy difference map, determine the first pixel value of the maximum value of the energy difference map in the j-th sub-block of the reference scan sub-image and the second pixel value of the energy difference map in the j-th sub-block of the corresponding scan sub-image according to the spatial coordinates of the maximum value; perform motion artifact detection based on the maximum value of the energy difference map, the first pixel value and the second pixel value in the following manner;

[0031]

[0032] in, Indicates the test results. represents the maximum value of the energy difference map, represents the first pixel value, represents the second pixel value, Represents or logical operation, represents AND logic operation, σ represents the preset threshold, Represents foreground energy.

[0033] Preferably, in S3, a three-dimensional matrix is ​​generated according to the motion artifact detection results of each sub-block in the reference scanning sub-image and the sub-blocks in the remaining scanning sub-images, and the motion artifact detection results of the multi-view rearranged image are output according to the three-dimensional matrix.

[0034] The present invention further provides a three-dimensional reconstruction method, comprising:

[0035] Acquire light-field microscopy data;

[0036] Rearrange the light field data to obtain a multi-view rearranged image.

[0037] Performing motion artifact detection on the multi-view rearranged image using the adaptive motion artifact detection method as described in any one of the above items;

[0038] If there are motion artifacts, 3D reconstruction is performed after removing the motion artifacts. Otherwise, 3D reconstruction is performed directly based on the multi-view rearranged images.

[0039] It can be seen from the above technical solutions that the present invention discloses an adaptive motion artifact detection method and a three-dimensional reconstruction method using the same. Compared with the prior art, the present application utilizes the four-dimensional information of the scanned light field image to perform fast and accurate automatic motion detection on the noisy image. Compared with manual judgment and manual intervention in the subsequent processing algorithm, it can not only improve the detection accuracy and efficiency of motion artifacts, but also improve the compactness and integrity of the scanning light field image processing process, thereby realizing true end-to-end processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0041] Figure 1 A flow chart of the adaptive motion artifact detection method provided by the present invention;

[0042] Figure 2This is a flowchart of an adaptive motion artifact detection method provided by the present invention. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] Although the scanning light field system can greatly improve the spatial resolution of light field data in static scenes, when shooting dynamic scenes, motion artifacts will inevitably appear due to the scanning characteristics, which will make the advantages of the scanning mode disappear, and the motion artifacts will affect the subsequent reconstruction process. Therefore, efficient and accurate detection of the motion of the sample is crucial for subsequent motion artifact removal and three-dimensional reconstruction.

[0045] In view of this, the present invention discloses an adaptive motion artifact detection method and a three-dimensional reconstruction method using the same, and the implementation process is described below through embodiments.

[0046] Embodiment 1

[0047] This embodiment discloses an adaptive motion artifact detection method. Figure 1 , mainly includes the following steps:

[0048] S1. Acquire a multi-view rearranged image, and decouple the multi-view rearranged image to obtain a scan sub-image sequence;

[0049] S2, dividing each scanning sub-image in the scanning sub-image sequence into blocks;

[0050] S3. Determine a reference scanning sub-image, and perform motion artifact detection on each sub-block in the reference scanning sub-image and the sub-blocks in the remaining scanning sub-images in turn.

[0051] In an exemplary embodiment, Figure 2 As shown, step S1 includes:

[0052] S11, after obtaining the multi-view rearranged image, noise pre-cleaning is performed preferentially;

[0053] Generally, the noise components of the scanning light field system are mainly Gaussian noise and background noise. In order to prevent these two noises from affecting the subsequent motion artifact detection algorithm, noise pre-cleaning is required. Pre-cleaning includes reducing Gaussian noise by averaging multiple perspectives and reducing background noise by subtracting a fixed value of background noise intensity.

[0054] Since Gaussian noise is random noise, it conforms to the independent and identically distributed assumption in different viewing angles of the scanned light field image. Therefore, multi-view averaging operation can be used to reduce the influence of Gaussian noise and enhance the signal. noise For example, the viewing angle averaging operation can be expressed as follows:

[0055]

[0056] in, Represents the rearranged image after perspective averaging, and its shape is H*W.

[0057] For background noise, its influence on the subsequent motion artifact judgment algorithm can be weakened by directly subtracting a fixed value of the background noise intensity.

[0058] In one embodiment, the background noise is removed on the basis of reducing the Gaussian noise; that is, after the viewing angle is averaged, the influence of the Gaussian noise on the image is weakened, and the remaining background noise is a composite noise with an intensity distribution near a fixed value, so the background noise is further removed according to the following formula:

[0059]

[0060] in, Represents the system background noise intensity value, I clean Represents the rearranged image after subtracting the background noise, and its shape is H*W.

[0061] S12, decoupling the multi-view rearranged image to obtain a scanned sub-image sequence;

[0062] In this embodiment, the rearranged image of the scanning light field system is formed by splicing each scanning sub-image in a scanning cycle according to the spatial scanning trajectory. The subsequent judgment algorithm is performed on each sub-image in a scanning cycle. Therefore, the noise-free image I obtained in the previous step is first clean Decoupled into a scanning sub-image sequence I with shape s*h*w according to the scanning mode scan , where s is the total number of scanned sub-images, which is actually the number of scan points in a scan cycle, and h and w are the pixel height and width of each scanned sub-image. , .

[0063] In a preferred embodiment, it also includes:

[0064] S13, performing motion compensation after obtaining the scanned sub-image;

[0065] Due to the scanning characteristics of the scanning light field system, each scanned sub-image will have a spatial sub-pixel offset, which is visually manifested as the relative jitter of the entire image when the scanned sub-images are observed continuously. In order to prevent the subsequent motion detection algorithm from mistaking this jitter for the movement of the sample in the image, scanning mode motion compensation is required to eliminate the jitter between the scanned sub-images.

[0066] The compensation method includes determining the coordinate offset of each scanning point in the scanning sub-image relative to the central scanning point. In order to eliminate the offset, each point sub-image is spatially affine transformed according to the horizontal and vertical coordinate offsets; the expression is:

[0067]

[0068] In the formula, represents the scanned sub-image sequence after motion compensation, , are the offsets of the scanning points in the column and row directions relative to the central scanning point, is the affine transformation operation, is the scan sub-image sequence.

[0069] In the present invention, S2 is to divide each scanning sub-image in the scanning sub-image sequence into blocks;

[0070] Since some samples may have only a small part of the structure moving while the rest is still, in order to prevent the small movement from being submerged in the overall stillness, a block-based motion judgment strategy is set. Each scanned sub-image of shape h*w is spatially divided into n*n sub-blocks. , and then perform the operation in step S3 on each block in turn.

[0071] Further, step S3 includes:

[0072] Determine a reference scanning sub-image, and perform motion artifact detection on each sub-block in the reference scanning sub-image and the sub-blocks in the remaining scanning sub-images in turn; the specific process is as follows:

[0073] S31, according to the jth sub-block in the reference scanning sub-graph, traverse the sub-blocks with the same sequence number j in the remaining s-1 scanning sub-graphs, where s is the total number of scanning sub-graphs;

[0074] S32, respectively obtaining an energy difference map between the j-th sub-block in the reference scanning sub-image and the j-th sub-block in each scanning sub-image; the energy difference map is the absolute value of the energy difference between the two sub-blocks;

[0075] S33, obtaining a motion artifact detection result of the j-th sub-block in each scanning sub-image relative to the j-th sub-block in the reference scanning sub-image based on the energy difference map; this process includes:

[0076] Determine a maximum value of the energy difference map, and determine a first pixel value of the maximum value of the energy difference map in the jth sub-block of the reference scan sub-image and a second pixel value of the energy difference map in the jth sub-block of the corresponding scan sub-image according to the spatial coordinates of the maximum value;

[0077] Based on the maximum value of the energy difference map, the first pixel value and the second pixel value, motion artifact detection is performed as follows;

[0078]

[0079] in, Indicates the test results. represents the maximum value of the energy difference map, represents the first pixel value, represents the second pixel value, represents the OR logical operation, and σ represents the preset threshold.

[0080] Scan the sub-image based on the first scanned sub-image, and use the first sub-block in it Take s as an example to illustrate, at this time, it is necessary to traverse all the first sub-blocks in the remaining s-1 scanning sub-images, and then calculate the absolute value of the difference between the pixel value of the first sub-block and the first sub-block in the reference scanning sub-image, and obtain the sub-block With the first sub-block of each scanned sub-image The corresponding energy difference diagram , where s is the total number of scanned subgraphs;

[0081] When judging the motion artifact of the first sub-block in the scanned sub-image relative to the first sub-block in the reference scanned sub-image; first find the maximum value in the energy difference map and the spatial coordinate position of the maximum value, and respectively determine the first pixel value of the spatial coordinate position in the first sub-block of the reference scanning sub-image , and the second pixel value in the first sub-block of the corresponding scan sub-image in the energy difference map ;

[0082] Based on the prior knowledge that sample motion is the movement of pixel positions on an image, we calculate the and and The ratio of the reference scan sub-image and the energy difference image can be used to determine whether there is motion in the first sub-block by the degree of pixel value change.

[0083] In an optional embodiment, when most of the sub-block is background and the pixel intensity is very low, a slight intensity change may cause and and The ratio of changes greatly, which leads to the identification of background fluctuations as sample motion.

[0084] Therefore, this embodiment proposes to add background elimination judgment on the basis of the above judgment conditions. This judgment can be calculated by and and and If the ratio is higher than the set threshold, it can be considered that the changed point is the foreground rather than the background. If the above two judgment conditions are met at the same time, it can be considered that the reference scan sub-image and the scan sub-image corresponding to the maximum value in the energy difference image have motion in the sub-block and the moving point is the foreground target rather than the background.

[0085] At this point, the detection process is:

[0086] Determine respectively a first pixel value of a spatial position of a maximum value of the energy difference map in a j-th sub-block of a reference scan sub-image and a second pixel value in a j-th sub-block of a corresponding scan sub-image of the energy difference map; and perform motion artifact detection based on the first pixel value and the second pixel value in the following manner;

[0087]

[0088] in, Indicates the detection result (1 represents motion artifacts, 0 represents no motion artifacts), represents the maximum value of the energy difference map, represents the first pixel value, represents the second pixel value, Represents or logical operation, represents AND logic operation, σ represents the preset threshold, represents the foreground energy of the benchmark scan sub-image, Indicates the foreground energy of the scanned sub-image corresponding to the maximum value; wherein the foreground energy is the average of the top 50 pixel values ​​of each scanned sub-image.

[0089] S34, let j=j+1, and repeat steps S31-S33 until all sub-blocks in the scanned sub-image are traversed.

[0090] In this embodiment, the traversal process is implemented by referring to Figure 2 , as follows:

[0091] Let i=2, j=1, where i is the serial number of the scan sub-image, the first scan sub-image (i=1) is the reference scan sub-image, and j is the serial number of the sub-block in the reference scan sub-image;

[0092] Sequentially perform single block motion detection on the j-th block at the first scanning position and the j-th block at the i-th scanning position;

[0093] After the detection is completed, if i < s, then let i + 1 return and continue to execute; otherwise, judge whether j < n 2 , if j < n 2 , then return and continue to judge the (j + 1)-th sub-block, otherwise output the motion determination result matrix.

[0094] In this embodiment, each time steps S31 - S33 are executed, the motion situation between two scanned sub-images based on a certain sub-block can be obtained. When all sub-blocks in the scanned sub-images are traversed, a three-dimensional matrix with the shape of (s - 1) * n * n can be obtained, and the values in the matrix are 0 or 1. Through this matrix, the detailed motion artifact detection results of the detected multi-view scanned light field image can be obtained.

[0095] Embodiment 2

[0096] This embodiment discloses a three-dimensional reconstruction method, and the steps include:

[0097] Obtain light field microscopy data, that is, obtain light field microscopy data with 4D information through a scanned light field microscopy system;

[0098] Rearrange the light field data to obtain a multi-view rearranged image. In this embodiment, the light field data obtained in the previous step is rearranged to obtain a multi-view rearranged image with the shape of V * H * W, where V is the number of views of the rearranged image, H is the pixel height of the rearranged image, and W is the pixel width of the rearranged image;

[0099] Use the above-mentioned adaptive motion artifact detection method to perform motion artifact detection on the multi-view rearranged image; specifically, send the multi-view rearranged image obtained in the previous step into the motion artifact detection program to judge whether there are motion artifacts in this image;

[0100] If there are motion artifacts in the detection result of the previous step, then first remove the motion artifacts in the corresponding rearranged image through a motion artifact removal network, and then send the rearranged image after removing the motion artifacts into a reconstruction algorithm for three-dimensional reconstruction. If there are no motion artifacts in the detection result of the previous step, then directly send the rearranged image into the reconstruction algorithm for three-dimensional reconstruction. It should be noted that this application does not limit the motion artifact removal network, and the application of any network that can implement the function of removing artifacts falls within the protection scope of this application.

[0101] The present invention innovatively uses the four-dimensional information of the scanned light field image to perform fast and accurate automatic motion detection on the noisy image. Compared with manual determination and manual intervention in subsequent processing algorithms, the present invention can improve the compactness and integrity of the scanned light field image processing flow and achieve true end-to-end processing.

[0102] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0103] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An adaptive motion artifact detection method, characterized in that: include: S1, obtaining a multi-view rearranged image, and decoupling the multi-view rearranged image to obtain a scanned sub-image sequence; S2, dividing each scanning sub-image in the scanning sub-image sequence into blocks; S3, determining a reference scanning sub-image, and sequentially performing motion artifact detection on each sub-block in the reference scanning sub-image and the sub-blocks in the remaining scanning sub-images, including: S31, according to the jth sub-block in the reference scanning sub-graph, traverse the sub-blocks with the same sequence number j in the remaining s-1 scanning sub-graphs; S32, respectively obtaining an energy difference map between the j-th sub-block in the reference scanning sub-image and the j-th sub-block in each scanning sub-image; S33, based on the energy difference map, obtaining the motion artifact detection result of the j-th sub-block in each scanning sub-image relative to the j-th sub-block of the reference scanning sub-image; including: determining the maximum value of the energy difference map, and determining the first pixel value of the maximum value of the energy difference map in the j-th sub-block of the reference scanning sub-image according to the spatial coordinates of the maximum value, and the second pixel value of the maximum value of the energy difference map in the j-th sub-block of the corresponding scanning sub-image; based on the maximum value of the energy difference map, the first pixel value and the second pixel value, performing motion artifact detection in the following manner: motion s =(diff max / E1)>σ||(diff max / E2)>s) Among them, motion s Indicates the test results, diff max represents the maximum value of the energy difference map, E1 represents the first pixel value, E2 represents the second pixel value, || represents an OR logical operation, and σ represents a preset threshold.

2. The adaptive motion artifact detection method according to claim 1, characterized in that: Before decoupling, the multi-view rearranged image is denoised, including: Gaussian noise reduction via multi-view averaging, and Reduce background noise by subtracting a fixed value from the background noise intensity.

3. The adaptive motion artifact detection method according to claim 1, characterized in that: In step S1, the scanned sub-image sequence is s*h*w, where s is the total number of scanned sub-images, h and w are the pixel height and width of each scanned sub-image, and the values ​​are 4. The adaptive motion artifact detection method according to claim 1, characterized in that: In step S1, motion compensation is performed after obtaining the scan sub-image, including determining the coordinate offset of each scan point in the scan sub-image relative to the central scan point, and performing affine transformation according to the coordinate offset; the expression is: In the formula, I transed represents the scanned sub-image sequence after motion compensation, Δ h , Δ w are the offsets of the scanning points in the column and row directions relative to the central scanning point, warpAffine is the affine transformation operation, and I scan is the scan sub-image sequence.

5. The adaptive motion artifact detection method according to claim 1, characterized in that: In S3, motion artifact detection is performed on each sub-block in the reference scan sub-image and the sub-blocks in the remaining scan sub-images in turn; including: S34, set j=j+1, and repeat steps S31-S33 until all sub-blocks in the scanned sub-image are traversed.

6. The adaptive motion artifact detection method according to claim 5, characterized in that: In S33, based on the energy difference map, a motion artifact detection result of the j-th sub-block in each scanning sub-image relative to the j-th sub-block in the reference scanning sub-image is obtained, including: Determine the maximum value of the energy difference map, determine the first pixel value of the maximum value of the energy difference map in the j-th sub-block of the reference scan sub-image and the second pixel value of the maximum value of the energy difference map in the j-th sub-block of the corresponding scan sub-image according to the spatial coordinates of the maximum value; perform motion artifact detection based on the maximum value of the energy difference map, the first pixel value and the second pixel value in the following manner; Among them, motion s Indicates the test results, diff max represents the maximum value of the energy difference map, E1 represents the first pixel value, E2 represents the second pixel value, || represents the OR logic operation, && represents the AND logic operation, σ represents the preset threshold, E signal represents the prospect energy, represents the foreground energy of the benchmark scan sub-image, Indicates the foreground energy of the scanned sub-image corresponding to the maximum value.

7. The adaptive motion artifact detection method according to claim 1, characterized in that: In S3, a three-dimensional matrix is ​​generated according to the motion artifact detection results of each sub-block in the reference scan sub-image and the sub-blocks in the remaining scan sub-images, and the motion artifact detection results of the multi-view rearranged image are output according to the three-dimensional matrix.

8. A three-dimensional reconstruction method, characterized in that: include: Acquire light-field microscopy data; Rearranging the light field data to obtain a multi-view rearranged image; Using the adaptive motion artifact detection method according to any one of claims 1 to 7 to perform motion artifact detection on the multi-view rearranged image; If there are motion artifacts, 3D reconstruction is performed after removing the motion artifacts. Otherwise, 3D reconstruction is performed directly based on the multi-view rearranged images.