CN107369135A - A kind of micro imaging system three-dimensional point spread function space size choosing method based on Scale invariant features transform algorithm - Google Patents
A kind of micro imaging system three-dimensional point spread function space size choosing method based on Scale invariant features transform algorithm Download PDFInfo
- Publication number
- CN107369135A CN107369135A CN201710478941.XA CN201710478941A CN107369135A CN 107369135 A CN107369135 A CN 107369135A CN 201710478941 A CN201710478941 A CN 201710478941A CN 107369135 A CN107369135 A CN 107369135A
- Authority
- CN
- China
- Prior art keywords
- spread function
- point spread
- dimensional point
- plies
- matching
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 21
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 17
- 238000003384 imaging method Methods 0.000 title claims abstract description 15
- 238000011084 recovery Methods 0.000 claims abstract description 21
- 238000001514 detection method Methods 0.000 claims abstract description 7
- 230000003287 optical effect Effects 0.000 claims description 9
- 238000004088 simulation Methods 0.000 claims description 6
- 238000007476 Maximum Likelihood Methods 0.000 claims description 3
- 230000002380 cytological effect Effects 0.000 claims description 2
- 230000000694 effects Effects 0.000 description 11
- 238000005516 engineering process Methods 0.000 description 9
- 238000004364 calculation method Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000009792 diffusion process Methods 0.000 description 1
- 230000000886 photobiology 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
-
- 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
- G06T2207/10056—Microscopic image
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Image Processing (AREA)
Abstract
A kind of micro imaging system three-dimensional point spread function choosing method based on Scale invariant features transform algorithm, comprises the following steps：Recovery emulation experiment is carried out to image using the three-dimensional point spread function of different radial dimensions and the number of plies, to original picture rich in detail, image after blurred picture and recovery, the detection of characteristic point is carried out respectively, image after restoring, blurred picture does Scale invariant features transform characteristic matching with original picture rich in detail respectively, the number of record matching point, and with the diameter of corresponding three-dimensional point spread function, the number of plies is corresponding, draw Feature Points Matching number and three-dimensional point spread function diameter, number of plies curve, matching is drawn again on this basis to count out increment and three-dimensional point spread function diameter, number of plies curve, and combine the existing structural theory of three-dimensional point spread function, three-dimensional point spread function space size is chosen.
Description
Technical field
The present invention relates to a kind of micro imaging system three-dimensional point spread function based on Scale invariant features transform algorithm is empty
Between size choosing method, this method be in Digital confocal microscope technology three dimensional biological microscopic image restoration handle an important ring
Section, belongs to digital image restoration processing technology field.The application of this method, can according to fast browsing, normal check and finely count
Different restore of calculation requires to need, from the angle of Scale invariant features transform algorithm, comprehensive three-dimensional point spread function structure
Theory, three-dimensional point spread function is chosen.
Background technology
Digital confocal microscope technology is based on common Photobiology microscope, configuration image detector, accurate mobile control
Mechanism processed and computer, using digital image processing techniques, restoration disposal is carried out to the biological specimen micro-image of collection, eliminated burnt
The influence defocused beyond face, to improve the resolution ratio of cell image.
Restoration disposal in Digital confocal microscope technology, deconvoluted restored method using three-dimensional microscopic image.It is three-dimensional
Point spread function characterizes microscope optical system, directly decides the effect for the restoration disposal that deconvolutes.Three-dimensional point spread function
The Energy distribution of reflection microscope optical system, recovery effect are better more exactly.On this basis, three-dimensional point spread function
Space size selection is bigger, and bigger comprising energy, recovery effect is better, while processing time is longer.Three-dimensional point spread function is
To pushing up double funnel structures, most energy concentrate on double funnels at middle part to the tiny area at top.Therefore, image is being carried out
During restoration disposal, rational selection is the three-dimensional point spread function to choose the certain space size of tiny area at this as origin
Image restoration processing is carried out, but with the gradual increase of space size, recovery effect lifting progressively eases up, and processing time
Increase sharply.Therefore, how according to different recovery requirements, comprehensive and balance consideration recovery effect and processing time, choose suitable
When three-dimensional point spread function carry out restoration disposal, be that three-dimensional microscopic image is deconvoluted restoration disposal needs in digital confocus technology
The major issue of solution.
The content of the invention
It is an object of the invention to provide a kind of micro imaging system three-dimensional point expansion based on Scale invariant features transform algorithm
Function space size choosing method is dissipated, considers the Scale invariant features transform algorithm between image and original picture rich in detail after restoring
Characteristic matching count out, with reference to three-dimensional point spread function structural theory, choose the three-dimensional point spread function of appropriate space size,
Restoration disposal is carried out to the microsection image of collection.
The present invention reaches above-mentioned purpose by the following technical programs：It is a kind of based on the micro- of Scale invariant features transform algorithm
Imaging system three-dimensional point spread function space size choosing method, including：
1) according to the parameter of micro imaging system, keep optical section layer to determine multiplication factor away from constant, make characterize it is aobvious
The large-space three-dimensional point spread function h of micro mirror optical system, interception is entered with its center to h, obtains a series of three for recovery
Tie up point spread function hM_r_z, wherein M expression multiplication factors, r expression radial dimensions, the z expression numbers of plies；2) with the clear cytological map of a width
As being used as original two-dimensional sample, the sequence two dimensional image that is mutually related is made, and one is built clearly with these sequence two dimensional images
Clear simulation sample 3-D view f, Fuzzy Simulation three-dimensional imaging image g is obtained with h and f convolution；3) use step (1) in it is a series of not
Same radial dimension, the different numbers of plies are the three-dimensional point spread function of corresponding different-energy size, different spaces respectively to 3-D view g
The restoration disposal that deconvolutes is carried out, restoration algorithm uses maximum likelihood method, and the image after recovery is represented with fi；
It is further comprising the steps of：
A, feature point detection and chi Feature Points Matching, the number of recording feature Point matching are carried out to f, g and fi；
It is described that feature point detection and Feature Points Matching are carried out to f, g and fi, refer to first detect feature to f, g and fi respectively
Point, i.e., satisfactory characteristic point is detected in metric space, then carried out fi and f characteristic point, g and f characteristic point special
Sign Point matching, the number of recording feature Point matching,
B, Feature Points Matching number and three-dimensional point spread function diameter, number of plies graph of relation are drawn；
C, matching is drawn to count out incremental rate curve figure, it is theoretical with reference to three-dimensional point spread function existing structure, three-dimensional point is expanded
Scattered function space is chosen.
The drafting matching is counted out and three-dimensional point spread function diameter, number of plies graph of relation, refers to count matching
The different-diameter of mesh and recovery, the three-dimensional point spread function of the different numbers of plies are corresponding, and using the number of plies as abscissa, characteristic point
It is ordinate with number, draws the curve under different-diameter.
The drafting matching is counted out incremental rate curve figure, theoretical with reference to three-dimensional point spread function existing structure, to three-dimensional point
Spread function space is chosen, and refers to calculate the increment counted out of matching, and diameter with three-dimensional point spread function, number of plies phase
Corresponding, using the number of plies as abscissa, Feature Points Matching number increment is ordinate, draws the curve under different-diameter, analyzes curve
And three-dimensional point spread function existing structure theory is combined, three-dimensional point spread function space is chosen.
The protrusion effect of the present invention is：
In the case where different recoveries require, the selection result of three-dimensional point spread function is directly decide in Digital confocal microscope technology
Three-dimensional microscopic image is deconvoluted recovery effect and processing time.How to choose, be always that Digital confocal microscope technology is to be solved
Problem.Correlation theory of the invention based on Scale invariant features transform algorithm, after original picture rich in detail, blurred picture and recovery
Image, first respectively carry out characteristic point detection, image, blurred picture after recovery are then made into chi with original picture rich in detail respectively
Spend invariant features conversion algorithm characteristics matching, the number of record matching point, and with diameter, the layer of corresponding three-dimensional point spread function
Number is corresponding, draws Feature Points Matching number and three-dimensional point spread function diameter, number of plies curve, draws match point again on this basis
Number increment and three-dimensional point spread function diameter, number of plies curve, and the existing structural theory of three-dimensional point spread function is combined, to three
Dimension point spread function space size is chosen.The it is proposed of this method, gone for the three-dimensional microscopic image of Digital confocal microscope technology
Convolution restoration disposal provides a kind of effective three-dimensional point spread function choosing method.
Brief description of the drawings
Fig. 1 is emulating image.
Feature Points Matching number and 3D-PSF diameters, the number of plies graph of relation when Fig. 2 is 40 times.
Feature Points Matching number increment and 3D-PSF diameters, the number of plies graph of relation when Fig. 3 is 40 times.
Fig. 4 is restoration result figure.
Embodiment
It is three-dimensional to the micro imaging system based on Scale invariant features transform algorithm of the present invention below by way of an example
Point spread function space size choosing method, is described in further detail according to the following steps：
1. three-dimensional point spread function makes
Arrange parameter：Microscope machinery optical tube length is 160mm；Optical source wavelength is 550nm；CCD parameters：1/3 inch, as
Plain value 640 × 480.
Microscope optical system multiplication factor M and digital aperture NA takes arranged below：
M=40 times of multiplication factor；Numerical aperture NA=0.6；
Three-dimensional point spread function layer takes 0.1 μm away from L, makes the 3D-PSF that space size is 21 × 21 × 21, and its radial direction is big
Small is 21 × 21, axial extent 21, with h40_21_21Represent.
With h40_21_21Space center's point (11,11,11) centered on, intercept space size successively to surrounding respectively as 3 × 3
× 3,3 × 3 × 5,3 × 3 × 7 ..., 3 × 3 × 21,5 × 5 × 5,5 × 5 × 7 ..., 21 × 21 × 21, totally 100 three-dimensional points expand
Function is dissipated, is named as h3_3、h3_5、h3_7…、h21_21。
3. three-dimensional samples emulating image f makes
For the original picture rich in detail of two dimension as initial sample, size is 151 × 151 using in Fig. 1 (a), folded by micro rotation
Add and make the three-dimensional artificial sample image f containing 21 width two dimensional images, size is 151 × 151 × 21.
4. Fuzzy Simulation three-dimensional imaging image g40Generation
Use h40_21_21Convolution algorithm is carried out with image f, generation obtains three-dimensional blurred picture g40, Fig. 1 (b) show g40Take
The two dimensional image of central core.
5. three-dimensional artificial image restoration
The 100 three-dimensional point spread functions obtained respectively with step 1 are to Fuzzy Simulation image g40Deconvolute at recovery
Reason, restored method use maximum likelihood method, and iterations is 500 times, and image is represented with fi after recovery.
6. Feature Points Matching
Carry out feature point detection and Feature Points Matching to f, g and fi, recording feature Point matching number, as shown in table 1.
Feature Points Matching number and three-dimensional point spread function diameter, the number of plies data during 1 40 times of table
7. draw Feature Points Matching number and three-dimensional point spread function diameter, number of plies curve
According to the data of table 1, using the number of plies as abscissa, matching is counted out as ordinate, and it is as shown in Figure 2 to draw curve.
8. draw matching number increment and three-dimensional point spread function diameter, number of plies curve
According to the data of table 1, matching points increment is calculated, i.e.,：To each column data in table, subtracted with the value of rear a line before
The value of a line, the value of the first row then subtract the Feature Points Matching number of blurred picture g and picture rich in detail f under corresponding multiplication factor,
Diameter of the obtained matching points increment with three-dimensional point spread function, number of plies drafting curve map, its result is as shown in Figure 3.
9. a pair three-dimensional point spread function space is chosen
At 40 times, as can be seen from Figure 3：When the number of plies is 9 layers, Feature Points Matching number increment reaches minimum first, 11
Layer, 13 layers when, in maximum, reach minimum for the second time at 17 layers.This is illustrated, when the number of plies is smaller, image restoration
Effect is constantly in increase state, and recovery effect has reached some stationary value when the number of plies is 9, and the number of plies continues to increase afterwards,
At 11 layers or 13 layers, recovery effect has reached a higher value again, and afterwards as the number of plies increases, the increase of image restoration effect is slow
It is slow even to keep constant.
Therefore, with reference to existing three-dimensional point spread function structural theory, the selection to three-dimensional point spread function is as follows：It is if multiple
It is " fast browsing " that original, which requires, and the number of plies of three-dimensional point spread function can be selected to be less than 9, and diameter is identical with the number of plies, such as
h40_7_7；Requiring to be " normally checking " if restoring, the number of plies that can select three-dimensional point spread function is 9, and diameter is identical with the number of plies,
Such as h40_9_9；Require to be " explication de texte " if restoring, the number of plies that can select three-dimensional point spread function is 11, diameter and the number of plies
It is identical, such as h40_11_11.The result that restoration disposal obtains of being deconvoluted using this method to emulating image is as shown in Figure 4.
Claims (3)
- A kind of 1. micro imaging system three-dimensional point spread function space size selection side based on Scale invariant features transform algorithm Method, including：1) according to the parameter of micro imaging system, keep optical section layer to determine multiplication factor away from constant, make characterize it is aobvious The large-space three-dimensional point spread function h of micro mirror optical system, interception is entered with its center to h, obtains a series of three for recovery Tie up point spread function hM_r_z, wherein M expression multiplication factors, r expression radial dimensions, the z expression numbers of plies；2) with the clear cytological map of a width As being used as original two-dimensional sample, the sequence two dimensional image that is mutually related is made, and one is built clearly with these sequence two dimensional images Clear simulation sample 3-D view f, Fuzzy Simulation three-dimensional imaging image g is obtained with h and f convolution；3) use step (1) in it is a series of not Same radial dimension, the different numbers of plies are the three-dimensional point spread function of corresponding different-energy size, different spaces respectively to 3-D view g The restoration disposal that deconvolutes is carried out, restoration algorithm uses maximum likelihood method, and the image after recovery is represented with fi；It is characterized in that, further comprising the steps of：A, feature point detection and Feature Points Matching, the number of recording feature Point matching are carried out to f, g and fi；It is described to f, g and fi Feature point detection and Feature Points Matching are carried out, refers to first detect characteristic point to f, g and fi respectively, i.e., symbol is detected in metric space Desired characteristic point is closed, fi and f characteristic point, g and f characteristic point are then subjected to Feature Points Matching, recording feature Point matching Number,B, Feature Points Matching number and three-dimensional point spread function diameter, number of plies graph of relation are drawn；C, matching is drawn to count out incremental rate curve figure, it is theoretical with reference to three-dimensional point spread function existing structure, to three-dimensional point spread function Chosen number space.
- A kind of 2. micro imaging system three-dimensional point spread function based on Scale invariant features transform algorithm according to right wants 1 Number space size choosing method, it is characterised in that the drafting matching is counted out and three-dimensional point spread function diameter, number of plies relation Curve map, refer to will matching count out it is corresponding from the three-dimensional point spread function of the different-diameter of recovery, the different numbers of plies, and with The number of plies is abscissa, and Feature Points Matching number is ordinate, draws the curve under different-diameter.
- A kind of 3. micro imaging system three-dimensional point spread function based on Scale invariant features transform algorithm according to right wants 1 Number space size choosing method, it is characterised in that the drafting matching is counted out incremental rate curve figure, with reference to three-dimensional point spread function Existing structure is theoretical, and three-dimensional point spread function space is chosen, and refers to calculate the increment that matching is counted out, and and three-dimensional point The diameter of spread function, the number of plies are corresponding, and using the number of plies as abscissa, Feature Points Matching number increment is ordinate, draw different Curve under diameter, analyze curve and combine three-dimensional point spread function existing structure theory, three-dimensional point spread function space is entered Row is chosen.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710478941.XA CN107369135A (en) | 2017-06-22 | 2017-06-22 | A kind of micro imaging system three-dimensional point spread function space size choosing method based on Scale invariant features transform algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710478941.XA CN107369135A (en) | 2017-06-22 | 2017-06-22 | A kind of micro imaging system three-dimensional point spread function space size choosing method based on Scale invariant features transform algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107369135A true CN107369135A (en) | 2017-11-21 |
Family
ID=60305001
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710478941.XA Pending CN107369135A (en) | 2017-06-22 | 2017-06-22 | A kind of micro imaging system three-dimensional point spread function space size choosing method based on Scale invariant features transform algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107369135A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108492345A (en) * | 2018-01-24 | 2018-09-04 | 西安理工大学 | A kind of data block division methods based on change of scale |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100074486A1 (en) * | 2006-11-22 | 2010-03-25 | Max-Planck-Gesellschaft Zur Forderung Der Wissenschaften E.V., A Corporation Of Germany | Reconstruction and visualization of neuronal cell structures with bright-field mosaic microscopy |
| CN102073993A (en) * | 2010-12-29 | 2011-05-25 | 清华大学 | Camera self-calibration-based jittering video deblurring method and device |
| CN103942765A (en) * | 2014-05-09 | 2014-07-23 | 广西大学 | Three-dimensional point spread function space size selection method based on restoration efficiency |
| CN104008539A (en) * | 2014-05-29 | 2014-08-27 | 西安理工大学 | Image super-resolution rebuilding method based on multiscale geometric analysis |
| CN104123704A (en) * | 2014-07-21 | 2014-10-29 | 广西大学 | Microscopy imaging system three dimension point spread function selection method based on energy distribution |
| CN104751425A (en) * | 2015-03-25 | 2015-07-01 | 北京工商大学 | Fluorescent microscopic image rebuilding method and system based on space variation point spread function |
| CN105488780A (en) * | 2015-03-25 | 2016-04-13 | 遨博（北京）智能科技有限公司 | Monocular vision ranging tracking device used for industrial production line, and tracking method thereof |
| CN105957036A (en) * | 2016-05-06 | 2016-09-21 | 电子科技大学 | Video motion blur removing method strengthening character prior |
| CN106067166A (en) * | 2016-06-03 | 2016-11-02 | 广西大学 | A kind of acquisition method of micro imaging system three-dimensional point spread function minimum space collection |
| CN106355567A (en) * | 2016-10-29 | 2017-01-25 | 广西大学 | Interval estimation-based method for selecting three-dimensional point spread functions of microscopic imaging system |
| CN106709879A (en) * | 2016-12-08 | 2017-05-24 | 中国人民解放军国防科学技术大学 | Spatial variation point diffusion function smoothing method based on simple lens calculating imaging |
-
2017
- 2017-06-22 CN CN201710478941.XA patent/CN107369135A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100074486A1 (en) * | 2006-11-22 | 2010-03-25 | Max-Planck-Gesellschaft Zur Forderung Der Wissenschaften E.V., A Corporation Of Germany | Reconstruction and visualization of neuronal cell structures with bright-field mosaic microscopy |
| CN102073993A (en) * | 2010-12-29 | 2011-05-25 | 清华大学 | Camera self-calibration-based jittering video deblurring method and device |
| CN103942765A (en) * | 2014-05-09 | 2014-07-23 | 广西大学 | Three-dimensional point spread function space size selection method based on restoration efficiency |
| CN104008539A (en) * | 2014-05-29 | 2014-08-27 | 西安理工大学 | Image super-resolution rebuilding method based on multiscale geometric analysis |
| CN104123704A (en) * | 2014-07-21 | 2014-10-29 | 广西大学 | Microscopy imaging system three dimension point spread function selection method based on energy distribution |
| CN104751425A (en) * | 2015-03-25 | 2015-07-01 | 北京工商大学 | Fluorescent microscopic image rebuilding method and system based on space variation point spread function |
| CN105488780A (en) * | 2015-03-25 | 2016-04-13 | 遨博（北京）智能科技有限公司 | Monocular vision ranging tracking device used for industrial production line, and tracking method thereof |
| CN105957036A (en) * | 2016-05-06 | 2016-09-21 | 电子科技大学 | Video motion blur removing method strengthening character prior |
| CN106067166A (en) * | 2016-06-03 | 2016-11-02 | 广西大学 | A kind of acquisition method of micro imaging system three-dimensional point spread function minimum space collection |
| CN106355567A (en) * | 2016-10-29 | 2017-01-25 | 广西大学 | Interval estimation-based method for selecting three-dimensional point spread functions of microscopic imaging system |
| CN106709879A (en) * | 2016-12-08 | 2017-05-24 | 中国人民解放军国防科学技术大学 | Spatial variation point diffusion function smoothing method based on simple lens calculating imaging |
Non-Patent Citations (6)
| Title |
|---|
| BO FU 等: "Motion-blurred SIFT invariants based on sampling in image deformation space and univariate search", 《IET COMPUTER VISION》 * |
| 何淑珍 等: "基于灰度梯度的散焦图像测距算法的研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 刘政 等: "基于图像深度信息的尺度不变特征算法误匹配点对剔除", 《计算机应用》 * |
| 杨凤娟: "三维点扩散函数空间大小选取方法的研究", 《万方数据库》 * |
| 蔡熠: "三维点扩散函数能量分布与选取方法研究", 《万方数据库》 * |
| 陈华 等: "基于能量分布的显微成像***3D-PSF选取方法", 《光学学报》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108492345A (en) * | 2018-01-24 | 2018-09-04 | 西安理工大学 | A kind of data block division methods based on change of scale |
| CN108492345B (en) * | 2018-01-24 | 2021-12-03 | 广州华资软件技术有限公司 | Data block dividing method based on scale transformation |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Bharat et al. | Resolving macromolecular structures from electron cryo-tomography data using subtomogram averaging in RELION | |
| Möckl et al. | Accurate phase retrieval of complex 3D point spread functions with deep residual neural networks | |
| Shroff et al. | Image formation analysis and high resolution image reconstruction for plenoptic imaging systems | |
| Di Criscienzo et al. | The ASTRODEEP Frontier Fields catalogues-III. Multiwavelength photometry and rest-frame properties of MACS-J0717 and MACS-J1149 | |
| CN104123704A (en) | Microscopy imaging system three dimension point spread function selection method based on energy distribution | |
| Holmes | Expectation-maximization restoration of band-limited, truncated point-process intensities with application in microscopy | |
| Harpsøe et al. | High frame rate imaging based photometry-Photometric reduction of data from electron-multiplying charge coupled devices (EMCCDs) | |
| Lee et al. | Depth extraction of three-dimensional objects using block matching for slice images in synthetic aperture integral imaging | |
| US8965133B2 (en) | Image processing apparatus and control method therefor | |
| US20160110324A1 (en) | Compression of cascading style sheet files | |
| FR2966951A1 (en) | SIMULATION METHOD FOR DETERMINING AERODYNAMIC COEFFICIENTS OF AN AIRCRAFT | |
| Harrison et al. | DAVIS: A direct algorithm for velocity-map imaging system | |
| CN107369135A (en) | A kind of micro imaging system three-dimensional point spread function space size choosing method based on Scale invariant features transform algorithm | |
| Yang et al. | JPEG steganalysis with combined dense connected CNNs and SCA-GFR | |
| CN111797971A (en) | Method, device and electronic system for processing data by using convolutional neural network | |
| Clarkson et al. | Quantifying the loss of information from binning list-mode data | |
| McCallum et al. | Simultaneous reconstruction of object and aperture functions from multiple far-field intensity measurements | |
| Xu et al. | Accelerating wavefront-sensing-based autofocusing using pixel reduction in spatial and frequency domains | |
| WO2009126427A2 (en) | Image descriptor quantization | |
| CN103942765A (en) | Three-dimensional point spread function space size selection method based on restoration efficiency | |
| CN106355567B (en) | The choosing method of micro imaging system three-dimensional point spread function based on interval estimation | |
| Barnard et al. | Total variation-based neutron computed tomography | |
| Li et al. | Method of super-resolution based on array detection and maximum-likelihood estimation | |
| Zhao et al. | Improved vector-extrapolation-based Richardson–Lucy algorithm used for wavefront coded imaging | |
| Richter et al. | Receptive field refinement for convolutional neural networks reliably improves predictive performance |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20171121 |