CN104123704A - Microscopy imaging system three dimension point spread function selection method based on energy distribution - Google Patents
Microscopy imaging system three dimension point spread function selection method based on energy distribution Download PDFInfo
- Publication number
- CN104123704A CN104123704A CN201410348095.6A CN201410348095A CN104123704A CN 104123704 A CN104123704 A CN 104123704A CN 201410348095 A CN201410348095 A CN 201410348095A CN 104123704 A CN104123704 A CN 104123704A
- Authority
- CN
- China
- Prior art keywords
- point spread
- spread function
- dimensional point
- dimensional
- energy
- 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
Landscapes
- Microscoopes, Condenser (AREA)
Abstract
Disclosed is a microscopy imaging system three dimension point spread function selection method based on energy distribution. The microscopy imaging system three dimension point spread function selection method based on the energy distribution includes following steps: using sequence three dimension point spread functions different in energy to perform an image restoration simulation experiment; obtaining a relation among the three dimension point spread functions different in energy, an image restoration effect and restoration time so as to calculate restoration efficiency, draw a restoration efficiency curve, calculate a second derivative, and obtain knee points of the curve; confirming an energy threshold value according to coordinates of the knee points, and then confirming an origin three dimension point spread function; selecting the intensity of three dimension point spread function energy and the size of a space corresponding to the three dimension point spread function energy of the intensity according to different requirements of general observational browse and analytical measurement. The microscopy imaging system three dimension point spread function selection method based on the energy distribution can overall consider the image restoration effect and processing time according to the different requirements of the observational browses and the analytical measurement, selects the corresponding three dimension point spread function, and performs restoration treatment on a collected sequence microscopic section image.
Description
One, technical field
The present invention relates to a kind of micro imaging system three-dimensional point spread function choosing method based on energy distribution, the method is an important step of three-dimensional biological micro-image restoration disposal in Digital confocal microscope technology, belongs to digital image restoration processing technology field.The application of the method, can browse the difference needs with analysis to measure according to overview, in conjunction with the comprehensive balance in recovery effect and processing time, considers, makes three-dimensional point spread function different-energy size and choosing of space size correspondingly.
Background technology
Digital confocal microscope technology be take common Photobiology microscope as basis, configuration image detector, accurate moving control mechanism and computer, adopt digital image processing techniques, the biological specimen micro-image gathering is carried out to restoration disposal, eliminate the focal plane impact defocusing in addition, to improve the resolution of cell image.
Restoration disposal in Digital confocal microscope technology, employing be the three-dimensional micro-image restored method that deconvolutes.Characterize the three-dimensional point spread function of microscope optical system, the effect of the restoration disposal that directly determining to deconvolute.Three-dimensional point spread function reflects the energy distribution of microscope optical system more exactly, and recovery effect is better.On this basis, the space size of three-dimensional point spread function is chosen larger, comprises energy larger, and recovery effect is better, and the processing time is longer simultaneously.The three-dimensional point spread function maximum space size that comprises whole energy is sample volume 2 times, this is a very huge space.Three-dimensional point spread function is to pushing up two funnel structures, the tiny area that most concentration of energy are located top at two funnels at middle part.Therefore,, when carrying out image restoration and process, reasonably choosing is to take the three-dimensional point spread function that this place chooses the certain space size of tiny area as initial point to carry out image restoration processing.Research shows, along with the increase gradually of space size, recovery effect promotes and progressively eases up, and the processing time increases sharply, and restores decrease in efficiency.Therefore, how to browse according to the observation the difference needs with analysis to measure, recovery effect and processing time are considered in comprehensive and balance, choosing different energy size and then choose correspondingly different spaces size three-dimensional point spread function and carry out restoration disposal, is the three-dimensional micro-image major issue that restoration disposal need to solve of deconvoluting in digital confocus technology.
Three, summary of the invention
The object of this invention is to provide a kind of micro imaging system three-dimensional point spread function choosing method based on energy distribution, the method can be browsed the difference needs with analysis to measure according to the observation, consider recovery effect and processing time, choose corresponding energy size and the three-dimensional point spread function of space size correspondingly, the sequence microsection image gathering is carried out to restoration disposal.
The present invention achieves the above object by the following technical programs: a kind of micro imaging system three-dimensional point spread function choosing method based on energy distribution, comprises the steps:
(1) according to micro objective enlargement factor, numerical-aperture parameters, optical source wavelength, camera target surface size and resolution parameter, and set an optical section layer distance, make one for representing that imaging process carries out the three-dimensional point spread function h of the sign microscope optical system of convolution algorithm
n, and make the sequence three-dimensional point spread function h for image restoration of different spaces size
i, difference called after h
3, h
5, h
7..., h
n-2;
(2) by each different spaces size three-dimensional point spread function h obtaining in formula (1) calculation procedure (1)
nand h
ienergy value E, and be normalized;
(3) with a clear high s/n ratio cell image as original two-dimensional sample, make the sequence two dimensional image that is mutually related, and with a clear simulation sample 3-D view f of these sequence two dimensional images structures;
(4) use h
nobtain Fuzzy Simulation three-dimensional imaging image g with f convolution;
(5) use the three-dimensional point spread function h of different-energy size in step (1)
ito the 3-D view g restoration disposal that deconvolutes, restoration algorithm adopts maximum likelihood method, records recovery effect and processing time simultaneously respectively;
(6) the deconvolute evaluation of recovery effect of restoration disposal, employing formula (2) improve signal to noise ratio (S/N ratio) ISNR, processing time unit adopts second, calculates different-energy three-dimensional point spread function h
ithree-dimensional imaging image g is carried out to the recovery efficiency q of restoration disposal;
(7), to the ENERGY E of each three-dimensional point spread function and restore accordingly efficiency q and carry out data fitting, obtain q-E relation curve;
(8) ask the second derivative d of q to E
2q/dE
2, finding out second derivative is zero knee point γ, by energy settings corresponding with flex point on rectangular coordinate transverse axis, is threshold energy E
γ, with this, determine " starting point three-dimensional point spread function ";
(9) choosing of three-dimensional point spread function, if restored image is browsed for overview, the three-dimensional point spread function of choosing " starting point three-dimensional point spread function " or slightly large space size carries out image restoration.If restored image, for research, analysis to measure, is chosen the three-dimensional point spread function of greater room.Effect requirements to restored image is higher, chooses the more three-dimensional point spread function of large space size.
Described for representing that imaging process carries out the three-dimensional point spread function h of the sign microscope optical system of convolution algorithm
n, be the n rank three-dimensional matrice that a space size is n * n * n, n is double-digit odd number.
The sequence three-dimensional point spread function h for image restoration of described making different spaces size
i, be with h
nspace center's point be initial point, to surrounding ascending respectively by 3 * 3 * 3,5 * 5 * 5,7 * 7 * 7 ..., (n-2) * (n-2) * (n-2) three-dimensional matrice of intercepting.
The described sequence three-dimensional point spread function h for image restoration
i, their normalized energy value is with h
nenergy value be that benchmark is normalized and obtains.
Described recovery efficiency q, the definition q=ISNR/t being proposed by the present invention calculates, and wherein t is with minimum space size three-dimensional point spread function h
3processing time be the normalization restoration disposal time that radix is normalized.
Described " starting point three-dimensional point spread function ", refers to that energy value E is greater than the three-dimensional point spread function of the minimum space of threshold energy E γ.
Outstanding effect of the present invention is:
The difference of the energy size of three-dimensional point spread function is chosen and the difference of corresponding space size is chosen, and is directly determining that in Digital confocal microscope technology, three-dimensional micro-image deconvolutes recovery effect and processing time.How to choose, be Digital confocal microscope technology problem to be solved always.The present invention is based on the research of three-dimensional point spread function energy distribution, proposed by the calculating of image restoration emulation experiment and recovery efficiency and knee point thereof, determine energy threshold, and then definite starting point three-dimensional point spread function, browse according to the observation the difference needs with analysis to measure, choose the method for different-energy size and the size of different spaces correspondingly three-dimensional point spread function.The proposition of the method, for the three-dimensional micro-image of the Digital confocal microscope technology restoration disposal that deconvolutes provides a kind of effective three-dimensional point spread function choosing method.
Four, accompanying drawing explanation
Fig. 1 emulating image.
Fig. 2 q-E matched curve.
Fig. 3 analogous diagram is restored result.
The actual cytological map of Fig. 4 restores result.
Five, embodiment
By example, technical scheme of the present invention is described in further detail below.
Micro imaging system three-dimensional point spread function choosing method based on energy distribution of the present invention, comprises the following steps:
1. three-dimensional point spread function is made
Parameters: microscope machinery optical tube length is 160mm; Optical source wavelength is 550nm; CCD parameter: 1/3 inch, pixel value 640 * 480.
Microscope optical system enlargement factor M and digital aperture NA get following setting:
Enlargement factor M=40 doubly; Numerical aperture NA=0.6;
Three-dimensional point spread function layer is got 0.3125 μ m apart from L, the 3D-PSF that making space size is 21 * 21 * 21, and its radial dimension is 21 * 21, axial extent is 21, with h
21represent.
With h
21space center's point (11,11,11) centered by, respectively to surrounding intercept successively space size be 3 * 3 * 3,5 * 5 * 5,7 * 7 * 7 ..., 9 and 21 * 21 * 21 such as 19 * 19 * 19 totally 10 three-dimensional point spread functions itself, called after h
3, h
5, h
7..., h
19.
2. three-dimensional point spread function energy calculates
According to the energy value E of formula (1) 10 three-dimensional point spread functions that calculation procedure 1 obtains respectively.The energy value of each three-dimensional point spread function is all with h
21energy value is that benchmark is normalized.
3. three-dimensional samples emulating image f makes
Using that two-dimentional original picture rich in detail is as initial sample in Fig. 1 (a), size is 151 * 151, rotates stack make the three-dimensional artificial sample image f containing 21 width two dimensional images by trace, and size is 151 * 151 * 21.
4. Fuzzy Simulation three-dimensional imaging image g
40generate
Use h
21carry out convolution algorithm with image f, generate and obtain three-dimensional blurred picture g
40, Fig. 1 (b) is depicted as g
40get the two dimensional image of central core.
5. three-dimensional artificial image restoration
10 three-dimensional point spread functions that obtain by step 1 are respectively to Fuzzy Simulation image g
40the restoration disposal that deconvolutes, restored method adopts maximum likelihood method, and iterations is 600 times.Simultaneously the recording processing time, processing time unit is second.
6. restore the calculating of experiment
According to formula (2), calculate respectively 10 three-dimensional point spread functions to image restoration experiment resulting improvement signal to noise ratio (S/N ratio) ISNR.And with minimum space size three-dimensional point spread function h
3processing time be that benchmark carries out time normalization, calculate each three-dimensional point spread function and 3-D view g carried out to the recovery efficiency q:q=ISNR/t of restoration disposal, t is the normalization restoration disposal time.
The recovery experimental calculation such as the space size size of 10 three-dimensional point spread functions, energy value E, recovery effect ISNR and recovery efficiency q the results are shown in table 1.Wherein space size size represents with pixel count.In table 1,3D-PSF is the english abbreviation of three-dimensional point spread function.
Three groups of three-dimensional point spread functions of table 1 restore experiment and result of calculation
According to table 1, mark the recovery efficiency q of 10 three-dimensional point spread functions and ENERGY E 10 be related to discrete point a, b ..., j, as shown in Figure 2.
7. restore efficiency q and ENERGY E and be related to matching
According to the table 1 pair relation of restoring between efficiency q and ENERGY E, carry out matching, obtain formula (3), the graph of relation of q-E matching as shown in Figure 2.
q＝1.58E
4-2.14E
3-E
2+1.7E-0.12 (3)
8. knee point position and " starting point three-dimensional point spread function " determines
According to formula (3), ask respectively the second derivative d of q to E
2q/dE
2, making second derivative is zero, tries to achieve point of inflexion on a curve λ, corner position is (0.807,0.1477), corresponding energy threshold E
γbe 0.807, as shown in Figure 2.
Check table 1, energy is greater than energy threshold E
γthe three-dimensional point spread function of minimum space be h
9, this three-dimensional point spread function is decided to be " starting point three-dimensional point spread function ".
9. three-dimensional point spread function chooses
If restored image is browsed for overview, choose " starting point three-dimensional point spread function " h
9or slightly the three-dimensional point spread function of large space size such as h
11carry out image restoration.If restored image, for research, analysis to measure, is chosen the three-dimensional point spread function of greater room such as h
15carry out image restoration.Effect requirements to restored image is higher, chooses the more three-dimensional point spread function of large space size.
Adopt " starting point three-dimensional point spread function " h
9and h
7, h
15emulating image is deconvoluted to result that restoration disposal obtains as shown in Figure 3, actual cell image is deconvoluted to result that restoration disposal obtains as shown in Figure 4.
Claims (6)
1. the micro imaging system three-dimensional point spread function choosing method based on energy distribution, is characterized in that, comprises the steps:
(1) according to micro objective enlargement factor, numerical-aperture parameters, optical source wavelength, camera target surface size and resolution parameter, and set an optical section layer distance, make one for representing the three-dimensional point spread function h of the sign microscope optical system of imaging process convolution algorithm
n, and make the sequence three-dimensional point spread function h for image restoration of different spaces size
i, difference called after h
3, h
5, h
7..., h
n-2;
(2) by each different spaces size three-dimensional point spread function h obtaining in formula (1) calculation procedure (1)
nand h
ienergy value E, and be normalized,
(3) with a clear high s/n ratio cell image as original two-dimensional sample, make the sequence two dimensional image that is mutually related, and with a clear simulation sample 3-D view f of these sequence two dimensional images structures;
(4) use h
nobtain Fuzzy Simulation three-dimensional imaging image g with f convolution;
(5) with different spaces size in step (1) be the three-dimensional point spread function of corresponding different-energy size respectively to the 3-D view g restoration disposal that deconvolutes, restoration algorithm adopts maximum likelihood method, records recovery effect and processing time simultaneously;
(6) the deconvolute evaluation of recovery effect of restoration disposal, employing formula (2) improve signal to noise ratio (S/N ratio) ISNR:
(7), to the ENERGY E of each three-dimensional point spread function and restore accordingly efficiency q and carry out data fitting, obtain q-E relation curve;
(8) ask the second derivative d of q to E
2q/dE
2, finding out second derivative is zero knee point γ, by energy settings corresponding with flex point on rectangular coordinate transverse axis, is threshold energy E γ, with this, determines " starting point three-dimensional point spread function ";
(9) three-dimensional point spread function chooses, if restored image is browsed for overview, the three-dimensional point spread function of choosing " starting point three-dimensional point spread function " or slightly large space size carries out image restoration, if restored image is for research, analysis to measure, choose the three-dimensional point spread function of greater room, effect requirements to restored image is higher, chooses the more three-dimensional point spread function of large space size.
2. the micro imaging system three-dimensional point spread function choosing method based on energy distribution according to claim 1, it is characterized in that, step (1) described for representing that imaging process carries out the three-dimensional point spread function h of the sign microscope optical system of convolution algorithm
n, be the n rank three-dimensional matrice that a space size is n * n * n, n is double-digit odd number.
3. the micro imaging system three-dimensional point spread function choosing method based on energy distribution according to claim 1, is characterized in that, the sequence three-dimensional point spread function h for image restoration of the described making different spaces of step (1) size
i, be with h
nspace center's point be initial point, to surrounding ascending respectively by 3 * 3 * 3,5 * 5 * 5,7 * 7 * 7 ..., (n-2) * (n-2) * (n-2) three-dimensional matrice of intercepting.
4. the micro imaging system three-dimensional point spread function choosing method based on energy distribution according to claim 1, is characterized in that, the sequence three-dimensional point spread function h for image restoration that step (1) is described
i, their normalized energy value is with h
nenergy value be that benchmark is normalized and obtains.
5. the micro imaging system three-dimensional point spread function choosing method based on energy distribution according to claim 1, it is characterized in that, the recovery efficiency q that step (6) is described, is calculated by q=ISNR/t, and wherein t is with minimum space size three-dimensional point spread function h
3processing time be the normalization restoration disposal time that radix is normalized.
6. the micro imaging system three-dimensional point spread function choosing method based on energy distribution according to claim 1, is characterized in that, " the starting point three-dimensional point spread function " that step (8) is described, refers to that energy value E is greater than threshold energy E
γthe three-dimensional point spread function of minimum space.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410348095.6A CN104123704A (en) | 2014-07-21 | 2014-07-21 | Microscopy imaging system three dimension point spread function selection method based on energy distribution |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410348095.6A CN104123704A (en) | 2014-07-21 | 2014-07-21 | Microscopy imaging system three dimension point spread function selection method based on energy distribution |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104123704A true CN104123704A (en) | 2014-10-29 |
Family
ID=51769103
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410348095.6A Pending CN104123704A (en) | 2014-07-21 | 2014-07-21 | Microscopy imaging system three dimension point spread function selection method based on energy distribution |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104123704A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| CN107369135A (en) * | 2017-06-22 | 2017-11-21 | 广西大学 | A kind of micro imaging system three-dimensional point spread function space size choosing method based on Scale invariant features transform algorithm |
| CN108088660A (en) * | 2017-12-15 | 2018-05-29 | 清华大学 | The point spread function measuring method and system of wide field fluorescence microscope |
| CN111157226A (en) * | 2020-03-04 | 2020-05-15 | 山东理工大学 | Method and device for measuring point spread function of microscope |
| CN111685788A (en) * | 2020-06-09 | 2020-09-22 | 广东明峰医疗科技有限公司 | Method for improving signal-to-noise ratio of PET |
| CN113228099A (en) * | 2019-01-09 | 2021-08-06 | 爱克发有限公司 | Method and system for computing point spread function of digital image detector system based on quantum noise measurement of MTF modulation |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1900971A (en) * | 2005-07-18 | 2007-01-24 | 北京航空航天大学 | Improved NAS RIF blind image recovery method |
| WO2008020109A1 (en) * | 2006-08-03 | 2008-02-21 | Consejo Superior De Investigaciones Científicas | Method for restoration of images which are affected by imperfections, device for implementation of this, and the corresponding applications |
-
2014
- 2014-07-21 CN CN201410348095.6A patent/CN104123704A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1900971A (en) * | 2005-07-18 | 2007-01-24 | 北京航空航天大学 | Improved NAS RIF blind image recovery method |
| WO2008020109A1 (en) * | 2006-08-03 | 2008-02-21 | Consejo Superior De Investigaciones Científicas | Method for restoration of images which are affected by imperfections, device for implementation of this, and the corresponding applications |
Non-Patent Citations (2)
| Title |
|---|
| CHEN HUA ET AL.: "Relationship between point spread function layers and biological microscopic image restoration based on the same space size", 《JOURNAL OF CHEMICAL AND PHARMACEUTICAL RESEARCH》 * |
| CHEN HUA ET AL.: "Studies on three dimensional point spread function energy distribution of biological microscopic imaging system", 《JOURNAL OF CHEMICAL AND PHARMACEUTICAL RESEARCH》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| CN106355567B (en) * | 2016-10-29 | 2019-06-04 | 广西大学 | The choosing method of micro imaging system three-dimensional point spread function based on interval estimation |
| CN107369135A (en) * | 2017-06-22 | 2017-11-21 | 广西大学 | A kind of micro imaging system three-dimensional point spread function space size choosing method based on Scale invariant features transform algorithm |
| CN108088660A (en) * | 2017-12-15 | 2018-05-29 | 清华大学 | The point spread function measuring method and system of wide field fluorescence microscope |
| CN108088660B (en) * | 2017-12-15 | 2019-10-29 | 清华大学 | The point spread function measurement method and system of wide field fluorescence microscope |
| CN113228099A (en) * | 2019-01-09 | 2021-08-06 | 爱克发有限公司 | Method and system for computing point spread function of digital image detector system based on quantum noise measurement of MTF modulation |
| CN111157226A (en) * | 2020-03-04 | 2020-05-15 | 山东理工大学 | Method and device for measuring point spread function of microscope |
| CN111685788A (en) * | 2020-06-09 | 2020-09-22 | 广东明峰医疗科技有限公司 | Method for improving signal-to-noise ratio of PET |
| CN111685788B (en) * | 2020-06-09 | 2022-07-29 | 广东明峰医疗科技有限公司 | Method for improving signal-to-noise ratio of PET |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104123704A (en) | Microscopy imaging system three dimension point spread function selection method based on energy distribution | |
| Broxton et al. | Wave optics theory and 3-D deconvolution for the light field microscope | |
| CN110441271B (en) | Light field high-resolution deconvolution method and system based on convolutional neural network | |
| JP4806630B2 (en) | A method for acquiring optical image data of three-dimensional objects using multi-axis integration | |
| EP2648206B1 (en) | Transmission electron microscope for reconstructing a three-dimensional structure and transmission electron microscope image reconstruction method | |
| JP2017063414A (en) | Method and apparatus for generating data representative of light field | |
| CN113256772B (en) | Double-angle light field high-resolution reconstruction system and method based on visual angle conversion | |
| CN106779055B (en) | Image characteristic extracting method and device | |
| Li et al. | Efficient assessment method of on-board modulation transfer function of optical remote sensing sensors | |
| JP2018538709A5 (en) | ||
| JP2015108837A (en) | Image processing apparatus and image processing method | |
| CN110070503A (en) | Scale calibration method, system and medium based on convolutional neural networks | |
| Lee et al. | Improving focus measurement via variable window shape on surface radiance distribution for 3D shape reconstruction | |
| JP2018512070A (en) | A method for determining the complex amplitude of an electromagnetic field associated with a scene | |
| Zhao et al. | Faster super-resolution imaging with auto-correlation two-step deconvolution | |
| CN107590787B (en) | Image distortion correction method of scanning electron microscope | |
| Viganò et al. | Tomographic approach for the quantitative scene reconstruction from light field images | |
| CN103942765A (en) | Three-dimensional point spread function space size selection method based on restoration efficiency | |
| Ohashi et al. | Joint estimation of high resolution images and depth maps from light field cameras | |
| Ghosh et al. | Three-dimensional block-based restoration integrated with wide-field fluorescence microscopy for the investigation of thick specimens with spatially variant refractive index | |
| EP3144888A1 (en) | An apparatus and a method for generating data representing a pixel beam | |
| Kong et al. | A light field measurement system through PSF estimation by a morphology-based method | |
| Mahmood et al. | 3D shape recovery from image focus using Gabor features | |
| CN114157853A (en) | Apparatus and method for generating a data representation of a pixel beam | |
| Cao et al. | Method based on bioinspired sample improves autofocusing performances |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20141029 |
|
| WD01 | Invention patent application deemed withdrawn after publication |