CN112485230A - Super-resolution microscopic imaging method and device based on active time modulation frequency mixing excitation irradiation - Google Patents

Abstract

The invention belongs to the technical field of optical microscopy, and particularly relates to a super-resolution microscopic imaging method and device based on active time modulation frequency mixing excitation irradiation. According to the invention, a spatial light modulation technology is adopted to modulate each element on a pixel matrix according to different frequencies, and finally, a model established by an algorithm is used for analyzing image information collected by a CCD detector through an inverted microscopic imaging system to obtain a sample information value covered in the image information, so that the function of super-resolution imaging is realized; according to the invention, the sample information can be finally obtained only by analyzing the frequency information contained in each pixel point through the model constructed by the algorithm, the sample does not need to be dyed, the device is convenient to construct, the operation is simple, the cost is low, and the method can be applied to the research of various optical super-resolution cell biological imaging.

Description

Super-resolution microscopic imaging method and device based on active time modulation frequency mixing excitation irradiation

Technical Field

The invention belongs to the technical field of optical microscopy, and particularly relates to a super-resolution microscopic imaging method and device.

Background

Human exploration of the microcosm originated from the invention of the first microscope (1665), and then the requirement of the microscope for resolution became higher with the continuous improvement of the manufacturing process of the microscope and the increasing curiosity of the microcosm, but the resolution of the optical microscope has been stopped around 200nm for decades due to the existence of diffraction limit. In order to study organism structures below 200nm, especially biological structures at the cellular level containing subcellular structures, the diffraction limit of an optical microscope must be broken so that the microscopic resolution is below 200 nm. The probel prize was awarded to Eric Betzig, Stefan w. Several winners skillfully design a method for avoiding diffraction limit, and research thereof breakthroughs the optical microscope into nanometer dimension.

However, super-resolution microscopy technologies such as PALM, STORM and nonlinear SIM have many limitations, and each technology needs a special fluorescent dye and an excitation mode. Whereas the STED technique has a wide range of dye selection but high stability requirements. There is no special requirement for dyes for SIM technology, but resolution of around 100nm is only possible. In recent years, methods for realizing super-resolution optical imaging based on optical fluctuation, such as the SOFI, and 3B derived therefrom, realize super-resolution images of around 50nm based on analysis of fluctuation of fluorescent dyes in time series. All the above technologies are realized by fluorescence technology, and there is no suitable super-resolution microscopy technology in the technology of transmitted light or scattered light imaging.

Disclosure of Invention

Aiming at the defects of the existing super-resolution technology, the invention provides a super-resolution microscopic imaging method and device based on active time modulation frequency mixing excitation irradiation.

According to the invention, a spatial light modulation technology is adopted to modulate each element on the pixel matrix according to different frequencies, and finally, a model established by an algorithm is used for analyzing the image information collected by the CCD detector through an inverted microscopic imaging system to obtain a sample information value covered in the image information, so that the function of super-resolution imaging is realized. The method comprises the following specific steps:

assuming that the modulation frequency v (x, y) of the modulation is set at the position (x, y) of the excitation light source array, the position (x) is₀,y₀) The molecular excitation spectrum of (a) is:

ΣI(x-x₀,y-y₀)·v(x,y)， (1)

where I is the intensity of light at position (x, y) versus position (x)₀,y₀) Influence distribution function of (1);

(1) and for each array pixel light source position, modulating by using low-frequency exciting light in a manner that:

I_{Laser_Power}(x_Laser,y_Laser,v)＝I_{Laser_Power}(x_Laser,y_Laser)·sin(2πv(x_Laser,y_Laser)) (2)

wherein, I_{Laser_Power}(x_Laser,y_Laser) Is the laser intensity at the (x, y) position, v (x)_Laser,y_Laser) Is the modulation frequency that modulates the laser amplitude according to a sinusoidal function.

In a single pixel (x)₀,y₀) The sample on the sample is influenced by different low-frequency light intensity modulation laser beams around, and the emission light intensity of the sample is as follows:

wherein, I_{Laser_Power}(x_Laser-x₀,y_Laser-y₀) Is the laser excitation position (x)_Laser,y_Laser) For position (x)₀,y₀) P (x) of the light intensity contribution function of₀,y₀) Is a sample pairThe intensity amplitude of the light.

Finally, through diffraction, the light intensity of any pixel point (x, y) is influenced by molecules in the surrounding diffraction range.

(2) Shooting a video for a period of time through pixel array modulation excitation; then, the two-dimensional image data (x, y) is developed into a three-dimensional or four-dimensional video image (x, y, ω) by analyzing each pixel by a time Fourier transform_x,ω_y) Wherein, ω is_x,ω_yRespectively, in the (x, y) direction, it can be seen that not only the position information (x, y) but also the two-dimensional frequency information (ω), is comprised here_x,ω_y)；

(3) Finally, through Fourier transformation, pixel information corresponding to each modulation frequency is extracted, an equation model is constructed, and therefore specific (frequency) information under a single pixel of the super-resolution image which is covered by diffraction is analyzed.

In step (3), the process of constructing an equation model and analyzing specific (frequency) information of the super-resolution image under a single pixel, which is covered by diffraction, is as follows:

the method comprises the following steps: the following assumptions were made from the actual optical system model:

(1) assuming that the change amount of the lens to the phase factor is zero, the lens is regarded as a thin lens;

(2) considering the imaging condition of incoherent light source due to different frequency of each light source, the imaging process is equivalent to an optical transfer function, and diffraction is caused by an entrance pupil or an exit pupil;

(3) the sample surface and the CCD receiving surface do not consider the loss of high-frequency information, and the diffusion matrix is approximate to a function in a Gaussian matrix form;

(4) the time function of each pixel can be completely recovered on the assumption that enough images are acquired;

step two: according to the assumption that the actual optical system is simulated, the specific frequency analysis algorithm process is as follows:

each pixel is written in the form of a matrix of pixels, if there is a value at the corresponding location, indicating that there is a corresponding frequency spread over the pixel. And setting that each initial pixel only has the corresponding frequency, mixing different frequency components on a matrix in a diffusion range through one-time diffusion convolution calculation, and carrying the upper weight factors by the different frequency components according to the corresponding positions of the diffusion matrix. The values at each pixel are as follows:

where (2 × a +1) is the dimension of the diffusion matrix, which is equal to 1/3 (if the pixel matrix is set to 100, the diffusion matrix is 33 × 33, i.e., a — 16) about the dimension of the pixel matrix; the size of the diffusion lift-off matrix D is (2a +1) × (2a +1), (x, y) are pixel positions;

and carrying sample information on the pixel position by the pixel on different positions obtained after the primary diffusion point diffusion function. After carrying the sample information, performing diffusion convolution calculation on all pixels again, and for each pixel, overlapping the contribution of the surrounding pixels to the pixel according to a diffusion matrix to obtain an actual frequency value on each pixel; and after two times of diffusion convolution calculation, adding all elements on the obtained pixels to obtain the final expression of each pixel containing the sample information.

Finally the values at all pixel positions are added:

where n x n is the image matrix dimension.

Then Fourier transform is carried out according to the designed frequency value, and then a proper range integral is taken for a given frequency, namely the following processing is carried out:

firstly, Fourier transform (formula (6)) is carried out, and then the screening property of the impact function is utilized to obtain the intensity information value (formula (7)) carried by the specific frequency from the integral frequency band; wherein, ω is_n,mChanging the frequency at the position of the matrix (n, m) into a delta shock function by utilizing cosine function Fourier change, wherein the screening property of the shock function can obtain amplitude information contained in one frequency band information through integration; a and b are respectively the minimum value and the maximum value of the modulation frequency band.

Each frequency value corresponds to an equation (diffraction superposition) that contains information for the sample at multiple locations. By processing the formula (5) by the formulas (6) and (7) (also called frequency analysis algorithm), the formula carrying the image information variable corresponding to each modulation frequency can be obtained, and since the dimension of the image matrix is n × n, n × n formulas containing unknown numbers are finally obtained. Analyzing all pixel information by the image time sequence set obtained by the CCD detector according to the formula (6) and the formula (7), obtaining an actual value corresponding to each modulation frequency information, forming an equation with the formula obtained by the corresponding modulation frequency obtained in the previous step, and forming a linear equation set by n equations; and solving the linear equation set to obtain a final result, and restoring a sample information value.

Based on the method, the invention also provides a super-resolution microscopic imaging device based on the active time modulation mixing excitation irradiation. The super-resolution microscopic imaging device comprises: the system comprises alaser light source 1, afirst lens 2, asecond lens 3, afirst reflector 4, asecond reflector 5, athird lens 6, afourth lens 7, athird reflector 8, a firstlinear polarizer 9, a half-wave plate 10, a spatial light modulator 11, a secondlinear polarizer 12, afifth lens 13, a half-mirror 14, anobjective lens 14, an objective table 16, afourth reflector 17 and aCCD detector 18; the components are connected in turn by optical paths to form a super-resolution microscopic imaging device; wherein:

laser output from thelaser light source 1 sequentially passes through thefirst lens 2 and thesecond lens 3 for beam expansion, and the expanded light sequentially passes through thefirst reflector 4 and thesecond reflector 5 and enters thethird lens 6 and thefourth lens 7 for secondary beam expansion; the twice-expanded laser enters a firstlinear polarizer 9 through athird reflector 8, and linear polarized light obtained through the first linear polarizer is rotated through a half-wave plate 10 until the linear polarized light is parallel to the long edge of the spatial light modulator; at the moment, the incident light can be modulated by the spatial light modulator 11, the emergent light enters thefifth lens 13 after being adjusted by the secondlinear polarizer 12, and parallel light is focused on the back focal plane of the lens; part of the light enters theobjective lens 15 through thehalf mirror 14 and is irradiated on the sample on thestage 16, and the light reflected on the sample passes through thehalf mirror 14 and thefourth mirror 17 and enters theCCD detector 18. A continuous time series of raw image sets, each pixel containing different frequency information and sample information, is captured (captured) by theCCD detector 18. And the generated original image sequence is subjected to a frequency analysis algorithm to extract frequency information of the original image set. The original image set also has intensity information data of the detected radiation of the sample, and the frequency analysis algorithm can convert the rule of the intensity information changing along with time into different frequencies with different amplitude intensity information through Fourier transform.

In the invention, after laser output from a laser light source is expanded twice, linear polarized light is obtained through a linear polarizer and is adjusted to a specific position through a half wave plate and then is incident to a spatial light modulator, the emitted modulated light passes through the linear polarizer and then is focused to the back focal plane of an objective lens through a lens, then partial light enters the objective lens through a half-mirror and is irradiated on a sample of an objective table, and reflected light carrying sample information enters a CCD detector through the half-mirror and a reflecting mirror. And analyzing the frequency information on each pixel point by the image information collected by the CCD detector through an algorithm constructed by a computer, and finally restoring the sample information.

According to the invention, the sample information can be finally obtained only by analyzing the frequency information contained in each pixel point through the model constructed by the algorithm, the sample does not need to be dyed, the device is convenient to construct, the operation is simple, the cost is low, and the method can be applied to the research of various optical super-resolution cell biological imaging. The method has the greatest advantage that the sample information can be restored only by algorithm analysis, so that super-resolution microscopic imaging is realized.

Drawings

Fig. 1 is a schematic diagram of the basic structure of the present invention.

Fig. 2 is an image input during simulation of the algorithm according to the present invention.

Fig. 3 is a diagram showing the result finally analyzed during verification of the algorithm simulation according to the present invention.

Fig. 4 is an image obtained by the algorithm of the present invention after two diffusion processes during simulation. Wherein, (a) t is 0, (b) t is 1, and (c) t is 2.

Detailed Description

As shown in fig. 1, the schematic structural diagram of a super-resolution micro-imaging device based on active time modulation mixing excitation irradiation includes: the system comprises alaser light source 1, afirst lens 2, asecond lens 3, afirst reflector 4, asecond reflector 5, athird lens 6, afourth lens 7, athird reflector 8, a firstlinear polarizer 9, a half-wave plate 10, a spatial light modulator 11, a secondlinear polarizer 12, afifth lens 13, a half-mirror 14, anobjective lens 15, an objective table 16, afourth reflector 17 and aCCD detector 18;

the laser beam output from thelaser light source 1 is expanded by thefirst lens 2 and thesecond lens 3, and the expanded light enters thethird lens 6 and thefourth lens 7 via the first reflectingmirror 4 and the second reflectingmirror 5 to be expanded for the second time. The twice expanded laser enters a firstlinear polarizer 9 through athird reflector 8, linear polarized light obtained through the firstlinear polarizer 9 rotates through a half-wave plate 10 until the linear polarized light is parallel to the long side of a spatial light modulator 11, at the moment, incident light can be modulated by the spatial light modulator 11, emergent light enters afifth lens 13 after being adjusted by a secondlinear polarizer 12, parallel light is focused on the back focal plane of the lens, partial light enters anobjective lens 15 through a half-mirror, the partial light irradiates on a sample of an objective table 16, and light reflected on the sample enters aCCD detector 18 through the half-mirror 14 and afourth reflector 17.

When the CCD detector receives the images, a time sequence of images is taken, and the two-dimensional image data (x, y) is developed into a three-dimensional or four-dimensional video image (x, y, omega)_x,ω_y). All pixel positions are correspondedThe information is subjected to Fourier transform, different frequencies are integrated to obtain information values corresponding to the frequencies, each frequency value corresponds to a formula (diffraction superposition) containing a plurality of position sample information, and finally a model established through a corresponding algorithm is used for solving.

In the following embodiment, in a solid laser with a center wavelength of 488 under a normal temperature environment, laser light output from thelaser light source 1 is expanded via thefirst lens 2 and thesecond lens 3, the expanded light enters thethird lens 6 and thefourth lens 7 via the first reflectingmirror 4 and the second reflectingmirror 5 to be expanded for the second time, focal lengths of the second lens and the fourth lens are two times of the first focal length and the third focal length, and a light spot is expanded twice each time due to expansion. The twice-expanded laser enters a firstlinear polarizer 9 through athird reflector 8, linear polarized light obtained through the firstlinear polarizer 9 rotates through a half-wave plate 10 until the linear polarized light is parallel to the long edge of the spatial light modulator 11, at the moment, incident light can be modulated by the spatial light modulator 11, emergent light enters afifth lens 13 after being adjusted by a secondlinear polarizer 12, parallel light is focused on the back focal plane of the lens, and partial light enters anobjective lens 15 through a half-mirror and irradiates on a sample on an objective table 16. The sample is selected to be quantum dots with the diameter of 15-20 nanometers, and can be excited by laser with the diameter of 488 nanometers. The 8. mu.M quantum dots were pipetted 8. mu.L onto a 0.17mm glass slide and placed on the sample stage. The light reflected on the sample passes through thehalf mirror 14 and thefourth mirror 17 to enter theCCD detector 18.

When the CCD detector receives the images, a time sequence of images is taken, and the two-dimensional image data (x, y) is developed into a three-dimensional or four-dimensional video image (x, y, omega)_x,ω_y). And carrying out Fourier transform on the information corresponding to all the pixel positions, integrating different frequencies to obtain information values corresponding to all the frequencies, wherein each frequency value corresponds to a formula (diffraction superposition) containing a plurality of position sample information, and finally solving through a model established by a corresponding algorithm. The specific algorithm is to calculate out a point spread function, set up an led array matrix (wavelength is not limited), the modulation frequency of each led is different, and set up 9 × 9 in the actual operation to be randomThe frequency matrix is as follows:

the PSF point spread matrix is set as follows (and can also be set to gaussian):

when the algorithm simulation light path is verified, an image is input as shown in fig. 2, and then the result of superposition of image information at various frequencies at different times can be seen after two times of diffusion, as shown in fig. 4. The 9 x 9 optical path algorithm verifies that the calculations are fast, and the individual stage time is shown in table 1 (unit: seconds).

TABLE 1

The above embodiments are used for explaining and understanding the technical solutions of the present invention, and do not limit the ideas and technical solutions of the present invention.

Claims (3)

Translated fromChinese

1.一种基于主动时间调制混频激发照射的超分辨显微成像方法，其特征在于，采用空间光调制技术对像素矩阵上的每个元素按照不同的频率进行调制，最终通过倒置显微成像系统对被CCD探测器收集的图像信息经过算法建立的模型进行解析，得到图像信息中所涵盖的样本信息值，从而实现超分辨成像的功能；具体步骤如下：1. a super-resolution microscopic imaging method based on active time modulation mixing excitation illumination, it is characterized in that, adopts spatial light modulation technology to modulate each element on the pixel matrix according to different frequencies, and finally by inverted microscopic imaging The system analyzes the image information collected by the CCD detector through the model established by the algorithm, and obtains the sample information value covered by the image information, thereby realizing the function of super-resolution imaging; the specific steps are as follows:假设在激发光源阵列的位置(x,y)处设置调的制频率为ν(x,y),则在位置为(x₀,y₀)处的分子激发频谱为：Assuming that the modulation frequency is set as ν(x, y) at the position (x, y) of the excitation light source array, the molecular excitation spectrum at the position (x₀ , y₀ ) is:ΣI(x-x₀,y-y₀)·v(x,y)， (1)ΣI(xx₀ ,yy₀ ) v(x,y), (1)其中，I为位置(x,y)处光强对位置(x₀,y₀)的影响分布函数；Among them, I is the distribution function of the influence of the light intensity at the position (x, y) on the position (x₀ , y₀ );(1)对于每个阵列像素光源位置处利用低频激发光进行调制，调制方式为：(1) The low-frequency excitation light is used for modulation at the light source position of each array pixel, and the modulation method is:I_{Laser_Power}(x_Laser,y_Laser,v)＝I_{Laser_Power}(x_Laser,y_Laser)·sin(2πv(x_Laser,y_Laser)) (2)I_{Laser_Power} (x_Laser ,y_Laser ,v)=I_{Laser_Power} (x_Laser ,y_Laser )·sin(2πv(x_Laser ,y_Laser )) (2)其中，I_{Laser_Power}(x_Laser,y_Laser)为在(x,y)位置处的激光光强，v(x_Laser,y_Laser)为按照正弦函数调制激光幅值的调制频率；Among them, I_{Laser_Power} (x_Laser , y_Laser ) is the laser light intensity at the (x, y) position, and v (x_Laser , y_Laser ) is the modulation frequency that modulates the laser amplitude according to the sine function;在一个单像素(x₀,y₀)上的样本，受到周围不同低频光强调制激光束影响，其发射光强为：A sample on a single pixel (x₀ , y₀ ) is affected by the modulation of the laser beam with different low-frequency intensity around it, and its emission intensity is:

其中，I_{Laser_Power}(x_Laser-x₀,y_Laser-y₀)是激光激发位置(x_Laser,y_Laser)对于位置(x₀,y₀)的光强贡献函数，P(x₀,y₀)是样本对于光强幅值的影响函数；Among them, I_{Laser_Power} (x_Laser -x₀ , y_Laser -y₀ ) is the light intensity contribution function of the laser excitation position (x_Laser , y_Laser ) to the position (x₀ , y₀ ), P(x₀ , y_{0 )} ) is the influence function of the sample on the light intensity amplitude;最终经过衍射，任一像素点(x,y)的光强都受到周围衍射范围内的分子影响；Finally, after diffraction, the light intensity of any pixel point (x, y) is affected by the molecules in the surrounding diffraction range;(2)经过像素阵列调制激发，拍摄一段时间的视频；然后，通过时间傅里叶变换分析各像素，将二维图像数据(x,y)发展成三维或四维视频图像(x,y,ω_x,ω_y)，其中，ω_x,ω_y分别是(x,y)方向上的频率，这里不仅包括位置信息(x,y),还包括二维频率信息(ω_x,ω_y)；(2) After pixel array modulation and excitation, a video is captured for a period of time; then, each pixel is analyzed by time Fourier transform, and the two-dimensional image data (x, y) is developed into a three-dimensional or four-dimensional video image (x, y, ω)_x , ω_y ), where ω_x , ω_y are the frequencies in the (x, y) direction, respectively, including not only the position information (x, y), but also the two-dimensional frequency information (ω_x , ω_y );(3)最终通过傅里叶变换，把各个调制频率所对应的像素信息提取出来，构建方程模型，从而解析出被衍射掩盖的超分辨图单像素下的具体频率信息。(3) Finally, through the Fourier transform, the pixel information corresponding to each modulation frequency is extracted, and the equation model is constructed, so as to analyze the specific frequency information under the single pixel of the super-resolution image masked by diffraction.2.根据权利要求1所述的超分辨显微成像方法，其特征在于，步骤(3)中，所述构建方程模型，解析出被衍射掩盖的超分辨图单像素下的具体频率信息的过程如下：2. super-resolution microscopic imaging method according to claim 1, is characterized in that, in step (3), described building equation model, resolves the process of specific frequency information under the super-resolution image single pixel that is masked by diffraction as follows:步骤一：根据实际光学系统模型进行如下假设：Step 1: Make the following assumptions based on the actual optical system model:(1)假设透镜对相位因子的改变量为零，将透镜看做薄透镜；(1) Assuming that the change amount of the lens to the phase factor is zero, the lens is regarded as a thin lens;(2)由于每个光源频率不同，考虑非相干光源成像的情况，将成像过程等效为光学传递函数，而衍射是由入瞳或者出瞳所引起的；(2) Since the frequency of each light source is different, considering the imaging of incoherent light sources, the imaging process is equivalent to an optical transfer function, and the diffraction is caused by the entrance pupil or the exit pupil;(3)样品面和CCD接受面不考虑高频信息的损耗，扩散矩阵近似为高斯矩阵形式的函数；(3) The sample surface and the CCD receiving surface do not consider the loss of high-frequency information, and the diffusion matrix is approximated as a function in the form of a Gaussian matrix;(4)假设采集到足够多的图像，能够完全恢复出每个像素上的时间函数；(4) Assuming that enough images are collected, the time function on each pixel can be completely recovered;步骤二：根据假设对实际光学系统进行模拟，具体频率解析算法的流程如下：Step 2: Simulate the actual optical system according to the assumption. The specific frequency analysis algorithm flow is as follows:将每个像素都写成像素矩阵形式，如果对应位置上有值，即表示有对应的频率扩散到此像素上；设初始每个像素上只有自己对应的频率，进行一次扩散卷积计算，将不同的频率成分混合到扩散范围内的矩阵上，并且不同频率成分按照扩散矩阵对应位置携带上权重因子；于是每个像素上的值如下：Write each pixel in the form of a pixel matrix. If there is a value in the corresponding position, it means that there is a corresponding frequency diffused to this pixel; suppose that each pixel has only its own corresponding frequency at the beginning, and perform a diffusion convolution calculation. The frequency components of are mixed into the matrix within the diffusion range, and different frequency components carry weighting factors according to the corresponding positions of the diffusion matrix; then the value of each pixel is as follows:

其中，(2*a+1)为扩散矩阵的维数，约等于像素矩阵维数的1/3；扩散举阵D的大小为(2a+1)*(2a+1)，(x,y)为像素位置；Among them, (2*a+1) is the dimension of the diffusion matrix, which is approximately equal to 1/3 of the dimension of the pixel matrix; the size of the diffusion matrix D is (2a+1)*(2a+1), (x,y ) is the pixel position;经过一次扩散点扩散函数后所得到的不同位置上的像素携带该像素位置上的样本信息；携带样品信息后，对所有像素再一次进行扩散卷积计算，对于每个像素，将其周围像素按照扩散矩阵对它的贡献进行叠加，得到每个像素上的实际频率值；经过两次扩散卷积计算后，将所得到的像素上所有元素相加，得到最终每个像素包含样品信息的表达式：The pixels at different positions obtained after passing through the diffusion point spread function once carry the sample information at the pixel position; after carrying the sample information, the diffusion convolution calculation is performed on all pixels again, and for each pixel, the surrounding pixels are calculated according to The diffusion matrix superimposes its contribution to obtain the actual frequency value on each pixel; after two diffusion convolution calculations, add all elements on the obtained pixel to obtain the final expression that each pixel contains sample information :

其中，n*n为图像矩阵维度；Among them, n*n is the image matrix dimension;再依此按照设计的频率值进行傅里叶变换，之后对某一给定频率取合适的范围积分，即进行如下处理：Then perform Fourier transform according to the designed frequency value, and then take an appropriate range integral for a given frequency, that is, perform the following processing:

即先傅里叶变换，再利用冲击函数的筛选性质从积分频段你中得到特定频率所携带的强度信息值；其中，ω_n,m为矩阵(n,m)位置上的频率利用余弦函数傅里叶变化变为δ冲击函数，冲击函数的筛选性质能够通过积分获得一个频段信息所包含的幅值信息；a,b分别为所取调制频段的最小值和最大值；That is, first Fourier transform, and then use the screening property of the shock function to obtain the intensity information value carried by a specific frequency from the integral frequency band; where ω_{n, m} is the frequency at the position of the matrix (n, m), using the cosine function Fourier The Liye change becomes a delta shock function, and the screening property of the shock function can obtain the amplitude information contained in a frequency band information through integration; a and b are the minimum and maximum values of the selected modulation frequency band, respectively;通过式(6)和式(7)对式(5)进行处理，称为频率解析算法，得每个调制频率所对应的携带图像信息变量的式子，因为图像矩阵的维度为n*n，所以最终获得n*n个含未知数的式子；将最终CCD探测器获得的图像时间序列集按照式(6)和式(7对所有像素信息进行分析，得到每个调制频率信息对应的实际的值，与前面得到的对应调制频率所获得的式子组成等式，一共有n*n个等式，按照线性方程组进行求解，得到最终的结果，还原出样本信息值。Equation (5) is processed by Equation (6) and Equation (7), which is called a frequency analysis algorithm, and the expression of the image information variable corresponding to each modulation frequency is obtained, because the dimension of the image matrix is n*n, Therefore, n*n formulas containing unknowns are finally obtained; the image time series set obtained by the final CCD detector is analyzed according to formula (6) and formula (7) to analyze all pixel information, and obtain the actual corresponding to each modulation frequency information. value, and the formula obtained from the corresponding modulation frequency obtained earlier to form an equation, there are a total of n*n equations, which are solved according to the linear equation system to obtain the final result and restore the sample information value.3.一种基于权利要求1或2所述方法的超分辨显微成像装置，其特征在于，包括：激光光源1，第一透镜2，第二透镜3，第一反射镜4，第二反射镜5，第三透镜6，第四透镜7，第三反射镜8，第一线偏振片9，二分之一波片10，空间光调制器11，第二线偏振片12，第五透镜13，半透半反镜14，物镜14，载物台16，第四反射镜17，CCD探测器18；这些部件依次光路连接组成超分辨显微成像装置；其中：3. A super-resolution microscope imaging device based on the method of claim 1 or 2, characterized in that, comprising: a laser light source 1, a first lens 2, a second lens 3, a first reflecting mirror 4, a second reflecting mirror Mirror 5, third lens 6, fourth lens 7, third mirror 8, first linear polarizer 9, half-wave plate 10, spatial light modulator 11, second linear polarizer 12, fifth lens 13 , the half mirror 14, the objective lens 14, the stage 16, the fourth mirror 17, the CCD detector 18; these components are connected in turn by the optical path to form a super-resolution microscopic imaging device; wherein:从激光光源1输出的激光依次经由第一透镜2和第二透镜3进行扩束，扩束后的光依次经由第一反射镜4和第二反射镜5进入第三透镜6和第四透镜7进行第二次扩束；两次扩束后的激光经过第三反射镜8进入到第一线偏振片9，经过第一线偏振片得到的线偏光通过二分之一波片10进行旋转，直到平行于空间光调制器的长边；此时入射光才能够被空间光调制器11调制，出射光经过第二线偏振片12调整后进入第五透镜13，把平行光聚焦在透镜背焦面处；再经由半透半反镜14使部分光进入物镜15，照射在载物台16的样品上，样品上反射的光透过半透半反镜14和第四反射镜17进入CCD探测器18；The laser light output from the laser light source 1 is expanded through the first lens 2 and the second lens 3 in sequence, and the expanded light enters the third lens 6 and the fourth lens 7 through the first reflecting mirror 4 and the second reflecting mirror 5 in sequence. Carry out the second beam expansion; the laser beam after the two beam expansion enters the first linear polarizer 9 through the third reflector 8, and the linearly polarized light obtained through the first linear polarizer rotates through the half-wave plate 10, Until it is parallel to the long side of the spatial light modulator; at this time, the incident light can be modulated by the spatial light modulator 11, and the outgoing light enters the fifth lens 13 after being adjusted by the second linear polarizer 12, and the parallel light is focused on the back focal plane of the lens Then, through the half mirror 14, part of the light enters the objective lens 15 and is irradiated on the sample on the stage 16, and the light reflected on the sample enters the CCD detector 18 through the half mirror 14 and the fourth mirror 17 ;由CCD探测器18拍摄一个连续时间序列的原始图像集合，其中每个像素都包含不同的频率信息和样品信息；生成的原始图像序列经过频率解析算法提取原始图像集的频率信息；原始图像集合还具有样品的检测辐射的强度信息数据，而上述频率解析算法将强度信息随时间变化的规律通过傅里叶变换转换为带不同幅值强度信息的不同频率。The CCD detector 18 captures a continuous time series of original image sets, in which each pixel contains different frequency information and sample information; the generated original image sequence is subjected to frequency analysis algorithm to extract the frequency information of the original image set; the original image set also It has intensity information data of the detected radiation of the sample, and the above-mentioned frequency analysis algorithm converts the law of intensity information changing with time through Fourier transform into different frequencies with different amplitude intensity information.

Images