Spectral microscopic imaging device and implementation method

文档序号:167945 发布日期:2021-10-29 浏览:21次 中文

阅读说明:本技术 一种光谱显微成像装置及实现方法 (Spectral microscopic imaging device and implementation method ) 是由 颜成钢 柴夏媛 吕彬彬 孙垚棋 张继勇 李宗鹏 于 2021-06-29 设计创作,主要内容包括:本发明公开了一种光谱显微成像装置及实现方法,包括依次设置的照明光源、载物台、显微物镜、视场光阑、闪耀光栅、带通滤波器、4F中继透镜、微透镜阵列、平移台、掩膜和CCD阵列;所述的4F中继透镜共有两组,分别设置在带通滤波器和微透镜阵列之间以及平移台和CCD阵列之间;本发明装置通过光路设计,达到多次曝光并通过算法即可得到清晰样本多路连续光谱通道的目的,可以获得观察样本多维清晰图像信息。去除了其中的一组不必要的4F中继透镜,使得结构更紧凑,光通量更大,像差变好。增加了平移台和掩膜,有效弥补了由于将图像升维之后降采样引起的像素降低问题。(The invention discloses a spectral microscopic imaging device and an implementation method thereof, wherein the spectral microscopic imaging device comprises an illumination light source, an objective table, a microscope objective, a field diaphragm, a blazed grating, a band-pass filter, a 4F relay lens, a micro lens array, a translation table, a mask and a CCD array which are sequentially arranged; the 4F relay lenses are divided into two groups and are respectively arranged between the band-pass filter and the micro lens array and between the translation stage and the CCD array; the device achieves the purpose of obtaining a clear sample multi-path continuous spectrum channel through multiple exposures and an algorithm through light path design, and can obtain multi-dimensional clear image information of an observed sample. An unnecessary group of 4F relay lenses are removed, so that the structure is more compact, the luminous flux is larger, and the aberration is better. The translation stage and the mask are added, and the problem of pixel reduction caused by down-sampling after the image is subjected to dimension lifting is effectively solved.)

1. The spectral microscopic imaging device is characterized by comprising an illumination light source (1), an objective table (2), a microscope objective (3), a field diaphragm (4), a blazed grating (6), a band-pass filter (7), a 4F relay lens (5), a micro-lens array (8), a translation table (12), a mask (13) and a CCD array (9) which are sequentially arranged; the 4F relay lenses (5) have two groups which are respectively arranged between the band-pass filter (7) and the micro-lens array (8) and between the translation stage (12) and the CCD array (9);

the illumination light source (1) irradiates the objective table (2), two-dimensional image information of a sample on the objective table is acquired in an imaging lens of the microscope objective (3), the two-dimensional image information is imaged on a plane where the field diaphragm (4) is located, and the two-dimensional image information irradiates the surface of the blazed grating (6); the blazed grating (6) disperses the transmitted light under different spectral wavelengths to different angles, and the band-pass filter (7) enables the spectral band to be recorded in the +1 level with the highest blazed grating brightness to pass through independently and shields other bands and light rays on other grating levels; at the moment, light after grating dispersion is converged on a plane where a micro-lens array (8) is located again through a 4F relay lens (5), light with different wavelengths is focused on a micro-lens focal plane behind the micro-lens array (8), a continuous spectrum is spread in a word along the grating dispersion direction, and a spread image is projected onto a CCD array (9) through a second 4F relay lens (5); numerical aperture matching is needed before and after the whole optical path system, namely the difference between the numerical aperture of the light projected onto the micro lens array (8) and the numerical aperture of the micro lens array (8) cannot exceed a set threshold value; the translation stage (12) and the mask (13) are used for sharpening the obtained unsharp image after down sampling.

2. The method for realizing the spectral microscopic imaging device according to claim 1, characterized by comprising the following steps:

the method comprises the following steps: light is emitted from the illumination light source (1) and is irradiated onto the objective table (2), and an observation object on the objective table (2) is illuminated; then, an imaging lens of the microscope objective 3 projects a real image of an object to be observed on a plane where a field diaphragm (4) is located, namely on a blazed grating 6; at the moment, the real image of the observed object is superposed with the groove surface of the blazed grating (6), and the real image of the observed object on the surface of the blazed grating (6) is subjected to dispersion;

step two: due to the band-pass characteristic of the band-pass filter (7), spectral wave bands L1-Ln to be recorded in +1 level with highest blazed grating brightness pass through independently and are converged on the micro lens array (8) again through the 4F relay lens (5);

step three: because the dispersion angles of real images of observed objects mapped on the surface of the blazed grating (6) are different, light with different wavelengths is converged on the micro-lens array (8) again, so that the real images have different emergence angles, the light is dispersed along one dimension on a micro-lens focal plane (10), the light penetrates through an irregular mask to form an image formed by overlapping the mask and the object images, the observed dispersed real images of the objects are imaged on a pixel array of a CCD (charge coupled device) array (9) through a 4F relay lens (5), and image information is obtained through a camera;

step four: the sub-pixel (11) area in the pixel array of the CCD array (9) corresponds to each microlens in the microlens array 8, the size of the sub-pixel (11) is N × N pixels, wherein N is an odd number, and 3 < N < 13, and emergent light passing through the microlens is projected onto a middle row of pixels of the sub-pixel 11, at this time, pixels at corresponding positions of the middle row in the sub-pixel (11) are recombined in a manner that ith pixels of (N +1)/2 th rows in the sub-pixel (11) corresponding to each microlens are combined into an ith image Ai according to the microlens position sequence, wherein i is 1, 2 … … N, so that a spectral image Ai corresponding to an object observed on the objective table (2) at a wavelength of λ i is obtained, wherein λ i is L1+ (i-0.5) × (Ln-L1)/N; recording image information;

step five: moving the mask 13 in the mask plane by using the translation stage 12 for k x k times, moving for 1 micron each time, recording the image information by adopting the method of the step four, repeating the k x k times in total, and recording k x k pieces of image information;

step six: obtaining N groups of k × k images, and performing the following operations on each group of single-wavelength images:

the input is J (k) original images, and an initial guess O is given to the object and illumination light at the beginning of the iterationj,PjWherein the initial guess illumination light is P1(x0,y0),j∈[1,J]:

(1) From the shifted x, y information, with P and the shifted value xj、yjTo indicate after movingMask information Pj=P(x0-xj,y0-yj) (ii) a Using formulasA simulated superimposed image is calculated, wherein,representing the superimposed image, OjRepresenting the object image corresponding to the j-th image, PjThe mask information after moving;

(2) obtaining the distance d from the camera to the image surface by using the point spread function PSF and the formula psij(x, y) ═ psf (d) × (x, y) the simulated wavefront was calculated, where ψjRepresents the simulated wavefront, phijRepresenting the superimposed image;

(3) calculating the simulated light Intensity of | ψ from the square of the light Intensity equal to the amplitudej(x,y)|2

(4) Upsampling (update) the real image with the simulated light intensity and expanding the pixels to m x m times;

(5) using formulasCalculating a wavefront for converting the image into an m x m image, wherein ψj' is the updated simulated wavefront;

(6) according to the distance d from the camera to the image surface, using a point spread function PSF and a formula Performing an inverse transformation equivalent to (3) to obtain a new true wavefront;

(7) substitution formula Updating the data of object 0, where conj is the conjugate, αobjIs the algorithm weight in rPIE;

(8) substitution formula Updating data of the illumination light P, wherein αpIs the algorithm weight in rPIE;

(9) repeating the steps (1) to (9), and adding one to J each time until J is equal to J;

(10) repeating the steps (1) to (11) n times untilIs small enough.

(11) Taking the object O obtained by final optimization as the complex amplitude of the image with improved resolution, and reconstructing the image into I ═ O-2

Step seven: for each group of single-wavelength images in the step six, obtaining the image I with improved resolutioniAs an image of the current wavelength.

3. The method for realizing a spectral microscopic imaging device according to claim 2, wherein further in step six, αobj=αp=1,n=6。

Technical Field

The invention relates to the field of spectral imaging, in particular to a multiple spectrum microscopic imaging device.

Background

Compared with the traditional imaging technology, the spectral imaging technology can record multi-dimensional spectral information and increase the abundance of recorded information by using a mask and an algorithm while shooting a scene two-dimensional image and increasing the dimension, overcomes the defect of pixel reduction after the dimension is increased, and is favorable for later analysis and processing. Due to the fact that the micro-lens array can enable the light source to spread a line with different spectrum wavelengths along the middle of the sub-pixels so as to independently collect information of each spectrum, image down-sampling can be achieved, the mask-based PIE technology can achieve a super-resolution technology, and the problem of low pixel caused by down-sampling is effectively solved. In the initial stage of the spectral imaging technology, the spectral information is acquired by using a traditional method, and then pixel compensation is performed by using PIE (pixel image enhancement), namely, after the two-dimensional spatial information and the spectral information at the corresponding wavelength are recorded by using a narrow-band filter, a clearer image is obtained by using a mask and an algorithm. The method has the advantages of high precision, easy realization and capability of ensuring the definition of the image while acquiring a plurality of spectral channels.

The rapid spectral microscopic imaging technology can make up for the situation that the definition of an image is still ensured after the acquisition of a plurality of spectral channels, so that spectral data are richer and pixels are more ensured. Therefore, the rapid spectral microscopic imaging technology can effectively solve the problem of pixel reduction after the image is subjected to dimension raising.

Disclosure of Invention

Aiming at the defects in the prior art, the invention provides a spectral microscopic imaging device. The method can record information of a plurality of continuous spectral channels and make up for the defect of pixel reduction after the dimension of the image is increased.

A spectral microscopic imaging device comprises an illumination light source (1), an objective table (2), a microscope objective (3), a field diaphragm (4), a blazed grating (6), a band-pass filter (7), a 4F relay lens (5), a micro-lens array (8), a translation table (12), a mask (13) and a CCD array (9) which are sequentially arranged. The 4F relay lenses (5) have two groups which are respectively arranged between the band-pass filter (7) and the micro-lens array (8) and between the translation stage (12) and the CCD array (9);

the illumination light source (1) irradiates on the objective table (2), two-dimensional image information of a sample on the objective table is acquired in an imaging lens of the microscope objective (3), the two-dimensional image information is imaged on a plane where the field diaphragm (4) is located, and the two-dimensional image information irradiates on the surface of the blazed grating (6). The blazed grating (6) disperses the transmitted light under different spectral wavelengths to different angles, and the band-pass filter (7) enables the spectral band to be recorded in the +1 stage with the highest blazed grating brightness to pass through independently and shields light on other bands and other grating stages. At the moment, light after grating dispersion is converged on a plane where the micro-lens array (8) is located again through the 4F relay lens (5), light with different wavelengths is focused on a micro-lens focal plane behind the micro-lens array (8), the continuous spectrum is spread in a word along the grating dispersion direction, and the spread image is projected onto the CCD array (9) through the second 4F relay lens (5). The numerical aperture matching is needed before and after the whole optical path system, namely the difference between the numerical aperture size of the light projected onto the micro lens array (8) and the numerical aperture size of the micro lens array (8) cannot exceed a set threshold value, and the numerical aperture size is close to the set threshold value as possible so as to avoid image overlapping confusion. The translation stage (12) and the mask (13) are used for sharpening the obtained unsharp image after down sampling.

By adopting the structure, because the micro-lens array performs sampling segmentation on the imaging in the visual field, the imaging of different spectrum channels can be focused in different pixel coordinates, and the pixels at corresponding positions in the sub-pixels are selected to be recombined to obtain corresponding spectrum information.

The invention has the following beneficial effects:

the device achieves the purpose of obtaining a clear sample multi-path continuous spectrum channel through multiple exposures and an algorithm through light path design, and can obtain multi-dimensional clear image information of an observed sample. An unnecessary group of 4F relay lenses are removed, so that the structure is more compact, the luminous flux is larger, and the aberration is better. The translation stage 12 and the mask 13 are added, so that the problem of pixel reduction caused by down-sampling after the image is subjected to dimension raising is effectively solved.

Drawings

FIG. 1 is a schematic structural diagram of a spectral microscopic imaging device according to the present invention.

Reference numerals: the device comprises an illumination light source 1, an objective table 2, a microscope objective 3, a field diaphragm 4, a 4F relay lens 5, a blazed grating 6, a band-pass filter 7, a micro-lens array 8, a CCD array 9, a micro-lens focal plane 10, a sub-pixel 11, a translation table 12 and a mask 13.

Detailed Description

The invention relates to a spectral microscopic imaging device, which is characterized in that a plurality of exposures are used for acquiring a plurality of continuous spectral information of a biological sample and obtaining a clear image, and the spectral microscopic imaging device comprises the following steps:

referring to fig. 1, a spectral microscopic imaging device includes an illumination light source 1, an objective table 2, a microscope objective 3, a field diaphragm 4, a blazed grating 6, a band-pass filter 7, a 4F relay lens 5, a microlens array 8, a translation table 12, a mask 13 and a CCD array 9, and is sequentially arranged from left to right; the 4F relay lenses 5 are divided into two groups and respectively arranged between the band-pass filter 7 and the micro-lens array 8 and between the translation stage 12 and the CCD array 9;

the illumination light source 1 illuminates an observed object on the objective table 2, the imaging lens of the microscope objective 3 images a real image of the observed object on the plane where the field diaphragm 4 is located, and then the real image is mapped on the surface of the blazed grating 6, and at the moment, the real image of the observed object is superposed with the groove surface of the blazed grating 6.

The real image of the observation object mapped on the surface of the blazed grating 6 is dispersed, and the band-pass filter 7 enables the spectral band to be recorded in the +1 level with the highest blazed grating brightness to pass through independently and to be converged on the micro-lens array 8 again through the 4F relay lens 5. Then, dispersion occurs along one dimension on the microlens focal plane 10, and a real image of the observed object after dispersion is imaged on a pixel array of the CCD array 9 through the 4F relay lens 5. A translation stage 12 and a mask 13 are arranged between the micro-lens array 8 and the CCD array 9 and are used for sharpening the obtained unsharp image after down-sampling.

A method for realizing a spectral microscopic imaging device comprises the following steps:

the method comprises the following steps: light is emitted from the illumination light source 1 to the stage 2, and the observation object on the stage 2 is illuminated. Then, the imaging lens of the microscope objective 3 projects the real image of the object to be observed on the plane where the field diaphragm 4 is located, that is, on the blazed grating 6. At this time, the real image of the observation object overlaps the groove surface of the blazed grating 6, and the real image of the observation object on the surface of the blazed grating 6 is dispersed.

Step two: due to the band-pass characteristic of the band-pass filter 7, the spectral bands L1 to Ln to be recorded in the +1 level with the highest blazed grating brightness pass through individually and are converged on the microlens array 8 again through the 4F relay lens 5.

Step three: because the dispersion angles of the real images of the observed objects mapped on the surface of the blazed grating 6 are different, light with different wavelengths is converged on the microlens array 8 again, so that the real images have different emergence angles, the light is dispersed along one dimension on the microlens focal plane 10, the light penetrates through an irregular mask to form an image formed by overlapping the mask and the object images, the observed dispersed real images of the objects are imaged on a pixel array of the CCD array 9 through the 4F relay lens 5, and image information is obtained through a camera.

Step four: the specific sub-pixel 11 area in the pixel array of the CCD array 9 corresponds to each microlens in the microlens array 8, the size of the sub-pixel 11 is N × N pixels, where N is an odd number and 3 < N < 13, and the emergent light passing through the microlens is projected onto the middle row of pixels of the sub-pixel 11, at this time, the pixels at the corresponding positions in the middle row of the sub-pixels 11 are recombined in such a way that the ith pixel of the (N +1)/2 th row in the sub-pixel 11 corresponding to each microlens is combined into the ith image Ai according to the microlens position order, where i is 1 and 2 … … N, so as to obtain the spectral image Ai corresponding to the object observed on the stage 2 at the wavelength λ i, where λ i is L1+ (i-0.5) (Ln-L1)/N. The image information is recorded.

Step five: the mask 13 is moved by the translation stage 12 in the mask plane a total of k x k times, each time 1 micron, and the image information is recorded by the method of step four, which is repeated a total of k x k times, and k x k pieces of image information are recorded.

Step six: obtaining N groups of k × k images, and performing the following operations on each group of single-wavelength images:

the input is J (k) original images, and an initial guess O is given to the object and illumination light at the beginning of the iterationj,PjWherein the initial guess illumination light is P1(x0,y0),j∈[1,J]:

(1) From the shifted x, y information, with P and the shifted value xj、yjTo represent the mask information P after movementj=P(x0-xj,y0-yj) (ii) a Using formulasA simulated superimposed image is calculated, wherein,representing the superimposed image, OjRepresenting the object image corresponding to the j-th image, PjThe mask information after moving;

(2) obtaining the distance d from the camera to the image surface by using the point spread function PSF and the formula psij(x, y) ═ PSF (d) x, y calculating the simulated wavefront, where ψjRepresents the simulated wavefront, phijRepresenting the superimposed image;

(3) calculating the simulated light Intensity of | ψ from the square of the light Intensity equal to the amplitudej(x,y)|2

(4) Upsampling (update) the real image with the simulated light intensity and expanding the pixels to m x m times;

(5) using formulasCalculating a wavefront for converting the image into an m x m image, wherein ψj' is the updated simulated wavefront;

(6) according to the distance d from the camera to the image surface, using a point spread function PSF and a formula Performing an inverse transformation equivalent to (3) to obtain a new true wavefront;

(7) substitution formula Updating the data of the object O, where conj is the conjugate, alphaobjIs the algorithm weight in rPIE;

(8) substitution formula Updating data of the illumination light P, wherein αpIs the algorithm weight in rPIE;

(9) repeating the steps (1) to (9), and adding one to J each time until J is equal to J;

(10) repeating the steps (1) to (11) n times untilIs small enough.

(11) Taking the finally optimized object 0 as the complex amplitude of the image with improved resolution, and then reconstructing the image into I ═ O-2

Step seven: for each group of single-wavelength images in the step six, obtaining the image I with improved resolutioniAs an image of the current wavelength.

Further, αobj=αp=1,n=6。

7页详细技术资料下载
上一篇:一种医用注射器针头装配设备
下一篇:一种小直径直侧视电动旋转对焦内窥镜

网友询问留言

已有0条留言

还没有人留言评论。精彩留言会获得点赞!

精彩留言,会给你点赞!