阅读说明：本技术 一种基于多角度照明反卷积的高分辨率显微成像方法 (High-resolution microscopic imaging method based on multi-angle illumination deconvolution ) 是由 张晓磊 左超 孙佳嵩 胡岩 沈德同 尹维 于 2021-07-26 设计创作，主要内容包括：发明公开了一种基于多角度照明反卷积的高分辨率显微成像方法。与传统明场显微镜采用的科勒照明方式不同,本系统采用LED阵列作为照明光源,通过依次采集不同照明角度下的物体图像,获取不同空间频率成分的物体信息,然后根据系统参数计算出不同照明角度对应的强度传递函数对图像进行最小二乘反卷积重构。相比传统明场显微镜同时获取所有空间频率信息的成像方法,本发明通过使用最小二乘反卷积的算法,对噪声具有更强的鲁棒性。(The invention discloses a high-resolution microscopic imaging method based on multi-angle illumination deconvolution. Different from a Kohler illumination mode adopted by a traditional bright field microscope, the system adopts an LED array as an illumination light source, obtains object information of different spatial frequency components by sequentially collecting object images under different illumination angles, and then calculates intensity transfer functions corresponding to different illumination angles according to system parameters to perform least square deconvolution reconstruction on the images. Compared with the traditional imaging method for simultaneously acquiring all spatial frequency information by using a bright field microscope, the method has stronger robustness to noise by using the least square deconvolution algorithm.)

1. A high-resolution microscopic imaging method based on multi-angle illumination deconvolution is characterized by comprising the following steps:

acquiring an original image, namely using a programmable LED array as an illumination light source of an imaging system, sequentially lightening LED units of the LED array, and acquiring a series of image sequences I under different illumination angles by a camera under synchronous trigger signals_m,n(x, y), wherein I is a light intensity image, (x, y) is a spatial coordinate, and m, n are position coordinates of the LED unit corresponding to the light intensity image in the LED array;

calculating the illumination light space frequency vector (u) corresponding to each LED unit_m,u_n)；

Step three, calculating a coherent intensity transfer function ATF corresponding to the LED unit with the LED array coordinate of (m, n)_m,n；

Performing least square deconvolution on the image sequence obtained in the step one, and firstly performing intensity image sequence I_m,nFourier transform is carried out on (x, y) to obtain an intensity map frequency spectrum sequence(u, v) are frequency domain coordinates, and then a least squares deconvolution is performed to obtain a reconstructed image I_deconv(x,y)。

2. The method as claimed in claim 1, wherein in step two, the illumination light spatial frequency vector (u) corresponding to each LED unit is used as the illumination light spatial frequency vector_m,u_n) The calculation formula of (2) is as follows:

the LED array comprises a programmable LED array and a plurality of LED units, wherein m is the mth LED unit in the X direction of the LED array, n is the nth LED unit in the Y direction of the LED array, h is the distance between the programmable LED array and a sample to be tested, d is the distance between the LED units, and lambda is the illumination wavelength.

3. The method for high-resolution microscopic imaging based on multi-angle illumination deconvolution as claimed in claim 1, wherein in the third step, the coherent intensity transfer function ATF_m,nThe calculation formula of (2) is as follows:

where (u, v) are frequency domain coordinates,for coherent illumination cut-off frequency, NA, of the objective lens_objIs the objective lens numerical aperture.

4. The high-resolution microscopic imaging method based on the multi-angle illumination deconvolution as claimed in claim 1, wherein in step four, the least square deconvolution formula is:

wherein, I_deconv(x, y) is the intensity map of the deconvoluted image, F^-1And (4) performing inverse Fourier transform.

Technical Field

The invention belongs to an optical microscopic imaging technology, and particularly relates to a high-resolution microscopic imaging method based on multi-angle illumination deconvolution.

Background

Since Leeuwenhoek observed cells for the first time three hundred years ago using a manual home-made microscope, human beings have continued to explore many in the field of microscopic imaging. Among them, high-resolution microscopic imaging for the purpose of better clarity and finer definition is one of the important points. In 1873, Abbe published the microscope diffraction imaging theory in German microscope report, scientifically elucidated the imaging process of the microscope, and first proposed that the optical resolution of the microscope is limited and limited by the objective value (Zhengxiang. modern microscopy imaging technology review [ J ] optics, 2015(6): 550-.

In bright field microscopic imaging under traditional kohler illumination, since diffracted light cannot completely pass through an objective lens, a point in an object space is imaged by a microscope and then is mapped into a diffraction spot, which results in reduction of resolution. The method of calculating the optical resolution of the system here often uses the rayleigh criterion, which states that when the diffraction maximum of one point coincides with the first diffraction minimum of another point, then just two points can be resolved. (Katsumasa Fujita, Follow-up review: recovery in the depth of super-resolution optical micro [ J ]]Microcopy, Volume 65, Issue 4, August 2016, 275-281) under the Rayleigh criterion, the optical resolution f of the microscope_RComprises the following steps:

as described above, since the optical resolution of the microscope is determined by the numerical aperture of the objective lens when the illumination light is constant, the objective lens having a larger numerical aperture is required to look finer at the object. However, the optical resolution cut-off frequency under the rayleigh criterion is only the resolution limit for direct observation by the human eye and not the theoretical resolution limit of the optical imaging system. The theoretical resolution limit f (Chao Zuo, Transport of interest evaluation: a tutorial [ J ], Optics and Lasers in Engineering, Volume 135,2020, 106-:

wherein, NA_illThe numerical aperture of the illumination source is the numerical aperture of the objective lens NA_objSmaller, NA_illUsually with NA_objEqual, the theoretical resolution limit f of the optical system at this time is:

however, the optical imaging system often cannot reach its theoretical resolution limit, because the transmission capability of the optical imaging system to information is attenuated with increasing frequency, and the attenuation to 0 is reached near the cutoff frequency, which may cause high-frequency information in a certain range to be submerged in noise. Therefore, the problem that the actual resolution of the optical imaging system is difficult to reach the theoretical resolution limit needs to be solved urgently. Here we propose a high resolution microscopic imaging method based on multi-angle scanning, which can solve the above problems.

Disclosure of Invention

The invention aims to solve the problem that the diffraction limit resolution cannot be achieved due to the fact that the signal-to-noise ratio at the cut-off frequency is too low under the traditional Kohler illumination.

The technical scheme of the invention is as follows:

a high-resolution microscopic imaging method based on multi-angle illumination deconvolution comprises the following steps:

acquiring an original image, namely using a programmable LED array as an illumination light source of an imaging system, sequentially lightening LED units of the LED array, and acquiring a series of image sequences I under different illumination angles by a camera under synchronous trigger signals_m,n(x, y), wherein I is a light intensity image, (x, y) is a spatial coordinate, and m, n are position coordinates of the LED unit corresponding to the light intensity image in the LED array;

calculating the illumination light space frequency vector (u) corresponding to each LED unit_m,u_n)；

Step three, calculating a coherent intensity transfer function ATF corresponding to the LED unit with the LED array coordinate of (m, n)_m,n；

Performing least square deconvolution on the image sequence obtained in the step one, and firstly performing intensity image sequence I_m,nFourier transform is carried out on (x, y) to obtain an intensity map frequency spectrum sequence(u, v) are frequency domain coordinates, and then a least squares deconvolution is performed to obtain a reconstructed image I_deconv(x,y)。

Preferably, in step two, the illumination light spatial frequency vector (u) corresponding to each LED unit_m,u_n) The calculation formula of (2) is as follows:

the LED array comprises a programmable LED array and a plurality of LED units, wherein m is the mth LED unit in the X direction of the LED array, n is the nth LED unit in the Y direction of the LED array, h is the distance between the programmable LED array and a sample to be tested, d is the distance between the LED units, and lambda is the illumination wavelength.

Preferably, in step three, the coherent intensity transfer function ATF_m,nThe calculation formula of (2) is as follows:

where (u, v) are frequency domain coordinates,for coherent illumination cut-off frequency, NA, of the objective lens_objIs the objective lens numerical aperture.

Preferably, in step four, the least squares deconvolution formula is:

wherein, I_deconv(x, y) is the intensity map of the deconvoluted image, F^-1And (4) performing inverse Fourier transform.

The invention has the beneficial effects that: (1) the programmable LED array is used for shooting a light intensity image sequence of an object to be detected under multi-angle illumination by using an angle scanning method, different spatial frequency components in incoherent illumination can be subjected to high signal-to-noise ratio and high dynamic range information acquisition, and the method has the characteristics of higher imaging resolution and better imaging effect. (2) Based on the actual parameters of the system, the light intensity transfer functions under different illumination angles are calculated, then deconvolution reconstruction is carried out on the shot light intensity graph sequence based on the least square method, the obtained result is more accurate, and the robustness to noise is stronger.

Drawings

FIG. 1 is a flow chart of a high resolution microscopic imaging method based on multi-angle illumination deconvolution

FIG. 2 is a graph of intensity images under different angle illumination, intensity image spectra and corresponding coherence intensity transfer function ATF_m,n。

Fig. 3 is a comparison of the composite intensity transfer function using this method and a conventional incoherent illumination intensity transfer function.

Fig. 4 is a comparison of imaging results using this method, using conventional incoherent imaging results.

Detailed Description

In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.

With reference to fig. 1, this embodiment is a high resolution microscopic imaging method based on multi-angle illumination deconvolution, including the following steps:

the method comprises the following steps: under multi-angle illuminationAnd collecting an intensity image. The programmable LED array is used as an illumination light source of the system, the LED units are sequentially lightened, the central unit of the LED array needs to be aligned to the optical axis of the imaging system, and the LED units emit monochromatic or white approximate plane waves with different angles to irradiate on a sample each time. Simultaneously, the monochrome or color camera starts exposure after receiving a synchronous trigger signal from the programmable LED array, acquires the image of the object to be measured under multi-angle illumination and forms a monochrome or color light intensity image sequence I_m,n(x, y). Wherein, I (x, y) is the light intensity image, and m, n are the position coordinates of the LED unit corresponding to the light intensity image in the LED array.

Step two: an illumination light spatial frequency vector (u) for each LED unit based on the position parameters of the programmable LED array_m,u_n) And (3) calculating:

the LED array comprises a programming LED array and a plurality of LED units, wherein m is the mth LED unit in the X direction of the LED array, n is the nth LED unit in the Y direction of the LED array, h is the distance between the programming LED array and a sample to be tested, d is the distance between the LED units, and lambda is the illumination wavelength.

Step three: calculating the coherent intensity transfer function ATF corresponding to each LED unit_m,n. As shown in FIG. 2, the coherent intensity transfer function ATF_m,nIntensity image I under multi-angle illumination_m,n(x, y) corresponding to each other, if the position parameters of the programmable LED array are calibrated accurately, the calculated coherent intensity transfer function ATF_m,nShould be correlated with the intensity image I_m,nThe spectral shape of (a) is consistent. Coherent intensity transfer function ATF_m,nThe calculation formula of (a) is as follows:

wherein (u)_m,u_n) Is the corresponding illuminating light spatial frequency vector, f, of the LED unit_cCut-off frequency for objective coherent illumination:

FIG. 2 is an intensity image spectrumCoherent intensity transfer function ATF_m,nAnd intensity image I under different angle illumination_m,n(x, y), it can be seen that at these several illumination angles, the intensity image spectrumShape and coherence strength transfer function ATF of_m,nThe matching situation is good.

Step four: according to the coherent intensity transfer function ATF obtained in the third step_m,nAnd performing least square deconvolution on the image sequence obtained in the step one to obtain a high-resolution imaging result of the sample to be detected. Firstly, Fourier transform is carried out on a shot intensity map sequence I (x, y) to obtain an intensity map frequency spectrum sequenceThen, performing least square deconvolution, wherein the deconvolution formula is as follows:

wherein, I_deconv(x, y) is the intensity map of the deconvoluted image, F^-1And (4) performing inverse Fourier transform.

Fig. 3 is a comparison of the composite intensity transfer function using this method and a conventional incoherent illumination intensity transfer function. It can be seen that the information transfer capability of the composite intensity transfer function of the method in the medium-high frequency region is higher than that of the traditional incoherent imaging.

Fig. 4 is a comparison of the imaging result obtained by using the method (right) and the conventional incoherent illumination imaging result (left), and it can be seen that the method has higher resolution and better imaging effect.

The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.